MacIntyre 2009.

Study characteristics
Methods	Prospective cluster‐RCT carried out in Sydney, Australia, to assess the use of surgical masks, P2 masks, and no masks in preventing ILI in households. The study was carried out during the 2 winter seasons of 2006 and 2007 (August to the end of October 2006 and June to the end of October 2007). “Gaussian random effects were incorporated in the model to account for the natural clustering of persons in households"
Participants	290 adults from 145 families. 47 households (94 enrolled adults and 180 children) were randomised to the surgical mask group, 46 (92 enrolled adults and 172 children) to the P2 mask group, and 52 (104 enrolled adults and 192 children) to the no‐mask (control) group.
Interventions	Use of surgical masks and P2 mask versus no mask. The P2 mask is described as very cumbersome. See Table 4 for details.
Outcomes	Laboratory: serological evidence Effectiveness: ILI (described as fever, history of fever or feeling feverish in the past week, myalgia, arthralgia, sore throat, cough, sneezing, runny nose, nasal congestion, headache) However, a positive laboratory finding for influenza converts the ILI definition into one of influenza. Safety: N/A
Notes	The study authors conclude that adherence to mask use significantly reduced the risk for ILI‐associated infection, but < 50% of participants wore masks most of the time. They concluded that household use of face masks is associated with low adherence and is ineffective for controlling seasonal respiratory disease. Compliance was by self‐report, therefore likely to be an underestimate. The primary outcome was ILI or lab‐positive illness. This showed no effect. Sensitivity analysis by adherence showed that under the assumption that the incubation period is equal to 1 day (the most probable value for the 2 most common viruses isolated, influenza (21) and rhinovirus (26)), adherent use of P2 or surgical masks significantly reduces the risk for ILI infection, with a hazard ratio = 0.26 (95% CI 0.09 to 0.77; P = 0.015). No other covariate was significant. Under the less likely assumption that the incubation period is equal to 2 days, the quantified effect of complying with P2 or surgical mask use remains strong, although borderline significant; hazard ratio was 0.32 (95% CI 0.11 to 0.98; P = 0.046). The study was underpowered to determine if there was a difference in efficacy between P2 and surgical masks (Table 5). The study conclusion appears to be a post hoc data exploration. Regardless of this, the study message is that respirator use in a family setting is unlikely to be effective as compliance is difficult unless there is a situation of real impending risk. Funding: the Office of Health Protection, Department of Health and Ageing, Australia, 3M Australia, and Medical Research Council (UK). Disclosure: Simon Cauchemez, PhD; Dominic E. Dwyer, BSc(Med), MBBS, FRACP, FRCPA, MD; Holly Seale, BSc, PhD; Pamela Cheung, RN; Gary Browne, MBBS; James Wood, BSc, PhD; and Zhanhai Gao, BSc, MSc, PhD, have disclosed no relevant financial relationships. C. Raina MacIntyre, MBBS, FRACP, FAFPHM, M App Epid, PhD, has disclosed that she has received grants for clinical research from 3M. Michael Fasher, MBBS, PhD, has disclosed that he has received grants for educational activities from and has served as an advisor or consultant to GlaxoSmithKline. Robert Booy, MBBS, FRACP, FRCPCH, MSc, MD, has disclosed that he has received grants for clinical research and educational activities from, and has served as an advisor or consultant to, CSL, Roche, Sanofi, GlaxoSmithKline, and Wyeth. All funding received is directed to a research account at The Children’s Hospital at Westmead, Sydney, Australia, and is not personally accepted by Dr. Booy. Neil Ferguson, FmedSci, DPhi, has disclosed that he has served as an advisor or consultant to Crucell Inc.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Participating households were randomised to 1 of 3 arms by a secure computerised randomisation process", but sequence generation not described.
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	"Study participants and trial staff were not blinded, as it is not technically possible to blind the mask type to which participants were randomised."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"However, laboratory staff were blinded to the arm of randomisation."
Incomplete outcome data (attrition bias) All outcomes	Low risk	143 of 145 randomised families were analysed; 2 families in the control group were lost to follow‐up during the study, for which no reasons were given.
Selective reporting (reporting bias)	Low risk	No differences between groups at baseline