Donaldson, Liam J. MD; Noble, Douglas J. MPH
Correspondence: Douglas J. Noble, MPH, World Health Organization, Patient Safety, Avenue Appia 20, CH-1211 Geneva 27, Switzerland (e-mail: firstname.lastname@example.org; email@example.com).
The authors alone are responsible for the views expressed in this publication, and they do not necessarily represent the decisions, policy, or views of the World Health Organization.
Imagine that a doctor or nurse were asked to perform a seemingly simple task. They would be invited to stand in a doorway opening onto a courtyard. They would see that the courtyard was paved with cobblestones having been made irregular in shape with the passage of time. It being in the depth of winter, they would note that the cobblestones were covered in thick frost. They would be asked to walk across the courtyard-a distance of some 15 m to a door on the other side. They would be asked to wear a pair of brand new shoes with shiny leather soles. Their task? To carry across the courtyard a large glass flask filled with concentrated sulfuric acid.
It is likely that the task would be accompanied by a degree of apprehension. More importantly, each individual who carried it out would have an awareness of the risks involved of dropping and losing the flask's contents or of having it fall and splash on to them with its dangerous and potentially disfiguring contents. Every step would be taken with the greatest of care. The whole task, done by doctors or nurses, would be executed with a high degree of vigilance.
When risks present a direct threat to our own personal safety, caution and attentiveness are paramount. Few of us in health care would undertake such a task carelessly or thoughtlessly. We would first carefully assess the risks to ourselves. Do we apply this same cautious risk-aware approach when we deal with our patients' care journeys? Would we apply the same standards, protecting them as we would ourselves? Would our courtyard doctor or nurse who perhaps treats thousands of patients a year even have the same heightened awareness of the risks of their clinical role?
In many areas of medicine, there are defined treatment processes that are followed. Examples include resuscitation protocols, procedures such as suprapubic catheterization, and medical treatment of hypertension. Good work has been done through traditional hazard analysis approaches and other techniques that detail risks in areas such as medication use. Yet, this is not universally true, and even when work has been carried out, it is often not followed. Risks are associated with all aspects of clinical care, and patients often come to harm because these risks are not anticipated, recognized, or understood.
Whereas health care professions are familiar with known complications, such as infection or bleeding after surgery, less attention has been given to systematically mapping the potential for risk at each stage of a process of care. This is in marked contrast to other high-risk industries.1
In Formula 1 racing, pit stop checks and changes have to be performed under extreme stress and time pressure-consequences of error are life-threatening. When compared with the clinical process of care relating to handover of critical information between operating theaters and intensive care units after pediatric congenital heart surgery, several major differences were noted.2 In Formula 1, the sequence of events was carefully coordinated and sequential with detailed risk analysis at each stage. By contrast, in the clinical process, events are not regularly in order and risk recognition is informal and often not given due attention. Reworking clinical protocols based on the Formula 1 model led to a measurable reduction in error.2
This approach is not common in health care. Yet, there are some examples of early systematic risk profiling. The error of delivering vincristine sulfate via an intrathecal and not intravenous route is a devastating error for all involved. Careful work in the United Kingdom identified more than 40 errors in the process of care leading to the final error.3 Further work made specific recommendations to address risk systematically at each stage,4 and this has now been implemented as policy for almost a decade. There are other examples but not usually before major accidents. Frequently, risk analysis is only performed after the catastrophic event.
In the United Kingdom, several types of errors repeated themselves in the delivery of radiotherapy during several years from 1982 to 2001. These involved mix-ups of computer readings that led to repetition of unrecognized harm that led to many deaths.5 An analysis of these serious errors in radiotherapy delivery led to the identification of a core process of care common to most treatment programs in the United Kingdom.6
This recognized the clustering of harm from incidents involving patient deaths and harm from overdosage and underdosage of radiation at only 2 points in a 7-step core process of care. This stimulated interest in establishing an internationally agreed process of care to which risks could be mapped and frequency could be determined to guide risk reduction efforts. This, in turn, led to the establishment of an international multiprofessional group hosted by the World Health Organization Patient Safety Programme.7 In this setting, the work was extended to examine other incidents worldwide. This enabled refinement of the conceptualization of the process of care and the way in which different risks mapped on to it. From this, the concept of a formal risk profile for radiotherapy was developed. It is the first health care risk profile to be published by the World Health Organization.8
The problem being addressed by the World Health Organization working group was the lack of a standardized process of care in radiotherapy to which risks could be systematically mapped. To help the group focus their efforts, a systematic review of all errors in radiotherapy delivery was completed to establish a burden of harm and to provide qualitative data on the nature of incidents.9 This revealed that more than 30 years, since 1976, 7741 incidents and near misses occurred in radiotherapy delivery worldwide. Although the research included gray literature, it was agreed likely that the true number due to underreporting was far higher. Although denominator data were not available internationally, such data were available in the United Kingdom. It has been estimated that the rate of occurrence of incidents in radiotherapy was 40 per 100,000 courses.10
The World Health Organization working group agreed by expert consensus on a 10-stage process of care (Fig. 1). This was a critical and time-consuming process. The main challenge was ensuring it was comprehensive, yet simple enough to map risks. It was the opinion of many group members that often these types of process-mapping exercises led to vastly complex nonlinear processes that became problematic to make operational on a day-to-day clinical basis.
At each stage of the process, risks were identified and prioritized. The planning and information transfer stages revealed the highest frequency of incidents. Dosage errors were a repeated theme in the analysis of all incidents. These types of incidents have potentially serious consequences for patients: overdosing leads to tissue damage and can induce cancer; underdosing leads to inadequate tumor control. Although human factors were important in causality, the human-machine interface was a recurring theme.
Incidents often ran unnoticed for many years. System safeguards such as independent checking go some way to rectify these problems. The top 3 risk-reduction strategies emerging from the risk profile were planning protocol checklists, independent checking, and competency certification.
Several challenges remain, but the initial risk profile has helped guide future efforts. There has been a strong interest in in vivo dosimetry (a measurement of radiation after the first dose in a course is delivered to a patient). If error margins are greater than 5% to 10%, subsequent doses can be altered. In addition, it offers the possibility of picking up significant harm at an early stage.
Risk profiling as a method of mapping processes and identifying points of risk could be extended to many other aspects of clinical medicine where a linear sequence of events occurs. Other examples include administration of medicines or intrapartum care. Some processes in clinical practice may not be sufficiently linear, with diverging clinical options leading to diverging risk possibilities, making the risk identification process more complex. For example, areas such as medication reconciliation in geriatric patients and surgical procedures, where findings dictate treatment, present multiple potential pathways, each with its own risks and each step potentially changing the risks at another. Although this may seem daunting, where the daily operational knowledge exists a risk profile is possible. Research is needed to develop this approach further, but especially when processes of care are linear, it should be possible to systematically define risk and produce usable tools that map out the highest risk areas and drive risk-reduction efforts.
Often in patient safety, we have lacked the evidence to prove that our approaches are beneficial and result in a measurable reduction in harm. This is slowly changing. Proving that traditional hazard analysis and risk profiling can make a difference is at an early stage. More research is needed not only in thinking through the process of care but also in showing that the heightened awareness of risk improves patient outcomes. Implementing such an approach to define risk requires more than a distant external risk expert to map the process of care and document the risks. Health care professionals need to be actively engaged.
Everyday, millions of health care professionals have in their hands not a flask of sulfuric acid but a process of care. Their hands regularly hold the lives of others, not themselves, in the balance. A heightened awareness on their part of the hidden dangers for the patient and for their standard of professional practice is urgently needed. Health care can be made safer but only if its risks are fully understood and anticipated and not just after a tragedy has occurred.
1. Lee RC, Kell KL, Newcomb C, et al. HTA Initiative #15 Quantitative Approaches to Patient Safety: Research in Risk Analysis and Risk Management as Applied to Radiotherapy. Alberta, Canada: Alberta Heritage Foundation for Medical Research; 2004.
2. Catchpole KR, De Leval MR, McEwan A, et al. Patient handover from surgery to intensive care: using Formula 1 pit-stop and aviation models to improve safety and quality. Paediatr Anaesth. 2007;17:470-478.
3. Toft B. External Inquiry into the Adverse Incident That Occurred at Queen's Medical Centre, Nottingham, 4th January 2001. London, UK: UK Department of Health; 2001.
4. Woods K. The Prevention of Intrathecal Medication Errors: A Report to the Chief Medical Officer. London, UK: UK Department of Health; 2001.
5. Donaldson LJ. Reducing harm from radiotherapy. BMJ. 2007;334:272.
6. Chief Medical Officer: On the State of Public Health: Annual Report of the Chief Medical Officer 2006. London, UK: UK Department of Health; 2007:34-39.
9. Shafiq J, Barton M, Noble D, et al. An international review of patient safety measures in radiotherapy practice. Radiother Oncol. 2009;92:15-21.
10. Towards safer radiotherapy. Royal College of Radiologists. 2008. ISBN: 978-1-905034-25-3.
© 2010 Lippincott Williams & Wilkins, Inc.