An ethnographic exploration of diarrheal disease management in public hospitals in Bangladesh: From problems to solutions.

INTRODUCTION: Diarrheal disease is one of the most common causes of hospital admission globally. The barriers that influence guideline-adherent care at resource limited hospitals are poorly defined, especially during diarrheal disease outbreaks. The objective of this study was to characterize challenges faced in diarrheal disease management in resource-limited hospitals and identify opportunities to improve care. METHODS: The study was conducted during a diarrheal disease outbreak period at ten public district hospitals distributed across Bangladesh. A rapid ethnographic approach included observations and informal interviews with clinicians, staff nurses and patients. In the first phase, observations identified common and unique challenges in diarrheal management at the ten sites. In the second phase, four hospitals were purposively selected for additional ethnographic study. Systematic observations over 420 total hours were collected from patient-clinician interactions (n = 76) and informal interviews (n = 138). Applied thematic analysis identified factors that influenced hospitalbased diarrhea management. RESULTS: Normalization of guideline deviation was observed at all ten sites, including prescription of non-indicated antibiotics and intravenous (IV) fluids. Conflict between 'what should be done' and 'what can be done' was the most common challenge identified. Clinical assessments and patient treatment plans were established at admission in a median of 2 minutes (n = 76), often without a physical examination (57%; n=43/76). Factors that prevented adherence to clinical guidelines included human resource constraints, conflicts of interests, overcrowding, and inadequate hygiene and sanitation in the emergency department and wards. CONCLUSION: This study identified challenges in hospital-based management of diarrheal disease and opportunities to improve care in seemingly change-resilient hospital settings. The results reveal important areas for intervention and policy engagement that may have additive benefit for both hospitals and their patients. These interventions include targeting barriers to clean-water, sanitation and hygiene that prevent clinicians from adopting guidelines out of concern for hospital acquired infections.

Field evaluation of a locally produced rapid diagnostic test for early detection of cholera in Bangladesh.

BACKGROUND: Cholera remains a substantial health burden in Asia and Africa particularly in resource poor settings. The standard procedures to identify the etiological organism V. cholerae are isolation from microbiological culture from stool as well as Polymerase Chain Reaction (PCR). Both the processes are highly lab oriented, labor extensive, time consuming, and expensive. In an effort to control for outbreaks and epidemics; an effective, convenient, quick and relatively less expensive detection method is imperative, without compromising the sensitivity and specificity that exists at present. The objective of this component of the study was to evaluate the effectiveness of a locally produced rapid diagnostic test (RDT) for cholera diagnosis. METHODS: In Bangladesh, nationwide cholera surveillance is ongoing in 22 hospitals covering all 8 divisions of the country since June, 2016. In the surveillance, stool samples have been collected from patients presenting to hospitals with acute watery diarrhea. Crystal VCTM (Span diagnostics, India) and Cholkit (locally produced RDT) have been used to detect V. cholerae from stool samples. Samples have also been sent to the main laboratory at icddr,b where the culture based isolation is routinely performed. All the tests were carried out for both direct and enriched stool samples. RDT sensitivity and specificity were calculated using stool culture as the gold standard. RESULTS: A total of 7720 samples were tested. Among these, 5865 samples were solely tested with Crystal VC and 1355 samples with Cholkit whereas 381 samples were tested with both the RDTs. In comparison with culture, direct testing with Crystal VC showed a sensitivity of 72% (95% CI: 50.6% to 87.9%) and specificity of 86.8% (95% CI: 82.8% to 90.1%). After enrichment the sensitivity and specificity was 68% (95% CI: 46.5% to 85.1%) and 97.5% (95% CI: 95.3% to 98.8%) respectively. The direct Cholkit test showed sensitivity of 76% (95% CI: 54.9% to 90.6%) and specificity of 90.2% (95% CI: 86.6% to 93.1%). CONCLUSION: This evaluation has demonstrated that the sensitivity and specificity of Cholkit is similar to the commercially available test, Crystal VC when used in field settings for detecting V. cholerae from stool specimens. The findings from this study suggest that the Cholkit could be a possible alternative for cholera endemic regions where V. cholerae O1 is the major causative organism causing cholera.