Western and coastal Kenya sick visit data 2019-2022

Antimicrobial resistant pathogens are a leading cause of morbidity and mortality worldwide, with overuse and misuse of antimicrobials being key contributors. We aimed to identify factors associated with antibiotic prescriptions among patients presenting to clinics in Kenya. We performed a retrospective, descriptive cohort study of persons presenting to outpatient clinics in Western and Coastal Kenya, including symptoms, physical exams, clinician assessments, laboratory results and prescriptions. We reviewed 1,526 visits among 1,059 people who sought care from December 2019-February 2022. Median age was 16 (IQR 6-35) and 22% were under 5. 30% of malaria RDTs were positive and 3% of dengue RT-qPCRs were positive. Antibiotics were prescribed in 73% of encounters overall and in 84% among children under 5. In 48% of visits antibiotics were prescribed without a provisional bacterial diagnosis. In the multivariable model, factors associated with increased odds of an antibiotic prescription were the clinic in Western Kenya (OR 5.1, 95% CI 3.0-8.8), age less than or equal to 18 (OR 2.1, 95% CI 1.4-3.2), endorsement of cardiorespiratory symptoms (OR 5.2, 95% CI 3.2-8.3), a negative malaria RDT (OR 4.0, 95% CI 2.5-6.8), and a provisional diagnosis that could be bacterial in etiology (OR 5.9, 95% CI 3.5-10.3). High rates of antibiotic prescriptions are common even when associated diagnoses are not bacterial. Compared to our 2014-2017 cohort, we found higher rates of antibiotic prescriptions among children. Improved diagnostics to rule in alternative diagnoses as well as stewardship programs are needed.

Methods
This data is a subset of a larger cohort study to determine arbovirus seroprevalence, seroconversion and factors which influence transmission in Western and Coastal Kenya (AI102918-08, PI: ADL). Written informed consent was obtained by all study participants, with parents/guardian consenting for children. This consent included use of clinical and laboratory data. Institutional Review Board approval for human subjects research was obtained from Stanford (#49683) and Technical University of Mombasa, Kenya (TUM ERC EXT/004/2019).

Participants received instructions to attend specific clinics if they experienced a fever during the study period, and these “sick visits” were performed by several medical officers based at them full time during the study period. There were two clinics included in the study, one in an urban area of Western Kenya and the other in an urban area in Coastal Kenya. The data file contains information that was collected during the sick visits. We analyzed this data to identify factors associated with antibiotic prescriptions among patients presenting to clinics in Kenya.

The data is in an excel file. The data was cleaned in this excel file prior to uploading to R studio for analysis. Some of the variables in this excel file come directly from the initial form that clinicians filled out during the sick visit, and other variables were created for the purposes of analysis.