Data collection framework for electronic medical record-based real-world data to evaluate the effectiveness and safety of cancer drugs: a nationwide real-world study of the Korean Cancer Study Group (KCSG).

Abstract

Background: Electronic medical record (EMRs) have the highest reliability among real-world data (RWD), but controlling for biases that may affect study outcomes remains challenging. This study aimed to establish a data collection framework of EMR-based RWD to evaluate the effectiveness and safety of cancer drug by conducting a nationwide real-world study based on Korean Cancer Study Group.

Methods: We selected ramucirumab plus paclitaxel (RAM/PTX) and trastuzumab-emtansine (T-DM1), which are currently used to treat advanced gastric cancer and breast cancer, respectively, under the national health insurance, and systematically collected EMR-based RWD at relevant institutions in South Korea. Investigator reliability was evaluated using the concordance rate between the recommended input value for representative fictional cases and the input value of each investigator. Reliability of collected data was evaluated twice during study period at three institutions randomly selected by statisticians using the concordance rate between the previously collected data and data collected by the independent investigator. Two statisticians independently analyzed the same data and compared their results to evaluate the data analysis reproducibility.

Results: Between the starting date of medical insurance coverage and December 2018, a total of 1,063 patients at 56 institutions in the RAM/PTX cohort and 824 patients at 60 institutions in the T-DM1 cohort were included. Mean investigator reliability for major case report form (CRF) variables in the RAM/PTX and T-DM1 cohorts was 73.5% and 71.9%, respectively. The most common CRF variable with a concordance rate < 70% between the recommended input value and the input value of each investigator was related to adverse events in both cohorts. Mean reliability of collected data for major CRF variables in the RAM/PTX and T-DM1 cohort was 90.0% for both cohorts in the first analysis, and 88.0% and 89.0% in the second analysis, respectively. The input data discrepancies between previously collected data and data collected by independent investigators were due to the input for variables that were not categorized and were in textual form in the EMRs (performance status, reason for discontinuation, adverse events, best response or disease progression, and survival data).

Conclusions: This real-world study provides a framework that ensures relevance and reliability of EMR-based RWD for evaluating the effectiveness and safety of cancer drugs. There is a need for a digital healthcare system in which EMR-based RWD can be structured, defined, formatted, and exchanged with an integrated computer system and converted into scientific data.

Clinical trial information: NCT04192734 and NCT04202328