BackgroundThis study examines health-related "hardship financing" in order to get better insights on how poor households finance their out-of-pocket healthcare costs. We define hardship financing as having to borrow money with interest or to sell assets to pay out-of-pocket healthcare costs.MethodsUsing survey data of 5,383 low-income households in Orissa, one of the poorest states of India, we investigate factors influencing the risk of hardship financing with the use of a logistic regression.ResultsOverall, about 25% of the households (that had any healthcare cost) reported hardship financing during the year preceding the survey. Among households that experienced a hospitalization, this percentage was nearly 40%, but even among households with outpatient or maternity-related care around 25% experienced hardship financing.Hardship financing is explained not merely by the wealth of the household (measured by assets) or how much is spent out-of-pocket on healthcare costs, but also by when the payment occurs, its frequency and its duration (e.g. more severe in cases of chronic illnesses). The location where a household resides remains a major predictor of the likelihood to have hardship financing despite all other household features included in the model.ConclusionsRural poor households are subjected to considerable and protracted financial hardship due to the indirect and longer-term deleterious effects of how they cope with out-of-pocket healthcare costs. The social network that households can access influences exposure to hardship financing. Our findings point to the need to develop a policy solution that would limit that exposure both in quantum and in time. We therefore conclude that policy interventions aiming to ensure health-related financial protection would have to demonstrate that they have reduced the frequency and the volume of hardship financing.

Background & objectives:Against the backdrop of insufficient public supply of primary care and reports of informal providers, the present study sought to collect descriptive evidence on 1st contact curative health care seeking choices among rural communities in two States of India - Andhra Pradesh (AP) and Orissa.Methods:The cross-sectional study design combined a Household Survey (1,810 households in AP; 5,342 in Orissa), 48 Focus Group Discussions (19 in AP; 29 in Orissa), and 61 Key Informant Interviews with healthcare providers (22 in AP; 39 in Orissa).Results:In AP, 69.5 per cent of respondents accessed non-degree allopathic practitioners (NDAPs) practicing in or near their village; in Orissa, 40.2 per cent chose first curative contact with NDAPs and 36.2 per cent with traditional healers. In AP, all NDAPs were private practitioners, in Orissa some pharmacists and nurses employed in health facilities, also practiced privately. Respondents explained their choice by proximity and providers’ readiness to make house-calls when needed. Less than a quarter of respondents chose qualified doctors as their first point of call: mostly private practitioners in AP, and public practitioners in Orissa. Amongst those who chose a qualified practitioner, the most frequent reason was doctors’ quality rather than proximity.Interpretation & conclusions:The results of this study show that most rural persons seek first level of curative healthcare close to home, and pay for a composite convenient service of consulting-cum-dispensing of medicines. NDAPs fill a huge demand for primary curative care which the public system does not satisfy, and are the de facto first level access in most cases.

No abstract

Abstractobjective Non-communicable diseases (NCD) are on the increase in low-income countries, where healthcare costs are paid mostly out-of-pocket. We investigate the financial burden of NCD vs. communicable diseases (CD) among rural poor in India and assess whether they can afford to treat NCD.methods We used data from two household surveys undertaken in 2009-2010 among 7389 rural poor households (39 205 individuals) in Odisha and Bihar. All persons from the sampled households, irrespective of age and gender, were included in the analysis. We classify self-reported illnesses as NCD, CD or 'other morbidities' following the WHO classification.results Non-communicable diseases accounted for around 20% of the diseases in the month preceding the survey in Odisha and 30% in Bihar. The most prevalent NCD, representing the highest share in outpatient costs, were musculoskeletal, digestive and cardiovascular diseases. Cardiovascular and digestive problems also generated the highest inpatient costs. Women, older persons and less-poor households reported higher prevalence of NCD. Outpatient costs (consultations, medicines, laboratory tests and imaging) represented a bigger share of income for NCD than for CD. Patients with NCD were more likely to report a hospitalisation.conclusion Patients with NCD in rural poor settings in India pay considerably more than patients with CD. For NCD cases that are chronic, with recurring costs, this would be aggravated. The cost of NCD care consumes a big part of the per person share of household income, obliging patients with NCD to rely on informal intra-family cross-subsidisation. An alternative solution to finance NCD care for rural poor patients is needed.

BackgroundMost healthcare spending in developing countries is private out-of-pocket. One explanation for low penetration of health insurance is that poorer individuals doubt their ability to enforce insurance contracts. Community-based health insurance schemes (CBHI) are a solution, but launching CBHI requires obtaining accurate local data on morbidity, healthcare utilization and other details to inform package design and pricing. We developed the “Illness Mapping” method (IM) for data collection (faster and cheaper than household surveys).MethodsIM is a modification of two non-interactive consensus group methods (Delphi and Nominal Group Technique) to operate as interactive methods. We elicited estimates from “Experts” in the target community on morbidity and healthcare utilization. Interaction between facilitator and experts became essential to bridge literacy constraints and to reach consensus.The study was conducted in Gaya District, Bihar (India) during April-June 2010. The intervention included the IM and a household survey (HHS). IM included 18 women’s and 17 men’s groups. The HHS was conducted in 50 villages with1,000 randomly selected households (6,656 individuals).ResultsWe found good agreement between the two methods on overall prevalence of illness (IM: 25.9% ±3.6; HHS: 31.4%) and on prevalence of acute (IM: 76.9%; HHS: 69.2%) and chronic illnesses (IM: 20.1%; HHS: 16.6%). We also found good agreement on incidence of deliveries (IM: 3.9% ±0.4; HHS: 3.9%), and on hospital deliveries (IM: 61.0%. ± 5.4; HHS: 51.4%). For hospitalizations, we obtained a lower estimate from the IM (1.1%) than from the HHS (2.6%). The IM required less time and less person-power than a household survey, which translate into reduced costs.ConclusionsWe have shown that our Illness Mapping method can be carried out at lower financial and human cost for sourcing essential local data, at acceptably accurate levels. In view of the good fit of results obtained, we assume that the method could work elsewhere as well.

Background: Community-based health insurance (CBHI) schemes are increasingly implemented in low-income settings. These schemes limit the coverage they offer both by the types of care considered, and by applying thresholds and/or caps to costs reimbursed. The consequences of these thresholds and/or caps on insurance coverage have hitherto been usually ignored, for lack of data on the distributions of healthcare costs or understanding of their impact on effective coverage levels. This article describes a theoretical model to obtain the distributions even without data collection in the field, and demonstrates the quantitative impact of thresholds and/or caps on claim reimbursements. Methods: This model applies to applications on healthcare expenditures in lowincome settings, following research methods examined in the Western world. We looked at hospitalizations and tests; we compared the simulated distributions to empirical data obtained through 11 household surveys conducted between 2008 and 2010 in rural locations (9 in India and 2 in Nepal). Results: We found that the shape of the distributions was very similar in all locations for both benefits, and could be represented by a model based on a lognormal distribution. The agreement between theoretical and empirical results was satisfactory (mostly within 10% difference). Conclusions: The model makes it possible to simulate the expected performance of the CBHI (represented by the percentage of costs or bills covered). The aim is to match costs with local levels of willingness-to-pay for health insurance. This model makes it possible to determine at the stage of package-design the optimal levels of thresholds and/or caps for each benefit-type included.

No abstract

Background: This study examines health-related "hardship financing" in order to get better insights on how poor households finance their out-of-pocket healthcare costs. We define hardship financing as having to borrow money with interest or to sell assets to pay out-of-pocket healthcare costs. Methods: Using survey data of 5,383 low-income households in Orissa, one of the poorest states of India, we investigate factors influencing the risk of hardship financing with the use of a logistic regression. Results: Overall, about 25% of the households (that had any healthcare cost) reported hardship financing during the year preceding the survey. Among households that experienced a hospitalization, this percentage was nearly 40%, but even among households with outpatient or maternity-related care around 25% experienced hardship financing. Hardship financing is explained not merely by the wealth of the household (measured by assets) or how much is spent out-of-pocket on healthcare costs, but also by when the payment occurs, its frequency and its duration (e.g. more severe in cases of chronic illnesses). The location where a household resides remains a major predictor of the likelihood to have hardship financing despite all other household features included in the model. Conclusions: Rural poor households are subjected to considerable and protracted financial hardship due to the indirect and longer-term deleterious effects of how they cope with out-of-pocket healthcare costs. The social network that households can access influences exposure to hardship financing. Our findings point to the need to develop a policy solution that would limit that exposure both in quantum and in time. We therefore conclude that policy interventions aiming to ensure health-related financial protection would have to demonstrate that they have reduced the frequency and the volume of hardship financing.