Document resource
Background Intravenous (IV) loop diuretics are a cornerstone of therapy for people hospitalised with heart failure (HF). There are several strategies that might enhance diuresis (protocol-based titration of loop diuretic, thiazide [or thiazide-like] diuretics, acetazolamide [ACZ], digoxin, steroids, and oral sodium chloride [NaCl]) but trial data are limited. We are planning a multi-arm, multi-stage randomised trial to test each potential strategy, in addition to usual care (UC), against a shared control group of UC alone – MAMMOTH-HF. Before designing the trial, some uncertainties need to be addressed.Methods We developed an online survey of HF specialists in the UK to help define UC for people hospitalised with HF; appropriateness of each proposed intervention; outcome measures; preparedness to participate in MAMMOTH-HF; and estimated recruitment. The survey was based on a clinical vignette of a typical person hospitalised with HF estimated to require a diuresis of several litres. It launched at the British Society for HF annual conference in November 2020 and ran until January 2026. We performed quantitative and thematic analyses.Results There were 115 responses from 26 different NHS trusts in the UK, 68 of which (59%) were fully completed. Most respondents (61%) selected 80mg IV furosemide twice daily as UC, followed by 80 mg IV bolus followed by 10 mg per hour continuous IV infusion (20%). Most respondents would be prepared to randomise a patient to a loop diuretic titration protocol (90%), thiazide (or thiazide-like) diuretic (91%), ACZ (72%), and digoxin (65%); but fewer to oral NaCl (37%), or steroids (20%) (figure 1). The most common reason for not randomising to oral NaCl was perceived lack of evidence that it enhanced diuresis (52%) followed by concerns of potential harm (48%). The most common reason for not randomising to steroids was concerns of potential harm (68%). The top three endpoints in order of perceived importance for our proposed trial were 1) relief from symptoms and signs of congestion; 2) rehospitalisation rate; and 3) diuresis (table 1). Although mortality and time to fitness for discharge were also commonly chosen. Most respondents believed the endpoint should be measured after 1-6 months (32%), or discharge (27%). Overall, 80% of respondents would be prepared to recruit a patient into our proposed trial. A perceived lack of staff was the most common reason for being unable to recruit. Respondents estimated being able to recruit 3 or more patients per month (62%) (figure 2).Conclusion This survey provides evidence of 1) what constitutes UC for people hospitalised with HF and congestion; 2) which diuretic strategies investigators were most willing to consider; 3) what to measure as the primary endpoint and when; and 4) what the estimated recruitment rate might be. The results will be invaluable for designing the trial protocol.