CODING MEDICAL CONCEPTS - A CONTROLLED EXPERIMENT WITH A COMPUTERIZEDCODING TOOL

In clinical routine there is a growing need to encode medical concepts with available standard coding systems. The coding process can be tim e consuming and may significantly add to daily paperwork, particularly regarding patients with multiple diagnoses and in busy clinical envir onments with a high turnover of patients. We have developed a generic computerised encoding tool - the PADS encoder - to ensure rapid, corre ct and complete coding of diagnoses in daily routine. The tool is inte grated into an electronic patient record system (PADS, Patient Archivi ng & Documentation System) and takes full advantage of the user friend ly Macintosh interface. The tool was tested in a controlled experiment by 18 clinicians who encoded a total of 666 medical concepts in each protocol (study protocol vs. control). The following positive findings were significantly associated with the use of the computerised coding tool: the number of correctly encoded medical concepts was higher (99 .55% vs. 86.1%)coding errors were lower (0% vs. 10.81%) more modifier codes were encoded correctly (increase by up to 43%) less coding error s were made (decrease by up to 43%) the overall rate of correctly enco ded and complete main and modifier codes was increased by 31.27% (97.2 9% vs. 66.02%) - coding time was reduced by 50% This paper presents da ta to suggest that a computerised coding tool can produce more complet e data of higher quality and can save time compared with the tradition al approach to encode medical concepts.