Increased risk of hospitalisation and death with the delta variant in the USA.

Increased hospitalisation risk with SARS-CoV-2 delta (B.1.617.2) variant infection compared with alpha (B.1.1.7) variant infection in unvaccinated people was reported by Katherine Twohig and colleagues in the UK (hazard ratio [HR] 2·32 [95% CI 1·29–4·16]) and Peter Bager and colleagues in Denmark (3·01 [2·02–4·50]). To corroborate these results in the USA, as well as to examine risk of intensive care unit (ICU) admission or death, we analysed a large cohort of unvaccinated COVID-19 cases within the Veterans Health Administration (VHA) using the previously described COVID-19 Shared Data Resource.

In the absence of available genomic confirmation of SARS-CoV-2 variant, we used two time periods: the first before substantial delta variant detection in the USA (Feb 1–May 15, 2021) and the second with clear dominance by the delta variant (July 10–Aug 31, 2021). Patient characteristics are described in the appendix. During the delta surge, patients were more likely to be younger and female, and there were lower rates of comorbidities. Comparing delta and pre-delta timeframes, and adjusting for age, race, ethnicity, gender, body-mass index, diabetes, hypertension, chronic obstructive pulmonary disease, cardiovascular disease, and kidney disease, there was a significant increase in the risk of hospitalisation (HR 1·93 [95% CI 1·84–2·03]), ICU stay (odds ratio 2·29 [2·12–2·47]), and death (HR 2·15 [1·93–2·39]). Stratifying by age (<50 years vs ≥50 years), we found that although not statistically different, the increase in risk of death with the delta variant seemed higher for those younger than 50 years (HR 3·31 [2·05–5·34]) than those aged 50 years and older (2·09 [1·88–2·34]).

These results show that in a population of unvaccinated VHA patients, infection with the delta variant conferred an approximately two-fold increased risk of hospitalisation, consistent with previous studies.1, 2 Furthermore, the concomitant increase in risk of ICU admission and death, despite controlling for numerous risk factors, is concerning. Better powered studies should examine whether the increase in risk is disproportionately large in those younger than 50 years. A limitation of the study is that the VHA population has high rates of comorbidities and might not be reflective of the US population. However, our analysis reports an increase in risk between variants within the same population, rather than providing an estimation of absolute risk. Considering that vaccination still protects against adverse outcomes due to the delta variant, evidence of delta variant infection increasing the risk of hospitalisation and death among all age groups should inform vaccination policies.

We declare no competing interests. This study used data created and maintained by the Veterans Health Administration, US Department of Veterans Affairs. These data are available to approved individuals upon request after fulfilling specified requirements.