Rapid systematic review of clinical risk factors for severe Respiratory Syncytial Virus infection in adults resulting in hospitalisation, intensive care unit admission or death

Methods

This review included peer-reviewed articles comprising study cohorts of adults (≥18 years) with laboratory-confirmed RSV infection that reported associations between predefined clinical risk factors and severe outcomes. Records were identified through searches in electronic databases (Embase, Cochrane Library and PubMed, including MEDLINE) and by screening reference lists of included articles. Effect estimates were interpreted using predefined thresholds. Relative effects of 1.1 to <1.5 were categorised as weak, 1.5 to <2.0 as moderate, 2.0 to

<4.0 as strong, and ≥4.0 as very strong. These thresholds were applied as a descriptive framework to support interpretation across heterogeneous studies and should not be taken to imply comparability of baseline risk, outcome prevalence, or clinical context. Priority was given to findings from low risk of bias studies and adjusted estimates.

The database search identified 1 594 records. After the removal of duplicates, 1 134 records remained. Title and abstract screening excluded 994 records, leaving 140 articles for full-text review. Of these, eight articles met the inclusion criteria. One additional relevant article was identified through reference list screening, resulting in a total of nine included articles.

Key findings

The evidence base was dominated by retrospective hospital-based cohort studies, focused on older (>60 years) adult populations. Studies were heterogeneous in terms of study design, population characteristics, definitions of exposures and outcomes, and analytical methods, limiting comparability and precluding definitive ranking of risk factors or quantitative synthesis by meta-analysis. With the exception of age, most studies assessed risk factors as binary variables, limiting the ability to examine gradients of risk or associations across levels of the risk factor (e.g. varying degrees of disease or differences in biomarker levels).

Demographic factors

Older age, particularly 65–75 years and above, was among the most frequently evaluated risk factors for severe RSV outcomes across the included studies. Some evidence, particularly from a low risk-of-bias study in older populations, suggests an increased risk with advancing age, especially for mortality, with higher risks observed in the oldest age groups. However, findings were not consistent across all studies, including those at low risk of bias, and several analyses showed no statistically significant association. Where associations were observed, they were often imprecise and not always robust to adjustment for confounding. The overall evidence base is limited, and the observed associations may in part reflect residual confounding by frailty and underlying health status. Taken together, older age is likely to be an important risk factor for severe RSV outcomes, particularly mortality, but uncertainty remains regarding the strength and consistency of this association. While not all studies showed consistent effects, and some estimates were imprecise, the overall evidence suggests that advancing age is the most important and policy-relevant risk factor. Sex was not supported as an independent risk factor in the more robust analyses.

Comorbidity factors

Among comorbidities, chronic lung disease, particularly chronic obstructive pulmonary disease (COPD), showed some evidence of increased risk. Moderate to strong associations were reported across several studies, especially for hospitalisation, although these were largely based on unadjusted analyses, some effects were attenuated after adjustment, and evidence for more severe outcomes was inconsistent and limited.

Neurological disease was associated with an increased risk in a limited number of studies, including low risk of bias studies reporting moderate adjusted effects. This suggests a plausible association, although the evidence base remains small.

Cancer and haematological malignancies were associated with increased mortality in a small number of studies, with some strong to very strong adjusted estimates reported. However, these findings are based on limited data.