Automated data extraction of electronic medical records: Validity of data mining to construct research databases for eligibility in gastroenterological clinical trials

Abstract

Background: Electronic medical records (EMRs) are adopted for storing patient-related healthcare information. Using data mining techniques, it is possible to make use of and derive benefit from this massive amount of data effectively. We aimed to evaluate validity of data extracted by the Customized eXtraction Program (CXP).
 Methods: The CXP extracts and structures data in rapid standardised processes. The CXP was programmed to extract TNFα-native active ulcerative colitis (UC) patients from EMRs using defined International Classification of Disease-10 (ICD-10) codes. Extracted data were read in parallel with manual assessment of the EMR to compare with CXP-extracted data.
 Results: From the complete EMR set, 2,802 patients with code K51 (UC) were extracted. Then, CXP extracted 332 patients according to inclusion and exclusion criteria. Of these, 97.5% were correctly identified, resulting in a final set of 320 cases eligible for the study. When comparing CXP-extracted data against manually assessed EMRs, the recovery rate was 95.6–101.1% over the years with 96.1% weighted average sensitivity.
 Conclusion: Utilisation of the CXP software can be considered as an effective way to extract relevant EMR data without significant errors. Hence, by extracting from EMRs, CXP accurately identifies patients and has the capacity to facilitate research studies and clinical trials by finding patients with the requested code as well as funnel down itemised individuals according to specified inclusion and exclusion criteria. Beyond this, medical procedures and laboratory data can rapidly be retrieved from the EMRs to create tailored databases of extracted material for immediate use in clinical trials.

Ulcerative colitis progression: a retrospective analysis of disease burden using electronic medical records

“…One of the strengths of the current study is the linkage of several databases including the most proximal EMR to provide a comprehensive and robust evaluation of data in a large cohort of patients with moderate to severe UC (37). However, there are some limitations that need to be considered when interpreting the results of the present study.…”

Automated algorithm for medical data structuring, and segmentation using artificial intelligence within secured environment for dataset creation

“…There are a few automatic data extraction methods that exist. A customized extraction program, an automated data extraction model, was evaluated for its efficiency and shows extracted data accurately of 97.5 % [17] . A series of MR images were first structured and then many computational analyses were performed on the structured data [18] , [19] .…”

A pragmatic methodology to extract anesthetic and physiological data from the electronic health record

“…In practice settings not utilizing paper anesthetic records, these data are now typically recorded continuously in the electronic health record (EHR) anesthesia information systems (AIMS) [1][2][3] and can be accessed and gathered post hoc by using a variety of methods that range from a manual data extraction approach from patients' EHR flowsheets-a process traditionally involving a human data analyst(s) visually reading and transcribing minute-by-minute anesthetic-physiological data by hand onto a separate spreadsheet software program, a time-consuming and error-prone process [4][5][6][7][8][9] -to more technically complicated data extraction approaches involving computer coding in languages such as Structured Query Language (SQL) that require additional expertise and resources. [10][11][12][13] Data can be accessed by relying on the support of (1) helpful intradepartmental tech savvy clinicians, (2) EHR company staff with computer coding expertise to increase access in a more scalable fashion, (3) data scientists in larger academic practices, and (4) through independent direct access to a variety of monitored variables in a perioperative data warehouse. At all times, potential limitations in data granularity and artifacts in the perioperative EHR and AIMS require careful examination, exploration, and analytics to avoid the hazards of "garbage in, garbage out" (GIGO) 3,14 in order to enhance data quality and mitigate inaccurate analyses and study interpretation.…”

“…Two common examples of anesthetic data of interest include (1) pharmacological drug administration such as expired end‐tidal sevoflurane concentration and (2) physiological hemodynamic data such as mean arterial blood pressure. In practice settings not utilizing paper anesthetic records, these data are now typically recorded continuously in the electronic health record (EHR) anesthesia information systems (AIMS) 1–3 and can be accessed and gathered post hoc by using a variety of methods that range from a manual data extraction approach from patients' EHR flowsheets—a process traditionally involving a human data analyst(s) visually reading and transcribing minute‐by‐minute anesthetic‐physiological data by hand onto a separate spreadsheet software program, a time‐consuming and error‐prone process 4–9 —to more technically complicated data extraction approaches involving computer coding in languages such as Structured Query Language (SQL) that require additional expertise and resources 10–13 . Data can be accessed by relying on the support of (1) helpful intradepartmental tech savvy clinicians, (2) EHR company staff with computer coding expertise to increase access in a more scalable fashion, (3) data scientists in larger academic practices, and (4) through independent direct access to a variety of monitored variables in a perioperative data warehouse.…”