Article Text

Download PDFPDF

Why even good physicians do not wash their hands
  1. Donald A Redelmeier1,
  2. Eldar Shafir2
  1. 1Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
  2. 2Department of Public Policy, Princeton, Princeton, New Jersey, USA
  1. Correspondence to Dr Donald A Redelmeier, Sunnybrook Health Sciences Centre, G-wing Rm 151, 2075 Bayview Ave, Toronto, Ontario, Canada M4N 1J5; dar{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Lapses in hand hygiene are a persistent problem that has resisted a simple solution

  • Behavioural decision science provides a framework for understanding some of the lapses

  • Affective factors include a lack of positive reinforcement and a missing sense of certainty

  • Cognitive factors include recurrent monotony, divided attention and faulty memory

  • Social factors include insufficient prestige with inadequate enforcement of norms

  • An awareness of behavioural factors helps explain past failures to improve hand hygiene

  • Behavioural insights may lead to new technologies and to more effective solutions


Hospital-acquired infections contribute to an estimated 1.4 million deaths worldwide, including about 100 000 annually at a cost of $30B in North America alone.1 Inadequate hand hygiene remains a frequent and modifiable contributing factor, as established from hospital outbreaks of methicillin-resistant Staphylococcus aureus, vancomycin-resistant Enterococcus, and other hospital-acquired infections.2 Much attention has been devoted to promoting greater hand-washing through persuasion, education or admonishment.3 ,4 The purpose of this viewpoint is to highlight the countervailing behavioural factors that help explain the ongoing shortfalls of hand-washing among well-intentioned and hard-working healthcare providers.

Physicians are often singled-out as a group with low rates of hand hygiene compliance. A classical study, for example, found that physicians had about half the rate of compliance when compared with nurses (30% vs 52%, p<0.001).5 One interpretation is that physicians are arrogant, sloppy, recalcitrant, incorrigible, inept or oblivious. An alternate interpretation is that individual practitioners and the surrounding medical culture consider hand hygiene compliance unimportant. Here, we explore a different account for why the problem might not be solved by changing the personalities or motivations of physicians. Specifically, we suggest that the circumstances surrounding physicians in hospitals are rife with behavioural factors conducive to failures of hand hygiene.

Perhaps the most fundamental principle discovered by behavioural research is that faulty decisions are often not caused by carelessness or misunderstanding. Instead, well-intentioned actions can fail to materialise because of misplanning, misremembering or a limited attention span encumbered by other concerns.6 The failure to turn intention into action means that outcomes are inadvertent rather than deliberate. A second insight in behavioural research is that the first principle is easily ignored. Instead, a standard impulse is to interpret suboptimal behaviour as a personal failure that requires explanations, threats, incentives or critiques.7 Hand washing, we propose, is a task where familiar behavioural factors conspire to undercut reliable patient care in hospitals by physicians who mean well, fail through natural limitations and are resented for it afterward.

Affective factors

Bacteria are invisible to the unaided eye and imperceptible to touch. Thus, physicians have no easy way to determine whether their hands are contaminated or whether their hand-washing is sufficient.8 Throughout, clean hands are rarely distinguished by patients, families, colleagues, superiors or the physicians themselves. Hand-washing is also sometimes awkward because it requires both hands to be empty and is clumsy for a physician carrying medical equipment or a hospital chart.9 The net result is that a good habit receives little reinforcement, involves recurring inconvenience and produces little feeling of accomplishment. Indeed, hand-washing feels like an empty routine if the patient subsequently dies from cancer, uraemia or another cause unrelated to infection.

Preventive acts mostly feel comforting when they offer a sense of certainty. For example, a perfect vaccine against one strain of influenza (offering no benefit otherwise) seems more appealing than partial protection against all pneumonia (despite the latter's greater net benefit).10 Hand-washing by physicians raises similar probabilistic vagaries due to the uncertain chain of associations that connect microbial load with pathogen transmission, patient colonisation and clinical infection.11 The fear of making a mistake can also influence professional behaviour because errors cause regret and regret aversion leads to diligence. However, the indeterminacy of the act of hand-washing and its distant connection to outcomes means that a failure of hand-washing does not raise the angst of a possible life-threatening mistake or feel like a clear source of culpability.

Cognitive factors

Hand-washing is never at the forefront of a physician's attention because it does not directly relieve pain, dyspnoea or other form of patient suffering. The high-speed large-stakes culture of hospitals may further lessen the salience of hand hygiene when compared with other priorities for patients in distress. In addition, a physician's recent history of hand-washing can be difficult to self-monitor because the memory relies on differential activation of recent and distant cues.12 Memories of closely repeated activities are hard to distinguish one from the other; moreover, hand-washing is a monotonous and frequently repeated procedure. Consequently, honest mistakes can arise when hand-washing for an earlier patient is misattributed to the current case. Honest misplanning can also happen when a psychiatrist or other busy physician intends to speak with a patient but not touch anything.

Hand-washing, furthermore, is often timed inconveniently since it is generally sequenced at room entrance—exactly when the physician's focus is on a new patient encounter. Linking hand-washing to room entrance would seem like a natural and effective prompt (as well as a feasible method for surveillance). Yet, the link is unfortunate since it demands a physician's attention exactly when his or her mind is absorbing a new, sometimes surprising and often demanding situation. The physician's full focus on the patient paradoxically deflects attention from ancillary activities such as hand-washing that might have been intended just moments earlier. Precisely because it is so simple (unlike the elaborate protocols for entering a silicon chip manufacturing clean room), hand-washing is typically budgeted no time nor allocated any planning.

Social factors

Hand-washing is not a glamorous activity that saves a patient's life. It garners no prestige in mainstream media and it is not a selection criterion for medical school admission. Physicians might pursue reputations for having great acumen, technical skill, bedside manner—but they are unlikely to pride themselves for a hand hygiene award. Surveillance by auditors is intended to motivate, but typically conveys only blameful and embarrassing negative feedback rather than constructive unambiguous positive encouragement.13 Empowering patients to check clinicians is another strategy to mitigate slips, yet this approach has mixed success and is unlikely to be effective for patients who are delirious, disabled or simply reluctant to question their physician.14

Social norms are fragile, require ongoing enforcement and quickly fade with repeated violations or when individuals perceive contradictory behaviours in peers. Young physicians, in particular, are especially prone to copy the behaviour of their superiors for both hand-washing and a lack of hand-washing.15 Powerful norms, such as those around vehicle speeds on highways or hard-hats inside construction sites, rely on easy detection and continuous enforcement for group adherence.16 Conformity in hand-washing, in contrast, is challenging since it is difficult to monitor and the violations are largely invisible. Lapses can also occur at arbitrary moments during patient care, such as when shaking hands or operating colonised equipment. When faulty adherence is common, errant acts are construed as the norm.


An awareness of behavioural factors has the potential to help improve programmes designed to increase hand-washing inside hospitals. Affective salience might be strengthened by vivid images relayed as individual feedback (eg, photographs showing bacterial growth from a physician's own hands). Cognitive load could be eased by strategically positioning hand-washing stations where individuals are idle (eg, elevators). Social norms may be reinforced through publicity campaigns encouraging hand hygiene by everyone in hospitals, including highly visible hand-washing by patients themselves. Combined strategies are also possible such as promoting a culture where all—nurses, doctors, patients—are encouraged to raise the issue of hand-washing without seeming confrontational, much like campaigns surrounding alcohol where ‘friends don’t let friends drive drunk’.17

The tendency to blame failures of hand-washing on insufficient motivation may sometimes be appropriate but may often be mistaken. A hypothetical physician with a single-minded dedication to hand-washing, of course, might avoid many of the behavioural factors that lead to hand hygiene failures. However, the affective, cognitive and social factors that interfere with intentions are regularly manifested among physicians who fall short in their genuine personal goals to exercise more, lose weight or save for retirement.18 Medical practice, moreover, requires physicians to think about patient suffering in multiple complex situations and people simply do not have an infinite supply of motivation. In many cases where behavioural factors interfere with sincere intentions, behavioural interventions can lead to better outcomes.19

Behavioural insights also help clarify why physicians sometimes have lower rates of hand-washing than other healthcare workers.20 The integrity of hand hygiene audits can be flawed and the fallibilities may affect different staff unequally.21 The rates of adherence among nurses, for example, may be overestimated due to a Hawthone effect, where easily recognised auditors bring about an illusory improvement.22 ,23 Social workers and other allied professionals, for their part, often enter patient rooms for regular visits rather than during emergencies.24 Food-handlers or administrative clerks stay away from patients and are typically required to wash their hands no more than once a shift. These groups of hospital staff do not experience the same confluence of behavioural factors that typically characterise active physicians.

Behavioural factors are sometimes so engrained that the greatest progress is achieved through an innovation in technology. Consider, for example, a soap containing a fast fading (few minutes) colour tint that is non-toxic, non-allergenic, fabric-safe and marketed exclusively for professional settings. Using such soap would immediately distinguish hands that were recently washed, thereby allowing easy visible confirmation (figure 1). Entering a patient's room without the right coloured hands would be akin to smoking in hospital—conspicuous, counter-normative and unacceptable. Fast fading colour soap would help with salience, memory, monitoring, gratification and promoting a readily verifiable norm. Together, such technology might address multiple behavioural contributors to why even good physicians do not wash their hands.

Figure 1

Example of fast fading colour soap. Three photographs depicting a hypothetical fast fading colour soap, intended to serve as a behavioural intervention to promote more hand-washing in hospitals. Images show dispensing of soap, resulting colour on washed hands, and subsequent fading after 5 min. Soap dries rapidly in a stain-free manner, whereas antiseptic effect could extend beyond subsequent colour fading. Ultimately, soap might foster a new social norm with hand-washing behaviours that are manageable, visible, identifiable, verifiable and salient.

Behavioural pitfalls, of course, do not justify poor hand hygiene among physicians any more than visual illusions justify flight errors in pilots. As with pilots, an improved understanding is intended to avoid such pitfalls. And while an understanding of behavioural factors is no panacea, insufficient awareness of human fallibility may explain persistent failures of hand hygiene and the disappointments of programmes to address these failures. Such fallibilities are a predictable outcome of human behaviour, yet they are rarely discussed in medical guidelines or self-report interviews.25 Insights into behavioural factors might help explain the relative ineffectiveness of past campaigns for sustained hand hygiene and the possible promise of new interventions. We cannot alter human psychology, but we might better understand behaviour to obtain improved patient outcomes.



  • Contributors Both authors contributed to the conception of the work, interpretation of data, drafting the manuscript, revising the content critically, approval of the version submitted and ongoing accountability for all aspects of the work including questions of accuracy or integrity.

  • Funding Division of Behavioral and Cognitive Sciences, Canadian Institute for Advanced Research, BrightFocus Foundation, Canada Research Chair in Medical Decision Science, Alfred P. Sloan Foundation.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.