Background
The economic burden of malaria
Country and Authors | Direct costs per capita per month (1999 US$) | Monthly total direct costs | Direct costs as % of income | |
---|---|---|---|---|
Prevention | Treatment | |||
Sri Lanka [8] | - | 1.91 | 1.91 | 2.0 |
Malawi [16] | 0.05 | 0.41 | 0.46 | 2.0 |
Tanzania [42] | 0.76 | - | - | - |
Zaire [47] | 0.97 | - | - | - |
Cameroon [45] | 1.29 | 2.05 | 3.34 | - |
Cameroon [41] | 1.74 | 2.67 | 4.41 | - |
Cameroon [41] | 2.10 | 3.88 | 5.98 | - |
Burkina Faso [44] | 0.09 | - | - | - |
Burkina Faso [44] | 0.93 | 1.18 | 2.11 | - |
Ghana [7] | - | 0.65 | 0.65 | - |
Nigeria [11] | - | 1.84 | 1.84 | 2.9 |
Strategy | Highly vulnerable | Vulnerable | Least vulnerable |
---|---|---|---|
Borrowing | Rarely borrowed cash because they were not creditworthy (too poor to pay back); fear of borrowing and being unable to pay back leading to bad reputation & gossip | A common strategy because they had moderate assets but still not enough to rely more on other sources of credit like shops or private providers | Not common because they had other sources of credit but they could easily borrow if need arose. |
Amount of money borrowed was small (KES 10) because their friends were equally poor | Could borrow up to KES 100–200 | Could easily borrow KES 5000 if need be because their friends were in a good economic situation | |
Credit from private providers | Not accessible by these households due to poverty | Could get treatment on credit but limited amounts depending on providers understanding of their economic status | Unlimited access to credit from providers because they were wealth, had permanent jobs & could easily pay by end of the month |
Credit from shops | Occasionally but small amounts to buy drugs | Had access to credit but could be denied when they asked for large amounts | Could acquire all goods on credit until end of the month |
Sale of assets (Goats & chickens) | Those that had assets sold them to pay for treatment or other needs but some had nothing to sell | Sold assets but usually to clear a debt at private providers. | Assets not sold to pay for treatment because there were other 'better' options |
Sale of labor on farms | Preferred but not used due to drought | A possibility but drought limited its use | Unlikely for these households to use the strategy |
Borrowing drugs | Preferred because they had no access to cash and were not required to pay back drugs | Not reported | Not reported |
Sharing drugs | A common strategy when drugs are borrowed or bought | Common for households with many children | Reported when more than one child fell ill at the same time |
Ignoring illness | A common strategy because they rarely had cash and access to other strategies was limited | Reported on two occasions because illnesses not perceived serious enough | Not reported |
Malaria, poverty and vulnerability at the household level: an analytical framework
Materials and methods
Study setting
Data collection and analysis
-
Household surveys: Maps indicating the location of every homestead and landmarks were drawn by hand to enable the random selection of survey households. A total of 294 households were visited in the wet season. The same households were then visited in the dry season (n = 285, 9 refusals). The survey gathered information on socio-demographic characteristics, direct and indirect costs, and expenditure and coping strategies. The questionnaire was administered to the household head or spouse and in his/her absence, another senior adult member of the household. The majority of the respondents were female since males were often absent. For illnesses reported among adults and children, efforts were made to interview the ill person and the primary carer respectively.
-
Case studies: 15 households were purposively selected to represent varying degrees of vulnerability and poverty in the community and thereby contribute to deepening theoretical understanding. Selection was based on indicators of socio-economic status (including poor and less poor households), cost burdens (including households with high and low costs) and coping strategies (Figure 1). Case study households were visited monthly for a year. Data on illness, treatment seeking patterns and cost burdens were collected once every month and updated until full recovery was reported. More in-depth information on other aspects highlighted in Figure 1 were captured in an additional five sets of visits organized over the follow-up period. Topics included factors that influenced the household's situation before the study started, the range of assets households have access to, social networks and their role in meeting malaria costs, and household debt and repayment. A final visit at the end of the study explored changes in asset composition over the study period, and perceived reasons for that change.
Results
Contextual and household level factors influencing vulnerability (Boxes A and B of Figure 1)
Treatment seeking behavior, economic pathways and coping strategies (Boxes C-E of figure 1; survey data)
Wet season (%) | Dry season (%) | P values | |
---|---|---|---|
Households reporting at least one malaria episode 2 wks before survey | 187 (63.6) | 104 (36.5) | <0.001 |
Number of ill individuals | 307 (14.2%) | 187 (8.8%) | <0.001 |
Individuals reporting malaria by age | |||
• <5 | 95 (31.0) | 70 (37.5) | 0.79 |
• 5–<10 | 62 (20.2) | 36 (19.3) | 0.80 |
• 10–<18 | 60 (19.5) | 34 (18.2) | 0.71 |
• 18–<35 | 37 (12.1) | 24 (12.8) | 0.78 |
• 35+ | 53 (17.3) | 23 (12.3) | 0.14 |
Actions taken within HH: | |||
• Herbs | 26 (7.5) | 19 (11.5) | 0.13 |
• Modern drugs already there | 22 (6.4) | 6 (3.6) | 0.21 |
• Prayers | 18 (5.2) | 9 (5.5) | 0.89 |
Actions taken outside HH: | |||
• Shops | 196 (56.7) | 84 (51.2) | 0.25 |
• Private clinic | 39 (11.3) | 10 (6.1) | 0.06 |
• Government | 30 (8.7) | 25 (15.2) | 0.00 |
• Healer | 3 (0.9) | 3 (1.8) | 0.39 |
12 (3.5) | 8 (4.8) | 0.47 |
Variable | Wet season | Dry season | p-value |
---|---|---|---|
Mean monthly expenditure per household in KES (median) | 271 (55) | 165 (40) | 0.13 |
Mean monthly direct costs as % of expenditure (median) | 7.1 (2.1) | 5.9 (1.4) | 0.58 |
Mean monthly indirect cost as % of expenditure (median) | 5.4 (0.0) | 2.1 (0.0) | 0.04 |
Mean direct costs as % of monthly expenditure
| |||
• Poorest | 11.0 | 16.1 | 0.47 |
• Very poor | 7.8 | 3.2 | 0.18 |
• Poor | 5.0 | 3.7 | 0.59 |
• Less poor | 6.8 | 3.3 | 0.37 |
• Least poor | 3.4 | 2.6 | 0.57 |
Indirect costs as % of monthly expenditure
| |||
• Poorest | 8.1 | 1.9 | 0.17 |
• Very poor | 5.7 | 3.3 | 0.47 |
• Poor | 3.4 | 3.5 | 0.89 |
• Less poor | 1.4 | 1.5 | 0.43 |
• Least poor | 1.6 | 0.5 | 0.9 |
Households adopting coping strategy
(%)*
|
n = 74
|
n = 42
| |
• Borrowing | 37 (50.0) | 29 (69.0) | 0.05 |
• Gifts | 29 (39.2) | 7 (16.7) | 0.01 |
• Sell labour | 21 (28.3) | 4 (9.5) | 0.02 |
• Sell assets | 6 (8.1) | 7 (16.7) | 0.22 |
• Credit from health care provider | 8 (10.8) | 6 (14.3) | 0.57 |
• Other (mixed) | 14 (18.9) | 12 (28.6) | 0.25 |
Exploring the links between malaria, poverty and vulnerability – case study findings
Vulnerability and poverty before and at the end of the research
-
Experienced a stressful event in the past and had not yet recovered. All had depleted their assets and were descending into poverty when the research started. Three reported the main cause of their economic decline to be malaria;
-
Accumulated large debts they had not been able to repay. Often these debts were accrued in their attempts to finance past treatment-seeking, especially hospitalizations and funeral expenses;
-
No source of regular income, jobs were insecure and unpredictable;
-
Limited asset bases and were 'struggling' to survive.
-
Experienced a stressful event in the past but still had assets that would enable them to cope with future costs. Two reported having had their livelihoods affected by malaria prior to the study;
-
At least one source of regular income, but no permanent employment;
-
A moderate asset base (all owned cows or goats).
-
Did not report any events with major impacts on their livelihoods. None reported major malaria episodes in the years preceding the study;
-
Had at least one member with a permanent source of income (thus secure income in the long-run)
-
Had accumulated assets, primarily cows and goats and had savings with financial institutions.
Status at the beginning | Status at the end of the research | ||
---|---|---|---|
Declined | Stable | Improved | |
Highly vulnerable | 5 | 1 | 0 |
Vulnerable | 1 | 4 | 1 |
Least vulnerable | 0 | 1 | 3 |
Factors influencing outcome at the end of research
Contextual factors influencing outcome (Box A of Figure 1)
Self reported malaria, and treatment seeking, payment and coping patterns (Boxes B-E of Figure 1)
Household | Self reported malaria | Number of times household used type of treatment | Average monthly cost burdens (%) | Outcome at end of the research | |||||
---|---|---|---|---|---|---|---|---|---|
Total episodes | Per capita episodes | Shops | Dispensary | Private | Herbs | Healer | |||
1 | 9 | 1 | 1 | 2 | 4 | 0 | 0 | 19.6 | Declined |
2 | 8 | 1 | 3 | 3 | 0 | 2 | 1 | 0.3 | Declined |
3 | 7 | 1 | 2 | 1 | 2 | 2 | 1 | 6.0 | Declined |
4 | 3 | 0 | 3 | 0 | 2 | 0 | 0 | 0.5 | Declined |
5 | 9 | 1 | 2 | 0 | 0 | 4 | 0 | 0.3 | Declined |
6 | 5 | 1 | 0 | 1 | 3 | 0 | 2 | 12.1 | Stable |
7 | 7 | 1 | 4 | 1 | 0 | 1 | 0 | 0.2 | Stable |
8 | 7 | 0 | 5 | 2 | 1 | 0 | 1 | 5.1 | Stable |
9 | 28 | 1 | 15 | 1 | 6 | 0 | 1 | 7.3 | Improved |
10 | 10 | 2 | 1 | 0 | 0 | 1 | 0 | 0.1 | Declined |
11 | 5 | 1 | 0 | 0 | 0 | 0 | 0 | 0.0 | Stable |
12 | 5 | 1 | 1 | 0 | 4 | 0 | 0 | 2.0 | Stable |
13 | 13 | 2 | 8 | 1 | 3 | 0 | 0 | 1.0 | Improved |
14 | 3 | 1 | 2 | 0 | 1 | 0 | 0 | 1.1 | Improved |
15 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.0 | Improved |
Total
|
119
|
48
|
12
|
19
|
10
|
5
|
Discussion
Seasonality of cost burdens
Factors contributing to malaria vulnerability
The relationship between malaria cost burdens and outcome
-
Exposure to risk of infection and cost burdens does not always lead to increased vulnerability or poverty: households' assets endowments can mediate potential impact through supporting cost management strategies. Households that had a good asset base were able to meet arising costs without depleting their resources and without risking any decline;
-
For poor and highly vulnerable households, the type of treatment-seeking behavior selected was often a coping strategy. Highly vulnerable households prevented costs by adopting treatment actions that did not require cash (for example herbs) and by not seeking any treatment. Therefore, low cost burdens do not necessarily imply less need but rather can indicate desperation or affordability barriers to seeking care. In a community characterized by high levels of poverty and food insecurity, seeking health care might not be 'important' because diverting even small amounts away from food towards shop drugs is difficult.
-
Household poverty and vulnerability is influenced by many factors and the costs of malaria are just one. Key factors identified in this study were natural factors (drought and hunger); economic factors (types of jobs and income levels); social and demographic factors (household composition and management of resources); and illness and health (number of reported illnesses). Based on this it is difficult, if not impossible, to associate a single factor to outcome. For example, some households sold assets to buy food, pay school fees for their children or to pay for the treatment of other illnesses (not malaria). What is clear is that high costs of treatment, together with accumulated debts and sale of animals led to increased poverty and vulnerability for some households and constrained improvements for others.