The effect of down-titration and discontinuation of heart failure
pharmacotherapy in older people: a systematic review and meta-analysis
Abstract
Aim: To investigate if interventions to discontinue or down-titrate
heart failure (HF)-pharmacotherapy are feasible and associated with
risks in older people. Methods: A systematic review and meta-analysis
were conducted according to PRISMA 2020 guidelines. Electronic databases
were searched from inception to March 8th 2023. Randomised controlled
trials (RCTs) and observational studies included people with HF, aged
>50 years and who discontinued or down-titrated
HF-pharmacotherapy. Outcomes were feasibility (whether discontinuation
or down-titration of HF-pharmacotherapy was sustained at follow-up) and
associated risks (mortality, hospitalisation, adverse drug withdrawal
effects [ADWE]). Random-effects meta-analysis was performed when
heterogeneity was not substantial (Higgins I2<70%).
Sub-analysis by frailty status was conducted. Results: Six RCTs
(536-participants) and 27 observational studies (810,499-participants)
across six therapeutic classes were included, for 3-260 weeks follow-up.
RCTs were conducted in patients presenting with stable HF.
Down-titrating a renin-angiotensin system inhibitor (RASI) in patients
with chronic kidney disease was 76% likely than continuation (Risk
Ratio [RR] 1.76, 95%CI 1.14-2.73), with no difference in mortality
(RR 0.64, 95%CI 0.30-1.64). Discontinuation of beta-blockers were
feasible compared to continuation in preserved ejection fraction (RR
1.00, 95%CI 0.68-1.47). Participants were 25% likely to re-initiate
discontinued diuretics (RR 0.75, 95%CI 0.66-0.86). Digoxin
discontinuation was associated with 5.5-fold risk of hospitalisation
compared to continuation. Worsening HF was the commonest ADWE. One
observational study measured frailty but did not report outcomes by
frailty status. Conclusions: The appropriateness and associated risks of
down-titrating or discontinuing HF-pharmacotherapy in people aged
>75 years is uncertain. Evaluation of outcomes by frailty
status necessitates investigation.