Cost of pediatric hospitalizations in Burkina Faso: A cross-sectional study of children aged <5 years enrolled through an acute gastroenteritis surveillance program

INTRODUCTION

Diarrheal illness is a leading cause of hospitalizations among children <5 years. We estimated the costs of inpatient care for rotavirus and all-cause acute gastroenteritis (AGE) in two Burkina Faso hospitals.

METHODS

We conducted a cross-sectional study among children <5 years from December 2017 to June 2018 in one urban and one rural pediatric hospital. Costs were ascertained through caregiver interview and chart abstraction. Direct medical, non-medical, and indirect costs per child incurred are reported. Costs were stratified by rotavirus results.

RESULTS

211 children <5 years were included. AGE hospitalizations cost 161USD (IQR 117-239); 180USD (IQR 121-242) at the urban and 154USD (IQR 116-235) at the rural site. Direct medical costs were higher in the urban compared to the rural site (140USD (IQR 102-182) vs. 90USD (IQR 71-108), respectively). Direct non-medical costs were higher at the rural versus urban site (15USD (IQR 10, 15) vs. 11USD (IQR 5-20), respectively). Indirect costs were higher at the rural versus urban site (35USD (IQR 8-91) vs. 0USD (IQR 0-26), respectively). Rotavirus hospitalizations incurred less direct medical costs as compared to non-rotavirus hospitalizations at the rural site (79USD (IQR 64-103) vs. 95USD (IQR 80-118)). No other differences by rotavirus testing status were observed. The total median cost of a hospitalization incurred by households was 24USD (IQR 12-49) compared to 75USD for government (IQR 59-97). Direct medical costs for households were higher in the urban site (median 49USD (IQR 31-81) versus rural (median 14USD (IQR 8-25)). Households in the lowest wealth quintiles at the urban site expended 149% of their monthly income on the child's hospitalization, compared to 96% at the rural site.

CONCLUSIONS

AGE hospitalization costs differed between the urban and rural hospitals and were most burdensome to the lowest income households. Rotavirus positivity was not associated with greater household costs.

Haiti Poliovirus Environmental Surveillance

Poliovirus (PV) environmental surveillance was established in Haiti in three sites each in Port-au-Prince and Gonaïves, where sewage and fecal-influenced environmental open water channel samples were collected monthly from March 2016 to February 2017. The primary objective was to monitor for the emergence of vaccine-derived polioviruses (VDPVs) and the importation and transmission of wild polioviruses (WPVs). A secondary objective was to compare two environmental sample processing methods, the gold standard two-phase separation method and a filter method (bag-mediated filtration system [BMFS]). In addition, non-polio enteroviruses (NPEVs) were characterized by next-generation sequencing using Illumina MiSeq to provide insight on surrogates for PVs. No WPVs or VDPVs were detected at any site with either concentration method. Sabin (vaccine) strain PV type 2 and Sabin strain PV type 1 were found in Port-au-Prince, in March and April samples, respectively. Non-polio enteroviruses were isolated in 75-100% and 0-58% of samples, by either processing method during the reporting period in Port-au-Prince and Gonaïves, respectively. Further analysis of 24 paired Port-au-Prince samples confirmed the detection of a human NPEV and echovirus types E-3, E-6, E-7, E-11, E-19, E-20, and E-29. The comparison of the BMFS filtration method to the two-phase separation method found no significant difference in sensitivity between the two methods (mid-P-value = 0.55). The experience of one calendar year of sampling has informed the appropriateness of the initially chosen sampling sites, importance of an adequate PV surrogate, and robustness of two processing methods.