An accuracy comparison between two methods of double data entry in Chinese medicine research

To compare two methods for data verification, provide reference for appropriate data entry´s verification. Methods: This study is about the traditional Chinese medicine (TCM) clinical research for influenza A (H1N1) therapy of the special project of TCM Industry. We used two methods for double data entry, and randomly selected 50 cases from two databases using two different data entry methods, respectively. We verified each variable with the original case report forms, recorded every error, and analyzed the reason of mistakes. Results: The error rates of the Interactive Double Data Entry verification was 15.17 per 10,000 data items (fields). The error rate of the consistent verification was 1.18 per 10,000 data items (fields). The interactive verification´s error rate was significantly higher than the consistent verification (P <; 0.05). Conclusions: The error rates of the consistency check were lower than the interactive verification under the same condition, interactive verification was more suitable for data in large clinical research, limited time and case analyzing required right after verification. Experienced data entry operators who are familiar with the contents of data, and have rigorous attitude are the keys to guarantee high quality data no matter what methods are chosen.