Table 2.

Characteristics of included models

Study characteristics	Park et al24	Lee et al25
Study population	Hospitalized adults with an RSV RTI presenting at the emergency department (n = 227); 133 (59%) community acquired, 94 (41%) healthcare‐associated. In total, 84 (37%) patients were immunocompromised (25 solid organ recipients, 9 patients, with HCT 50 using immunosuppressants/ corticosteroids) and 42 (19%) had a chronic pulmonary disease	Hospitalized adults with an RSV RTI (n = 607). In total, 83 (13.7%) patients were immunocompromised and 216 (36%) had a chronic pulmonary disease
	Exclusion: ≤18 y, outpatient treatment, RSV diagnosis >48 h after admission, concurrent infections at other sites	Exclusion: none
Primary outcome	Life‐threatening RSV‐infection (admission to ICU, need for ventilator care or in‐hospital death; n = 34, 15.0%)	30‐d mortality (n = 55, 9.1%), 60‐d mortality (n = 72, 11.9%)
Patient identification and data collection	Identification using RSV positive PCR assays; retrospective data collection	Identification of RSV positive viral antigen immunofluorescence assay tests; retrospective data collection
Inclusion location	ED of a 2700‐bed tertiary care hospital in Seoul, South Korea	Three acute care, general public hospitals in Hong Kong, China
Inclusion period	October 2013‐September 2015	January 2009‐December 2011
Modeling technique	Multivariable logistic regression analysis with stepwise backward variable selection	Multivariable Cox proportional hazards analysis with stepwise backward variable selection
Variable selection for multivariable analysis	Variables with P  ≤ .05 in univariate analyses of association with life‐threatening infection. Exclusion of variables in causal pathway (confusion, saturation); subjective symptoms (dyspnea); correlated variables (smoking history, correlated with chronic pulmonary disease)	Variables with P  ≤ .1 in univariate analyses of association with mortality. Inclusion of demographics, comorbidities, cardiorespiratory complications, ventilation requirement, bacterial superinfection, and corticosteroid use
Variables included in multivariable analysis a	Lower RTI, chronic pulmonary disease, bacterial coinfection, fever ≥38°C, rhinorrhoea, CRP, procalcitonin, RSV type A and B, antimicrobial use b , ribavirin use b	Age, gender, major systemic comorbidity, chronic pulmonary disease exacerbation, cardiovascular complications, pneumonia, need for ventilatory support, bacterial coinfection, urea, total white cell count, systemic corticosteroid use
Variables in final model a	Lower RTI, chronic pulmonary disease, bacterial coinfection, fever ≥38⁰C	Age >75 y, male gender, pneumonia, need for ventilatory support, bacterial coinfection, urea
Missing data handling	Not described	Not described

Abbreviations: CRP, C‐reactive protein; DFA, direct fluorescent antibody; ED, emergency department; HCT, hematopoietic cell transplant; ICU, intensive care unit; PCR, polymerase chain reaction; RSV, respiratory syncytial virus; RTI, respiratory tract infection.

^aTable S2 for definitions.

^bNo further definition or details given.