Impact of a diagnostic decision support system on GP clinical documentation

Abstract

Listen

Translate article

BackgroundKostopoulou and colleagues designed and evaluated a diagnostic decision support system (DSS) that presented GPs with differential diagnoses to consider at the start of the consultation. The DSS was integrated with the electronic health record (EHR) and evaluated in simulated consultations (34 GPs consulting with 12 actors), where it was found to improve diagnostic accuracy.AimTo evaluate the impact of the DSS on GP clinical documentation.MethodSecondary data analysis. The analysis dataset consisted of all data items that the 34 GPs recorded during their consultations with the 12 actors (408 consultations). Each GP had conducted six consultations with the EHR alone, and six with the integrated DSS. The 12 patient scenarios had a pre-defined set of clinical cues and differential diagnoses. For each patient scenario, a Delphi panel of five GPs identified which cues were consistent with each differential diagnosis. The data items recorded, either in free text or code, were counted. These were referenced against the Delphi panel’s matrix of cues and diagnoses. For each GP, it was estimated what proportion of the recorded data was consistent with the diagnosis that they finally gave.ResultsUsing the DSS increased the total amount of information documented (b = 3.58 [95% confidence interval {CI} = 2.81 to 4.35] P<0.001) and reduced the proportion of documentation that was consistent with the GP’s final diagnosis (b = −0.08 [95% CI = −0.11 to −0.05] P<0.001), suggesting less bias.ConclusionUsing the DSS led to more complete and less biased documentation. This has implications for the use of routine EHR data to create clinical prediction rules.