Objective To assess the cost-effectiveness of three monitoring strategies for optimising medical therapy in chronic heart failure (CHF).

Design This analysis was based on six randomised controlled trials. Costs were measured from a UK NHS perspective and estimated for patients’ lifetime. The health outcome was the quality-adjusted life-year (QALY).

Setting Hospital and community.

Patients Patients with CHF.

Interventions Serial measurement of natriuretic peptide (NP) by a specialist, clinical assessment by a specialist, and usual care in the community.

Main outcome measures Costs, QALYs, and incremental cost-effectiveness ratio (ICER).

Results Serial NP measurement by a specialist was the most cost-effective option in patients with CHF due to left ventricular systolic dysfunction (LVSD), showing an ICER of £3304 compared with clinical assessment. Serial NP measurement by a specialist was strongly favoured in patients with CHF from any cause, for all patients (ICER of £14 694 compared with clinical assessment by a specialist) and for the age subgroup ≤75 years (ICER of £2517 compared with usual care). However, serial NP measurement by a specialist was dominated (less effective and more costly) by alternative strategies in the subgroup age >75 years with CHF from any cause. Clinical assessment by a specialist of patients >75 years of age with CHF from any cause was cost-effective compared with usual care (ICER of £11 508).

Conclusions Serial measurement of NP concentration by a specialist is the most cost-effective strategy for CHF due to LVSD and from any cause, except in the subgroup of patients >75 years with CHF from any cause, where treatment guided by NP measurement may be harmful and not cost-effective.