Randomised crossover trial comparing the performance of Clinical Terms Version 3 and Read Codes 5 byte set coding...

Prior research has shown that coding systems influence the accuracy of electronic health records. It was already known that Read Codes 5 byte set is widely used in primary care. Yet, uncertainty remained about whether newer systems like Clinical Terms Version 3 improve coding precision. No prior work had resolved how these systems compare in real-world use. This gap motivated a study to evaluate their performance in general practice. That uncertainty drove the need for a controlled trial comparing both systems. No prior work had measured both accuracy and coding time together. This study aimed to address that gap directly.

The aim was to compare Clinical Terms Version 3 and Read Codes 5 byte set in coding electronic patient records. The specific problem was to assess which system provides greater accuracy and consistency. The motivation came from the need to improve primary care documentation. No prior work had tested these systems in a crossover trial. This study sought to measure exact matches and coding time. The goal was to determine if Clinical Terms offers better performance. No prior work had evaluated clinician preferences in this context. This study aimed to provide evidence for system selection.

A randomised crossover trial was conducted across 10 general practices in Norfolk. Participants were 10 general practitioners working in urban, suburban, and rural settings. Each clinician coded patient records using both systems after randomisation. The study collected concepts from 100 patient encounters. Exact matches were measured by comparing coded choices to original terms. Identical coding choices between paired clinicians were also recorded. Time per term was tracked to assess efficiency. The trial design allowed direct comparison of both systems under real conditions.

Clinical Terms Version 3 had 70% exact matches compared to 50% for Read Codes (P < 0.001). This difference was significant for each of the 10 participants individually. The pooled proportion of exact and identical matches was 0.58 for Clinical Terms versus 0.36 for Read Codes (P < 0.001). A total of 995 unique concepts were collected from the records. Coding time with Clinical Terms was 30 seconds per term, not significantly longer than Read Codes. Exact matches were consistently higher in Clinical Terms across all settings. The confidence intervals showed strong separation between the two systems. These results suggest Clinical Terms outperforms Read Codes in capturing concept meaning.

The authors propose that Clinical Terms Version 3 performs significantly better than Read Codes 5 byte set. They suggest that improved coding accuracy is achievable with Clinical Terms in primary care. The study found no significant difference in coding time between the two systems. The authors claim that Clinical Terms captures concept meaning more effectively. They propose that this system could enhance electronic patient record accuracy. No prior work had demonstrated such consistent performance differences. The authors suggest that this finding supports adoption of Clinical Terms in practice. These conclusions are based on the observed exact match and identical match rates.