"Has the patient been coughing?" Do Ugandan providers ask patients tested for malaria about cough history? Secondary analysis of 2011 data from the Integrated Infectious Disease Capacity Building Evaluation (IDCAP)

Abstract

Introduction: In 2013, malaria caused an estimated 584,000 deaths worldwide, with 90% of those deaths occurring in Africa. The Integrated Management of Childhood Illness (IMCI) strategies have been effective at improving the health outcomes of malaria patients (Schellenberg et al., 2004). Recent case management algorithms guide health care providers to ask about cough history and perform laboratory testing for evidence of infection with malaria by either microscopy or a malaria rapid diagnostic test (RDT) in all patients suspected of having malaria. Many health workers do not adhere to these guidelines uniformly. Little is known about the influence of performing a parasitological diagnosis on the clinical practice of taking a history of cough in low- and middle-income countries. Method: We performed a secondary analysis using data from a cluster randomized trial to conduct a cross-sectional study using 36 Ugandan level IV health facilities to test the hypothesis that practitioners would be more likely to record cough history in patients with a negative malaria test result compared to patients with a positive malaria test result. We used a generalized linear model to perform a multivariate regression clustered by health facility and adjusted for patient age, gender, emergency triage status, facility, intervention arm, and month of visit. Result: A total of 138,285 patients were tested for malaria. Cough history was performed for 80,266 (58%) patients. In the univariate analysis, cough history was assessed in a similar percentage of patients with a negative malaria test 53,017/92,683 (57%) versus a positive malaria test 27,249/45,602 (60%), and this difference was not statistically significant (relative risk [RR] 0.96, 95% CI 0.88-1.04, P=0.313). After adjusting for confounding factors, the percentage of patients with a documented cough history was the same across malaria test results (adjusted RR 0.99, 95% CI 0.91-1.08, P=0.833). Conclusion: This analysis revealed no difference in the relative risk that a provider would perform a cough history in patients with a negative malaria test result versus positive malaria test result. The frequency of patients asked about history of cough by a health provider was a little over half, despite the IMCI and other WHO guidelines suggestion that cough history precede other examinations and should be asked in all patients. This information has important implications for infectious disease training programs as well as clinical guidelines.