Leveraging real-world data for continuous evaluation of computational clinical practice guidelines

Abstract

Objectives: There is a bidirectional interaction between clinical practice guidelines and clinical care, with each informing the other. Structural signalling of trends in guideline adherence in clinical practice is essential for advanced updates. Recent advances in computable care guidelines allow automated evaluation using real-world registry data. Here, we assess the feasibility by evaluating adherence to Dutch endometrial cancer (EC) guidelines.

Methods: This retrospective cohort study uses real-world data of EC patients from the Netherlands Cancer Registry (NCR) between January 2010 and May 2022. The Dutch guideline for EC was parsed into clinical decision trees (CDTs). Primary outcome was guideline adherence for multiple (sub)populations, with secondary outcomes encompassing adherence trends, recommendation implementation pace, non-adherent treatment strategies and impact of additional non-guideline-based patient and tumour characteristics on adherence.

Results: The Dutch EC guideline was parsed into 10 CDTs, revealing 22 patient and disease characteristics and 46 interventions. NCR data were mapped to CDT data items. Four CDTs were successfully populated with NCR data, and 21 602 cases were assessed. Adherence levels were computed, which showed a mean adherence of 82.7% (range 44-100%). Three statistically significant trends in adherence were identified: two increasing trends in the 'non-adherent' compared with the 'adherent' group, and one decreasing trend.

Discussion: This study introduces a novel framework for continuously evaluating (non-)adherence to cancer guidelines. Future efforts should focus on the inclusion of health outcome measurements.

Conclusion: Through the integration of real-world data with a computer-interpretable guideline, we effectively calculated various facets of adherence to guidelines for EC.

Keywords: Common Data Elements; Decision Trees; Electronic Data Processing; Health Information Interoperability; Medical Informatics Computing.

Figure 1. Screenshot of the guideline-based clinical decision algorithm for adjuvant treatment of patients with endometrioid carcinoma. A visual representation of the guideline-based clinical decision tree for the adjuvant treatment of endometrial cancer version 2011 is depicted in the screenshot. The root node, positioned at the top, is succeeded by 8 (white) nodes, representing disease and patient characteristics. Users can input values for these nodes, guiding them to the corresponding recommendation represented in a (blue) leaf. These leaves incorporate the applicable guideline recommendations for the specific subpopulation. FIGO, Fédération Internationale de Gynécologie et d'Obstétrique; LVSI, lymphovascular space invasion.

Figure 2. Screenshot of the Alertness prototype dashboard using synthetic data. Screenshot of the Alertness prototype dashboard exposing adherence levels of ‘Staging diagnostics’, using synthetic data for demonstration purposes (as a replacement for the real-world data from the NCR). From the total population of 20 000 cases, 19 099 are eligible for evaluation in the decision algorithm of the ‘Staging diagnostics’ phase in the care pathway. For all guideline recommendations being part of this phase (under ‘Recommendations’), the upper bar indicates data availability and the lower bar distinguishes in adherent, non-adherent and ‘other’ cases. The population in scope is defined under ‘Paths’, accompanied by the recommended interventions under ‘Data items’. Implemented interventions from the dataset are displayed under ‘Adherence’ (guideline recommended interventions) and ‘Other interventions’ (guideline non-recommended interventions). Finally, at the top right of the screenshot, the preset ‘Alert settings’ are accessed, including the tailored adherence level or range. CA-125, cancer antigen 125