Professionals from various cadres in the health sector raise concerns regarding the poor quality of clinical coding leading to lack of evidence-based practice. Assessing the quality of the clinical coding in one of Nairobi City County's major hospital would be a step towards establishing the exact gaps in quality of the coding process and outcome. Training the professionals would also foster better clinical coding practice in one of the major facilities nationally. Method: The study aimed at

more »

... ablishing the quality of clinical coding within Mbagathi County Referral Hospital, and thereafter determined the effect of training on the established clinical coding quality. An interventional trial study design was used, with a quality of clinical coding checklist used classify codes assignment or lack of which. The sample included 320 patient files selected randomly from a month-long list of patients. Results: The study found out that the overall baseline code quality was slightly above average given that majority (55%) of the code assignment were good as established by a composite score of the various coding quality attributes assessed. Given the need for training based on the low quality, a training intervention was then conducted based on the needs identified. An indexing database was also installed for the coders to use in encoding the codes assigned. Code quality improved to 77% after the training. Code completion was excellent at the facility, as established from the 97% of the files that were completely coded at baseline and later improved to 99%. Notably, also, is that the hospital improved its coding of procedures and death certification by 32 and 53% respectively. The hospital also started using the indexing tool that was introduced as an intervention. Conclusions: The health facility could act as a good benchmark for code completion. However, code completion without accuracy in the code assignment invalidates the overall quality of coding. Code accuracy improved with the training almost immediately after the interventions. More practice would for sure lead to better clinical coding accuracy.