Intended for healthcare professionals


High reliability in health care

BMJ 2010; 340 doi: (Published 19 January 2010) Cite this as: BMJ 2010;340:c84
  1. Charles Vincent, professor of clinical safety research,
  2. Jonathan Benn, lecturer in patient safety and quality improvement ,
  3. George B Hanna, professor of surgical sciences
  1. 1Imperial Centre for Patient Safety and Service Quality, Imperial College of Science Technology and Medicine, Department of Biosurgery and Technology, St Mary’s Hospital, London W2 1NY
  1. c.vincent{at}

    Examples from other industries should be informative, not prescriptive

    High reliability organisations achieve high levels of safety and performance in the face of considerable hazards and operational complexity.1 The original studies by the Berkeley Group, which looked at nuclear power, naval aviation, and air traffic control, have been influential and inspired much comment and interpretation. High reliability organisations are frequently referenced as models to which health care should aspire, particularly because the environments and challenges are similar.1 2 3

    Meeting the challenges of high reliability operations requires accountability, strong basic procedures, multiple procedural checks, and continual communication between operators.2 3 For example, during critical operations on naval carriers, multiple checks and observations by different people ensure that dangerous conditions are detected rapidly. “Buddy” systems, in which individuals monitor each other’s performance, are used to guard against unsafe actions. High reliability organisations also engage in varied training and simulation activities for a broad range of operational scenarios (such as deck fires on aircraft carriers) to prepare for crises and foster a flexible problem solving approach. Although there is a strong emphasis on protocol and procedure, staff of all levels of seniority have the authority to interrupt operations. For example, the landing signal officer on a carrier, who has a relatively junior role, has the authority to abort a landing attempt if safety is compromised.3

    The original descriptive studies of these organisations are inspiring, but we need to be cautious about extrapolating their conclusions to health care. Firstly, while studies highlight a wide range of characteristics said to be important to reliable performance, it is not clear which are the most important. Secondly, however insightful subsequent authors have been, they have compounded these problems by selectively looking at the aspects they considered most important and have also offered new interpretations and terminology.4 5 6 7 The range of alleged high reliability concepts is now enormous. Thirdly, theoretical abstractions abound, but few empirical studies have been done since those of the original Berkeley Group. Fourthly, the field has remained resolutely descriptive with few attempts to measure the characteristics of high reliability organisations or relate them to substantive safety outcomes.

    A particular worry is that health care has been selective about the lessons of high reliability organisations and has neglected the role of basic procedures. Many of the organisations studied are solely military or include military personnel, which brings an acceptance and adherence to routines and procedures. In contrast, much of the literature in health care has focused almost exclusively on the response to the hazardous and unexpected, and it has neglected the solid foundations of training, procedure and standardisation, shared discipline, and commitment to working as a team.8

    Health care is sometimes contrasted unfavourably with high reliability organisations, although it is just as demanding and complex an environment.9 The problem is not that health care is not reliable or resilient at all, but that huge variability exists within teams, within organisations, and across the system. The hospital that contains centres of excellence may have other units in which outcomes are poor or even dangerous. Many instances of high reliability exist in health care, in the everyday behaviour of clinical staff and at a unit and hospital level. In surgery, for example, a small number of units around the world have achieved almost zero mortality during operations for gastric cancer and other conditions that are difficult to treat.10

    The study of high reliability organisations has encouraged optimism about what can be achieved in health care and pointed to a much more proactive approach to safety than the more familiar reactive learning from incidents and adverse events.11 The challenge now is to take this diffuse set of ideas, refine them, consider their implications, and test them in a healthcare context. Rather than health care being treated as the poor relation in high reliability terms, it should be seen as an ideal environment for testing and implementing some of these concepts.12 Importing systems from high reliability organisations into clinical practice without considering how the task fits into the unique characteristics of the healthcare system is likely to be unproductive and potentially destabilising.

    One solution is to look more carefully and systematically at high performance within health care, drawing on other industries for ideas and inspiration, but not as beacons of reliability that we should simply emulate. Individuals, teams, and organisations in health care that already embrace this perspective provide a means of understanding the nature of reliability and resilience, and they can be an inspiration to others.


    Cite this as: BMJ 2010;340:c84


    • Competing interests: All authors have completed the Unified Competing Interest form at (available on request from the corresponding author) and declare: (1) No financial support for the submitted work from anyone other than their employer; (2) No financial relationships with commercial entities that might have an interest in the submitted work; (3) No spouses, partners, or children with relationships with commercial entities that might have an interest in the submitted work; (4) CV undertakes occasional paid consultancy work on safety related issues but JB and GBH have no non-financial interests that may be relevant to the submitted work.

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