Background Studies about health care seeking behaviour and healthcare utilisation give a good insight about factors that may have programmatic and policy implications. In a developing country like India people spend a substantial proportion of their incomes on medical treatment. WHO estimates reveal that every year 25 million households are forced into poverty by illness and the struggle to pay for health care. We explore health care seeking behaviour and related out-of-pocket (OOP) expenditure among urban slum households in Davangere, Karnataka.

Methods A cross-sectional study was conducted between March 1st and August 31st, 2015, in urban slums of Davangere. Among 38 slums in Davangere, 22 residential slums formed our sampling frame where a list of houses with under-fives was collected from the Anganwadi registers. In each slum 10 houses were selected through systematic random sampling; thus a total of 220 houses were studied. Data were collected on health care seeking behaviour and expenditure on under-five illnesses. To avoid recall bias, a one-month recall period for any under-five child morbidity was considered.

Findings Among 160 under-fives studied, 30% reported illnesses in the past one month. The mean age of the mothers of the under-fives was 26±1.3 years. Majority (81 %) of the mothers belonged to class IV socio economic class (as per modified BG Prasad SES classification); 15% of mothers had more than one under-five child. The most frequent illness reported was diarrhoea (38%), followed by fever (31%) and common cold (29%). Most of the households consulted a healthcare provider (73%), either an allopathic practitioner (51%) or a complementary-medicine practitioner (22%). Delay in seeking health care was on average 1.8±0.4 days. Reasons for delay in health seeking were high cost and distance from the households. Mean expenses incurred on treatment for the under-five illnesses was INR 550±125. Most households (89%) bear the expenses out-of-pocket, either by borrowing money from a neighbour or a relative (59%), or by selling household belongings (32%). More than half of the households spent 17–40% of their monthly income in case of an under-five illness. Most expenses were on drugs (71%), followed by transport. We found a strong statistical association between the amount of money spent and the both the delay in care seeking and the type of healthcare provider approached. No significant statistical association was found between the expenses incurred and household income, or gender of the under-five. Only a minority of the mothers (23%) were part of self-help groups in the community; some of the households (28%) had savings in a small-scale daily deposit (Pigmy savings) scheme.

Discussion In our study population, but OOP spending and prevalence of catastrophic health expenditure should be considered as high. Irrespective of their income, households were spending a substantial amount on health care for their ill under-fives: 17–40% of their monthly income is considerably higher than what was observed in a similar study in a Bijapur (also Karnataka) urban slum 6 years earlier (15–22%). Catastrophic health expenditure (defined as spending 5–20% of total household income on health care) occurred in 15–20% of our study population, lower compared to the Bijapur study.

Conclusions In the absence of institutionalised financial protection for vulnerable urban slum populations, safety nets in the form of self-help groups or community-based health insurance schemes could be first steps to delink impoverishment from health care seeking behaviour. In the long run, accessible and affordable public primary care services, embedded in the urban slums and of good quality, are needed to reduce the inequitable burden that slum populations suffer today.

No competing interest.