Evaluation of the effectiveness of an algorithm to predict COPD exacerbations in primary care: an interrupted time series analysis.

Abstract

To evaluate the impact of the Clinical Exacerbation Score for Health Search (CEX-HScore) on the frequency of registered chronic obstructive pulmonary disease (COPD) exacerbations and referrals to specialists in Italian primary care using an interrupted time series (ITS) analysis. Using data from an Italian federated network of general practitioners (GPs), we conducted an ITS analysis on GP-level (aggregated) indicators to assess changes in registration of moderate-to-severe COPD exacerbations and pulmonologist referrals before (up to January 2023) and after (up to June 2024) the implementation of CEX-HScore as a decision support system in GPs software. The launch of the CEX-HScore on 1 January 2024 did not yield statistically significant increases in indicators quantifying monthly referral rates or coded exacerbations in the general COPD population. Instead, among incident COPD cases, a significant increase of 45% per month (P < .001) in the proportion of coded exacerbations out of total exacerbations was observed postintervention. These findings suggest that the implementation of CEX-HScore may enhance the coding accuracy of COPD exacerbations in newly diagnosed patients. Further data points are needed to robustly evaluate its long-term effectiveness on broader clinical outcomes.