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

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

Objective

Routinely collected electronic health records using artificial intelligence (AI)-based systems bring out enormous benefits for patients, healthcare centers, and its industries. Artificial intelligence models can be used to structure a wide variety of unstructured data.

Methods

We present a semi-automatic workflow for medical dataset management, including data structuring, research extraction, AI-ground truth creation, and updates. The algorithm creates directories based on keywords in new file names.

Results

Our work focuses on organizing computed tomography (CT), magnetic resonance (MR) images, patient clinical data, and segmented annotations. In addition, an AI model is used to generate different initial labels that can be edited manually to create ground truth labels. The manually verified ground truth labels are later included in the structured dataset using an automated algorithm for future research.

Conclusion

This is a workflow with an AI model trained on local hospital medical data with output based/adapted to the users and their preferences. The automated algorithms and AI model could be implemented inside a secondary secure environment in the hospital to produce inferences.

Behind the scenes: Key lessons learned from the RELIEVE-AKI clinical trial

Continuous kidney replacement therapy (CKRT) is commonly used to manage critically ill patients with severe acute kidney injury. While recent trials focused on the correct dosing and timing of CKRT, our understanding regarding the optimum dose of net ultrafiltration is limited to retrospective data. The Restrictive versus Liberal Rate of Extracorporeal Volume Removal Evaluation in Acute Kidney Injury (RELIEVE-AKI) trial has been conducted to assess the feasibility of a prospective randomized trial in determining the optimum net ultrafiltration rate. This paper outlines the relevant challenges and solutions in implementing this complex ICU-based trial. Several difficulties were encountered, starting with clinical issues related to conducting a trial on patients with rapidly changing hemodynamics, low patient recruitment rates, increased nursing workload, and the enormous volume of data generated by patients undergoing prolonged CKRT. Following several brainstorming sessions, several points were highlighted to be considered, including the need to streamline the intervention, add more flexibility in the trial protocols, ensure comprehensive a priori planning, particularly regarding nursing roles and their compensation, and enhance data management systems. These insights are critical for guiding future ICU-based dynamically titrated intervention trials, leading to more efficient trial management, improved data quality, and enhanced patient safety.

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

BACKGROUND/AIMS

Traditional manual methods of extracting anesthetic and physiological data from the electronic health record rely upon visual transcription by a human analyst that can be labor-intensive and prone to error. Technical complexity, relative inexperience in computer coding, and decreased access to data warehouses can deter investigators from obtaining valuable electronic health record data for research studies, especially in under-resourced settings. We therefore aimed to develop, pilot, and demonstrate the effectiveness and utility of a pragmatic data extraction methodology.

METHODS

Expired sevoflurane concentration data from the electronic health record transcribed by eye was compared to an intermediate preprocessing method in which the entire anesthetic flowsheet narrative report was selected, copy-pasted, and processed using only Microsoft Word and Excel software to generate a comma-delimited (.csv) file. A step-by-step presentation of this method is presented. Concordance rates, Pearson correlation coefficients, and scatterplots with lines of best fit were used to compare the two methods of data extraction.

RESULTS

A total of 1132 datapoints across eight subjects were analyzed, accounting for 18.9 h of anesthesia time. There was a high concordance rate of data extracted using the two methods (median concordance rate 100% range [96%, 100%]). The median time required to complete manual data extraction was significantly longer compared to the time required using the intermediate method (240 IQR [199, 482.5] seconds vs 92.5 IQR [69, 99] seconds, p = .01) and was linearly associated with the number of datapoints (rmanual  = .97, p < .0001), whereas time required to complete data extraction using the intermediate approach was independent of the number of datapoints (rintermediate  = -.02, p = .99).

CONCLUSIONS

We describe a pragmatic data extraction methodology that does not require additional software or coding skills intended to enhance the ease, speed, and accuracy of data collection that could assist in clinician investigator-initiated research and quality/process improvement projects.

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

BACKGROUND

Ulcerative colitis (UC) is a debilitating inflammatory bowel disease. Present knowledge regarding UC disease progression over time is limited.

OBJECTIVE

To assess UC progression to severe disease along with disease burden and associated factors.

METHODS

Electronic medical records linked with Swedish national health registries (2005-2015) were used to identify disease progression of UC. Odds of all-cause and disease-related hospitalization within 1 year were compared between patients with disease progression and those without. Annual indirect costs were calculated based on sick leave, and factors related to UC progression were examined.

RESULTS

Of the 1,361 patients with moderate UC, 24% progressed to severe disease during a median of 5.2 years. Severe UC had significantly higher odds for all-cause (OR [odds ratio] 1.47, 95% CI [confidence interval]: 1.12-1.94, P < 0.01) and UC-related hospitalization (OR 2.47, 95% CI: 1.76-3.47, P < 0.0001) compared to moderate disease. Average sick leave was higher in patients who progressed compared to those who did not (64.4 vs 38.6 days, P < 0.001), with higher indirect costs of 151,800 SEK (16,415 €) compared with 92,839 SEK (10,039 €) (P < 0.001), respectively. UC progression was related to young age (OR 1.62, 95% CI: 1.17-2.25, P < 0.01), long disease duration (OR 1.09, 95% CI: 1.03-1.15, P < 0.001), and use of corticosteroids (OR 2.49, 95% CI: 1.67-3.72, P < 0.001).

CONCLUSION

Disease progression from moderate to severe UC is associated with more frequent and longer hospitalizations and sick leave. Patients at young age with long disease duration and more frequent glucocorticosteroid medication are associated with progression to severe UC.