Catastrophic out-of-pocket payments for households of people with severe mental disorder: a comparative study in rural Ethiopia

In a recent global monitoring report by the World Health Organization (WHO) and the World Bank, the incidence of catastrophic health expenditure was shown to have increased globally [1]. Indeed, each year, 100 million people face poverty and another 800 million people suffer due to catastrophic out-of-pocket (OOP) health expenditures [1]. Out-of-pocket payments make up a significant proportion of the total expenditure on mental health care in most low-income countries [2]. In 43% of countries in the African region, OOP spending is the main source of financing for mental health services [3].

In 2013/14, total health expenditure was approximately 4.7% of Ethiopian gross domestic product (GDP), with per capita health expenditure amounting to US$28.65 [4]. Household out-of-pocket expenditure represents the largest financing source for health care, accounting for about 33% of total health expenditure [4]. Although a fee waiver program was introduced as a measure to protect the poor from the negative effects of out-of-pocket payments in Ethiopia, the coverage remains low in many regions [5]. Hence, chronic illnesses with ongoing need for care, as is seen with severe mental disorders (SMD) such as schizophrenia and bipolar disorder, exert a heavy economic burden on vulnerable households in accessing mental healthcare [6]. Evidence indicates that the extent of out-of-pocket payments for mental health problems varies from country to country and according to the methodology used [7‐9]. In an Indian study, the percentage of women with severe depression who had experienced catastrophic expenditure was 14.6%, compared to 4.9% for the non-cases [7]. Similarly, in a study conducted in Bangladesh, 15.1% of households with a person with mental illness encountered increased household expenditure that resulted in distress financing [8].

Out-of-pocket health care spending imposes a substantial financial burden on households in many countries and leads the household to sacrifice consumption of other items, to incur debts, or to sell productive resources [10, 11]. Furthermore, health care expenditures have been reported as one of the major pathways into poverty and an important reason why many households remain poor [12‐15].

Although studies have reported the impact of out-of-pocket expenditure on households; the investigation of household costs of disease is incomplete if coping strategies employed by households are ignored [16, 17]. Evidence suggests that in most African countries, the health financing system is too weak to protect households from health shocks (illness). Hence, using coping strategies to finance health care is common [18]. These strategies include informal transfers (in cash and in-kind) from relatives and friends [19], selling assets [20] or obtaining loans [14, 21], withdrawing children from school [20, 22] and taking on extra work [23]. Evidence from middle-income Asian countries, including China [24], Vietnam [25, 26], and Thailand [27], indicates that illness and associated medical expenditures have a significant negative impact on household basic consumption. In the Ethiopian context, Asfaw and Braun [28] and Dercon et al. [29] show that when the head of the household is the person who is unhealthy, non-medical consumption declines sharply, with reductions ranging from 15 to 35%.

The study reported in this paper is part of the cross-country Emerald programme (emerging mental health systems in low- and middle-income countries) which sought to provide rigorous, population-based evidence about the adequacy and fairness of mental healthcare financing [30].

In this study, we investigated health expenditure and financial coping strategies in relation to SMD in rural Ethiopia. We hypothesized that the incidence and intensity of catastrophic health expenditure and associated coping strategies would be significantly higher in households including a person with SMD compared to households with no affected family member. We also explored the factors associated with catastrophic health expenditure and coping strategies.

A flow diagram of participant recruitment and inclusion in the study is presented in Additional file 1: Figure S1.

In our study, the incidence rate of catastrophic health expenditure at various thresholds of household total expenditure and capacity to pay was consistently higher for households of a person with SMD compared to households without a person with SMD. In both households with and without SMD, the level of catastrophic health spending was high compared to the average for the African region of 11.4% [47]. Furthermore, at a 25%+ threshold of total expenditure, the proportion of households experiencing catastrophic expenditure was almost two times higher for households with SMD compared with households without an affected person. Based on the threshold of 40%+ of consumption-based capacity to pay, our result remains significantly higher for households of a person with SMD.

Similarly, Zuvekas and Selden [48] in USA, Patel et al. [7] in India and Zergaw et al. [49] in Ethiopia reported that families with one or more member experiencing mental health problems were more likely to have high out-of-pocket spending compared with the control groups. Moreover, this study is consistent with findings from an earlier study from the USA that reported people with bipolar disorder incurred over four times greater costs compared with the non-bipolar group [50]. Likewise, in São Paulo [51] and in the USA [52], having a mental disorder was strongly correlated with higher health care payments. The higher burden of catastrophic health payments in households of a person with SMD can be explained by the financing structure of the available health system, the scarcity of locally available treatment services, and the marked disabling nature of the disorder that may increase health service use which in turn leads to high OOP expenditure.

We found evidence of regressive health spending, whereby households in the lowest quintile of socio-economic status paid 38.0% of their capacity to pay compared to only 16.0% for the highest quintile. The differences found for catastrophic health spending based on capacity to pay in relation to mental health status were significant and also followed a pattern linked to socio-economic status of the household. These results indicate that catastrophic health spending affects more households of a person with SMD and the poorest households. In the absence of financial protection (fewer than 3% of households in our study were enrolled in a social protection scheme), catastrophic out-of-pocket payments will further increase the number of households of a person with SMD falling into poverty, as well as potentially act as a deterrent from using services. Having an older household member was also associated with catastrophic health expenditure. In other studies, the presence of elderly members in the household was also found to lead to catastrophic health expenditure [53‐55]. This may be because older adult members require more health care that may result in higher health expenditure. Nonetheless, there was no statistically significant difference between gender, educational level of the head of the household, residential location, children in the household and incurring catastrophic health expenditure. Previous studies on non-mental health problems reported mixed findings [53, 56‐59].

This study has shown that mental health status, demographic and economic factors play an important role in the coping strategy adopted by households in Sodo. Our study indicated that the probability of withdrawing children from school was significantly higher for households of a person with SMD. A number of studies have found that children are taken out of school during times of illness in order to take up the slack in the workload in their families [12, 20, 22, 60]. A recent population-based cohort study in Ethiopia demonstrated that maternal common mental disorder was associated with child school dropout and absenteeism [61]. For people with SMD, the impact on child education may be more enduring due to the tendency for the illness to persist or recur. From an economic point of view, such action in response to healthcare costs will have a long-term impact on investment in human capital which may have serious implications for the education of future generations.

We had shown previously in the same sample that households of people with SMD are more likely to be severely food insecure than comparison households [62]. Our finding that households of a person with SMD were more likely to reduce their daily food consumption to cope with healthcare costs indicates that catastrophic health costs are one pathway towards food insecurity. When OOP payments comprise a large share of household budgets, households are at risk of sacrificing current consumption and experience food insecurity to pay for medical costs [60, 63‐65]. Previous study on coping strategies for cost of illness in Ethiopia found that roughly one-third of households had to sacrifice other essential spending to pay for medical treatment [23]. Therefore, programmes that reduce out-of-pocket payments may indirectly improve access to food and nutrition.

Severe mental disorders are often chronic or recurrent, with a need for ongoing medical care for the best outcome. However, the results indicate that households of a person with SMD were more likely to adopt cost prevention strategies, like reducing medical visits, compared to comparison households, at the time when they most needed to access services. This strategy may threaten the person’s recovery from illness. Studies on non-communicable diseases reported similar findings [60, 66, 67].

Selling of assets in households of a person with SMD tended to be significantly lower than comparison households. This is explained by the possession of fewer assets that could be sold by these households [68]. Likewise, despite greater need, borrowing or taking a loan was not a common strategy adopted by households of a person with SMD. This might be because households with a person with SMD have less capacity to get loans because they cannot guarantee repayments. Nonetheless, previous studies on coping strategies for cost of illness in Ethiopia have found that roughly one-third had to adopt a cost management strategy, such as borrowing, selling assets or mortgaging a crop [23, 69]. Empirical evidence shows that borrowing and selling assets to cope with medical costs ranged from 23% of households in Zambia to 68% in Burkina Faso [18]. In India, half of the out-of-pocket expenditure made by households for psychiatric disorders came from loans [70]. This is not in line with our findings. Studies in the literature reported that coping strategies that can be adopted for financial stress are likely to be context and country specific [71, 72].

A transfer of either goods/money from friends and relatives was reported to be a common strategy in households of people with SMD in our study. Empirical evidence from low-and middle-income countries indicates that assistance from relatives and friends is very common when household members face illness, and that this acts as an insurance mechanism rather than as credit [73].

The least common coping response by households of a person with SMD to the financial stress in our data was the use of savings. This may reflect the fact that households of a person with SMD rarely have savings. Gresenz and Sturm [74] found that, compared to mentally healthy individuals, mentally ill individuals are 75% less likely to have any savings.

Our study is a population-based study of people with a clinically confirmed diagnosis of SMD using standardized measures. Most previous studies of health expenditure in people with mental health problems in LMICs have focused on depression/anxiety [7] or relied upon facility-based studies which are subject to selection bias [75]. The inclusion of matched comparisons without the condition in our study enabled us to estimate the net effect of SMD on out-of-pocket health expenditures. This approach is recommended by WHO in estimating economic consequences of a disease [76]. Furthermore, we estimated expenditures in a comprehensive and systematic way and used households as the preferred unit of analysis for assessing out-of-pocket payments because decisions about treatment and coping are based on negotiations within the household.

Limitations of the study are that self-reported household consumption expenditures may not be accurate due to recall bias. We did not include costs associated with treatment seeking, like transportation, which may lead to underestimation of the financial impact of out-of-pocket payments for health care on households. The study was based on a cross-sectional survey, so the temporal association between severe mental disorder and catastrophic expenditure was not established. Hence, future research could examine how OOP burdens vary over time with service utilization patterns.

Catastrophic out-of-pocket health expenditures in households of a person with SMD at any threshold were high and associated with financial coping strategies which may lead to poorer health outcomes, entrenchment of poverty, and intergenerational disadvantage. Therefore, policy interventions aimed at addressing the high burden of OOP payments among vulnerable households should be in place. Such mechanisms can include: subsidies, cash transfers or exemption mechanisms for accessing health care and the scale up of integrated mental health care, as well as risk pooling mechanism such as community-based health insurance [77].