Differences in health care seeking patners for childern less than five years of age who died of severe diarrhea compared to those with non-fatal severe diarrhea in selscted semi Urban areas of Karachi, Pakistan

B ACKGROUND Improving survival from childhood diarrhea still remains a challenge for most developing countries. Despite reductions in overall number of deaths due to diarrhea, there is insufficient progress in countries like Pakistan and children less than five years of age continue to die of diarrhea each year in many low income countries. Prevention of diarrhea in these settings requires both effective prevention and management strategies and methods to change behavior to improve access to care. It is therefore imperative to study health care seeking behavior patterns and their impact for addressing better future strategies. Hence we proposed to see the differences in health care seeking patterns for children less than five year of age who died of severe diarrhea compared to those with non-fatal severe diarrhea METHODS This was a mixed method study including a matched case-control study done at selected semi-urban communities of Karachi and focus group discussions using qualitative methods. Cases were identified from the ongoing surveillance system which captured diarrheal deaths and included children 0-59 months old who had died after developing severe diarrhea. Their age and neighborhood matched controls included children 0-59 months with severe diarrhea who survived after a severe diarrhea episode. The caretakers were interviewed through a questionnaire to inquire about patterns of health care seeking behavior. Analysis was done using statistical analysis system (SAS) version 9.1 and also on statistical package for social scientist (SPSS) version 11.5. Conditional logistic regression was performed to identify factors associated with severe diarrhea mortality in children less than five year of age. RESULTS Seventy four cases of fatal diarrhea and 148 matched controls with non-fatal diarrhea were included in the study. The results of the multivariable model showed that the odds of having appropriate care were 80% less in children with severe fatal diarrhea than in children with severe non-fatal diarrhea after matching for age and neighborhood. (MORadi = 0.20, 95% CI: 0.05, 0.90). Time (one hour or more) taken to reach health facility (MORadi = 5.0 95% CI 1.3, 19.1), and refusal to comply with advice for hospital admission (MORadi = 18.2 95% CI 4.0, 82.41) were also found to be associated with death in children with severe diarrhea. Other factors associated with fatal diarrhea in children were; having blood in stools (MORadi = 10.7 95% CI 2.2, 52.6), no boiling of drinking water (MORadi = 7.6 95% CI 2.5, 23.5) and use of public place as source of drinking water (MORadj = 12.6 95% CI 1.8, 86.6). Another factor of home care feeding which was independently associated with diarrheal mortality in children showed that children with fatal severe diarrhea were more likely to be given nothing to drink (crude MOR= 5.4 95%CI 1.4, 21.7) and nothing to eat (crude MOR= 2.9 95%CI 0.9, 9.3) compared to being given usual or more than usual to drink and eat, however this became insignificant in the final multivariable model. Additional findings from the qualitative study showed switching and simultaneous treatment from traditional, spiritual and modern healers by the caretakers of children 0-59 months with severe diarrhea were common rather than following the advice of one provider. Moreover limited decision making power of mothers, lack of belief in oneself for taking care of the child, inadequate knowledge of prevention of diarrhea illness, lack of awareness about water boiling, self treating child at home resulting in delay in seeking formal care and misconceptions regarding ORS were factors associated more commonly with caretakers of children with severe fatal diarrhea than nonfatal diarrhea. CONCLUSION The results of the study show that lack of appropriate care seeking, refusal to admit the child in hospital and delays in time to reach the health facility are significant determinants of diarrheal deaths among children with severe diarrhea. Other factors associated with fatal diarrhea in children were not boiling drinking water, use of public sources of drinking water and presence of blood in stools of the child. Caretakers from these low-income areas showed switching and simultaneous treatment from traditional, spiritual and modern healers, lack of belief in oneself and blamed their inadequate knowledge for their child death. Social marketing through health education promotion campaigns at the community level and mass media campaigns for oral rehydration therapy and for importance of seeking appropriate care should be implemented in order to achieve a substantial decline in diarrheal disease mortality in Pakistan