Authored by: Momanyi, Lydia B
Antibiotics are commonly prescribed drugs globally but most of their use is irrational. This causes wastage of scarce health care resources, poses increased risk of adverse drug reactions, emergence of resistance and poor treatment outcomes. There is scarce published local data on patterns of antibiotic prescribing in Kenya and hence the impetus for the present study.
To describe the antibiotic prescribing patterns among the patients admitted to Rift Valley Provincial General Hospital in Kenya.
This was a cross-sectional study. A total of 179 participants were recruited from all inpatient departments through stratified proportionate random sampling. All admitted patients on oral or parenteral antibiotic therapy were eligible. Data was abstracted from the patients‟ medical records into predesigned structured forms. This data included socio-demographic characteristics of patients, the name and regimen of antibiotic, diagnosis, ward or department type and adherence to the prescribing guidelines. The raw data was coded, entered into Microsoft excel version 2010 to create a database and then exported to STATA version 13 for analysis. Prevalence of antibiotic use was determined using proportions. Data was presented in frequency tables and figures for important variables. Associations between predictor variables such as socio-demographic factors and outcome variables such as the type of antibiotics, prevalence of use, guideline compliance and rational prescribing were determined using Chi square. Stepwise backward binary logistic regression was done to determine the independent predictors of rational antibiotic prescribing and guideline compliance. Statistical significance was set at 95% confidence level and values with p≤0.05 were considered statistically significant.
The prevalence of antibiotic prescribing at Rift Valley Provincial General Hospital was 54.7%. Among the department/ward type, the highest prevalence of antibiotic prescribing was found in critical care unit and isolation ward, both at 100%. Obstetrics and gynecology department had the least prevalence at 20.8%. Penicillins (46.9%) followed by cephalosporins (44.7%) were the most prescribed antibiotic classes. Benzylpenicillin (52.0%) was the most prescribed penicillin while ceftriaxone (91.0%) was the most prescribed cephalosporin. There was rational prescribing in 33.9% (n=121) of all the 357 antibiotic encounters. The use of an antibiotic for prophylaxis or treatment of a neonatal infection was 5 times more likely to be rational as compared to other antibiotic uses (COR=5.84, 95% CI=2.05-16.64, p=0.001). Compliance to prescribing according to the guidelines was observed in 45.8% (n=82) of the study population. There was a negative association between age and compliance to guideline such that the prevalence of guideline compliance declined as participants‟ age increased (COR=0.67, 95% CI=0.54-0.84, p<0.001). Antibiotic prescribing among the neonates was 4 times more likely to be in accordance to the established guidelines as compared to adults (COR=0.23, 95% CI= 0.08 – 0.69, p=0.009).
There was a high prevalence of irrational antibiotic prescribing and low compliance to local and international guidelines at Rift Valley Provincial General Hospital. Recommendations: Prescribers should be encouraged to improve on their antibiotic prescribing habits by following the treatment guidelines. Further research is, however, necessary to find out the patients, prescribers and hospital contextual factors that may be impacting on poor antibiotic prescribing habits as well non-adherence to treatment protocols.