Doctor, do you have a moment? National Hand Hygiene Initiative compliance in Australian hospitals

To the Editor: We read with interest the recent contribution by Azim and McLaws regarding hand hygiene in Australian hospitals.1 However, we have concerns regarding the methodology and interpretation of this study.

First, the authors conclude that public reporting of a national unadjusted average hand hygiene compliance rate “conceals” true performance.1 In fact, Australia is unique worldwide in the extent to which hand hygiene data are made available to all stakeholders.2 In addition to the national average, Hand Hygiene Australia publicly reports national data each audit period, stratified by profession, indication, profession plus indication, ward type and hospital peer group.3 Furthermore, compliance at each hospital is available (http://www.myhospitals.gov.au). Importantly, institutional, regional and jurisdictional coordinators have access to detailed real-time results.

Second, the authors do not acknowledge several major limitations in their attempt to quantify any association between hand hygiene and Staphylococcus aureus bloodstream infection:

• cross-sectional studies have weak capacity to explore causal relationships;
• the authors did not account for potential confounders, including antimicrobial use and colonisation pressure;
• using aggregate organisation-level data to predict patient-level outcomes renders the analysis vulnerable to the ecological fallacy, by which real patient-level associations are missed;4 and
• no information is provided regarding sample size estimation or study power.

Third, their discussion regarding the estimated cost of auditing hand hygiene practice extrapolates nationwide the unpublished and unvalidated input parameters from a single hospital and uses four audit periods to estimate annual cost, whereas there are actually three. Such inexact cost estimates risk doing more harm than good to rational dialogue regarding health policy decisions.

Finally, we agree that medical hand hygiene compliance should improve further. However, the authors do not acknowledge several important advances since the start of the National Hand Hygiene Initiative. Medical hand hygiene compliance has improved markedly (from 46% [8 252/17 830] in 2009 to 68% [51 018/75 105] in 2014);3 undergraduate and postgraduate education regarding use of alcohol-based hand rubs has finally become a routine component of most curricula; and important organisations such as the Australian Medical Association and Royal Australasian College of Surgeons now have active education programs. Similarly, we agree that Moments 1 and 2 are important to highlight, since they routinely rate 8%–10% lower for compliance nationally than the “after” moments (Moments 3, 4 and 5);3 but these latter moments are also important if we are to avoid contamination of health care workers' hands and, subsequently, the hospital environment.5