• TABLE OF CONTENTS
HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 Abstract
 Introduction
 Social and behavioral responses...
 Internal family structure and decision...
 Political and economic dimensions...
 GUSII: The land, its people and...
 Study site characteristics and...
 Findings and discussion
 Quantitative analysis of the ethnographic...
 Appendices
 References
 Biographical sketch














Title: Lay people's responses to illness
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100652/00001
 Material Information
Title: Lay people's responses to illness an ethnographic study of anti-malaria behavior among the Abagusii of southwestern Kenya
Physical Description: Book
Language: English
Creator: Nyamongo, Isaac K
Publisher: University of Florida
Place of Publication: Gainesville Fla
Gainesville, Fla
Publication Date: 1998
Copyright Date: 1998
 Subjects
Subject: malaria -- lay people -- behavior -- health-seeking
Anthropology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Abstract: ABSTRACT: Malaria is re-emerging with a vengeance after near eradication a generation ago. In Africa, it kills between 1 and 2 million people, mostly children, each year and many more are incapacitated leading to enormous human and economic loss. I present data on responses to malaria among the Abagusii, an agricultural, Bantu-speaking people who inhabit the fertile rainy highlands of southwestern Kenya. Data on knowledge of common illnesses in Bomorenda, the site for the study, and malaria symptoms, were collected using systematic ethnographic methods. Malaria-focused narratives from 35 informants yielded additional ethnographic information of lay people's responses. The study reveals that Abagusii have multiple notions of malaria causation. A majority (85.7%) consider the mosquito as the major cause of malaria. Other causes are eating sugary foods (57.1%) and witchcraft (34.3%). They assign naturalistic causation as well as associate malaria with environmental dangers. These findings support the indigenous contagion theory. Abagusii recognize the same symptoms as those in the clinical case definition of malaria. Using symptom salience as a measure of importance, headache (salience = .447), shivering (.393), fever (.226), vomiting (.188) and paining joints (.158) are the top five symptoms.
Abstract: Transition state probability results show that the longer the illness lasts, the more likely that the illness will be treated outside the home (transition probability =.772). Over 82% of lay people report self-treatment as a first choice (t). The percentage of people who use self-treatment drops to 12.5% at the time of second choice (t+1) and zero at time three (t+2) while those who seek treatment outside the home increases. Illnesses that last long are regarded as serious and patients prefer taking those illnesses to private or public health care facilities where they are likely to get specialized attention (Fisher's exact test p <.0001, Cramer's V = .719). Lay people in Gusii purchase a variety of drugs for malaria management from local shops. A cognitive map of informants reveals that they arrange these drugs along a dimension based on age of the patient and along a malaria--analgesic drugs dimension. Although informants have good knowledge regarding drug dosage, sometimes they get wrong information about the administration of different drugs. This has implications for the immediate management of malaria and for the long-term effects of improper use including the development of drug resistant parasites. Three quadratic assignment procedure (QAP) analyses indicate no gender differences with regard to lay people's responses to malaria-focused ethnographic interviews and similarity among illnesses and malaria drugs. The r-square for the three QAP analyses range between 0.72 and 0.88.
Abstract: A biocultural model is used to show that ecological and cultural factors play an important role in sustaining mosquito density in Bomorenda. Farming practices and type of houses constructed provide optimal conditions for Anopheline mosquitoes that transmit Plasmodium parasites. The utility of the biocultural model is assessed and policy implications drawn.
Thesis: Thesis (Ph. D.)--University of Florida, 1998.
Bibliography: Includes bibliographical references (p. 297-304).
System Details: System requirements: World Wide Web browser and PDF reader.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Isaac Keango Nyamongo.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains xii, 321 p.; also includes graphics.
General Note: Vita.
 Record Information
Bibliographic ID: UF00100652
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 50751626
alephbibnum - 002424945
notis - AMD0025

Downloads

This item has the following downloads:

master ( PDF )


Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
        Page iv
        Page v
    Table of Contents
        Page vi
        Page vii
        Page viii
        Page ix
    Abstract
        Page x
        Page xi
        Page xii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
    Social and behavioral responses to malaria infection in Kenya
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
    Internal family structure and decision making: Its relevance for treatment and control of disease
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
    Political and economic dimensions of health care delivery in Kenya: Implications for the national malaria control
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
    GUSII: The land, its people and social systems
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
    Study site characteristics and data collection
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
    Findings and discussion
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
    Quantitative analysis of the ethnographic data from Bomorenda
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
        Page 265
        Page 266
        Page 267
        Page 268
        Page 269
        Page 270
        Page 271
        Page 272
        Page 273
        Page 274
        Page 275
        Page 276
        Page 277
        Page 278
    Appendices
        Page 279
        Page 280
        Page 281
        Page 282
        Page 283
        Page 284
        Page 285
        Page 286
        Page 287
        Page 288
        Page 289
        Page 290
        Page 291
        Page 292
        Page 293
        Page 294
        Page 295
        Page 296
    References
        Page 297
        Page 298
        Page 299
        Page 300
        Page 301
        Page 302
        Page 303
        Page 304
        Page 305
        Page 306
        Page 307
        Page 308
        Page 309
        Page 310
        Page 311
        Page 312
        Page 313
        Page 314
        Page 315
        Page 316
        Page 317
        Page 318
        Page 319
        Page 320
    Biographical sketch
        Page 321
Full Text











LAY PEOPLE'S RESPONSES TO ILLNESS: AN ETHNOGRAPHIC STUDY OF
ANTI-MALARIA BEHAVIOR AMONG THE ABAGUSII OF SOUTHWESTERN
KENYA












By

ISAAC KEANGO NYAMONGO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998































To my mother, the Late Rhoda Bosibori Nyamongo















ACKNOWLEDGEMENTS


Many people have contributed in innumerable ways to this dissertation. First, I

would like to thank my informants in Bomorenda who braved my incessant questioning

and who were kind enough to sit through the long interviews and triad questionnaires that

sometimes seemed confusing. I also would like to thank my research assistants-Judith

Moraa Onsomu, Vincent Onditi Ombasa, Tom Ondimu Machuka, and Risper Abuya.

Though they all contributed immensely, I would like to single out Judith, who always took

charge whenever I was away from the field. She brought into the team her anthropological

training. Special thanks to Karani, a student of Anthropology at the Institute of African

Studies, University of Nairobi, who, with his friend, transcribed the taped interviews.

My appreciation to Prof. Paul N. Nkwi, President, Pan-African Anthropological

Association (PAAA). Prof. Nkwi was critical in organizing training workshops for the

Network of African Medical Anthropologists (NAMA) between 1992 and 1993 in

Yaounde and Douala, Cameroon. It is at those workshops that I met my mentor, Prof. H.

Russell Bernard.

In the US, I would like to thank the following: George Ngong Mbeh, Ken

Sturrock, Gery Ryan, Girma Hundie and Sharon Morrison. George has been a steady

pillar of support. I used him on numerous occasions as a sounding board. He has read the









whole manuscript. Ken read several sections of the dissertation and was kind enough to

offer suggestions. Gery was very helpful during the development of the research proposal.

Girma and Sharon have been a constant source of encouragement.

This dissertation would not have come this far but for the wonderful support from

my dissertation committee. My dissertation chair, Prof. H. Russell Bernard, has taught me

a lot inside and outside class. I have benefited in many ways from his enthusiasm, wide

range of interests, excellent scholarship and teaching abilities. He has been a mentor in the

fullest sense. Prof. Marvin Harris has influenced and shaped my thinking on theory. Dr.

Leslie Sue Lieberman and Dr. Della McMillan have both been very helpful in many ways.

Dr. Martin D. Young, my external committee member, made available to me his work on

malaria spanning over four decades.

Financial support from the Wenner Gren Foundation for Anthropological

Research, the Center for African Studies, University of Florida, the Deans' Committee,

University of Nairobi, and the Department of Anthropology, University of Florida, is

gratefully acknowledged. The Wenner Gren Foundation supported my graduate studies

through their Developing Countries Training Fellowship program. The Center for African

Studies gave me the initial grant to carry out a preliminary field survey and the Deans'

Committee, University of Nairobi, provided the grant for the study. The Department of

Anthropology, University of Florida, extended me a graduate assistantship during the

write-up period. I am also grateful to the Office of the President, Republic of Kenya, for

issuing me a permit to carry out the research.









Finally, I would like to thank my family for they have supported me in many ways.

My father has shown keen interest in my education. He has supported me through this

entire process. My wife, Mary, and our two sons-Jared and Jansen, have given me support

in many different ways-intellectual, social, emotional-than I can enumerate. Although

this work has kept us apart halfway round the world their support has been steadfast.















TABLE OF CONTENTS
page


A CK N O W LED G EM EN TS ......................................................... ........... iii

A B STR A C T ........................................................................................ .x

CHAPTERS

1 IN TR OD U CTION ............................................................................. 1

Biocultural Theoretical Approach................................................ ..... 3
Outline of the Dissertation ....................................................... ............ 9


2 SOCIAL AND BEHAVIORAL RESPONSES TO MALARIA
INFECTION IN KENYA ........... ...... ...... 12

Introduction .................................................................................. 12
The Biology of M alaria ....................................... . .......... ............. 14
Cause, Transmission, and Epidemiology of Malaria............................. 14
The Life Cycle of the Malaria Parasite...................................... ..... 17
Impact of Malaria on the Population.................................................. 19
Effects on the Local Economy ........................... ................................ 19
Effects on Pregnancy............................................................... 21
Effects on Genes: Adaptive Advantages of Sickle-cell Trait.........................27
Manifestation of Malaria and Lay People's Responses............................... 31
Manifestation of Malaria .................................. ............................ 31
Clinical Categories of Malaria..................................................... 32
Recognition of Malaria by Lay People.............................................. 34
Lay People's Responses to Malaria.................................... ........... 36
Factors Affecting Health Behavior................................... ................. 39
Indigenous Concepts of Disease Causation......................................... 40
C ost of T reatm ent.................. ........................................ ............42
H household C om position ................................................... ............. 44
D decision M akers .................................................. .......... ........... 45
The Emics and Etics of People's Response to Malaria................................ 45
R recognition of M alaria .................................................... .......... 46
C ause of M alaria............................................................ ...........46









Management of Malaria .................................... ............... ........... 47
D disease Progression ........................................................ ........... 47
Conclusion .................................................................................. .. 47


3 INTERNAL FAMILY STRUCTURE AND DECISION MAKING:
ITS RELEVANCE FOR TREATMENT AND CONTROL OF DISEASE ........... 54

Introduction .......................................................... ....................... 54
Determinants of Health Seeking Behavior............................................... 56
Illness Characteristics and Perceived Seriousness............................... 57
Lay People's Knowledge and Categorization of the Illness........................ 58
H health C are E expenses ...................................................... .......... 59
Distance to the Health Care Facility............................................. 61
Social Networks of Patients and Caretakers....................................... 62
Religious Affiliation of Patients and Caretakers.................................. 64
Health Care Behavior: A Case Study from Meru District, Kenya................... 65
Studies on D decision M making .............. .. ......................... ... ............... 67
Decisions Under Risk and Under Uncertainty..................................... 68
Approaches in the Study of Decision Making..................................... 68
Framing Decisions: Its Influence on Decision Making.......................... 73
Decision Making: External and Internal Factors................................... 75
Decision Models and their Prediction Power...................................... 79
Decision Making and the Family.................................................. 82
Therapy Management Groups ................................. ................ .......... 85
C conclusion ....................................................................... .......... 87


4 POLITICAL AND ECONOMIC DIMENSIONS OF HEALTH CARE
DELIVERY IN KENYA: IMPLICATIONS FOR THE NATIONAL
MALARIA CONTROL.................................. ..... ............ 93

Introduction ................................ .................................................. 93
The State and the Provision of Health Care............................................ 94
Disease and Community Intervention................................... .............. 98
History of Health Care in Kenya...................... .................... ..... .... .. 100
Pre-Colonial Phase ................................................................... 101
Colonial Phase ............................................................. .......... 103
Post-Colonial Phase ....................................................... .......... 107
Kenya and the Bamaki Initiative................................................ 109
The Health Care and Administrative Structure.................................... 111
M alaria Control-A Global View ...................... ............ .................... 113
Florida ...................................................................... .......... 115
M auritiu s ................................................................... .. . . .... 116
M alaria Control in K enya .................................................... .......... 118









History of Malaria Control in Kenya............. ......... .... .............. 119
Political and Economic Factors: Implications for Malaria Control
in Kenya ................................................................. ........... 126
Conclusion ...................................................................... .......... 126


5 GUSII: THE LAND, IT'S PEOPLE AND SOCIAL SYSTEMS..................... 134

Introduction ..................................................................... .......... 134
G eo-Clim atic Factors ......................................................... .......... 135
Abagusii Totems and Lineage .............................................................. 137
The Gusii Social Order ........................................ ............... ........... 143
Social Structure ................................................ ........... ........... 143
Living A rrangem ents ................................................. ............... 144
Sacrifices: Keeping Misfortunes at Bay.......................................... 146
Changing Life Situations and Decision Making in Gusii........................... 148
Therapy Management in the Gusii Family............................................... 150
What Place Do I have in Gusii? Doing Native Ethnography.......................... 152


6 STUDY SITE CHARACTERISTICS AND DATA COLLECTION.................165

Introduction ..................................................................... .......... 165
The Study Site................................... .............. ...... .......... 166
Administrative Characteristics of the Suneka Division......................... 166
Health Care Facilities in Suneka Division............................................. 168
Characteristics of the Study Population............................ ................... 169
Data Collection ............................................................ .... ........... 173
Freelist of Illnesses in Bomorenda............................ ...... ............ 174
Freelist of Malaria Symptoms ............................. ............................ 175
Triad Tests for 16 Illnesses.................... ..................................... 176
Pile Sorts Using Drugs for Malaria Management............................... 178
Malaria Focused Narratives ............................. ............................... 180
Field Work Dynamics .................................................................... 183


7 FINDINGS AND DISCUSSION.................................................... 191

Introduction .............................................................. ... ... .. .......... 19 1
Knowledge of Illnesses Common in Bomorenda..................................... 192
K now ledge of M alaria ............................................. ........... ........... 196
M alaria R recognition ........................................... ........... .......... 196
Vomiting: Expelling Malaria.................................................... 198
Causes of Malaria .................................................................... 200
Treatm ent of M alaria ..................................................... ........... 209

viii









M management of Enlarged Spleen................................................. 218
Management of Malaria: Treatment Choices ........... ... ........... 220
Factors Affecting Utilization of Health Care Services in Bomorenda.............. 226
Cost of Treatment ........................................................ ........... 227
Availability of Drugs ..................................................... .......... 228
Friendly and M motivated Staff...................... ............... ................. 228
Intensity of the Disease .................................. ................. ......... .. 230
Knowing the Health Care Provider................... ................ ............ 230
Information Exchange and Treatment of Malaria......................... ........... 234
Accounting for Malaria Increase in Bomorenda.......................... ........... 235


8 QUANTITATIVE ANALYSIS OF THE ETHNOGRAPHIC
DATA FROM BOM ORENDA............................................... .. 251

Testing for Differences Among Informants.............................. ............. 252
Treatment Transitions-How Does Duration of Illness Influence Use of
Health Care Alternatives? ............ ............... ........... 254
Maintaining Good Credit Record with Private Health Care Providers............. 257
M alaria, Sugary Foods and Use of Pills............................ ......... .......... 258
Malaria Symptoms, Illness Recognition and Treatment Seeking.................... 259
Predicting Outcomes from Informants' Characteristics............................. 260
Conclusion ................................................................... ............ 262
Summary of the Findings ................................................... 262
Utility of the Biocultural Model and Policy Implications........................ 267


APPENDICES

A TRIADS QUESTIONNAIRE FOR 16 ILLNESSES IN
BOM OREND A ............................................................ ......... 279

B MALARIA HEALTH CARE DECISIONS-GRAND TOUR
SURVEY GUIDE ........................................................ ........... 282

C FILL-IN FORM FOR RECORDING MALARIA INTERVIEWS.................. 284

D CODES FOR MALARIA NARRATIVES ........................... ............... 285

E QUOTES FROM INFORM ANTS........................................ ..............289

REFERENCES ........................................................................ ........... 297

BIOGRAPHICAL SKETCH............................ ................................. 321















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


LAY PEOPLE'S RESPONSES TO ILLNESS: AN ETHNOGRAPHIC STUDY OF
ANTI-MALARIA BEHAVIOR AMONG THE ABAGUSII OF SOUTHWESTERN
KENYA


By

Isaac Keango Nyamongo

December 1998

Chairperson: Dr. H. Russell Bernard
Major Department: Anthropology

Malaria is re-emerging with a vengeance after near eradication a generation ago. In

Africa, it kills between 1 and 2 million people, mostly children, each year and many more

are incapacitated leading to enormous human and economic loss. I present data on

responses to malaria among the Abagusii, an agricultural, Bantu-speaking people who

inhabit the fertile rainy highlands of southwestern Kenya. Data on knowledge of common

illnesses in Bomorenda, the site for the study, and malaria symptoms, were collected using

systematic ethnographic methods. Malaria-focused narratives from 35 informants yielded

additional ethnographic information of lay people's responses.

The study reveals that Abagusii have multiple notions of malaria causation. A

majority (85.7%) consider the mosquito as the major cause of malaria. Other causes









are eating sugary foods (57.1%) and witchcraft (34.3%). They assign naturalistic

causation as well as associate malaria with environmental dangers. These findings support

the indigenous contagion theory. Abagusii recognize the same symptoms as those in the

clinical case definition of malaria. Using symptom salience as a measure of importance,

headache (salience = .447), shivering (.393), fever (.226), vomiting (.188) and paining

joints (.158) are the top five symptoms.


Transition state probability results show that the longer the illness lasts, the more

likely that the illness will be treated outside the home (transition probability =.772). Over

82% of lay people report self-treatment as a first choice (t). The percentage of people who

use self-treatment drops to 12.5% at the time of second choice (t+1) and zero at time

three (t+2) while those who seek treatment outside the home increases. Illnesses that last

long are regarded as serious and patients prefer taking those illnesses to private or public

health care facilities where they are likely to get specialized attention (Fisher's exact test p

< .0001, Cramer's V= .719).

Lay people in Gusii purchase a variety of drugs for malaria management from local

shops. A cognitive map of informants reveals that they arrange these drugs along a

dimension based on age of the patient and along a malaria-analgesic drugs dimension.

Although informants have good knowledge regarding drug dosage, sometimes they get

wrong information about the administration of different drugs. This has implications for

the immediate management of malaria and for the long-term effects of improper use

including the development of drug resistant parasites. Three quadratic assignment









procedure (QAP) analyses indicate no gender differences with regard to lay people's

responses to malaria-focused ethnographic interviews and similarity among illnesses and

malaria drugs. The r-square for the three QAP analyses range between 0.72 and 0.88.

A biocultural model is used to show that ecological and cultural factors play an

important role in sustaining mosquito density in Bomorenda. Farming practices and type

of houses constructed provide optimal conditions for Anopheline mosquitoes that transmit

Plasmodium parasites. The utility of the biocultural model is assessed and policy

implications drawn.















CHAPTER 1
INTRODUCTION


Malaria is a world-wide problem. It was almost eradicated a generation ago, but

has come back from the brink with a vengeance. In Africa alone it kills between 1 and 2

million people every year, mostly children. Over 100 million people are exposed to the

malaria causing parasite each year in Africa and over 500 million are classified as at risk.

In addition to deaths, morbidity from malaria causes great economic harm. It is estimated

that between 3 and 7 days of work are lost per case of malaria, causing a loss of about 0.8

billion dollars (Shephard et al. 1991). I discuss these problems in greater detail in Chapter

two.

Despite early efforts to eradicate malaria, the disease is re-emerging in many areas

where it was thought to have been controlled. Resurgence of malaria is due to many

factors-environmental, genetic and cultural. The world over, human activities are

causing changes in the global environment creating conditions conducive to the

development of the mosquito, the vector responsible for the transmission of Plasmodium

spp., the malaria causing parasite. Computer simulations reveal that an increase of

between 0.25 C and 0.5 C in the global temperature could lead to an increase in the

incidence of malaria on the order of 50% to 100% in regions of low endemicity and a

reduction of between 15% to 30% in areas of high endemicity, the latter effect being the

result of increase in immunity (Janssen and Martens 1996: 26-27). Based on these

1








2
projections, a best case scenario would result in an overall increase of malaria in the

range of 20% while the worst case scenario would result in a 85% increase in malaria

cases annually. These predictions assume a steady state situation in demographic, social

and economic development.

Governmental and non-governmental bodies have instituted programs to improve

health care in rural communities where these services are weak. For their part,

governments are dealing with the problem of malaria through enhancement of

institutional capabilities and in the training of community based health care providers.

These people are responsible for providing health care to communities located in rural

areas, far from health care facilities. In certain remote areas they are the only people who

provide western health care. While the programs recognize the importance of

community-based approaches to the control of malaria, they have been less successful in

their implementation and sustainability. They fail to sufficiently consider the social and

cultural context of malaria (Tanner and Vlassoff 1998).

Governments face many problems in their endeavors to control malaria, including

lack of adequate funding. More important, however, is the emergence of parasites

resistant to drugs used for malaria treatment. Due to inadequate capacity to deal with

malaria in developing countries, lay people1 devise strategies to combat the malaria

problem at home (home case management). Home case management practices include

treatments administered at home as well as the search for and selection of outside health




' I use the term lay people throughout this dissertation to refer to non-health
professionals involved in day to day health problems.









care providers. Treatments offered at home include use of herbal remedies, over-the-

counter pharmaceuticals, and dietary changes (Ryan 1998). For them to start

appropriate treatment, lay people rely on correct diagnosis of the specific health

problem.

Actual treatment seeking behavior relies on several factors. This dissertation

explores these factors using data from Bomorenda in Gusii. Specifically, the focus is on

lay people's reaction when they suffer malaria.

In the next section I discuss the biocultural model and its relationship to the

medical ecology model; components of the biocultural approach; critiques

(predominantly from the critical medical anthropology group) leveled against both the

biocultural and medical ecology models; and the reasons why I have adopted the

biocultural approach in this study. I place the biocultural model within the cultural

materialist paradigm (Harris 1979).



Biocultural Theoretical Approach



There is a vast literature on determinants of disease rates and factors that

influence health care behavior of patients (Suchman 1965; Mechanic 1968, 1969; Colson

1971; McKinley 1973; Fosu 1981; Young 1980, 1981a; Mwabu 1986; Mathews and Hill

1990; Bentley 1992; Snow et al. 1994; Ryan 1995, 1998; Ryan and Martinez 1996). The

factors that influence health care include knowledge regarding specific illnesses (Fabrega

1971, 1975; Lipowsky et al. 1992), presence and accessibility of health care facilities









(Snow et al. 1994), and the cost of services (Gould 1957;Yoder 1989). Ecological

factors such as weather patterns, temperature, terrain and infrastructural factors,

people's beliefs and practices also influence disease patterns. These factors fit into a

three tier framework: infrastructural, structural, and superstructural (cultural). I discuss

these broad categories using a biocultural theoretical approach. The biocultural

approach places emphasis on both the cultural and the biological aspects of people's

response to health and suffering (Armelagos et al. 1992). The biocultural model borrows

heavily from the medical ecology model; the primary difference between these two

models is in the orientation of the researchers.

According to Brown et al. (1996) studies in disease ecology include at least three

levels of causation: (1) a microbiological level, which focuses on agents of disease acting

within the human body; (2) a cultural ecological (or microsociological) level, which

focuses on how individual behaviors-encouraged or constrained by sociocultural

factors-put people at risk for contracting certain diseases; and (3) a political ecological

(or macrosociological) level, which focuses on historical factors involving interactions

between human groups and how these factors shape people's access to resources.

Brown et al. (1996: 189) point out that studies in disease ecology are a "biocultural

enterprise" which "allows and requires a bridging of the biological and cultural

paradigms in anthropology." The malaria situation in Bomorenda can be explained

following these three levels. Although the data did not focus directly on the

microbiological level, it has been long established that malaria is the product of

interaction between parasites, the human host and cultural factors. I have devoted a









section of Chapter two to the biology of malaria-the interaction between the disease

causing agents, the parasites, and human beings.

There is no major difference between the ecological model and the biocultural

model (Townsend and McElroy 1992). Townsend and McElroy argue that any

differences which exist reflect technical or disciplinary affiliation. The ecological model

is used primarily by biological anthropologists to analyze biocultural responses to

disease (Armelagos et al. 1992). The "biocultural" term is used more often by cultural

anthropologists while the "ecological" term is used by biological anthropologists. Wiley

(1992) makes a finer distinction between the two however. According to Wiley the

biocultural orientation adds an historical perspective in the analysis while medical

ecological models do not necessarily consider such a diachronic perspective.

Ecological anthropologists place emphasis on the adaptive nature of human

beings; they see the environment, which is composed of the biotic, abiotic, and cultural

components as a stress causing agent to which we adapt. Ecological models allow the

integration of biological, demographic, and socio-cultural systems in accounting for

human patterns of adaptation (Townsend and McElroy 1992). This adaptive response

may be biological or behavioral. In the case of malaria, Livingstone (1958) showed the

development of the sickle-cell trait to be a biological response to the increased incidence

of malaria in tropical Africa. This arose from changes in the environment caused by new

agricultural practices. Behavioral and technological responses to malaria include: use of

drugs to treat the illness, use of mosquito nets, insecticides, and repellents, along with

construction of houses with protective screens, to prevent illness. Genetic responses, of









course, take many generations to stabilize in a population and hence are rare. The most

common responses are behavioral and technological. The general adaptive responses are

discussed further in Chapter two.

The medical ecology and the biocultural models have been heavily criticized by

some medical anthropologists who offer a competing approach (see for example Singer

1989; Singer 1990; Singer et al. 1992; Baer 1997). Critical medical anthropology theory

faults the medical ecology theory for placing emphasis on the individual rather than

looking for causal variables at the political and economic level. They argue that health

issues need to be understood within the context of political and economic forces that

influence human relationships, shape social behaviors, condition collective experiences,

re-order local ecologies, and situate cultural meaning (e.g. Singer et al. 1992). They

place greater significance on the structural level (political and economic forces), while

ecological theorists place greater emphasis on the infrastructural level-the interaction

between human beings and the natural environment-and then on the structural level and

cultural level. Scheper-Hughes (1990) calls for the radicalization of medical knowledge

and practice so as to focus on the afflicted. This implies that critical medical

anthropologists should see themselves as championing the cause of the patient. In Baer's

(1997: 1568) words, critical medical anthropologists, whether "in academia or in a

clinical setting, need to become proponents of 'patient power.'"

Critical medical anthropologists accuse medical ecologists of "stopping short of

'real analysis' in that their work does not focus on the political and economic origins of

illness" (Wiley 1992: 219). Those critical medical anthropology theorists who fault








7
medical ecological theory (e.g. Singer 1989) criticize it for being too biomedical and too

adaptationist. They use a narrow and limiting definition of adaptation and seem to

equate adaptation with "the perfect fit" between an organism and its environment (Wiley

1992). Although their criticism holds true of some past formulations of the medical

ecology model, "[the] limitations do not characterize the emergent biocultural

perspective" (Armelagos et al. 1992 :38).

The emergent paradigm of medical ecology is holistic and encompasses political

and economic forces. Political and economic forces have their base in the environment

(the infrastructure). The infrastructure influences the structure and the ideology of a

people (superstructure). Thus a model (theory) that places greater emphasis on the

structure or superstructure would have lower explanatory power than one that is based

on the infrastructure. Such a model would take into account factors such as terrain,

altitude, and weather conditions that are part of the infrastructure.

I do not imply that critical medical theory is of no use. Since the medical ecology

model places greatest emphasis on the infrastructure and then on the structure and

superstructure, it must have a higher explanatory power following the principle of

infrastructural determinism (Harris 1979). I find the materialist approach and the

principle of infrastructural determinism particularly useful in explaining the malaria

situation in Bomorenda.

Even critical scholars in the field admit that their approach gives "scant attention

to ecological factors" (Baer 1996: 129). There is no arguing that ecological factors

influence the incidence of disease around the world. The fact that many infectious









diseases are in the tropical regions has less to do with political and economic factors

initially. It is the prevailing ecological factors, such as warm climate and rainfall, that are

of first-order significance. Political will and economic muscle only come into play to

mobilize the resources that will bring success to the struggles of those who live in such

environments.

In what follows, I apply the biocultural model to analyze the malaria situation in

Bomorenda of Kisii District in Kenya, the major components of which are presented in

Figure 1.1. The model takes into account the cost of services, the distribution of health

care facilities, and the drug supply, as well as the cost of remuneration to health care

professionals, along with local beliefs regarding malaria, lay people's health care

practices, land inheritance rights and farming patterns, the type of local houses, and

parasite resistance to drugs. The same model allows me to analyze these factors which

are critical to long-term malaria control within a historical context that integrates both

ethnomedical and biomedical perspectives (Fabrega 1975; Kleinman and Mendelsohn

1978).

The utility of the biocultural model as a tool for understanding "coping behavior

within the context of ecological-political-cultural systems" (Armelagos et al. 1992) is

analyzed for the conclusion of Chapter eight. Based on this analysis I conclude that the

search for health care is indeed, tied to infrastructural, structural, and ideological factors.

There is no getting around the importance of culture here: the views of professionals

regarding malaria causation and treatment regimen sometimes differ from the views of

lay people. These differences influence the search for health care and the outcomes of









health care choices. By placing these cultural differences within a more comprehensive

cultural materialist and ecological explanation, we achieve greater understanding of the

dynamics of response to malaria in Gusii.



Outline of the Dissertation


The next Chapter provides a review of the literature about responses to malaria

around the world and, in particular, in Kenya. This review includes a critique of single

country studies that examine the cost of malaria, the effect of malaria on pregnancy, the

problem of compliance and the emergence of drug resistant parasites as well as how lay

people recognize and react to malaria. This literature shows that the family plays a key

role in health care. I discuss the family's role in the provision of health care to its

members in Chapter three.

Ill health affects the family's resources particularly its financial base. Those who

are sick withdraw from active participation in economic activities and resources are

redirected toward the sick individuals in order to restore their health. This in turn affects

other social aspects of the family. When patients/caretakers consider a case of malaria to

be serious, they may opt for choices that destabilize the family's economic and social

base in the short term. For less serious illnesses patients/caretakers may hold on longer

so as to assess disease progression before taking action. The social support patients

receive is found within the family or within the patient's social network. Health care

decisions are made, usually, within these circles. These same decisions are constrained








10
by a wide variety of macro-economic and institutional factors over which they have no

control (Figure 1.1).

This macro-economic and institutional context is described for Kenya in Chapter

four. In Chapter five, I provide background information on the Abagusii people-their

history, social systems, and, in particular, their actions and beliefs to keep misfortunes at

bay-whose decision making process provides the focus of the analysis presented in

Chapters six, seven, and eight. I discuss the special challenges of studying ones own

community in Chapter five. Chapter six deals with the methods used in data gathering,

processing, and analysis of the research which was conducted over a 10-month period

between February 1997 and November 1997 in Bomorenda, Kisii district. In Chapter

seven and Chapter eight I present the study findings and conclusions.













Main effect

,Beliefs



Patient
Type of houses behavior
Farming patterns Ecology
Land inheritance rights/regulations


CULTURAL-ECOLOGICAL



Renumeration of health care providers
Distribution of drugs
Accessibility Patient
Distribution of health care facilities behavior
Cost




POLITICAL-ECOLOGICAL


Drug resist
Parasites Parasite
Immunity transmissi1
Genetics


MICRO-BIOLOGICAL


Fiumre 1 1: A Biociultural Model of Malaria in Bomorenda















CHAPTER 2
SOCIAL AND BEHAVIORAL RESPONSES TO MALARIA INFECTION IN KENYA


Introduction


Malaria was one of the first vector borne diseases to be subject to widespread

efforts of control. One reason is the disease's huge economic impact on families and on

nations. Unfortunately, however, reduction in the incidence of malaria during the late

'50s and early '60s have been wiped out and now humans in malaria-endemic areas are

under great stress (Oaks et al. 1991:1). Malaria has increased worldwide, especially in the

last two decades due to the emergence of parasites resistant to chloroquine (the most

widely used drug for the control of malaria) and of mosquitoes resistant to insecticides

(Table 2.1).

In many countries where the disease is widespread, control programs are in

operation. As in the 1950s and 1960s, WHO is the driving force behind many of these

programs, which include providing people with insecticide-impregnated bed nets and

prophylactic medicine, and spraying against mosquitoes. Although research shows that a

combination of these methods can offer real protection against the disease, people have

not readily accepted these programs and compliance is often dismal. Helitzer-Allen et al.

(1994) argue that anti-malaria campaigns can only be successful if there is client demand.

To date, however, there is little understanding of the combined effect of infrastructural,








13
structural and cultural factors on client demand. Client demand can be manipulated better

by understanding local knowledge about illness and health.

It seems obvious that the study of illness behavior should help us design more

effective health delivery and control programs everywhere. While this is an often-

invoked assertion, it is rarely demonstrated. Illness behavior refers to any behavior

associated with conditions that cause individuals to concern themselves with disease and

to seek help (Mechanic 1968). By disease I mean the aggregate symptoms (abnormalities

of the body organs or organ systems) that lay people recognize to identify an ailment. I

use the word "infection" instead of "disease" to mean the presence of parasites in the

body without necessarily there being outward symptoms. Thus, a person can be infected

with parasites (or germs) that cause a particular disease and yet, to lay people, such a

person might be considered healthy because outward and recognizable symptoms are not

manifest.

People recognize a particular disease at different stages depending on familiarity

with symptoms and the progression of the illness (Feierman 1985). In some cases a

disease may be put in a different category depending on who is doing the classification.

The result is that people seek help at different times in their sickness and some never seek

help at all.

Mechanic (1968: 116) long ago recognized and emphasized the need to study

illness behavior more broadly taking the population in general, and not those who seek

care at specialized agencies. There are those who prefer not to go for treatment or who

use self-treatment at home. Therefore, studying illness behavior by selecting those who

seek help from general practitioners excludes those in the general population who do not









consult the general practitioner. This is particularly true of illnesses that are prevalent

(Mechanic 1968: 116). This calls for health care research that focuses on the larger

community through the study of household and individual responses to disease.

In this Chapter I examine factors that influence people's response towards

malaria. The chapter is divided into five sections. First, I review the biological

background of the disease. This is followed by a review of the literature on the impact of

malaria on individual populations. In the third section, I discuss the manifestation of

malaria and the responses of people in the search for treatment. Next, I discuss how

various factors influence the health care seeking behavior of lay people. The factors I

consider include: diagnosis of disease, cost, household composition, and decision makers.

Finally, I put the discussion into a broader theoretical perspective and I apply the emic-

etic1 distinction to account for differences in people's response to malaria.



The Biology of Malaria


Cause, Transmission, and Epidemiology of Malaria

Malaria is one of the oldest and most serious tropical diseases. The probability

and rate of its transmission is affected by variables associated with parasites, vectors,

hosts, and the environment. The hosts (humans) are particularly complex. Biological,

demographic, behavioral, cultural, and social variables all affect the transmission of

malaria parasites to humans.




1 In this dissertation I have used emic to imply the folk perceptions about illness, illness
causation and their interpretation of the symptoms. I use the term etic to imply the
biomedical interpretation of illness causation and symptoms.









There is a diversity of malaria parasites that infect human beings. Out of 120

known Plasmodium species Nevill (1990) identifies four types: P. falciparum, P. vivax,

P. malariae, and P. ovale to be of major concern to human beings because of their

association with malaria. P. falciparum is responsible for the majority of malaria cases. In

some areas of tropical Africa it accounts for over 90% of the infection (Beausoleil 1986).

The vector for the malaria causing parasite is the mosquito. There are more than

2500 known species of mosquitoes. Only a subgroup of 50 to 60 species belonging to the

genus Anopheles are capable of transmitting the Plasmodium parasites to humans. The

key vectors, particularly in East Africa, are the blood sucking female mosquitoes of four

species: A. gambiae, A. funestus, A. melas, and A. arabiensis (Meuris et al. 1986).

Nearly 40% of the world's population live in malaria-infested environments.

Malaria is endemic2 in many parts of Asia, Africa, Central and South America, Oceania

and certain Caribbean Islands. Occasionally imported cases of malaria are reported in

Europe. Imported cases have been reported in Britain (Pryce et al. 1993), and in

Switzerland (Steffen et al. 1993). In Africa about 500 million people are at risk of malaria

infection. Four-fifths of these people are in sub-Saharan Africa where about one million

people, mainly children, die from malaria each year (Meuris et al. 1986; Nevill 1990). In

Sub-Sahara Africa, poor utilization of services together with poor communication makes


2Endemicity is defined in terms of parasite rates, spleen enlargement rates (determined by
palpating the spleen to determine whether it is enlarged) and vector seasonality and
abundance. Based on these criteria four types of endemicity are recognized. They are
holoendemic (if the spleen rates in the 2-9 years age-group are > 75%), hyperendemic
(for rates from 50 74%), mesoendemic (for rates from 10 49%), and hypoendemic (for
rates < 10%). Holoendemic areas have a transmission more or less continuous throughout
much of the year. Hyperendemic areas have high rates of stable malaria with seasonal
increases in both morbidity and mortality. Mesoendemic areas have seasonal unstable
malaria. Hypoendemic areas have transmission occurring only in limited periods of the
year (Kenya, Republic of 1992 and Roberts 1974).








16
reporting of malaria cases problematic. The total number of people who die from malaria

and malaria-related complications is probably much higher than currently estimated (see

for example Najera et al. 1992).

One of the key determinants in the epidemiology of malaria is the emergence of

parasites resistant to chloroquine and mosquitoes resistant to insecticides. In 1961 Young

and Moore reported for the first time chloroquine resistant P. falciparum in a patient from

Colombia (Young and Moore 1961). Since then strains of drug resistant P. falciparum

have been reported in many areas (see Draper et al. 1988-East Africa; Alvar et al. 1987-

Equatorial Guinea; Raccurt et al. 1986-Cameroon; Van-der-Kaay et al. 1984-Central

Africa). Apart from Western Sahara, Morocco, Algeria, Tunisia, Libya, Egypt, South

Africa and Lesotho, all the other Africa countries have reported drug resistant P.

falciparum. The whole of the Indian sub-continent and southeast Asia, and South

America (excluding Argentina, Chile, Paraguay, and Uruguay) have parasites resistant to

chloroquine (Center for Disease Control 1990: 8). With 2.13 million cases reported, India

accounted for at least 40% of the total number of cases reported to WHO excluding

Africa in 1992 (WHO 1994, 1997; Table 2.1).

The leading cause of malaria in Kenya is P. falciparum. In some parts of the

country (e.g. the Kano Plains in Nyanza Province), P. falciparum accounts for at least

80% of malaria infections. P. malariae is responsible for about 10%-15% of the

infections, and P. ovale for 5% of the infections (Roberts 1974: 307; Spencer et al. 1987).

P. vivax is occasionally reported from the coastal area (Kenya, Republic of, 1992).









Three clearly defined epidemiological situations are found in Kenya. These are:

(1) endemic, (2) epidemic or seasonal malaria, and (3) no malaria transmission (Roberts

1974) (Table 2.2). This distribution corresponds to altitude and rainfall patterns.



The Life Cycle of the Malaria Parasite

Advances in knowledge regarding the life cycle of the malaria parasite were made

in the latter part of the 19th century (Oaks et al. 1991). In 1898 Giovanni Battista Grassi,

Amico Bignami, and Guiseppe Bastianelli, in Italy, documented the transmission of the

human malaria parasites. Soon after, they described the developmental stages of the two

most important malaria parasite species: P. falciparum and P. vivax. Before them, in

1880, Laveran, a French army surgeon in Algeria, for the first time saw and described

malaria parasites in human red blood cells. And, in 1897 Ronald Ross in India found a

developing form of the malaria parasite in the body of a mosquito that had previously fed

on the blood of a malaria patient.

The Plasmodium parasite has three phases of development in the mosquito and

two in the human host (Figure 2.1). It is transmitted into humans in the sporozoite forms

in the saliva of infected female mosquitoes. The sporozoites then invade the liver cells.

Within 5 to 15 days the sporozoites develop into schizonts. This period varies by species.

It takes about 7 days for P. falciparum, 6 to 9 days for P. vivax or P. ovale, and 14 to 16

days for P. malariae (Jetten and Takken 1994: 3). Each of the schizonts contains 10,000

to 30,000 merozoites. The merozoites are released and they themselves invade the red

blood cells. The pre-erythrocytic development in the human host is known as the intrinsic

incubation period.









In the red blood cells each merozoite matures into a schizont with 8 to 32 new

merozoites. The red blood cells eventually rupture to release the merozoites into the

blood stream. The merozoites can then again invade new red blood cells. It is this

rupturing that is associated with fever and it signals the clinical onset of malaria (Oaks et

al. 1991). Disease symptoms are caused by the asexual parasite stages present in the

human host.

Malaria parasites can remain in the human host for a long time. These can cause

malaria after a lapse of many months and, sometimes years. In patients with P. vivax and

P. ovale this phenomenon, which caused by dormant liver-stage forms of the malaria

parasites, is known as relapse. They can remain dormant up to 4 years before resuming

development and releasing merozoites into the bloodstream. In patients with P.

falciparum and P. malariae recurrence of malaria is due to recrudescence. Recrudescence

is caused by surviving blood-stage parasites from earlier infections (Oaks et al. 1991: 27).

Some merozoites in red blood cells differentiate into sexual forms, the

gametocytes, which may be ingested by mosquitoes. Once in the mosquito, gametocytes

leave the red blood cells to initiate the process of fertilization. Male and female gametes

fuse to form a zygote. Within 12 to 48 hours the zygotes elongate to form ookinetes (the

fertilized forms of the malaria parasite in the mosquito's body). The ookinete penetrates

the wall of the mosquito's stomach and becomes a oocyst. Within a week or more

depending on the plasmodium species and the ambient temperature, the oocyst forms

more than 10,000 sporozoites. The period of development of malaria parasites outside the

human host is known as the extrinsic period. When the oocyst ruptures, the sporozoites








19
migrate to the mosquito's salivary glands, ready to be injected into a human host, and the

life-cycle is completed.



Impact of Malaria on the Population


Many studies of malaria focus on the economic impact on families. These studies

concentrate on the direct costs (measured by amount of money spent) required for

treatment and transportation (Vosti 1990; Sauerborn 1991; Ettling 1991; Sherphard et al.

1991; Jayawardene 1993), and on the indirect costs-measured by lost agricultural

production time and the indebtedness that comes from medical expenses (Conly 1975;

Sauerborn 1991; Jayawardene 1993). Studies on the direct costs of malaria are by far the

more frequent, perhaps because direct costs are easier to measure than indirect ones.

Other studies have focused on malaria's effects on mothers during pregnancy (Cot

et al. 1992; Steketee et al. 1988; Kaseje et al. 1987; McGregor 1984; Bray and Anderson

1979), on adaptive advantages conferred by the sickle cell trait and immune responses in

malaria-infested environments (Madrigal 1990; Esposito et al. 1988; Marsh et al. 1988;

Fleming et al. 1985; Bienzle et al. 1972; Livingstone 1958; Allison 1954a, 954b), and on

the use and effectiveness of mosquito repellents-such as coils, body smears, and

mosquito nets (Snow et al. 1988).



Effects on the Local Economy

Conly (1975) studied the effects of malaria on the family's economy in eastern

Paraguay. She examined 28 new settler families over a 2-year period in a malaria

endemic region. This area had limited malaria control activities. Conly's study revealed:








20
(1) a decline in the rate of land clearing; (2) diminished harvests of subsidiary crops; (3) a

reduction in the amount and efficiency of farm work done; (4) a reduction in reciprocal

labor exchange; and (5) a shift in agricultural priorities. According to Conly, malaria was

also the probable stimulus for migration among many young men. Conly estimated a loss

of between 5 and 15 days of work for each malaria episode.

In Africa, it is estimated that between 3 and 7 days of work are lost per case of

malaria. On average about 2.1 days (approximately US $ 1.70 per capital per year, the

average cost of goods and services produced per day was US $0.82 in 1987) of output per

person are lost. In 1987, a case of malaria cost US $ 9.84 ($ 1.83 in direct costs and $

8.01 in indirect costs). The average value of goods and services produced per day was $

0.82 (Shephard et al. 1991). The economic burden of malaria to Africa was US $ 0.8

billion (Table 2.3). This was predicted to rise to US $ 1.7 billion in 1995 (Shephard et al.

1991). In Kenya, it is estimated about 174 million (Pp X Do, Table 2.3) working days per

year are lost among the 15-60 years age-group (46% of the population). The estimates

assume a disability due to malaria of 15 days per year (Kenya, Republic of 1992: 19).

Kenya's per capital GNP is US $ 270.00 per person per year (World Bank 1995). The

country, therefore, looses about US $ 129 million (Pp X Do X Vgs, Table 2.3) every year

due to lost working days because of malaria. This is a great burden for a developing

country whose economy relies primarily on agricultural production.

And this is surely an underestimate. Some work days are lost because relatives

and friends come to visit those suffering from malaria. In Kenya, at least, lost school days

and lost work days for mothers taking care of sick children and other family members are

not counted (Kenya, Republic of 1992: 19). In a 1994 study in Uasin Gishu district the









rate of school absenteeism among primary school pupils ranged between 17.6% and

54.4% for some days. The rates were higher in Class 1 and 2 (between 6 and 8 years)

than in the rest of the Classes. Students missed school because: (1) they had malaria; (2)

they were taking care of household duties of a sick relative; (3) they left to take care of

the home while adults were away attending funerals of close relatives or friends.

Sometimes the entire school missed an afternoon in order to join the community in a

funeral which may have been malaria related (Some 1994).

Immediately following a malaria episode work output is low before people regain

their full strength. This produces further, unknown, losses. We need better-informed

measures for the economic losses due to malaria. Researchers should determine all the

people linked to a sick person. Once this has been done, it should be easier to follow

them up and record their activities in relation to the sick person. The sick should also be

followed throughout a single episode and their activities accurately recorded. Following

this strategy we should be able to better determine the direct and indirect economic losses

that families incur as a result of malaria or other illnesses. Unfortunately, I have not been

able to do that here because of logistical problems, which I have discussed in Chapter six.



Effects on Pregnancy

Women may suffer from various problems during pregnancy. For example,

increased parasitemia3 occurs during pregnancy (Bray and Anderson 1979). Evidence

suggests a relationship between parity and parasitemia. McGregor (1984) reports that in a

sample of primigravidae women in Gambia, 64% were infected. For those in their second


3Parasitemia is defined as the level of parasites present in the blood.








22
or third pregnancy, 39% were infected; among women who had already been through at

least three pregnancies, 20.9% were infected. Elevated levels of P. falciparum are

associated with: 1) high maternal morbidity; 2) high maternal mortality; 3) spontaneous

miscarriage; 4) still births; 5) premature deliveries; and 6) low birth weight children

(Steketee et al. 1988; McGregor 1984).

It is not clear why parasitemia increases during pregnancy. Pregnancy appears to

depress pre-existing acquired immunity (Brabin 1983). Beer and Billingham (1978) argue

that increased production of hormones occurs during pregnancy. Some of these

hormones, particularly cortisone, can exert immunosuppressant effects in certain

experimental conditions.

Women in their second trimester are considered to be at particular risk for

complications due to malaria infection. Brabin (1983) observes that infection rate is at its

highest during this time. Plasmodium parasites multiply in the body and attack the red

blood cells. Hemoglobin in the red blood cells is broken down leading to anemia. In a

state of increased red blood cell destruction the maternal system is under pressure to

produce more red blood cells. This demand sometimes cannot be met in poor health (or

nutritional) conditions that are characteristic of most tropical countries. Under these

circumstances malaria related complications are high.

In 1986 WHO recommended early recognition and treatment of malaria and

chemoprophylaxis using effective antimalarial drugs throughout pregnancy. Following

this recommendation, malaria programs in endemic regions stepped up antimalarial drug

distribution to pregnant women. However, few studies have measured the clinical value

and cost-effectiveness of chemoprophylaxis. In Gambia malaria chemoprophylaxis








23
administered by traditional birth attendants resulted in reduced parasitemia, fewer cases

of anemia, and fewer low birth weight babies-but only in primigravidae women

(Greenwood et al. 1989). Parasitemia may have been reduced significantly in

primigravidae women because more of them are infected with the malaria parasites.

Reduced parasitemia then led to fewer anemia cases and fewer low birth weight babies.

The effect of prophylaxis on placental infection has been shown to be less

effective when administered for less than one month (Cot et al. 1992). In many parts of

the third world women do not go to prenatal clinics until late in their third trimester-if

they go at all. In some areas (e.g. the Mara region of Northern Tanzania) other factors

that hinder chemoprophylaxis programs may add to the complexity of the problem

(MacCormack and Lwihula 1983). These factors include: (1) inconsistent drug supply;

(2) problems in drug distribution; (3) poor communication, particularly in rural areas

(where the majority of women live); (4) inadequate staffing of health care facilities; (5)

poor (or lack of) community participation; and (6) drug side effects.

Similar factors have been identified in Saradidi, Kenya (see Kaseje et al. 1987).

One would question the effectiveness of prophylaxis at this late period in pregnancy.

During this stage parasitemia appears to decrease even without intervention (Brabin

1983). Thus, achieving maximum benefits at minimum cost requires proper timing.

However, proper timing alone is not enough. Pregnant women must be willing to

participate in prophylaxis programs. In areas where problems due to drug supply and

distribution have been overcome, most expectant mothers who are malaria asymptomatic

do not take antimalarial drugs for prophylaxis. Due to lack of recognizable symptoms

they may not readily associate antimalarial prophylaxis with long term health benefits.









This is a problem of illness-disease distinction. In endemic areas constant

reinfection may cause a change of patients' behavior towards malaria and the use of

medicine. Patients start to regard the medicines used as ineffective. The bitter taste of the

pills and the negative side effects of the drugs (e.g. chloroquine-induced itching and

miscarriages-Kaseje et al. 1987) also influences people's behavior. In their study, Kaseje

and colleagues found that 4% of the women from the Kano Plains of western Kenya

reported that they fear using chloroquine for prophylaxis because "they believed

chloroquine caused stillbirths and abortions" (Kaseje et al. 1987 : 81). Compliance

becomes erratic and patients stop taking drugs as soon as malaria symptoms disappear,

preferring to keep the stock of remaining pills for later episodes.

In some cases pregnant women who are identified as having malaria decide

against use of anti-malaria drugs. In Sri Lanka (Jayawardene 1993) and in Kenya (Kaseje

et al. 1987) pregnant women do not take anti-malarial drugs because they fear that using

these drugs would be harmful to the unborn child. The women in Sri Lanka refrained

from seeing a therapist even to obtain advice for fear that they would be asked to take

anti-malaria drugs. Their concern about the unborn child is genuine. In many developing

countries where abortion is illegal, young girls have been known to use, often with

disastrous results, a large dose of chloroquine tablets to induce abortion. Brabin (1991)

provides a detailed review of malaria in pregnant women.

Compliance is the degree to which a patient's behavior corresponds with

therapeutic recommendations. It encompasses a wide range of behavior which include

missing medication doses, taking too much medication, ceasing medication, and failing to

go for follow-ups at the recommended intervals (Elixhauser 1991).









Compliance is a problem everywhere. Many patients in the US fail to complete

the 10-day regimen for most antibiotics. When symptoms disappear after 3-4 days,

people stop taking pills. In coastal Kenya, Mwenesi (1993) found that 55% of the patients

who were seen at health care centers did not follow instructions for taking malaria drugs.

Mothers exiting clinics were asked about instructions given to them. Most did not

describe correctly what they were told by those dispensing drugs, which included

directions that the patients should complete the dose. The mothers did not ask for

clarification for fear that health workers would reprimand them (Mwenesi et al. 1995a).

Eighteen percent of households surveyed had anti-malaria drugs at home, but none had a

complete course (Mwenesi 1993). It is quite likely that the patients in the surveyed

households did not comply with the treatment regimen. Non-compliance reduces the

potential benefits of treatment and it exacerbates disease (Elixhauser 1991).

Patient compliance raises important theoretical issues. The explanation by patients

of their behavior varies between groups. In most developing countries non-compliance

results from the perceived need to preserve medicine for future use. This behavior

minimizes the cost that families incur to treat subsequent illness cases judged to be

similar. In the developed world non-compliance is the result of other factors. In the case

of non-compliance, I would expect to have similar justification between different groups

since we are seeking an explanation to account for one specific behavior. But this is not

the case.

Structural differences between developed and developing countries provide an

explanation for non-compliant behavior. In the model shown in Figure 2.2 I identify

infrastructural and structural factors that influence non-compliance behavior. These









factors present what I consider to be major determinants of patient non-compliance

behavior. In the developed countries forgetting to take prescribed medicine and clearing

of the symptoms is a post-hoc justification for behavior. In the developing countries the

reasons for non-compliance include keeping of medicine for self future use or for use by

other family members.

How do the infrastructure and the structure influence compliance? Developing

countries are characterized by large families with a network of extended relatives. The

cost of treatment is relatively high. Therefore, the strategy adopted by many families for

minimizing treatment costs is to keep medicine for future use. Availability and

accessibility are key issues also. Medical care facilities are few and far between. This

may cause a further increase in the cost of treatment.

In the developed world, families are small and the cost of treatment is high.

However, health care is available and accessible to most people. Furthermore having

health insurance off-sets the high cost of treatment. Under these circumstances patients

may stop taking medicine as soon as the symptoms disappear. They will go back to the

health care provider should the condition reappear.

Despite problems of non-compliance, research has shown that programs can be

evolved that lead to increased acceptance of prophylactic treatment (see Helitzer-Allen et

al. 1994). Using knowledge gained from previous community based research Helitzer-

Allen and her colleagues developed an intervention program in Malawi. Earlier research

had focused on local concepts of malaria and issues regarding malaria prevention and

treatment during pregnancy. They used this information to evolve an intervention that








27
included a change in the health education message given during antenatal clinics and the

distribution of sugar-coated chloroquine pills.

Use of the new health education message led to a 45% increase in chloroquine

utilization over the baseline rate while provision of sugar-coated chloroquine pills led to a

65% increase in chloroquine utilization. Utilization of chloroquine was measured by

detecting in the urine the level of the parent chloroquine compound and the desethyl

metabolite of the drug. If the compound and its metabolite were present the test was

recorded as positive and negative if absent. A combination of the new health message and

the sugar-coated pills suggests an additive effect (Helitzer-Allen et al. 1994). This result

may, however, be skewed because only 44% of the 1,035 women enrolled for the study

returned for follow-up. Some of the women were not included in the final analysis

because their follow-up date fell beyond the study period and most of those who returned

for the follow-up had negative urine chloroquine during the enrollment screening. The

presence of chloroquine in the urine may have been due to nurses putting emphasis on

using the chloroquine pills provided. It may also be that the women knew that their urine

would be sampled and may have complied because of fear that the health workers would

be angry with them if it was discovered that the chloroquine pills were not used (Helitzer-

Allen et al. 1994). Other factors such as packaging, pill color, shape, and size influence

compliance.



Effect on Genes: Adaptive Advantage of Sickle-cell Trait

Distribution of the sickle-cell trait. A map of the world showing the distribution of

sickle cell trait superimposed on that showing the distribution of malaria suggests a









relationship between the two. Sickle cell anemia, a condition resulting from having a

double dose of abnormal hemoglobin, was first described in the Western medical

literature in 1910. A Chicago physician, James B. Herrick, observed the condition in the

blood of a 20 year old black student from the West Indies (Durham 1991: 105). Herrick's

suggestion that the sickle cell disease is caused by a change in the composition of the red

blood cells later proved to be correct.

By 1950, the literature on the distribution of the sickle cell trait among human

populations was accumulating. The literature enabled Frank B. Livingstone to discuss the

general distribution of the sickle cell gene in the Old World and in immigrant populations

(Livingstone 1958). Data from this period raised an important question: why was there a

clear pattern of the sickle cell trait distribution? People who are homozygous (Hbs Hbs)

for the sickle cell trait rarely reproduce. Most die before reaching reproductive maturity.

Therefore there would be a constant loss of genes each generation. Livingstone (1958)

concluded that for the sickle cell genes to attain a frequency of 0.1 to 0.2 in the

population a mechanism compensating for the loss must exist.

It is now known that individuals who are homozygous dominant (HbA HbA) suffer

higher rates of malaria infection compared to those who are heterozygous (HbA Hbs)

(Allison 1954a, 1954b; Garlick 1960; Gilles et al. 1967; Raper 1955). Those who are

homozygous recessive (Hbs Hbs) die because of complications related to sickling of the

red blood cells. The sickle cell in a heterozygous state reduces the severity of malaria and

resistance to infection is enhanced (Ringelhann et al. 1976). Heterozygosity leads to a

condition known as heterozygouss advantage." In malaria endemic regions such as West









Africa, high levels of the sickle cell trait are maintained because of a strong selection

pressure from malaria against the normal hemoglobin (Haldane 1948).

The malaria hypothesis. The geneticist J.B.S. Haldane was among the first to

propose a balancing selection pressure to explain hemoglobin polymorphism. Haldane

(1948) pointed out that an unusually high rate would be required for mutation alone to

account for the frequencies of another hemoglobin disorder, alpha thalassemia (Cooley's

anemia), found in some Mediterranean populations. The homozygous condition for the

thalassemia gene is lethal while the heterozygous one leads to mild anemia. In the

malaria hypothesis, Haldane (1948: 270) argued that red blood cells with the thalassemia

trait were more resistant to attacks by the sporozoa which cause malaria. Malaria was

prevalent in the Mediterranean region (Italy, Sicily, and Greece).

The malaria hypothesis has been applied to explain the sickle cell trait

distribution. Some researchers claim that the sickle cell trait confers certain adaptive

advantages in malaria environments (e.g. Allison 1954a, 1954b; Madrigal 1990; Fleming

et al. 1985). From Togo (Bienzle et al. 1972) and East Africa (Allison 1954b) local level

correlation between malaria infection and the rate of the sickle cell trait in the population

has been shown to be high. In Togo, two Ewe population groups, one from the mountains

and the other from the plains, were selected for this study. The incidence of the sickle cell

trait was lower (5%) in the mountain group than in the group from the malarious lowland

(23%) (Bienzle et al. 1972). Allison (1954b) found a clear relationship between the sickle

cell trait and malaria among 35 ethnic groups from East Africa. In areas where

falciparum malaria is hyperendemic sickle cell trait levels varied from 14% to 40.5%

compared to < 6% from other areas. Areas that had considerable differences in malaria









endemicity had levels of between 7-10%. Less than 5% of the Abagusii were found to

have the trait.

Livingstone (1958) associated agricultural practices to the incidence of

mosquitoes. He observes that A. gambiae require warm, sunlit ponds of fresh water for its

reproduction. The cutting down of the forest in parts of West Africa created conditions

conducive for A. gambiae breeding. Therefore the spread of agriculture became

responsible for the spread of the selective advantage of the sickle cell gene. As a result of

selective advantage the sickle cell trait spread. This is known as the Livingstone

hypothesis (Durham 1991: 125). Livingstone hypothesized an association between

cultural (agriculture) and genetic evolution and he showed that gradients of sickle cell

gene frequency in West Africa correspond to geographical patterns in the spread of

agriculture. Livingstone's 1958 paper is a landmark in the development of medical

anthropology, providing a clear and detailed example of the association between culture,

biology, and disease (Johnston and Low 1984: 224).

How the sickle cell trait protects individuals from the effects of falciparum

malaria is not well understood. It appears that the malaria parasites do not thrive well in

red blood cells with sickle cell trait. Also, infected red blood cells have a tendency to

sickle when oxygen supply is low. Sickled cells are disposed of faster by the

macrophages and other cells of the reticulo-endothelial system (Bruce-Chwatt 1980: 59).

Other genetic factors such as glucose-6-phosphate dehydrogenase (G6PD)

deficiency might exert a protective effect against P. falciparum malaria. Greene (1993)

reviews the literature on G6PD deficiency and falciparum malaria. Like the sickle cell








31
trait, epidemiological studies reveal that there is a general covariation in the distribution

of G6PD deficiency andfalciparum malaria around the world.

Allison and Clyde (1961) studied 532 children aged between 4 months and 4

years in Tanzania in areas with holoendemicfalciparum malaria. Parasite rates and

densities were found to be lower in G6PD deficient children than in the G6PD normal

children. In another study, Gilles et al. (1967) reported protection by G6PD deficiency

againstfalciparum malaria in 100 Nigerian children (4 months to 4 years). They

compared children admitted to hospital suffering from severe malaria with a control

group over a three year period. The frequency of G6PD deficiency was significantly

lower among subjects with severe malaria compared to the frequency of the trait in the

control group (Gilles et al. 1967). The strength of these two studies is in the group of

subjects' used-they were young so that the development of natural immunity was not

complete. Literature on the protective effect of G6PD deficiency against falciparum

malaria is also available for African American soldiers in Vietnam (Butler 1973; Kar et

al. 1992).

Hemoglobin polymorphism has implications for lay people's response to malaria.

It affects infection frequency and severity. Later in this chapter I discuss how disease

frequency and severity influence lay people's responses to malaria.



Manifestation of Malaria and Lay People's Response


Manifestation of Malaria

Clinically, malaria is characterized by the presence of all or a combination of

some of the following symptoms: chills, severe headache, fever, general body weakness,









painful joints, excessive sweating, vomiting, nausea, dizziness, convulsions (in severe

cases), and anorexia. If the parasites continue unabated they destroy many red blood cells

leading to malaria-related anemia. Anemia has serious consequences, especially in

pregnant women (Steketee et al. 1988; McGregor 1984).

A more serious form offalciparum malaria, cerebral malaria, may present

additional neurological symptoms. It begins with headache and may progress to

convulsions, delirium, and altered consciousness ranging from mild confusion to coma.

Cerebral malaria can lead rapidly to unconsciousness and, in survivors, a remarkable

recovery into full awareness. Indeed this is cited as one of the distinguishing features of

cerebral malaria (Oaks et al. 1991). Cerebral malaria kills between 10 and 50 percent of

its victims depending on the level of endemicity, the definition of the disease, the level of

care available, and the age of the patient.



Clinical Categories of Malaria

Malaria attacks with varying degree of severity. The extent of severity depends on

several factors. They include the age of the individual, their general health and immunity

to malaria. Molyneux and Marsh (1993) identify three concepts associated with malaria

severity. First, the etiological concept involves distinct pathological entities such as

cerebral malaria. These must be separated from other non-cerebral forms of malaria.

Severity may also be defined in terms of the chances of a particular episode having a

greater than a predetermined chance of leading to death. This is known as the prognostic

concept (Molyneux and Marsh 1993). It is influenced by the need to identify, in a

hospital setting, groups that have bad prognoses, so that they can be closely monitored.









This approach excludes those subjects who do well in a hospital but who, nevertheless,

have a poor prognosis outside the hospital. Finally, in some cases severity may be defined

in terms of the load on the health care system. Under this category anyone who requires

in-patient treatment is included. The assumption here is that only severe cases of malaria

get hospital admission.

Defining severity is arbitrary. Individuals suffering from malaria may be

classified as severe when they are not. To overcome this problem the following

guidelines are used to identify clinical categories of malaria (Molyneux and Marsh 1993:

7 8):

Not malaria. This is characterized by the absence of parasites from a specified

number of oil-immersion fields on a thick blood film. The common stipulation is 100

microscope fields.

Asymptomatic infection. This is characterized by the presence of parasites in the

blood film, but the person is otherwise 'well'. The term 'well' is used here subjectively. It

should be clarified before any studies are undertaken. It should include objective

assessments such as the ability to attend school or work. The patient's temperature should

be used whenever possible.

Uncomplicated malaria. In uncomplicated malaria the patient has suggestive

symptoms or fever, with asexual forms of P. falciparum parasitemia at a density above a

level already chosen before the study. The minimum density in tropical Africa is usually

set at between 1,000 and 10,000 parasites per p1l. Care should be exercised such as taking

into account the study season and the distribution of the parasitemia in the study

population.








34
Severe or complicated malaria. In complicated malaria the presence of some or all

of the following characteristics may be chosen as the criteria for a particular study:

cerebral malaria; severe anemia-a high level of parasitemia; prostration; convulsions;

acute renal failure; pulmonary edema/respiratory stress syndrome; shock; and

spontaneous or prolonged hemorrhage. If unexpected death occurs in a patient in whom

malaria is the only significant finding at autopsy, the deceased should be included in the

severe or complicated malaria category.



Recognition of Malaria by Lay People

The symptoms used by lay people to diagnose and classify malaria are many and

not necessarily the same as those used in biomedicine. In a study of Liberian children and

mothers, Jackson (1985) found that, for the study population, malaria symptoms

consisted of a constellation of bioculturally defined signs. The expressions used by the

respondents were: body cold all over, head hurting too bad, body hot all over, weak body,

all bones hurting, sweating a lot, hurting belly, belly sore to touch, loss of appetite,

throwing up, body jerks, being sick-nauseated, turning eyes-dizziness, and other

symptoms. Eleven malaria signs and symptoms were reported for children and fourteen

for mothers (Jackson 1985).

In another study in rural Ghana, Agyepong (1992) reports that the Adangbe

recognize malaria as a symptom complex locally known as asra. Asra is characterized by

headache, a rise in body temperature, chills, bitterness of the mouth, yellow eyes, deeply

colored urine, loss of appetite, body aches and pains, weakness and easy fatigability,

vomiting, pallor of the palms and soles, and cold sores around the mouth.








35
Jayawardene (1993) has given the following graphic description of malaria by one

of his informants in Mahaweli Scheme in Sri Lanka: "The fever comes, the fever goes.

Suddenly we sweat, feel very cold and the fever leaves. You take a Panadol, feel better

for a day, the fever drops and then it rises. That's malaria" (Jayawardene 1993: 1171).

During a preliminary field trip to Gusii in June and July 1994, I found that they

also rely on similar symptoms to recognize malaria. They identify malaria through the

following symptoms: (1) feeling cold (usually accompanied by shivering and then a rise

in body temperature); (2) vomiting yellowish-green liquid (esosera); (3) joints aching; (4)

feeling tired and weak (usually accompanied by dizziness); (5) headache; (6) stomach

ache (amatema); (7) lack of appetite; (8) unusual heart beat; and (9) drooping eyes. For

children symptoms may also include dullness, crying a lot, and intermittent involuntary

spasms during sleep.

Ugandan women report that infants, children, adults and pregnant women exhibit

different malaria symptoms (Kengeya-Kayondo et al. 1994). For infants, the symptoms

include raised body temperature, refusing to suck, crying all the time, vomiting, sores in

and around the mouth, general weakness, jaundice, palpitations and loss of

consciousness. For children the main symptoms are raised body temperature and lack of

appetite. Headache, general weakness, and feeling thirsty were also mentioned for this

group. In adults, body weakness, feeling cold, and pain in joints were reported as the

main signs of malaria, while in pregnant women miscarriage, vomiting, general

weakness, a lot of heat in the stomach, and feeling cold were reported as the main

features.









Malaria has symptoms that can, sometimes, lead to misdiagnosis by lay people.

Even experts need to perform blood smear tests in order to confirm the presence of

malaria parasites. In developing countries blood smear tests are rarely done. Medical

personnel rely on clinical diagnosis and they use presumptive treatment. Across the

world, however, studies show that lay people generally identify malaria using similar

symptoms.



Lay People's Responses to Malaria

Whatever the criteria for recognizing disease, lay people's therapeutic choices are

determined, in part, through recognition of the disease symptoms (Foster and Anderson

1978; Scrimshaw and Hurtado 1988) and perception of disease seriousness based on the

recognized symptoms. In many non-western societies illnesses thought to be the result of

supernatural agents, or causes beyond the control of western medicine, are treated using

traditional medicine. (See Helitzer-Allen [1989] and Fivawo [1993] for examples on

cerebral malaria.) People generally rely on a variety of local therapeutic systems to

resolve their health problems.

Several treatment alternatives are available. They include the application of a

home remedy, self-medication with pharmaceuticals bought over-the-counter on the open

market, herbal therapies provided by traditional healers, and therapies obtained from

health centers or hospitals (Colson 1971; Young 1981b; Hunte and Sultana 1992). A

patient can also choose not to seek any therapeutic intervention. Lay people choose from

these treatment alternatives based on the perceived effectiveness of the particular choices.









However, infrastructural and structural forces also influence health care

utilization. Snow et al. (1994) found that the Giriama of the Kenya coast consult a variety

of health care sources for a single childhood illness. Up to five different therapies are

used: shop-bought drugs, traditional healers, government dispensary, private medicine,

and home remedies (herbs and prayers). About 72% (272) of the 376 respondents

reported using over-the-counter drugs to self-treat fevers, 51% (194) mentioned using

health care facility, 8% (31) mentioned self treatment using home remedies (tepid

sponging, traditional medicine, and prayers). Only one respondent mentioned consulting

a traditional healer to treat fevers (Snow et al. 1992). Snow and his colleagues also found

that measures of household socio-economic resources such as radio and mosquito nets

were higher in the children admitted to hospital while distance from the nearest bus stop

appears to influence utilization of hospital care. Children whose homes were located

further away from a bus stop were less likely to use hospital care (Snow et al. 1994).

Feierman (1981) identifies eleven sources of health care available to lay people in

Northeastern Tanzania. These are: (1) full-time healers, whose primary responsibility is

to provide care for the sick; (2) part-time practitioners some of whom perform a set of

inherited treatments (they provide their services upon request); (3) specialists who live

outside the village and provide care to serious cases of spirit-induced illnesses; (4) old

women who serve as village midwives; (5) common herbal cures used at the household

level; (6) private shops that sell (non-prescription) medicine over the counter; (7)

outpatient clinics or mission hospitals; (8) free government dispensary; (9) government

hospital; (10) stocks of unused pills kept in the household from previous hospital or








38
dispensary visits; and (11) health screening and treatment available from the researchers

doing research in the area.

In theory, treatment options are available to all seekers. However, in practice, not

all are utilized. The search for therapy may follow any one of a number of therapeutic

alternatives available (Young 1981b; Hunte and Sultana 1992) although the treatment

outcomes may be unknown to patients. Patients do not always get the anticipated

outcome and they have no sure way to determine the type of treatment alternative that

will yield the desired state or the best results.

Though unable to pre-determine treatment outcomes, patients still must prioritize

their decisions. They must first order the alternatives available according to some rules of

preference and they must decide on a strategy with a perceived good chance of leading to

the desired results (Fjellman 1976). If a particular treatment choice fails, patients or the

persons) responsible for their health must make new choices. As time passes, and if the

illness persists, the patient becomes desperate and receptive to therapy suggested by

others (Feierman 1981; Agyepong 1992).

Deciding what treatment option to take does not always follow the same sequence

in the same individual during different episodes nor need it be the same in different

individuals. The decisions made can be considered as part of a chain reaction. Decisions

made at time t+1 depend partly on decisions that were made at time t and their outcome.

Further, they also depend on conditions prevailing at time t+1. (The researcher is faced

with the task of mapping out the chain of decisions made and, by inference, the behavior

of the patients at each of these levels-a difficult but not an impossible task.) The

decision-maker is faced with a game of probabilities, a game informed by decisions and









outcomes of time t and the prevailing circumstances at time t+l. It appears that health

care decisions are a Markov process with transition-state probabilities at each step. For a

detailed discussion on the use of Markov models see Bailey (1964), Anderson et al.

(1976), and Isaacson and Madsen (1985).

Markov process modeling has been used to study health status switching of

infants in Kampala, Uganda (Biritwum and Odoom 1995). The main purpose of the study

was to obtain estimates of the transition probabilities between wellness and sickness,

from month to month of children aged from 0 to 18 months. Markov process modeling

can be used to measure disease prevalence (new illnesses arising or those continuing

from previous illnesses) and to assess the expected impact of a health improvement

program (Biritwum and Odoom 1995). It is possible to extend Markov models to health

care decisions.



Factors Affecting Health Behavior


Several factors operate on the family level to influence the choice of medical care.

They include the socio-economic status of the households; the educational status of the

decision-makers; the type of households; the number of children in each household; the

people whom patients (or decision-makers) know and can trust to give them good advice

(social networks); and personal experiences with earlier malaria episodes. Each of these

factors influence medical care decisions that affect disease progression.

Moving beyond the household, there are a wide variety of factors-distance of

health care facility from the patient's home, the presence of good access roads to these

facilities, and the availability of affordable transportation-that influence lay medical care









decisions. These factors in turn affect health care behavior of people at the household

level. In this section I focus on how health care-seeking behavior of patients might be

influenced by indigenous concepts of disease causation, the cost of treatment and the

effect of household composition as well as by the role played by decision makers.



Indigenous Concepts of Disease Causation

Among the Abagusii, mosquito bites are widely believed to cause malaria. Some

people believe that malaria can be caused by eating too much maize, or eating too much

sugarcane, and drinking "bad water." In Uganda, Kengeya-Kayondo et al. (1994) found

that women believe malaria can be caused by drinking water that has not been boiled, by

environmental conditions, and by vectors such as mosquitoes and by other illnesses.

Some people associate malaria with causes beyond human control. For example, it is

reported that some communities in Malawi, Tanzania and Kenya (see Helitzer-Allen

1989; Fivawo 1993; Mwenesi et al. 1995b) attribute a type of fever accompanied by

convulsions to spirits, witchcraft or to 'animals' and 'worms' which enter the patient.

This fever usually turns out to be cerebral malaria.

Mwenesi et al. (1995b) carried out a study in Kilifi, Kenya to determine whether

convulsions and anemia are: (a) recognized as symptoms of childhood illnesses, and (b)

perceived as life-threatening and how these symptoms are managed. They interviewed

883 mothers (608 Mijikenda, 152 Luo, and 123 from other communities). About 56% (or

498) of mothers said convulsion was a childhood illness. Eighty percent of these mothers

(about 400 of 498) believed convulsions to be non-preventable, 43% mentioned








41
avoidance of mosquitoes while 19% mentioned wearing charms and amulets as means of

preventing convulsions.

The Mijikenda and Luo have local names for childhood convulsions (Mijikenda-

nyago, dege, nyuni, and nyama wa dzulu; Luo-oriere). Among the Mijikenda

convulsions are attributed to a figurative animal or bird which enters children. It frightens

them and induces fits in the process. The 'animal' may also reside in the child's mother.

The child gets frightened upon seeing the mother's eyes (Mwenesi et al. 1995b). Luo

informants attributed convulsions to intestinal worms which migrate into the child's head.

Initial treatment of convulsions involves sponging the child using the mother's

urine, or that of a close female relative in the absence of the mother. The child is then

taken to a traditional healer who decides which of the two 'animals' is causing

convulsions, the child's or the mother's. According to Mwenesi et al. (1995b) if the

'animal' resides in the child, an herbal preparation, some to drink and the rest for bathing

the child, is given as treatment. However, if the 'animal' dwells in the mother, both the

mother and child are treated with herbal preparations, in addition to using charms and

amulets.

The Luo use dried and crushed roots which the child sniffs. Sniffing causes

sneezing which should get the 'worms' out of the head. No over-the-counter drugs were

used by the mothers in the study population to treat convulsions (Mwenesi et al. 1995b).

In fact, anti-malarial drugs were withheld or withdrawn from children with fits (Mwenesi

et al. 1995a). The mothers did recognize, but did not link, convulsions to malaria

(Mwenesi et al. 1995b).









In another study Ramakrishna and Brieger (1987) quote a mother in Nigeria as

saying, "Yesterday I thought my child was having malaria, but today when the

convulsion started, I knew it was another disease." Though the disease had signs of

cerebral malaria, the mother identified it as ile tutu, a nonmalarious condition. Under

these circumstances biomedical treatment may be abandoned midway because the

malarious condition has advanced to a state not associated with malaria.



Cost of Treatment

Western medical care in many developing countries is quite expensive and the

distribution of the medical facilities and personnel uneven.4 In Kenya, high medical fee

limits patients' health care alternatives. Poor households are likely to wait for a longer

period before action is taken. If the disease persists they may attempt to use cheaper

options, first, trying non-prescriptive medicines bought over the counter in shops and

small kiosks and if this fails, going to the health center or hospital.

Families in developing countries spend between 2-5% of their income on private

medical care (Gomes 1993). However the difference in health expenditure between the

poor and the rich on a typical illness episode is large in some countries. In Kenya, the

poorest 20% of the population spend 64% of household income on an illness episode

compared to 1% spent by the highest quartile (Gomes 1993).

The most common non-prescription medicines that Abagusii buy over the counter

are Panadol, Hedex, Action (these fall under the general category of pain relievers),

Cafenol, Aspirin (to control body temperature), Malariaquin, Dawaquin, Chloroquine,


4 In Kenya, as in the U.S., urban areas have a higher concentration of doctors than the
rural areas.









and Fansider. The choice of the drugs bought is determined by their cost. Malariaquin,

the cheaper of the drugs and the one most widely used has become less effective against

malaria due to chloroquine resistance of P. falciparum, the most common malaria-

causing parasite. If the first treatment choice fails, people seek alternative treatment from

government hospitals, clinics, mission hospitals as a next step, and sometimes a few of

the patients seek care from indigenous healers. Cost of malaria treatment is a primary

factor in Abagusii's illness behavior.

Sudden and life threatening attacks, for example from cerebral malaria, may force

the decision-maker to disregard all other options in favor of taking the patient directly to

hospital. For example, Molyneux et al. (1989) observe in a study of 131 Malawian

children that mothers wait longer (averaging 47 hours) for non-cerebral malaria but then

take children to the hospital within 8 hours of developing cerebral malaria. In serious

cases economic factors are taken into account after the initial crisis is brought under

control. The threat to life is enough reason to cause urgent decisions and immediate

action.

Mothers in coastal Kenya wait for 3 days before visiting an health facility.

Reasons given for waiting include: perception that the illness is mild; partner being

absent; other important matters to attend to; lack of someone to mind the ill child's

siblings; lack of money for transport; and use of over-the-counter drugs (Mwenesi et al.

1995a).









Household Composition

The composition of the household may influence household resource allocation in

the home. It is known that in some societies children of particular sex are given

preferential treatment while in polygynous households the husband may tend to like and

provide more for the younger wife. Indeed, in Bomorenda a woman informant told me

explicitly that "in a polygynous homestead, the husband is more affectionate to the

younger wife" (see quote [7-10]). She further stated that this is manifest in the time and

resources the husband spends on the younger wife. Differential allocation of resources by

the husband strains the financial base of other households curtailing, in the process,

people's health care choices for treatment of malaria and other health care problems.

Such households may opt for less expensive treatment alternatives such as using cheaper

drugs, and buying or taking an incomplete dose of prescribed medicine.

Studies have shown that in groups where male children are given preference over

their female siblings, more resources are spent on the health care of the male children

(Vlassoff et al. 1995). Differential expenditure on household members leads to an

imbalance in the health status in the family and the community. For instance, where male

children are favored the caretakers may take them to a health care provider earlier than

they would for female children.

The relationship between household composition, size of land, and malaria

incidence is discussed in Chapter seven in detail with data from the field. A direct effect

of reduced plot sizes is on malaria incidence. People now live much closer than they did

three decades ago. This has reduced the distance mosquitoes must travel between

households and individuals. This implies that chances of being bitten by an infected








45
mosquito have increased in the same period. If the rate of bites is directly proportional to

the probability of getting malaria, then people are more likely to have malaria now than

in the past. There is, in fact, increased malaria incidence in Gusii.



Decision Makers

Decisions regarding health care choices are not always taken by one person.

Usually the decision making process involves a period of informal consultations. Advice

given during these consultations is not binding, but is taken into account when decisions

are made. Frequently the person consulted is one who is knowledgeable about the disease

in question and with whom a social connection already exists. People are likely to consult

with and follow the advice of those they already know and can trust (see Feierman 1981).

The decisions may also reflect household experiences with earlier malaria

episodes. For instance Jayawardene (1993) has shown that for rural Sri Lanka subsequent

malaria episodes are more likely to be managed outside the home.



The Emics and Etics of People's Response to Malaria


It is clear that, all over the world, people's recognition of and response towards

disease varies according to the prevailing circumstances (see for example Foster and

Anderson 1978; Scrimshaw and Hurtado 1988; Molyneux et al. 1989; Jayawardene

1993). How does one account for differences in people's behavior towards illness? I will

use the emic and etic distinction (Harris 1979) to group people's responses to malaria. I

explain this in terms of recognition of malaria, cause of malaria, management of malaria,

and disease progress.









Recognition of Malaria

Lay people recognize two forms of malaria according to its gravity (I use gravity

to imply threat to life). The life threatening form closely parallels cerebral malaria-a

serious and often fatal form. The less threatening form is mild but can be sometimes

debilitating. Patients with mild malaria may go about doing light work. In this respect

there is consensus between the lay and medical perceptions about the forms of malaria.

Some lay people do not classify cerebral malaria as malaria. Instead, it is put into

categories of diseases that are considered to be the result of human actions-diseases that

do not have a biological origin or due to supernatural forces. Human actions may be, as in

the case of Gusii informants, the result of strained social relations or differential resource

allocation resulting in mental instability. Mental instability manifests itself in the form of

cerebral malaria. However, from the professional's position, cerebral malaria falls under

the same category as other forms of malaria. It sets out in the same way as non-cerebral

malaria. However, cerebral malaria has a different trajectory. It progresses faster and has

serious consequences if untreated.



Cause of Malaria

Lay people say that mosquitoes cause malaria. From a medical point of view

malaria is caused by Plasmodium parasites transmitted by mosquitoes. Because of its

style of presentation some lay people associate cerebral malaria with evil spirits or

witchcraft. When a patient has convulsions lay people often do not link the condition to

malaria. The tendency is to link convulsions to other causes. The cause of malaria from

the medical perspective is Plasmodium irrespective of its style of presentation.









Management of Malaria

When one has frequent malaria attacks, lay people may seek non-biological

explanations. In doing so malaria is removed from the realm of biologically caused

diseases. The medical explanation for such malaria may be ascribed to chloroquine

resistant Plasmodium parasites or to non-adherence to treatment regimen by the patient.

The latter is a question of the patient's disease management. It is within the prerogative

of patients to make decisions as to whether to take the prescribed dosage after the disease

symptoms have gone.



Disease Progression

Disease progress affects people's perceptions. When malaria progresses into the

cerebral form, patients may abandon or change treatment in favor of a new form of

treatment. Stopping or changing treatment occurs when patients or the therapy

management group give the disease a new classification. Treatment choices depend upon

the new classification. The emics of the lay people as well as of professionals evolve as

conditions and knowledge change. For example, a professional may reclassify malaria as

chloroquine resistant. In that case the new treatment regimen uses different drugs. This is

usually a shift from chloroquine-based drugs to, usually more expensive, sulphur-based

ones. The latter include Fansidar and Metakelfin.



Conclusion


Studies of malaria focus on specific populations (e.g. pregnant women and

children) or on the influence of a few constraints. The constraints include cost of









treatment, availability of transportation, distance to the nearest health care facility, and

the quality of health care provided. Only a few researchers have attempted to develop

models of people's response to malaria using a combination of infrastructural, structural,

and cultural factors which could be used by health planners to facilitate the integration of

control strategies.

Lay people make treatment choices that correspond to their emics. Social and

behavioral responses include waiting for a time, stopping or changing treatment regimen,

and consulting with kin or friends. Waiting for a time while observing the disease has its

purpose. It enables people to spend economic resources on other pressing activities. Folk

views do not always agree with the professional views, however. Those being studied

may see things in a completely different way from the researcher (outsider). For effective

control, malaria programs need to look into the folk/professional distinction. This will

help us determine how lay people's knowledge and the professionals' knowledge

interplay to influence malaria transmission in communities of interest.

Any improvement at the family level of the malaria situation, however small, will

have far reaching benefits to families in areas where malaria is endemic. Benefits will

come in the form of increased productivity, and lower spending on malaria management.

Economic gains resulting can then be channeled to areas such as education and providing

employment. At the national level, the millions of dollars spent on malaria management

can be channeled into other areas of the national economy. Therefore, local benefits due

to control of malaria can have a wider ripple-effect.















Seuial Para-ites


Mosquito


oC3


Human


Figure 2.1: Life Cycle of the Malaria Parasite (Plasmodium spp.)


















I ETIC EXPLANATION I


SUPERSTRUCTURE
Emic explanations for non-
compliance e.gs
1. Forgetfulness/Time
2. Symptoms cleared
3. Condition improved


ETIC EXPLANATION
......................................................


Figure 2.2: A Conceptual Model to Account for Patient Non-compliant Behavior


INFRASTRUCTURE/
STRUCTURE
Examples:
1. Socio-economic factors
2. Family size
3. Cost of treatment
4. Availability
5. Accessibility


SUPERSTRUCTURE
Emic explanations for non-
compliance e.gs
1. Forgetfulness/Time
2. Symptoms cleared
3. Condition improved
4. Medicine kept for future use


INFRASTRUCTURE/
STRUCTURE
Examples:
1. Socio-economic factors
2. Family size
3. Cost of treatment
4. Availability
5. Accessibility
















Table 2.1: Number of malaria cases reported, by WHO region (x 1000), 1983


WHO Region 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 d

Africabc 3,168 4,422 13,207 17,927 20,588 24,712 29,381 12,302 8,994 8,384 2,590 27,644

Americas 831 932 911 951 1,018 1,120 1,114 1,058 1,231 1,188 984 1,115

South-East Asia 2,731 3,003 2,501 2,685 2,834 2,791 2,942 2,970 3,087 3,078 3077 3,514

Europe 73 62 57 47 28 25 21 14 16 22 50 91

Eastern Mediterranean 304 335 391 612 608 434 528 586 541 309 292 321

Western Pacific 1,842 1,410 1,177 1,307 1,145 1,002 1,071 1,032 968 733 674 2,121
Total (excluding
Africa) 5,781 5,742 5,037 5,602 5,633 5,372 5,676 5,660 5,843 5,330 5,077 7,162


This information extracted from WHO Weekly Epidemiological Reports No. 42, 1994, Pp. 309 314;
No. 43, 1994, Pp. 317 321; No. 44, 1994, pp. 325 330, No. 36, Pp. 272.
aThe information provided does not cover the total population at risk in some instances
bMainly clinically diagnosed cases
CIncomplete figures
d The 1994 data include, for the first time, both slide confirmed and clinically diagnosed malaria cases for all regions.


- 1994a (WHO 1994: 4, 1997)













Table 2.2: Malaria epidemiology in Kenya by type and area (adopted from Roberts 1974)

Classification Spleen rate Area
(age 2-9 yrs)
(1) Endemic
(a) holoendemic > 75% Coast Province, coastal area; Tana River, Kano
plains, Taveta.
(b) hyperendemic 50 74% North Nyanza, Bungoma, Busia, Simba Hills (Coast)
Machakos, Kitui, Thika, parts of North Nyanza,
(c) mesoendemic 10 49% Murang'a and Embu below 1,300 m.
Meru, Pokot, Samburu, Isiolo, Baringo
(d) hypoendemic < 10%
(2) Epidemic variable Highland over 1,600 m. with high rainfall and dry
areas with exceptional rainfall: Masailand, Nandi,
Kericho, Kisii, NorthEastern Province, Eastern Kitui,
Londiani, Elgeyo.
(3) No None At altitude over 2,000 m: Aberdares, Mt. Kenya, and
transmission __Mt. Elgon.



















Table 2.3: Estimated economic cost of malaria in Sub-Saharan Africaa


Population Populationb Days of output lost GNPb/yr Average value of Loses in US $ Estimated
(millions) (15 60 yrs) /person /yr due to (US $) goods & services due to malaria/yr clinical malaria
(millions) malaria (15 60 yrs) produced (US $) (Pp X Do X Vgs) cases per year
(Pp) (Do) /person/day (Vgs)
SubSahara
n Africa 559 274 (49% of 12d 0.82d 800 million 100 million
pop.)

Kenya 25.3 11.6 (46% 15' 270.00 270/365 129 million 5.8 million'
of Pop.) (0.74) (reported 1989)
comprises all countries excluding Algeria, Egypt, Libya, Morocco, and Tunisia.
bWorld Bank (1995)
CIn 1987, a case of malaria in Africa cost US $ 9.84 ($ 1.83 in direct costs and $ 8.01 in indirect costs).
By 1995 it was estimated to rise to US $ 16.40 (Shephard et al. 1991).
d, e Shephard et al. (1991)


'Kenya, Republic of (1992)















CHAPTER 3
INTERNAL FAMILY STRUCTURE AND DECISION MAKING:
ITS RELEVANCE FOR TREATMENT AND CONTROL OF DISEASE


Introduction



The family is central in health care decisions. Between 70%-90% of illness

management takes place outside formal health care system. Self-treatment within the

family provides a considerable part of illness management (Turk and Kerns 1985). As a

culture bearing unit the family influences health behavior of its members and yields data

that can be used in studying health care behavior.

In this chapter I discuss the impact of internal family structure, decision making,

and therapy management groups on the treatment and control of disease. I use family

structure to imply the social organization, the roles played by family members and the

hierarchy of authority in decision making. Authority may be gained through accumulation

of economic power. Throughout the chapter, I examine the influence of infrastructural,

structural, as well as cultural forces on health care behavior.

Infrastructure refers to the underlying forces such as technology of subsistence,

techno-environmental relationship, ecosystem, work patterns, demography, mating

patterns, fertility and medical control of demographic patterns. Infrastructural factors also

include the distance of health care facilities from the homes of patients, the presence or










absence of roads to these facilities and the availability of affordable transportation.

Structural forces influence the domestic as well as the political economy. These forces

include household income, education of decision makers, number of children in each

household, the people whom patients (or decision makers) know and trust to give them

advice (social networks), past experiences with disease, and whether households are

monogamous or polygynous. Cultural (superstructural) factors include people's beliefs

about disease, people's health care practices, and how people give meanings and context

to their illness experiences. For a detailed discussion of the differences between

infrastructural, structural, and superstructural (cultural) factors see Harris (1979).

Knowing the influence of infrastructural, structural, and cultural factors on health

care seeking is critical. It assists researchers to design new and better predicting models to

study patient health care seeking behavior, and in the design of new low-cost disease

control technologies. The information also assists planners in planning public health

facilities and in the design of public health programs.

In the second part of this chapter I critique the literature on determinants of health

care seeking behavior. I discuss how these determinants influence health care seeking

below. In the section following it, I use a case study from Kenya to show factors that

might influence patients' consultation of health care providers. This is followed by a

review of studies on decision making (e.g. Young 1981b; Mathews and Hill 1990; Ryan

and Martinez 1996) and on therapy management in Africa (e.g. Janzen 1978; Feierman

1985; Feierman and Janzen 1992). I use these studies to examine the health behavior of

patients and their decisions within the context of the family, and the relevance of

understanding the internal family structure to treatment and control of diseases such as









malaria. In section four I discuss the role played by the therapy management groups in

health care decisions.



Determinants of Health Seeking Behavior


There is an enormous literature on the determinants of health-seeking behavior

(Suchman 1965-United States; Mechanic 1968 and 1969-United States; Colson 1971-

developing countries; McKinley 1973-Arbadeen, Scotland; Fosu 1981-Ghana; Young

1980 and 1981b-Mexico; Mwabu 1986-Kenya; Mathews and Hill 1990-Costa Rica;

Bentley 1992-India; Snow et al. 1994-Kenya; Ryan and Martinez 1996-Mexico). For

reviews see Mechanic (1969), Colson (1971), Young (1976), Kroeger (1983), and most

recently, Ryan (1995). The studies identify factors which determine treatment choices

made by patients or those who take care of them. However, few studies on decision

making at the household level have been carried out in Africa (Doyal 1989). Ryan's (1995,

1998) study in Cameroon is the most recent to address health care decision making in

Africa.

Determinants of health seeking behavior include the following: (1) illness

characteristics and its perceived seriousness (Suchman 1965; Mechanic 1969; Colson

1971; Young 1980), (2) lay people's knowledge and categorization of the illness (Fabrega

1971, 1975; Lipowsky et al. 1992), (3) expenses that are likely to be incurred for each

treatment choice (Gould 1957), (4) distance from the health care facility (Snow et al.

1994); social networks of the patient and the care takers (Zola 1973); and religion

(Mechanic 1963).










Illness Characteristics and Perceived Seriousness

People everywhere recognize, and respond differently to, diseases that are serious

from those that are not. In a sample of New York patients Suchman (1965) found pain (in

66% of the respondents) to be an important sign of things gone wrong. Fever or chills

(17% of the respondents) and shortness of breath in 10% of the respondents were other

main indicators. However, Suchman sampled patients who either (1) required at least

three physician visits and who were incapacitated for more than four consecutive days, or

(2) required hospitalization for one or more days (Suchman 1965). We have no idea how

the results or their interpretation might have been affected had Suchman included patients

who did not visit a physician or were not hospitalized.

In Malawi, Molyneux et al. (1989) found that mothers whose children had cerebral

malaria waited less in taking their children in for medical help than did mothers whose

children had non-cerebral malaria. In fact, mothers took children to the hospital within 8

hours if they suspected that the children had cerebral malaria, while mothers whose

children had non-cerebral malaria waited for an average of 47 hours. Thus disease severity

(or at least the perception of severity) affects the health-seeking behavior of patients or the

decisions made by those who take care of the patients.

Perception of severity of illness is subjective. It depends on factors like

recognizable symptoms and a person's prior experience with that or other illnesses. Quite

often people have different perceptions about an illness. In fact, some people perceive the

same illness differently at different times.










Lay People's Knowledge and Categorization of the Illness

There is now ample evidence to suggest that disease categorization influences

health-seeking behavior (Fabrega 1971, 1975; Fosu 1981; Helitzer-Allen 1989; Lipowsky

et al. 1992; Fivawo 1993). In Ghana, lay people classify diseases along a continuum from

those diseases due to natural agents to those diseases thought to be a result of

supernatural agents. In the middle are diseases that embrace both the natural and the

supernatural agents (Fosu 1981). Diseases considered being the result of supernatural

forces (but which according to the etic classification are not) may cause patients to seek a

supernatural intervention.

In some communities cerebral malaria is thought to be due to supernatural agents.

For example, some communities in Malawi (see Helitzer-Allen 1989) and in Tanzania (see

Fivawo 1993) attribute cerebral malaria to spirits or witchcraft. Patients with cerebral

malaria in these communities rarely utilize hospital care. Instead, they opt for a

supernatural intervention.

In Colombia, Lipowsky and colleagues (1992) found that lay people and traditional

healers use the hot-cold classification system of disease, body conditions, medicines, and

foods (Lipowsky et al. 1992). They classify malaria medicinal plants into two categories:

hot plants and cold plants. To reduce body temperature patients used "cold" medicinal

plants while "hot" medicinal plants were used to reduce inflammation of the liver and

spleen. (see Fabrega 1971, 1975; Frake 1961 for more examples of disease categorization

and health care-seeking behavior).










Health Care Expenses

There are studies that document how expenses incurred for each treatment

selected influence people's health seeking behavior (e.g. Gould 1957; Yoder 1989; Snow

et al. 1994; Mwabu et al. 1995). In Swaziland, following an increase of up to 400% in

hospital user fees at the government hospitals, average attendance in all health care

facilities dropped by about 17% (Yoder 1989). This increase was designed to even

charges that patients pay in government health facilities and mission hospitals where the

average outpatient fee remained about US $ 1.00 per visit. The drop was 32.4% in

government hospitals. Utilization increased by about 10.2% in mission hospitals. Yoder

also observed a decline in patient visits for childhood diseases (16%), Bacillus Calmette

Guerin (BCG) and Diptheria, polio, and Tetanus (DPT) immunizations (19%), and in

programs designed to reduce dehydration in children (24%). A study in Kenya by Mwabu

et al. (1995) reveals a similar trend in health care utilization. After introducing cost sharing

in government health facilities patient attendance dropped by 50%. When these charges

were suspended patients moved to government facilities from the private health sector

over the next 7 months.

When the cost of health care increases, some patients will drop out of the health

care system while others will alter their health care seeking behavior. In the Swaziland

study, at least three types of patients were expected to change their health-seeking

behavior. First, there are the low-income patients for whom the fee is no longer affordable.

Next are, patients who decide their ailment is not serious enough to justify the costs.

Finally, there are patients who make multiple visits are likely to reduce the number of









60
visits if they pay for each visit. Up to 34% of the overall decline in attendance was among

patients who had previously paid the least for health care (Yoder 1989).

In Figure 3.11 present a model of the relationship between people's rating of the

likelihood of getting well for various medical options and the cost incurred. I view each

health care alternative as having a diminishing effect. Thus for any particular health care

alternative once the maximum likelihood of getting well is reached any further increase in

cost is of no value to the patient. Increasing the cost does not lead to an increase in the

likelihood of healing. Therefore, if patients are given several health care options which

have varying degrees of success, they should opt for those options which offer the best

chance of getting well at the least cost.

However, if the chance of healing for all of available treatment options remains the

same, patients will tend to take into consideration the cost involved in each of the options.

Patients will tend to cost minimize while maximizing returns (success). Yoder (1989) has

shown that when the fee paid by patients is the same between the mission and government

hospitals, but the perceived quality of health care is different, patients will opt for the

sector providing the higher quality care. In Swaziland, this will be mission hospitals.

Mission hospitals provide better quality health care than do government run hospitals

(Yoder 1989).

The model I present in Figure 3.1 has limitations. It does not account for personal

preferences nor does it account for other factors which may have an additive effect on

people's choice making. Following the model, a person should choose the most likely to

succeed treatment at the least price. However, a person may use other criteria to choose

treatment. For example, if cost is not a consideration in a treatment choice, a patient may









61
opt for an expensive treatment for other reasons although the likelihood of success is the

same for all the available alternatives. The reasons why people opt for an expensive

treatment may include the desire to maintain the power differential as a means of

controlling forces of production.



Distance to the Health Care Facility

Health care seeking behavior is affected by distance and other factors such as

availability of affordable transport. Distance affects health care seeking behavior in the

following manner. First, transportation cost that is incurred when living further off the

nearest health care facility will be high. This cost will be proportionately higher in rural

areas where patients live further from the main roads and where there is less competition

among transport operators. Second, people in remote and difficult to access areas will

tend to allocate more time to health seeking (and visits to patients in hospitals) thus taking

time available for other activities such as farming. Hence distance will tend to hinder

utilization of some health care resources.

Distance has been studied to identify critical thresholds for different levels of health

care. It has a direct relationship to the cost incurred. Morrill and Earickson (1968) observe

that demand for a hospital declines as costs of reaching it increase. Cost is a function of

distance, such that if the distance to be traveled is big the cost of access will tend to be

high also. Consequently building more clinics should help in reducing the distance traveled

by patients. The alternative is to construct more roads. Building more and better roads

does not reduce the mosquito population but brings in more competition in the transport

sector. Increased competition, in turn, lowers fares as competitors struggle to capture the










market. Therefore, more clinics and roads should help to reduce the cost of accessing

health care facilities, and lead to an increase in care utilization of hospital (and other health

care) facilities.

In Chogoria, Kenya, construction of a new road has led to an increase in outpatient

hospital attendance (Airey 1992). In 1983 patients traveled between 11.5 and 13.6 km. to

a hospital. Since the construction of the new road, travel distance has been reduced by

about 39%. The road led to increased competition among transport operators.

Competition in turn led to reduced fare charges. This reduction had a dramatic effect on

outpatient hospital attendance-an increase of 78% compared to 37% increase for inpatient

attendance.

To explain the differential increase in hospital utilization between outpatients and

inpatients Airey (1992) considered the cost of health care in Chogoria. The fee charged for

health care services is higher than transport costs. Transport costs are a minor proportion

of the total costs of inpatient hospitalization. There are hidden additional costs incurred by

inpatients due to lost days of work. Since the new road reduced both distance and the fare

costs, a higher number of outpatients were more responsive than inpatients (see also Snow

et al. 1994).



Social Networks of Patients and Caretakers

A social network refers to a finite set or sets of actors and the relationship between

them (Wasserman and Faust 1994). It is composed of friends, relatives, and acquaintances

(Shelley 1992). The presence of relational information is a critical and defining feature of a

social network. Networks play a role in health care decisions that patients make. They










facilitate the transfer of information, and they are a mechanism by which social pressure

can be exerted on individuals to take action about their health or that of others.

The decision to seek medical care may be initiated by persons in the social life of

the patient and by institutions that may demand medical attention (Stoeckle et al. 1963).

Patients themselves make decisions based on their knowledge of particular doctors and

medical institutions, and by opinions among their network alters (cf. the lay referral

system, Freidson 1961: 267).

Zola (1973) outlines conditions under which decisions may be made in a sample of

Massachusetts General Hospital patients. His analysis of patient decisions reveals that the

decisions are based on extra-physical factors such as ethnic group membership. The

decision to seek medical care was triggered by: (1) the occurrence of interpersonal crisis;

(2) the perceived interference i //h the social or personal relations; (3) urging from

relatives and friends; (4) the perceived interference with vocational or physical activity;

and (5) symptom recognition. Italians tended to seek medical aid when their symptoms

interfered with social or personal relations and Irish patients tended to go for medical care

when they received approval of others (Zola 1973).

Zola (1973) gives the following account of an Italian patient, 18 year old John Pell,

in his senior year of high school. John had headaches over his left eye and pain in and

around his right, artificial, eye for almost a year. Pell claimed little general difficulty until

asked whether the symptoms affected how he got along. "The last few days of school it

bothered me so that I tried to avoid everybody ... and I wanted to go out ... (but) I get the

pains at 7 or 7.30 p.m." When Pell saw an announcement of an upcoming Prom, and

noticing the starting time of 8 p.m., he immediately went to the school nurse (Zola 1973:










684). In another account, the O'Briens, an Irish family suffering from myopia claimed

difficulty in seeing. Mrs. O'Brien's decision to seek medical care was a response to her

husband's urging. Several months later her husband went in for medical care at her urging

(Zola 1973: 684).



Religious Affiliation of Patients and Caretakers

Religious affiliation is a factor in illness behavior. Mechanic (1963) studied illness

behavior among four groups of people: Jews; Protestants; Catholics; and those who had

no particular preference to any religion. The groups were each divided into two social

classes: high and low. Mechanic found differences in illness behavior between Jews,

Protestants, and Catholics. In the higher social group more Jews (78%) than Protestants

(55%), Catholics (46%), and non-denominational (52%) reported a higher tendency to

visit a physician. These differences remained the same in the lower social group also.

However, they were less pronounced.

Religion may cause patients to seek medical intervention or restrict them from

visiting hospitals. For instance, I have observed that the Akorino (a religious sect in

Kenya) urge their faithful to seek medical intervention through prayers. They advocate

faith healing. Among the reformed Islamic movement (known as Halali Sunna) in coastal

Kenya treatment decisions are strongly influenced by religious ideology. In the Halali

Sunna movement, adherence to a religious ideology has led to illness being attributed to

nonmystical causes (Beckerleg 1994). Under Halali Sunna "the possibility of invasion of a

person by another human or a spirit is rejected ..." (Beckerleg 1994: 299). For a sick

person daily reading of the Quran is considered adequate protection.








65
Affiliation to a particular religious group may also be a factor in the use of certain

health care facilities. In Gusii, Seventh Day Adventist (SDA) followers travel a long

distance to visit Kendu Mission Hospital run by the SDA church. Patients often pass

health care facilities provided by the government at Nyamira and Kisii District Hospitals,

and by other denominations (the Lutheran Church at Tabaka Mission Hospital) (personal

observation). However, due to lack of appropriate data on quality and cost of health care

at these institutions it is difficult to discuss the specific factors causing this behavior. Their

behavior might be a function of the cost involved, and perhaps quality of the services.

Differences in illness behavior and religion may relate to underlying factors such as

low level of drunkenness and alcoholism in some religious groups such as among the Jews

(Mechanic 1963) and SDA followers. Differences may also be due to other factors

associated to the health advocacy of the specific religious groups.



Health Seeking Behavior: A Case Study from Meru District, Kenya



In a study on rural household decision making in Meru district, Kenya, Mwabu

(1986) found that patients use multiple health care providers. According to Mwabu this is

because: (1) generally patients are unable to choose with certainty the provider who will

cure them; (2) successful treatment of some illnesses require more than one health care

provider; and (3) patients believe that in order to get cured they need treatment from more

than one provider (Mwabu 1986). Mwabu does not, however, tell us which illnesses

require more than one health care provider nor the priority that patients give to each of the

three reasons. Mwabu (1986) found that patients' visit patterns vary according to the type








66
and the stage of the illness. For example, in cases of diarrhea, malaria, leprosy, swellings,

tuberculosis, and heart problems patients went to mission clinics more frequently than to

other health care providers.

Mwabu calculated the conditional probabilities1 (Table 3.1) of choosing health care

providers. Results reveal that: (1) the probability of a patient returning to government

health clinic for follow-up is small; and (2) patients or care takers (the therapy

management group2) are independent of health care providers in making health care

decisions. Mwabu (1986) points out that conditional probabilities for the diagonals should

be 1.0 or close to 1.0 if health care providers had a significant influence on patients' visit

decisions. Conditional probabilities for the diagonals would be 1.0 if patients returned to

the same health care provider for subsequent treatments. He assumes that providers would

ask patients to return to the same health care facility. This is not always the case. Health

care providers consider many factors before making recommendations to patients. We

should not assume that their tendency is to ask patients to come back to the same facility

for follow-up visits. Some health care facilities are better equipped than others. Health

care providers will ask patients to go to better equipped facilities if the patient's condition

warrants. Thus health care providers in government health centers might refer patients to a




1 Conditional probabilities provide information about an event (E2) occurring given that
another event (Ei) has already occurred. Probabilities are re-evaluated in the light of partial
information about the outcomes. The conditional probability of B occurring given that A
[p(BIA)] has occurred is defined by the equation p(BIA) = p(AnB)/p(A) where p(A) is the
probability of A occurring and p(ArB) is the probability of A and B occurring together.

2The therapy management group is that group of people who make critical health care
decisions for a sick person. They may be relatives or close friends. I discuss at length the
role of the therapy management group later on in this Chapter.









government hospital. On the other hand, patients might want to visit mission or private

clinics because of the better services they provide.

What we know is that patients or their care takers and health care providers are

always faced with situations under which decisions must be made. Both the patients (or

the therapy management groups) and the health care providers must make decisions on the

best course of action, one that would yield best results. It is assumed that choices are

governed by the desire to maximize benefits and minimize the risks (Siminoff and Fetting

1989). This is the classical approach (of expected utility) used in decision making widely

utilized by economists. In the next section I highlight the key issues in decision making

with an emphasis on the health care seeking behavior.



Studies on Decision Making



The traditional view of decision making is that decisions are made in a series of

steps that involve: (1) problem definition and analysis, (2) the search for and formulation

of alternatives, (3) selection of maximizing alternatives, (4) decision implementation, and

(5) follow-up. This conceptualization of the decision making process is founded upon two

assumptions (Duncan 1973: 2). The first assumption is that decisions are made in

accordance with some predetermined goalss. The second assumption is that maximization

is the objective of the decision maker.

Duncan (1973) states that conditions which bring about maximization may not

hold true though decision makers may be tempted to think that they do. These

assumptions are relevant in many cases. However, they oversimplify the problems that










decision makers have to overcome as well as the criteria that decision makers take into

account. For example, in a decision making process where maximization is used as the

criterion several alternatives may be accessible or acceptable, but all alternatives may not

be available to the decision maker. In such circumstances the payoffs may not be the

maximum possible for the problem being solved.



Decisions Under Risk and Under Uncertainty

Decisions are made under conditions of risk and uncertainty (Luce and Raiffa

1957). When decisions are made under risk, the probabilities of the outcomes are known

to the decision maker. If a choice is made between getting $2,000 outright and taking a

50-50 gamble to get $0 or $5,000, such a decision is made under risk (Fishburn 1988).

However, sometimes decisions are made under conditions in which the

probabilities of succeeding are unknown to the decision maker. Decisions made under this

condition are reached under uncertainty. In medical decision making choices are made

under uncertainty. Patients do not know the probability of success for each of the available

options. Under these circumstances patients (or care takers) make use of individual

preferences.



Approaches in the Study of Decision Making

Decision making studies aim at constructing theories about how decisions are

made by an individual or within a social group, such as the family. Garro (1986) reviews

the various methodological and theoretical approaches employed by anthropologists to










study decisions to seek help and how patients make treatment choices from available

alternatives.

Two theories, normative and descriptive, have been proposed in studies on

decision making (Garro 1986). In normative decision theory, decisions come from rational

behavior of human beings. People evaluate the alternatives available to them, and

ultimately come up with a choice which best suits the prevailing circumstances. From the

observed behavior mathematical models can then be constructed and used to predict how

people are likely to make choices under new situations. This approach is motivated by a

desire to produce accurate predictions of decisions being made.

Some researchers have argued that normative approaches are not psychologically

realistic. Quinn (1978), for example, argues that in everyday life people do not make use

of complex mathematical probabilities in order to arrive at the most economical choice and

one that will produce the required results. Instead, decision makers use strategies known

as heuristics which simplify the decision making process. According to Quinn such

strategies eliminate the need for recall, summarization, and computation (Quinn 1978).

This second approach is known as descriptive decision theory. Under this theory people

make decisions that differ substantially from decisions that they should make according to

normative theory (Mathews 1982).

What these approaches have in common is that their ultimate aim is to tell us

decision outcomes, actual (descriptive decision theory) or expected (normative decision

theory). Normative decision theory does not tell us about the cognitive processes that take

place in a person's mind. In normative decision theory the cognitive process is a "black

box."










To overcome this shortcoming descriptive decision theory researchers (e.g.

Gladwin 1989; Young 1981a) have suggested the use of decision models (also known as

process models) built using information elicited from informants. To build a model,

informants should share a cultural domain. A cultural domain can be anything from shared

knowledge about disease symptoms, various types of cures available, to types of health

care providers and the disease they treat. Caution should be exercised in delimiting

boundaries within which a cultural domain may be studied. It is difficult to build a stable

model if the boundaries for a cultural domain are not well defined. Weller and Romney

(1988) treat this topic in detail.

If indeed a criterion for decision making exists, one can construct a stable model

from a small group of informants who share a cultural domain. This model can then be

tested on a new set of informants from the same cultural group as the first set of

informants. The new group's decisions are compared against the model's predicted

outcomes. According to Gladwin (1989), good models predict at least 85% of the

decisions. Ryan and Martinez's model predicted 84% of childhood diarrhea treatment

decisions made by Mexican mothers (Ryan and Martinez 1996).

It is appropriate to ask whether informants agree on what decisions are reasonable

under different conditions. This is considered under the domain of consensus theory.

Romney et al. (1986) have developed techniques that can be applied to determine the level

of agreement about the choices that people make. Consensus theory makes three

assumptions. First, participants share cultural knowledge. In the case of health care

behavior the shared cultural knowledge is the appropriate set of health care choices

available for any particular health problem. The second assumption is of local










independence: experts within a group will produce the same answers without having to

consult each other. The third assumption is that there should be homogeneity of items-the

questions should be of the same difficulty level and from the same domain. Borgatti

(1992) has developed a suite of computer programs, ANTHROPAC, that can be used to

assess the level of agreement among respondents. Consensus analysis can identify

culturally correct answers to a given set of questions. ANTHROPAC can also estimate the

competence of each respondent relative to other group members.

Once the three assumptions in consensus theory are satisfied, it can be shown

mathematically that competent informants will tend to agree (Romney et al. 1986). These

informants represent the cultural knowledge of a group people. Boster (1985) asked

Aguaruna Jivaro informants to identify manioc plants growing on two experimental

gardens. He walked them one by one through the gardens, stopping at each plant to ask

what kind of manioc the plant is. The first garden had 61 different varieties of manioc. The

plants on this garden required very fine discrimination. The second garden had 15

common varieties. Each variety was represented six times (total of 90 plants). Boster

analyzed his data by comparing whether a pair of informants agreed on the identity of

particular plants. He then computed two measures of agreement. Proportion of agreement

indicated the amount of agreement between pairs of informants while overall agreement

provided a measure of each informants' average proportion of agreement with the rest of

the population (Boster 1985: 181). Boster concluded that people who agree with one

another on manioc names tend to know more about manioc (p. 192). Those people, it

turned out, were related to one another.









A third approach in the study of health care is the use of explanatory models

(Kleinman 1980). This approach focuses on people's lived experiences with an illness.

Explanatory models (Kleinman 1980; Good 1986) produce rich understanding of how

people experience illness but are not meant to predict what people will do during an

illness. An explanatory model is a set of beliefs about the etiology and onset of symptoms,

the pathophysiology, and severity of illness, and the type of sick role and treatment of

illness (Kleinman 1978). According to Good (1986: 167) explanatory models are "frames

provided by culture that we do things with." These frames can be (and sometimes are)

applied widely by patients to include other issues of concern to them. Therefore, when

these frames are being interpreted by the physician they must be put placed within a wider

social context of the patient.

Explanatory models provide "a means of exploring patients' understanding of their

conditions, for explicitly comparing and contrasting the perspectives of close avenues of

care seeking, for research into the micro-level changes in patients' understandings, and for

comparisons across cultures and across ethnic groups" (Good 1986: 165). The purpose is

to understand how patients interpret and deal with illnesses. That is, there is a cultural

meaning placed on an illness by patients. For example, in the early New England puritan

world and in the twentieth-century Africa witchcraft symbolizes fear. In both societies

witchcraft emerged as a "major explanatory model of malignant illnesses" and it offered "a

magical means to exert control over seemingly unjust suffering and untimely death"

(Kleinman 1988: 19). It is evident from the foregoing that the main focus of explanatory

models is on how the afflicted represent and interpret their suffering. Nevertheless, we










would like to predict people's behavior as a result of illness. Predicting behavior is

important as it lets us forecast the likely demands on health care resources.



Framing Decisions: Its Influence on Decision Making

The manner in which a decision is framed determines the decision making process.

McNeil et al. (1982) and Tversky and Kahneman (1981, 1988) used vignettes (see Box

3.1 for an example) to show how problem framing can lead to different decision choices.

In their study McNeil and his colleagues provided respondents with two reference frames:

the survival frame and the mortality frame and a choice of going through surgery or

radiation therapy. In the survival frame, of 100 people going through surgery, 90 live

through the post-operation period, 68 are alive at the end of the first year, and 22 are alive

at the end of five years. For radiation therapy, of 100 people going through this treatment,

all live through the post-operation period, 77 live past the first year, and 22 live past the

fifth year. In the mortality frame the exact information found in the survival frame was

repeated but was negatively framed. Thus of 100 people having surgery, 10 die within the

post-operation period, 32 die by the end of the first year, and 66 die by the fifth year. Of

the 100 people who have radiation therapy none die in the post-operation period, 23 die

within one year while 78 die by the end of the fifth year. This framing of the problem

produced a marked effect on the respondents' decisions.

Tversky and Kahneman (1988) found that when problems were framed negatively

(for example when there is risk of death) people tend to make choices that sometimes

involve higher risks. However, when problems were framed positively (for example where

the risk of death is minimized) people tend to be risk averse.











Imagine that the US is preparing for the outbreak of an unusual Asian disease, which is
expected to kill 600 people. Two alternative programs to combat the disease have been
proposed. Assume that the exact scientific estimates of the consequences of the programs
are as follows:

Problem 1 (N = 152)
If program A is adopted, 200 people will be saved. [72%]
If program B is adopted, there is 1/3 probability that 600 people will be saved, and 2/3
probability that no people will be saved. [28%]

Problem 2 (N = 155)
If program C is adopted 400 people will die. [22%]
If program D is adopted, there is 1/3 probability that nobody will die, and 2/3 probability
that 600 people will die. [78%]

(The number of respondents (N) is given in each case. The percentage who chose each of
the options is indicated in brackets. In problem 1 the outcomes were positively framed
(lives saved). The majority of the respondents were risk averse. Problem 2 was negatively
framed (lives lost). The majority of the respondents were risk seeking.)

Box 3.1: Example of a vignette used by Tversky and Kahneman (1981) to investigate the
effect of problem framing on decision making.


In an earlier study Tversky and Kahneman (1981) showed the effect of certainty on

people's decision making. In a hypothetical epidemic problem they found that respondents

preferred an 80% chance to lose 100 lives to a sure loss of 75 lives. However, when the

probabilities were reduced by a factor of 10 people's preferences were completely

reversed. Thus respondents preferred a 10% chance to lose 75 lives to an 8% chance of

losing 100 lives. Outcomes perceived with certainty are given more weight relative to

uncertain outcomes (Eraker and Politser 1982).

The manner in which decisions are framed can lead to making of different choices

by the actors. Since framing of decisions is mostly subjective, people will tend to frame

similar situations differently. If this is done, a prevailing medical situation can lead to









75
different treatment seeking decisions that will have various medical outcomes. In fact as a

result of decision framing people may opt for choices with fewer benefits leaving other

choices that have better returns. This is consistent with the Miller-Lyer illusion in which

the shorter of two lines appears longer depending on how the lines are framed.

Decision framing can also determine who gets involved in the decision making

process. Where a medical condition is sudden and life threatening fewer people are likely

to be involved in decision making. This is because little time is available to make wider

consultations. The circle of people involved will increase after this initial period, as time

becomes available to consult, and as more of them get to know that there is a medical

problem that requires attention. In chronic and potentially fatal medical conditions more

people from a social group may be involved in decision making. However ultimate

decisions are made by the person (or a limited group of people) who has juridical authority

(Feierman and Janzen 1992: 18).



Decision Making: External and Internal Factors

Many anthropological studies on decision making focus on factors external to the

decision maker. Nardi (1983) notes that when agricultural decisions are being made

external factors like the weather conditions, labor supply, and landholdings are considered.

However, the way people frame decisions depends on a number of external and internal

factors. According to Nardi (1983), people's values, beliefs, aspirations and ambitions are

rarely considered.

Nardi argues that external factors are subject to change from time to time. Our

focus on ever changing external factors also implies that decisions made are based on









76
conditions current in people's lives. "This focus on the immediate present tends to deflect

attention from the more stable, enduring factors in the decision process: the decision

maker's internal attitudes, beliefs, and world view" (Nardi 1983). Internal factors can be

influenced by external factors. Attitudes and beliefs are learned from society. Once they

are learned, attitudes and beliefs get internalized by individuals. The extent to which these

attitudes and beliefs influence behavior depends on how the people have interacted with

the external factors. People's behavior is, therefore, more likely to be influenced by

external factors rather than by internal factors. Most external factors are structural and

infrastructural, while internal factors (attitudes, and beliefs) are superstructural.

Kayser-Jones (1995) studied factors that influence decisions to send patients to

nursing homes in the United States. She found that external factors in the decision making

process include the following: unpleasant nursing home environment; lack of supportive

services and equipment; and the practice of medicine by telephone. Internal factors include

values, attitudes, beliefs, goals, plans, and expectations held by decision makers.

The unpleasant nursing home environment. The environment that surrounds

nursing homes is seen as depressing. The staff who work in nursing homes are

overworked. The elderly people are poorly groomed and dressed in bathrobes and

slippers, and are usually restrained in wheelchairs. This reputation of nursing homes puts

fear in many elderly people. They fear being admitted to a nursing home. In addition, the

unpleasant working conditions makes most nurses prefer working in other health care

institutions.

Lack of supportive services. Nursing homes do not offer the latest high-technology

care. Most nursing homes do not provide basic supportive services such as laboratory,










pharmacy, and X-ray. Many elderly patients admitted to these homes usually come from

the high-technology atmosphere of an acute care hospital. Due to lack of supportive

services in nursing homes many may not want to be admitted.

Practice of medicine by telephone. Physicians prefer to provide services by

telephone without the benefit of a clinical evaluation. This is because they go through a lot

of paperwork in order to get reimbursed for visits to nursing homes. Providing services by

telephone and the unavailability of the physician may have adverse consequences. First, the

physicians do not get to see their patients frequently, even as the patient's condition

deteriorates; second, the physicians are usually busy and therefore telephone conversations

are hurried and brief; and lastly, in order to provide diagnosis and treatment of residents,

physicians must rely on the nurse's assessment of the resident.

Life/biographies and goals. In addition to three key internal factors (goals, plans,

and expectations, see Nardi 1983) that influence decision making, Kayser-Jones (1995)

expands the conceptualization of internal factors to include life/biographies and goals.

Goals are considered as the aims and aspirations held by people, plans are the means by

which we achieve those goals, and expectations are forecasts about the future (Nardi

1983; Kayser-Jones 1995). Kayser-Jones views goals and plans as the basic aspects of the

decision process that provide people guidelines needed to make choices. When physicians

attend to patients they have to deal with them and their families as people who respond to

various health situations and who have a life/biography that influences their goals, plans,

and expectations. Similarly, physicians and other health care professionals rely on their

past experiences (biographies). These experiences influence the decision making

processes.









78
Nardi (1983) and Kayser-Jones (1995) point out that external and internal factors

are important in decision making studies. However, they do not tell us the relative

importance of each of these factors. Under what circumstances would external factors be

more important than internal factors? Or, if a person were to make a choice when is it

necessary to drop either of the two? Cultural materialists argue that infrastructural factors

take priority over structural factors in explaining people's behavior, and that structural

factors are more important than superstructural ones. In Harris' words


... infrastructural determinism ... provides a set of priorities for the formulation and
testing of hypothesis about the causes of socio-cultural phenomena. Cultural
materialists give highest priority to the effort to formulate and test theories in
which infrastructural variables are the primary causal factors. (If they fail to
determine such causal factors, then theories are formulated in which the structural
variables are the primary causal factors.) Cultural materialists give still less priority
... to the behavioral superstructure. (Harris 1979: 56, parenthesis mine)

Many studies have shown the priority of infrastructural and structural factors (e.g.

Handwerker 1986; Iverson 1992--demographic transition) over superstructural factors in

explaining aggregate phenomena. At the level of individual decision-making, however,

emic cultural factors play a very important role. External factors are either infrastructural

or structural while the internal factors are superstructural. If we can account for behavior

using external (infrastructural and structural) factors then we ought to do that since they

are the primary causal factors.

It is far easier to change external than internal factors. Once external factors are

changed, people are more likely to change their internal factors (values, beliefs, and

attitudes). For example, increased employment opportunities for women (Handwerker

1986; Iverson 1992) have led to rapid demographic decline in many countries. This is in








79
contrast to areas where women's access to employment outside the home is not available.

People's attitude towards having large families have altered as infrastructural and

structural changes take place. For a discussion of how this takes place see Bernard and

Pelto (1987).



Decision Models and their Prediction Power

People appear to use ordered criteria when they make treatment decisions. Young

(1981a) developed a decision tree model to predict treatment choices in a Mexican village.

The model incorporated the following criteria: (1) perceptions about the seriousness of an

illness, (2) knowledge regarding available home remedies, (3) faith in the likelihood of a

cure, and (4) availability of money and transportation. Young's model predicted 95% of

the time for first treatment choices, and 84% of the time for the second treatment choices.

Young's model made two important contributions to our knowledge about illness

treatment choice. First, Young differentiated those treatment choices which were usually

made and agreed upon as strategies of first resort from those that were not. Second, he

accounted for the sequence in which treatment options were likely to be made. Individuals

employed ordering strategies in their choice of treatment. Ordering strategies enabled

those individuals to rank options on the basis of cost or likelihood of cure. Ranking

strategies/options is important to decision makers because it simplifies the decision making

process, although, as Duncan (1973) points out, the best strategies may not be available to

decision makers.

Mathews and Hill (1990) developed a model to study treatment choices in Costa

Rica. Their model predicted only 62% of the decisions correctly using nine rules. Errors in








80
the prediction of the model resulted from ethnic differences in the interpretation of illness

symptoms, preference for curing sources, and due to a constant change in the population

composition of the community.

In another study, Ryan and Martinez (1996) found that rural Mexican mothers

consider several factors in deciding what treatment to give in case of infantile diarrhea.

The mothers gave the following list of factors: duration of the diarrhea episode; perceived

cause (mothers in San Jose perceive diarrhea to be caused by dirty food, teething,

empacho (the sticking of the food in the stomach or intestines), heat, green fruit, worms,

presence of mucus in the stool, presence of blood in the stool, whether the child had fever,

color of the stool, whether the child had a dry mouth, whether the child had dry eyes,

whether the child was vomiting, and whether the child had swollen glands.

Mothers in San Jose choose treatment modality from seven alternatives that they

know. The treatment alternatives include: teas, carbonated beverages, rice water, sugar-

salt solutions (SSS), pills, physical manipulations, and western medical personnel (Ryan

and Martinez 1996). By looking at the pattern of circumstances for all the seven

treatments, Ryan and Martinez built a model which had just six rules and three constraints.

This model gave a postdiction of 89% of the treatments. On a new group of mothers the

model predicted 84% of the treatments correctly.

Weller et al. (1997) have reassessed the model of Young (198 la) and that of

Mathews and Hill (1990). She compared the models' predictive power due to chance.

Young's four treatment options could be predicted by chance 59% of the time. In the case

of Mathews and Hill's data, treatment behavior could be predicted by chance 23% of the

time. However, Weller and her colleagues concluded that Young's model still has a 88%









81
better than chance3 accuracy rate while Mathews and Hill's model has a 51% better than

chance accuracy rate.

I have argued in this chapter that infrastructural and structural conditions have a

greater effect on people's health care seeking behavior than superstructural conditions.

Included in infrastructural conditions is distance of health care facility from the homes of

patients, the presence or absence of roads to these facilities, and availability of affordable

transportation, while structural conditions include household income, education of

decision makers, number of children in each household, the patients' social networks, and

past illness experiences. Yet, the decision model built by Young (1981 la) had three

superstructural factors and only one structural factor as the main determinant of the health

care behavior. If infrastructural or structural factors have greater effect, we would expect

to see more of these factors in Young's model. Young's model provided a high prediction

power of people's illness behavior. In fact, the model's accuracy was 88% better than

chance (Weller et al. 1997). Does this violate the principles of cultural materialism?

Young (198 la) identified patient perceptions about illness severity, patient

knowledge regarding available home remedies, faith in the likelihood of a cure, and

availability of money and transportation as the main criteria used by patients. Three of the

criteria appear to belong to the superstructure. Perception about illness severity is

influenced by illness characteristics such as temperature, length of period the patient has

been sick, and the patient's own physical reaction to the illness all which are infrastructural


3 Better than chance is calculated as the difference between observed accuracy rate and
chance accuracy rate divide by 1 minus chance accuracy rate. For Young (1981a) this
equals [.95-.59]/[1-.59] = .88 and for Mathews and Hill their model's predictive power =
[.62-.23]/[1-.23] = .51 (Weller et al. 1997).










or structural factors. Knowledge regarding home remedies is influenced by social

networks, earlier contacts with health care facilities or health care providers. Thus home

remedies are derived from structural factors which leaves faith alone as the only

superstructural factor. Faith appears not to have a direct underlying infrastructural or

structural forces. If this is the case, then Young's model is consistent with the first

principle of cultural materialism. Similarly, Ryan and Martinez's (1996) as well as

Mathews and Hill's (1990) models can be shown to be driven by underlying infrastructural

and structural factors. Cognitive models do not violate the infrastructure and structure

priority principle of cultural materialism, though cognitive anthropologists rarely connect

their theories to a materialist base (for an exception see Bernard and Pelto 1987).



Decision Making and the Family

Decisions and treatment choices must be made by the patient or by those in charge

of the patient. A patient is part of a "family." Therefore the family shares their illness

experiences. The family helps in the decision making so as to come to an acceptable

treatment choice. Families may be considered as groups composed of members who have

mutual obligations to provide a broad range of emotional and material support,

particularly at times of crises or threatening events (Dean et al. 1981). Dean and his

colleagues argue that social support buffers the impact of stressful life events, and may

also directly influence the occurrence of various disorders.

According to Turk and Kerns (1985) families do not necessarily consist of blood

relatives. Members do not have to live together, but they must have "mutual obligations"

to support each other. However, Turk and Kerns do not specify whether, for example,










children have mutually obligatory roles in the family. They further point out that families

have (a) structure, (b) functions and assigned roles, (c) modes of interacting, (d)

resources, (e) a life cycle, (f) a history, and (g) a set of individual members with unique

histories.

The structure of the family refers to characteristics of the individual members that

make up the family unit. These characteristics include gender, age, spacing and number of

family members. Function refers to tasks the family performs for society and its members

while assigned roles refer to responsibilities, expectations, and rights of the individual

members. A person's responsibilities will determine their behavior when they or a member

of the family is sick. Further, responsibilities may overlap. For example, mothers may be in

charge of the health of the children as well as managers within the household. The mode

of interaction determines the style adopted by family members to deal with the

environment and with one another with regard to problem solving and decision making.

The family members' general health, social support and skills, personality

characteristics and financial support constitute the resources. Social support includes the

relationships of the patients to their families, the medical staff, friends, neighbors, co-

workers, and other social agencies while personality characteristics include age,

intelligence, cognitive development, philosophical or religious beliefs, and previous coping

experiences (Turk 1979). These factors influence each individual's coping with disease.

According to Turk, patients who have satisfactory adjustment employ a wider range of

personal and environmental resources. However, it does not imply that patients with

satisfactory adaptive responses are always successful in coping with disease. Thus

resources influence the interpretation of and coping with disease.










Family history includes socio-cultural factors and prior illness experiences and

modes of coping with stress. A family's past experiences determine the manner in which

they respond to new illness episodes. With each new illness phase, the family goes through

a life cycle. In each of these phases family members have roles and obligations that relate

to a person's station in life. Roles played by various family members are transient.

Members of a family assume new roles as they move into a new social classification within

the family or society. The roles "assigned" to (or assumed by) members determine the

amount of time they spend out in the field farming, taking care of animals, or doing any

other activities.

Family members are exposed to certain disease causing agents depending on the

type of activity they do. For example, women who cook in poorly ventilated and smoky

kitchens (using cow-dung) are more likely to suffer upper respiratory infections than are

those who cook in well ventilated kitchens. Snow et al. (1988) have calculated the relative

risk of mosquito bites faced by people at night. The greatest risk is between 5.00 a.m. and

6.00 a.m. Risk is calculated as the proportion of each hour that people spend outside their

houses multiplied by the amount of biting in each hour. In societies where a certain group

of people (e.g. women milking cows in the early morning and late in the evening) are

outside at this time their likelihood of suffering from malaria is increased. Their chance of

having a mosquito bite is greatly increased as a result of the activities they perform

between 5.00 a.m. and 6.00 a.m. Families are made up of individual members each of

whom has unique experiences. Their experiences affect illness behavior, including decision

making and treatment choice.










Across Africa-indeed, in most of the world-the sick are cared for within the

framework of the family. In the next section I will discuss the importance of the family as

the primary therapy management group.



Therapy Management Groups



Therapy managers in Africa constitute the heart of African healing. Therapy

management fulfills two functions. The first function deals with authoritative diagnosis and

control over the treatment and, the second function is supportive care (Feierman and

Janzen 1992: 18). Authoritative diagnosis and control over the treatment is usually in the

hands of one person (or a limited group) who has juridical authority over the patient.

Adult patients sometimes make their own therapy management decisions. However, the

extent of their independence is determined by the nature of their illness. If the disease is

serious, a therapy management group may take over the therapy management decisions.

Even then adult patients can not be forced to accept treatments against their will. If the

therapy management group is unanimous in their choice of treatment, it may be provided

without the knowledge of the patient.

Supportive care may be provided by anyone: neighbors, old friends, passers-by,

and distant relatives, but their suggestions will be accepted only if allowed by authority

holders. Unlike juridical authority, supportive care is distributed more widely (Feierman

and Janzen 1992: 18).

In many households decision making is sometimes left to individuals who are

regarded as the central figures. These individuals make all major decisions on what is to be










done in the family (see for example Silberschmidt 1992). However, granting decision

making powers to one person is not a universal phenomenon. In some communities

intrafamilial negotiation contributes a lot to family decisions.

Feierman (1985) discusses ways in which family structure and responsibility of

family members affect medical activity. Within the extended family, the structure of the

therapy managing groups (Janzen 1978) allows input from a wider circle of relatives.

Relatives in this circle contribute in terms of disease recognition (diagnosis), treatment,

and choice of health care. However, Feierman (1985) argues that where therapy

management groups are involved, it is difficult to determine their composition and the

contribution of each group member. The groups are not well defined.

Group composition is fluid with ever changing membership. Some members within

the group express opinions, which are never acted upon, and yet these members are

considered as part of the therapy management group. The fluid nature of the therapy

management group hides the primary difference between the sets of people, who have

jural responsibility for the patient's welfare and the other which may provide voluntary

assistance but have no rights or obligations in the transactions.

Therapy management groups have one other feature. They do not have a

distinctive institutional hierarchy (Feierman 1985). Instead the groups are embedded

within the general patterns of control over domestic and community affairs. The

implication of this is clear. Because of the link between lay therapy managers and

generalized authority in the domestic and community arena, factors that act on the local

community will also affect healing.










In rural areas of the non-industrialized world, many people migrate to towns in

search of jobs. Migration of people may have altered the structure of the family and the

therapy managing groups but such groups do still get involved in decision making.

However, their involvement is limited (Pearce 1993: 151). Their involvement might

depend on factors such as their closeness to the patient, and the seriousness of the disease.

Close relatives such as siblings will usually be informed at the earliest opportunity.

Although they might not get involved in direct decision making, they help in the

implementation of the decisions by making contributions towards treatment costs or social

support.

Kinfolk play a central role in therapy management. In some communities such as

the Abagusii of Southwestern Kenya, key decision makers may be overruled by certain

members within the immediate family. I discuss circumstances under which this might

occur among the Abagusii in Chapter five.



Conclusion



In this chapter I have laid out factors that influence people's health care seeking

behavior. I have reviewed literature pertinent to decision making, on the determinants of

health care seeking behavior, and on the role of the family and therapy management

groups.

What emerges is a complex relationship between the determinants of health care

behavior, the family, and the therapy management groups. The relationship between the

various determinants of health behavior can be broken down into blocks that consist of









infrastructural, structural, and superstructural components. But still we remain with the

primary question: Can we model illness behavior more accurately given our knowledge

about the various factors that influence it? Predicting people's illness behavior is

important. It lets us forecast the likely demands on health care resources.

Three models have been widely used to study illness behavior of patients. These

are: explanatory models (Kleinman 1980; Good 1986); determinants models (Mechanic

1969; Ryan 1995); and process models (Young 1981b; Ryan 1995; Ryan and Martinez

1996). They are a good beginning point in the study of illness behavior.

Explanatory models produce a rich emic understanding of people's illness

experience. They provide a way for exploring the patients' understandings of their

conditions, a means for explicitly comparing and contrasting the perspectives of clinicians

and patients, [and] for investigating how cognitive orientations influence avenues of care

seeking (Good 1986). Explanatory models, however, do not predict patient health seeking

behavior.

Determinant models produce a rich etic understanding of people's behavior, but

they do not provide us with enough emic information. Determinants of health behavior

include the following: patient-based factors; provider-based factors; care taker

perceptions; social and demographic factors; availability of funds; distance from the health

care provider; social networks; and biological signs and symptoms (Ryan 1995).

Determinant models predict about 20%-30% of people's illness behavior and they provide

some understanding of people's illness behavior.

Process models have high prediction rates, typically about 80% (Gladwin 1989).

They provide high emic understanding of people's behavior but give us little etic




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs