TL;DR
This study explored the feasibility of using a Clinical Decision Support System (CDSS) to improve anticoagulation management for patients with non-valvular atrial fibrillation (AF) in Chinese primary care settings. The CDSS significantly increased the proportion of patients receiving guideline-directed therapy, suggesting it's a promising tool for enhancing care in these underserved settings.
What Problem Does This Paper Address?
Atrial fibrillation (AF) is a common heart condition that significantly increases the risk of stroke. Effective anticoagulation therapy is crucial for preventing these thromboembolic events. However, in China, the rate of appropriate anticoagulation treatment in primary care settings is notably low, often due to general practitioners' (GPs) heavy workloads and a lack of specialized knowledge in AF management. This gap leads to a higher incidence of strokes and associated healthcare burdens. The study aims to address this by evaluating the effectiveness and feasibility of a CDSS designed to assist GPs in making better-informed anticoagulation decisions for AF patients in community health centers.
Methods
This was a prospective cohort study with a semistructured interview component, designed as a feasibility study for a future cluster-randomized controlled trial. Two community health centers in Shanghai were selected: one as the "software group" where GPs used the CDSS, and another as the "control group" where standard care was provided.
The CDSS was integrated into the hospital information system and automatically assessed stroke and bleeding risks using established scores (CHADS2, CHA2DS2-VASc, HAS-BLED, ORBIT, ATRIA). It then provided treatment and follow-up suggestions based on these risks. GPs in the software group used the CDSS for initial patient assessments and during follow-up visits, with the final treatment decisions resting with the GP.
The primary outcome measured was the proportion of patients prescribed guideline-directed antithrombotic therapy. Secondary outcomes included consultation frequency and the incidence of adverse clinical events (bleeding, stroke, death, hospitalization). A semistructured interview was conducted with the GP in the software group to assess their acceptance of the CDSS and gather suggestions for improvement.
Datasets
The study utilized medical records from 84 patients who completed the follow-up period. These records included demographic information, medical history (hypertension, diabetes, heart failure, stroke/TIA history), current medications, laboratory findings (INR values for warfarin users), and clinical event data. No new datasets were introduced; the study relied on existing patient data collected within the participating community health centers.
Key Insights and Ideas
- CDSS as a Bridge for Primary Care: The study highlights the potential of CDSS to bridge the knowledge and resource gap in primary care settings, empowering GPs to manage complex conditions like AF more effectively.
- Significant Improvement in Guideline Adherence: The CDSS intervention led to a substantial increase in the proportion of patients receiving appropriate anticoagulation, demonstrating its direct impact on clinical practice.
- GP Acceptance is Crucial: The positive reception of the CDSS by the participating GP underscores the importance of user-friendly and clinically relevant tools for successful implementation.
- Feasibility for Scale-Up: The findings suggest that promoting such CDSS solutions in Chinese community health centers is not only feasible but also necessary to improve patient outcomes.
Key Results
- Primary Outcome: The proportion of patients prescribed guideline-directed antithrombotic therapy was significantly higher in the software group (75.5%) compared to the control group (38.7%). The adjusted odds ratio for the software group was 8.071 (p < 0.001).
- Patient Demographics: Eighty-four patients completed the follow-up, with a mean age of 75.71 years. While basic demographics were similar, the software group had a slightly higher average age and a higher proportion of patients with a HAS-BLED score of ≥3.
- Adverse Clinical Events: The incidence of adverse events was lower in the software group (5.7%) than in the control group (12.9%), although this difference was not statistically significant due to the small sample size.
- GP Acceptance: The GP in the software group found the CDSS helpful, strengthening their confidence in prescribing anticoagulants and educating patients. They noted its convenience and potential to assist GPs with less AF experience.
- Dropout Rate: The study experienced a dropout rate of 21.5%, with the primary reason being patients following up in other hospitals.
Limitations and Future Work
The authors acknowledge several limitations:
- Baseline Differences: The software and control groups had some differences in age and HAS-BLED scores at baseline, which could influence results.
- Patient and GP Preferences: Individual patient and GP preferences might have influenced treatment decisions, particularly for lower-risk patients.
- Socioeconomic Data: The study lacked data on patients' socioeconomic status, which could be a confounding factor.
- High Dropout Rate: The significant dropout rate may have affected the generalizability of the findings.
- Specialized GPs: The GPs involved were in AF special consulting rooms, potentially making them more experienced and interested in AF management than the average GP. This limits generalizability to all primary care settings.
Future work will involve a larger, cluster-randomized controlled trial to further validate the CDSS's effectiveness. The protocol has been refined to address dropout rates by adjusting inclusion/exclusion criteria and expanding the trial to all general practice clinics. The CDSS itself will be updated to include more comprehensive assessments (e.g., fall risk, cognition) and improved warfarin dosage management.
Summary updated