To illustrate a R language-based method for the extraction of data from thoracic surgery-based electronic medical records.

The information on lymph node metastasis among patients, who consecutively received surgery for primary lung or bronchial malignant disease between Aug 2019 and Aug 2022, was extracted from the pathological report. Initially, the pathological text was loaded on the EXCEL sheet. The total sample was evenly divided into four groups, which were managed by four independent researchers. The primary data were initially treated manually, then by regular expression under R language (version 4.2.2) environment, between which the time-consumed and the accuracy rate was compared, respectively. Additionally, the overall sample was managed by another independent researcher and the time consumed as well as accuracy was recorded.

During the study period, a total of 9 154 patients received surgery in our hospital, of which 8 579 (93.7%) patients underwent systematic lymph node dissection or sampling. Among the four independent researchers, one each treated data of approximately 2 288 patients. The time consumed by hand was 100, 120, 115 and 130 hours, respectively (average: 116.25 hours) with the corresponding accuracy rate of 96%, 94%, 97% and 95%, respectively, (average: 95.5%). Upon switching to the R language-based method, the time consumed was reflected by 60, 30, 50 and 75 min, respectively, (average: 55 min) with the accuracy rate calculated as 99%, 100%, 100% and 98%, respectively (average: 99.25%). There was a statistical significance observed not only in time consumed but also in accuracy rate with the R language-based method favored by both P values of less than 0.001. When all sample were processed by another researcher under R language environment, the total time consumed was 45 min with an accuracy rate of 99%.

R language-based method is an effective method for handling text materials derived from electronic medical records, by which the efficiency and quality for constructing database of specific thoracic disease are improved.