High-dose budesonide for early COVID-19 - Authors' reply.

We thank Ivan Berezowski and colleagues for highlighting the importance of the PRINCIPLE trial finding a safe, effective, and inexpensive community repurposed medication that shortens COVID-19 illness and reduces the need for hospitalisation and use of oxygen. Most participants (85%) had up to 10 days’ illness duration (63% fewer than 7 days in the concurrent population). Inclusion of those almost recovered would reduce rather than increase the chance of showing an effect. In addition, if people without obesity incorrectly reported as people with obesity (32% self-reported a body-mass index >35, but only 27·4% of those were eligible on this criterion alone), this would also probably bias the results towards the null because obesity can be associated with worse outcomes. For patient-reported recovery, asking participants how they feel is appropriate. Indeed, we have reported three treatments not benefiting patient recovery,3, 4, 5 with one tending to worsen patient recovery. Furthermore, several well validated patient-reported outcomes were also used, including the WHO-5 Wellbeing Scale, with differences favouring inhaled budesonide statistically significant at days 7, 14, and 28. Other measures of recovery were modifications of scales used in several large-scale clinical trials shown to be highly responsive to change. All measures showed benefit—while people were recovering, they felt less ill; once recovered they stayed well more often (10% absolute difference, nearly 50% relative difference in sustained recovery over 28 days); and they used fewer health-care resources.

All patients in the PRINCIPLE trial were symptomatic and treatment adherence was high (more than 80% used the inhalers for at least 7 days) suggesting acceptability and wide applicability. The difference in numbers was the exclusion of participants without follow-up information or asymptomatic on day 0. The limitations of the trial were acknowledged in relation to participant ethnicity, although it was representative of the overall UK population, and the exclusion of patients with chronic obstructive pulmonary disease on inhaled steroids.

We agree with Brian Lipworth and colleagues that our findings might be due to systemic anti-inflammatory effects of budesonide rather than local (lung and naso-pharyngeal) effects. Given the pragmatic nature of the trial we cannot ascertain mechanisms such as central effects from the high dose of budesonide. However, importantly, the treatment showed no appreciable adverse effects and should therefore be recommended for early community use in symptomatic COVID-19 patients older than 50 years. Repeating the trial using lower doses of inhaled steroid is not currently planned.

FDRH reports occasional consultancy fees from BMS, Pfizer, Novartis, Bayer, and Boehringer Ingelheim, unrelated to this Correspondence. BRS reports grant money paid to their employer (Berry Consultants) from the University of Oxford, for the sponsor's grant from the Medical Research Council, per the statistical design and analyses for the PRINCIPLE trial. MB reports grants from AstraZeneca and Roche; university honoraria from AstraZeneca, GSK, Cipla, and Boehringer Ingelheim; participation on data safety monitoring board or advisory board for AstraZeneca; and he is a scientific adviser to AstraZeneca's eosinophil strategy board, ProAxsis, and AlbusHealth. L-MY and CCB declare no competing interests.