Automated data extraction: merging clinical care with real-time cohort-specific research and quality improvement data

Although prohibitively labor intensive, manual data extraction (MDE) is the prevailing method used to obtain clinical research and quality improvement (QI) data. Automated data extraction (ADE) offers a powerful alternative. The purposes of this study were to 1) assess the feasibility of ADE from provider-authored outpatient documentation, and 2) evaluate the effectiveness of ADE compared to MDE.

A prospective collection of data was performed on 90 ADE-templated notes (N=71 patients) evaluated in our bowel management clinic. ADE captured data were compared to 59 MDE notes (N=51) collected under an IRB-exempt review. Sixteen variables were directly comparable between ADE and MDE.

MDE for 59 clinic notes (27 unique variables) took 6months to complete. ADE-templated notes for 90 clinic notes (154 unique variables) took 5min to run a research/QI report. Implementation of ADE included eight weeks of development and testing. Pre-implementation clinical documentation was similar to post-implementation documentation (5-10min).

ADE-templated notes allow for a 5-fold increase in clinically relevant data that can be captured with each encounter. ADE also results in real-time data extraction to a research/QI database that is easily queried. The immediate availability of these data, in a research-formatted spreadsheet, allows for rapid collection, analyses, and interpretation of the data.

IV.

Retrospective Study.