START-HF: a multi-centre evaluation of centralised pharmacist-led remote GDMT optimisation in HFrEF

Background

Guideline-directed medical therapy (GDMT) reduces mortality and hospitalisation in heart failure with reduced ejection fraction (HFrEF), yet real-world uptitration remains slow and incomplete due to outpatient constraints and infrequent follow-up. Centralised pharmacist-led remote models offer scalable solutions, but evidence from multi-hospital regional networks is limited.

Purpose

To evaluate whether a centralised pharmacist-led remote GDMT optimisation service across five hospitals shortened time to quadruple therapy achievement and increased optimisation rates in HFrEF compared with historical standard care.

Methods

We conducted a pre-specified multi-centre observational study comparing a centralised pharmacist-led remote service (September 2024–September 2025; n=192) with a historical control cohort from the same five hospitals (January 2023–August 2024; n=174). Eligible patients with HFrEF post-discharge or following outpatient review submitted daily remote monitoring data (blood pressure, heart rate, symptoms); pharmacists performed protocoled asynchronous uptitration twice weekly. The primary endpoint was time to maximum tolerated GDMT, defined as achievement of target or maximum tolerated doses of all four pillars (ACEi/ARNi, beta-blocker, MRA, SGLT2i). Heart failure hospitalisation at 90 days was a pre-specified secondary endpoint.

Results

Baseline characteristics were well matched: mean age 59.4 vs 61.4 years, male sex 75.5% vs 69.7%, mean LVEF 29.8% vs 28.7%. Median time to GDMT optimisation was 52 days (IQR 30.5–68) in the remote service versus 179 days (IQR 98.5–311) in controls (p<0.001), an absolute reduction of 127 days (>4 months). The proportion achieving optimal quadruple GDMT by 90 days was significantly higher in the remote cohort compared with historical care (93.2% vs 23.0%; p<0.001).

Heart failure hospitalisation occurred in 0/192 (0%) versus 5/174 (2.9%) patients at 90 days (p=0.024). No serious adverse events attributable to remote titration were recorded.

Conclusion

A centralised pharmacist-led remote GDMT optimisation service across five hospitals reduced time to quadruple therapy achievement by over four months and significantly increased the proportion of patients reaching optimal medical therapy, compared with historical standard care. This was associated with fewer heart failure hospitalisations and no attributable serious adverse events. The asynchronous hub-based model is scalable across regional networks and offers a pragmatic strategy to close the persistent gap between guideline evidence and clinical practice. Prospective randomised evaluation is warranted.