Unmasking the evidence about masks

Australian infection control strategies for pandemic influenza are influenced by world authorities — the Centers for Disease Control and Prevention (CDC) and the World Health Organization. WHO guidelines1 take into account the lack of health resources in many communities, and focus on affordability as well as reductions in infection risk. CDC guidelines2 presuppose a well resourced health sector and are aimed at achieving zero risk. The different approaches of the two organisations are manifest in their conflicting recommendations for the type of face mask to use in routine care of patients with influenza: CDC recommends the N95 respirator (equivalent to the P2 mask used in Australia), while WHO recommends the cheaper surgical mask. By giving the world free access to their guidelines, the organisations have saved countries the cost of guideline development. Yet, the gain in risk reduction with the adoption of the CDC’s recommendation is unknown and may not be cost-effective. Conversely, those who opt for the WHO guideline might not appreciate that health care workers (HCWs) in well resourced settings are unlikely to accept a strategy if they perceive it to be significantly riskier than the more costly alternative. The important question is whether either guideline is based on the best evidence and relates the potential risk reduction to the cost involved.

A potted history of CDC’s change in preference from surgical to P2 masks may help those seeking well informed policies about the use of masks for routine patient care. The history of the use of masks by HCWs has been classified into three eras: development and testing (1905–1920); “awareness of the importance of masks” (1920–1940); and the “unimportance of masks secondary to antibiotics” (1940 and beyond).3 We nominate a fourth era, “over-importance of masks” (1990s to the present), which was set in motion by changes to CDC guidelines.4 The CDC’s decision to recommend P2 masks instead of surgical masks for routine care of patients with tuberculosis (TB) was prompted by an unusual outbreak of multidrug-resistant TB in HCWs.5 The CDC made the change despite acknowledging that (1) P2 masks were manufactured to filter industrial, non-pathogenic aerosols and tested to filter out 95% of 0.3 μm sodium chloride particles, not airborne or droplet-sized bioaerosols; (2) some surgical masks were also capable of filtering out 95% of 0.3 μm sodium chloride particles; and (3) the protective efficiency of P2 masks against specific pathogens was unknown. The revised recommendation instigated a widespread non-evidential assumption of a link between wearing masks and preventing aerosolised transmission of pathogens based on particle size: that is, it was assumed that P2 masks prevent disease transmission by airborne particles (≤ 5 μm in size), and surgical masks prevent transmission by droplet particles (> 5 μm in size).4 With neither laboratory nor in vivo efficiency data to compare mask types, why were P2 masks advocated to protect HCWs against TB? The answer lies in the principles of evidence-based medicine.

Despite its deceptive moniker, evidence-based medicine values not only research evidence, but costs and the “needs and values” of stakeholders, including clinicians. CDC leaders moved from surgical masks to P2 masks to solve the “problem of merging scientific and theoretical data into a sound infection control approach for the protection of HCWs against tuberculosis”.6 The problem was not resolved on evidential grounds because the evidence was simply not there. Rather, the solution prioritised the needs and values of clinicians concerned that the surgical mask permits transmission of multidrug-resistant TB because it allows a gap between the face and mask. As a consequence, this revision bred a legacy that associates mask type with particle size rather than with HCWs’ needs and values.6

Confronted with a similar debate about influenza transmission, the Australian Department of Health and Ageing commissioned us to review the protectiveness of masks,7 antiviral prophylaxis and vaccination, and to develop evidence-based infection control algorithms8 for the protection of HCWs during a pandemic. Development of the algorithms was informed by the following considerations.

• Proper use of a mask is more protective than not using a mask, but research findings show that neither the P2 nor surgical mask type is statistically superior. This concurs with findings from a review9 as well as results of a randomised control trial showing non-inferiority of surgical masks.10

• The effectiveness of antiviral prophylaxis against future strains cannot be tested. However, antiviral prophylaxis is effective against seasonal influenza when administered in the first 48 hours of illness and if the drug-resistance of the virus is low.11

• Given the efficacy of seasonal vaccines, a novel pandemic vaccine may provide a similar level of protection.

• Implementing the recommendation that HCWs wear P2 masks for routine patient care in a pandemic has several difficulties, including cost and fit-testing. The need for fit-testing was added to both CDC and WHO infection control guidelines1,2 after occupational-acquisition of the severe acute respiratory syndrome (SARS) by HCWs. However, there is no evidence that fit-testing affords higher levels of protection than a comfortably fitting mask.

• During our study of stakeholders’ needs and values, local clinicians unanimously disputed the recommendation for the use of surgical masks for routine care despite the recommendation being only one part of our multi-tool infection control strategy, which included antiviral prophylaxis, vaccination and a face shield for eye protection.8 The clinicians told us that they were taught, when training in the management of SARS, that P2 masks give superior protection.

Nearly a century ago, the recommendation for use of the gauze mask came with a warning that it should not provide the wearer with an “unwarranted feeling of security”, but should be considered as one part of an infection control process.12 Research evidence7 suggests that a surgical mask plus face shield, rather than a P2 mask, is sufficient protection against pandemic influenza, but will this measure serve as sufficient protection for a health care system that needs healthy HCWs to manage a pandemic-sized caseload? Conventional evidence-based principles give equal importance to research evidence, economic cost, and needs and values of stakeholders. When research is inconclusive, principles should be prioritised. WHO guidelines1 prioritise research evidence within the limitations of resources of different socioeconomic settings; CDC guidelines2 prioritise the needs and values of clinicians within the limitations of research evidence. The new infection control algorithms8 compensate for weak research evidence — they remove the “over-importance” given to masks by prioritising needs and values of HCWs, while presenting masks as just one component of an infection control strategy. Without seminal research evidence of influenza being transmitted exclusively by airborne transmission, and in the absence of studies testing for superior protection of P2 masks, it would be prudent for health care executives to view providing P2 masks to HCWs, not as an additional cost, but as an additional investment in the continuity of health service provision — and in the full knowledge that, on today’s evidence, P2 masks provide a level of protection equivalent to that of surgical masks.