Review 4: "Outcomes of SARS-CoV-2 Omicron infection in residents of Long-Term Care"

RR:C19 Evidence Scale rating by reviewer:

• Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.

***************************************

Review:

This work investigates the risk of severe outcomes in UK Long-Term Care Facilities (LTCFs) residents infected with the SARS-CoV-2 Omicron variant. The typical resident in LTCFs is a person aged > 85 years, with high levels of comorbidity and having received a booster vaccine. Outcomes are compared between the pre-Omicron period (Delta) and Omicron period. The main result was that the hazards of hospitalisation and death were lower during the Omicron period than during the Delta period. The adjusted Hazard Ratio is 0.50 for the hospital admission.

A strength of this study is that it was nested within an existing cohort study. This allowed rapid data linking testing, vaccination, hospitalisations, and deaths. In addition, residents were tested at least monthly for SARS-CoV-2, which limits diagnostic bias.

However, there are limitations. The main one is that the definition of the pre-Omicron and Omicron periods is not optimal. The subsample where the variant was confirmed by sequencing or SGTF showed that 9% of the cases in the Omicron period were, in fact, Delta cases. It would have been preferable to interpose a washout period to increase the probability that the cases that occurred during the Omicron period were indeed Omicron cases.

Another limitation is that the analysis was carried out very shortly after the introduction of Omicron. This is quite justifiable, as there was a need to gain knowledge about the severity of this new variant quickly. However, as the Delta and Omicron cases are not contemporary, the consequence is that the follow-up time is longer for Delta cases than for Omicron cases. This may lead to bias. The authors report this limit for the outcome of "death" It is possible that this also influences the estimation of the Hazard Ratio for the hospital admission. We would hope that the authors could update the results of this study with longer follow-up times.

The results presented (Table 2) for the multivariate analysis are those of a Cox model without interactions. However, it is written that there is an interaction between period and booster vaccination status. This contradiction raises questions about the validity of the model.

This work is clearly written. The authors cite relevant literature and do not overstate their conclusions. I do not have any concerns about the article from an ethical perspective.