Developing a Standards-Based Information Model for Representing Computable Diagnostic Criteria: A Feasibility Study of the NQF Quality Data Model

Abstract

The lack of a standards-based information model has been recognized as a major barrier for representing computable diagnostic criteria. In this paper we describe our efforts in examining the feasibility of the Quality Data Model (QDM)–developed by the National Quality Forum (NQF)–for representing computable diagnostic criteria. We collected the diagnostic criteria for a number of diseases and disorders (n=12) from textbooks and profiled the data elements of the criteria using the QDM data elements. We identified a number of common patterns informed by the QDM. In conclusion, the common patterns informed by the QDM are useful and feasible in building a standards-based information model for computable diagnostic criteria.

Introduction

For decades clinicians have recognized that their practices greatly benefit from scientifically-based diagnostic criteria, iteratively improved through ongoing research [1]. With recent advances in computerized patient records systems there is an urgent need for producing computable and standards-based diagnostic criteria that could be utilized more effectively in clinical and translational applications, ultimately improving patient care. The lack of a standards-based information model has been recognized as a major barrier for achieving computable diagnostic criteria. The objective of this study is to describe our efforts in developing a standards-based information model for computable diagnostic criteria.

Methods

We established a library of diagnostic criteria for a collection of diseases and disorders. We profiled these diagnostic criteria using QDM elements. National Quality Forum (NQF) developed the Quality Data Model (QDM), which is an information model for representing electronic health record-based quality “eMeasures” [2]. We identified individual criteria from each disease or disorder and profiled them with QDM elements. We analyzed the distribution of QDM elements and identified common patterns for representing the diagnostic criteria. We developed an information model for representing diagnostic criteria by integrating QDM and ICD-11 content models in a Semantic Web-based framework.

Results

We searched electronic medical textbooks from the Mayo Clinic Library website using keywords “Diagnostic Criteria” and selected a collection of diagnostic criteria from 12 diseases or disorders. In total, 139 individual criteria were identified and used as a test set for further analysis. We profiled all 139 criteria using QDM elements in terms of QDM categories, datatypes, and attributes. In total we identified eight QDM datatypes and their attributes from eight distinct QDM categories; of these QDM datatypes, five are most commnly used: “Laboratory Test, Performed;” “Diagnostic Study, Performed;” “Diagnosis, Active;” “Physical Exam, Performed;” and “Symptom, Active.”

Discussion

The study is motivated by the requirement of the WHO ICD-11 revision project, in which the creation of diagnostic criteria for the diseases and disorders in ICD-11 has been proposed as a key component for its revision. In this study we used a data-driven approach to evaluate the feasiblity of using QDM to represent diagnostic criteria. We developed a standards-based information model informed by the QDM and ICD-11 content models. We envision that the QDM-based information model and tools for representing quality measures could be used to tackle the challenges in achieving computable diagnostic criteria. This study suggests that the common patterns informed by QDM are useful and feasible in building a standards-based information model for computable diagnostic criteria.