Treatment Outcomes and Associated Clinical Laboratory Parameters among Patients Hospitalised with COVID-19 at a Private Hospital in Kiambu County, Kenya DOI: https://dx.doi.org/10.4314/ajhs.v38i2.4

Abstract

Background: COVID-19 was a novel infection whose outcomes were not well established to be associated with any factors during the start of the pandemic that it caused globally. This research explored the relationship between clinical laboratory tests and mortality among patients hospitalized with a positive COVID-19 test at a tertiary hospital in Kiambu County, Kenya.

Methods: A retrospective analytical cross-sectional study was conducted among 154 COVID-19–positive patients admitted to a private tertiary hospital in Kiambu County, Kenya, to identify laboratory tests associated with mortality. Data was analysed using SPSS version 24, with descriptive statistics summarizing clinical and outcome variables. Binary logistic regression assessed associations between laboratory parameters and mortality, while significant factors were included in multivariate models to control confounders. Kaplan–Meier analysis evaluated time to death and discharge, with statistical significance set at p < 0.05.

Results: The mean age of patients admitted with a diagnosis of COVID-19 was 54.43(SD±17.1) years (54.43 years). Almost two-thirds of patients had comorbidities (n=93; 60.4%). In-hospital death was reported for 23.4% (n=36) of the investigated patients, and 76.6% (n=118) were discharged alive. Logistic regression revealed that the clinical laboratory parameter that predicted mortality was elevated C-reactive protein (adjusted odds ratio 1.067; 95% CI: 1.014–1.123; p=0.013).

Conclusion: Elevated C-reactive protein levels were identified as a significant predictor of mortality among hospitalised COVID-19 patients, highlighting the prognostic value of inflammatory markers in disease outcomes. Despite this, the majority of patients recovered and were discharged alive, suggesting generally favourable clinical outcomes in this setting.