This Article Abstract Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Williams, M V Search for Related Content PubMed PubMed Citation Articles by Williams, M V

Improving patient safety in radiotherapy by learning from near misses, incidents and errors

Faculty of Clinical Oncology, The Royal College of Radiologists, 38 Portland Place, London W1B 1JQ, UK

Correspondence: Dr Michael Williams, Faculty of Clinical Oncology, The Royal College of Radiologists, 38 Portland Place, London W1B 1JQ, UK. E-mail: michael-williams{at}rcr.ac.nhs.uk

	Abstract

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Radiotherapy incidents involving a major overdose such as that which affected a patient in Glasgow in 2006 are rare. The publicity surrounding this patient's treatment and the subsequent publication of the enquiry by the Scottish Executive have led to a re-evaluation of procedures in many departments. However, other incidents and near misses that might also generate learning are often surrounded by obsessive secrecy. With the passage of time, even those incidents that have been subject to a public enquiry are lost from view. Indeed, the report on the incident in Glasgow draws attention to strong parallels with that in North Staffordshire, the report of which is not freely available despite being in the public domain. A web-based system to archive and make available previously published reports should be relatively simple to establish. A greater challenge is to achieve open reporting of near misses, incidents and errors. The key elements would be the effective use of keywords, a system of classification and a searchable anonymized database with free access. There should be a well designed system for analysis, response and feedback. This would ensure the dissemination of learning. The development of a more open culture for reports under the Ionising Radiation (Medical Exposure) Regulations (IR(ME)R) is essential: at the very least, their main findings and recommendations should be routinely published. These changes should help us to achieve greater safety for our patients.

	Introduction

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Radiotherapy is seen as a mysterious procedure by patients and the public alike. Treatments vary in their intent, and in the number and size of fractions administered. Because radiation is involved and because it can neither be seen nor felt, there is an air of mystery that adds to the perception of danger.

Staff recognize radiotherapy as a high risk procedure because the process is complex and involves a large number of different steps and staff groups: each handover or data transfer offers the opportunity for miscommunication and error. The risks are compounded because radiotherapy is usually a repetitive procedure. Treatment is administered as a series of fractions and error can be perpetuated [1, 3–5]. Alternatively, a dosimetry or calculation error can affect a series of patients [2, 6–9]. Despite this, radiotherapy has a good safety record and patient injury caused by error is rare.

Several international organizations have reviewed major radiotherapy incidents [8, 9] and one might conclude that it is now difficult to invent a new error. Indeed, recently reported errors from Leeds [3–5] and Glasgow [1] were repetitions of previous mistakes, namely omission of a wedge compensator during treatment [8] incorporated when a computer treatment planning system was introduced [2]. However, it is important to understand the fine detail of the mechanisms by which a latent error was established [10, 11]: a series of mistakes then resulted in the delivery of an incorrect treatment. As technology develops, there will be new possibilities for mistakes, but the root cause usually lies not in the individual but in organizational procedures, structure and governance [10–15]. These details and their continual review in the light of incidents and non-conformances are the key to safety [10, 14, 16].

	Open reporting and the "no blame" culture

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Open reporting of near misses, incidents and errors is established as a key safety step in other high risk industries [12, 14]. Aviation has seen a dramatic fall in the incidence of accidents over the last 30 years, and this has been attributed to the safety and reporting culture [14, 16]. Anaesthesia has made safety a priority and has replicated the achievements of aviation, significantly reducing the risk of death from accidents [17]. Near misses do not have legal implications and thus there are fewer barriers to reporting. In addition, near misses are 3–300 times more common than adverse events and offer learning opportunities without patient injury [16]. However, medicine has lagged behind other industries in establishing a safety culture [12, 14, 16]. Reducing risk requires a determination to achieve improvement, open reporting and a "no blame" culture [14, 16–20]. It has been observed that all reporting is in some sense voluntary and that, to be effective, it needs to be linked to a system for analysis, response and feedback [18].

For the NHS, these principles were enunciated in the two publications "Organisation with a memory" [12] and its successor "Building a safer NHS for patients" [13], which led to the establishment of the National Patient Safety Agency (NPSA). The difficulty for radiotherapy is that we are far removed from the ideal solution that these reports describe. We are still locked into a "past" where attention focuses on individual error, leading to scapegoating and fear of reprisals; adverse events are regarded as "one-offs", subject to a short-term fix; and other staff do not hear the outcome of investigations that are considered to be only relevant to the team involved [12]. We are still a long way from the "future" where individuals are held accountable for their actions within a system approach to hazards, blame-free reporting is encouraged with regular feedback, the potential for replication is recognized and lessons are disseminated to others [12].

There is still a strong blame culture: indeed, the report on the Glasgow incident devoted considerable effort to apportioning blame to different individuals [1]; although managerial failures were documented, there was little emphasis on overall responsibility at a high level [1]. It has been argued that leaders should take personal responsibility for the safety of the processes and systems in which individuals work if sustained improvement is to be achieved [20].

	Internal reporting of radiotherapy incidents and the NPSA

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Radiotherapy departments in the UK are required to have a quality management system [21] and, within this, quality assurance programmes specific to the various processes are required, implicitly or explicitly, by UK legislation. It is usual to seek accreditation by an outside body. Quality assurance systems require internal reporting and audit of incidents and non-conformances: such events are reviewed at weekly or monthly meetings. Trends can be identified and remedial action put in place. In England, these reports are fed into NHS Trust risk reporting systems and then forwarded to the NPSA, where they are anonymized and fed into a database held on a separate server [22]. The difficulty with this laudable system is that it contains some 50 000 oncology incidents, half of which are trips, slips and falls. These are not usually informative about the radiotherapy process and, although important to the patient and the hospital, are essentially a byproduct of collecting data through risk management systems. The database is not searchable and there is no use of keywords. It may be difficult to deduce whether a dose error refers to a chemotherapy or a radiotherapy event. This may explain why the NPSA stated that it possessed no data on any adverse incidents that could have been of assistance to the panel reviewing the Leeds incident [4]. Only a single clinical oncologist has access to the NPSA database, and feedback mechanisms have yet to be established. If it is to be useful in the future, these limitations need to be addressed: as it is anonymized, there is no reason why it should not be open to professional or even public scrutiny.

	Reports under the Ionising Radiation (Medical Exposure) Regulations

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
In the UK, radiotherapy errors that result in a dose greater than intended are reportable to the inspectorate under the Ionising Radiation (Medical Exposure) Regulations (IR(ME)R) [23]. The regulations were laid before parliament to fulfil our obligations under a European Council directive [24]. This gave member states the responsibility for implementing a system of inspection to enforce the provisions of the directive. In particular, there is a requirement that:

Member states shall ensure that all reasonable steps to reduce the probability and the magnitude of accidental or unintended dose of patients from radiological practices are taken, economic and social factors being taken into account.

The main emphasis in accident prevention should be on the equipment and procedures in radiotherapy, but some attention should be paid to accidents with diagnostic equipment. ((Article 11) [24])

One can draw two conclusions about the implementation of this directive through the IR(ME)R in the UK. First, there seems to be an inappropriate focus on very small misexposures in diagnostic practice, which clearly was not the intention of those drafting the directive. Raising the reporting and external investigation threshold for incidents to encompass only those carrying a risk of carcinogenesis of greater than 1 in 10 000 has been proposed [25]. Second, the wording "accidental or unintended doses" includes both over- and underdoses, but only the former are reportable in the UK, despite the major detriment that can follow an underdose [2, 7]. Interestingly, despite the legal position, the new Healthcare Commission form permits the reporting of underdoses.

Enquiries under the Freedom of Information Act have elicited the fact that over the last 6 years there have been 211 radiotherapy incidents reported under the IR(ME)R [26, 27]. The data do not include underdoses (which are not reportable under these regulations) or overdoses that were recoverable by revising the radiotherapy plan or did not breach the reporting threshold. There is no public or professional access to any of these data other than under the Freedom of Information Act.

Investigations under the IR(ME)R are conducted under criminal law and usually under caution [1, 23]. The results of the IR(ME)R investigations are therefore confidential and are not disseminated unless unusual steps are taken. This is not a specified feature of the core European legislation [24] as the details of the legislation, regulation and administrative provisions are a matter for each country (Article 14(1) of [24]). In Denmark, things are very different: the Danish Act on Patient Safety came into force on 1 January 2004 [28]. It obliges frontline personnel to report adverse events to a national system. The purpose is to learn, not punish, and the act contains a paragraph protecting staff from sanctions: an individual who reports an adverse event cannot as a result of that report be subjected to investigation or disciplinary action by the employer, the Board of Health, or the Court of Justice [28]. Similar protection is offered in Australia and New Zealand, and in other industries. In Denmark, hospital owners are obliged to act on the reports, and the national Board of Health is obliged to communicate the learning nationally. The punitive system adopted by the UK was a matter of choice, made doubly strange by the earlier publication of "Organisation with a memory" [12].

	Reports of inquiries into major incidents

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Major incidents in which a patient comes to harm or where there is a system failure, such that a series of patients are exposed to risk, are investigated to determine what went wrong. But it is also essential that the results be made public and the lessons disseminated. The Scottish Executive is to be congratulated for publishing the report on the investigation into the unintended overdose in Glasgow [1]. This action had a huge impact on the radiotherapy community in the UK and has certainly banished complacency. In contrast, in 2004 a radiotherapy incident occurred in Leeds whereby a patient received treatment to the breast without a wedge compensator in place for a series of 14 fractions [3, 4]. An alert was published by the Department of Health [5] and disseminated through the risk management system, rather than directly to radiotherapy departments. The document is striking for its brevity and it had little impact. This has been identified as a common failing of such alerts [29]. The only report that has been published deals with "involuntary automaticity", the psychological mechanism proposed by the authors to explain the repetitive error made by a series of staff members [3]. The issue of how the incorrect data were entered into the treating linear accelerator in the first place was not addressed at all. A full enquiry was commissioned by the Trust, but the report has been kept confidential. Access under the Freedom of Information Act reveals that it contains significant recommendations for national practice [27]; for example, on workload, training, working practices, manual data transfer, checking procedures and the replacement of outdated computer programs [4]. It is very disheartening that the Trust that commissioned it has not seen fit to publish the report. However, this behaviour is inkeeping with the findings of the Leeds report itself, which states that:

Very little information seems to be collected or publicly shared on radiotherapy adverse events either nationally or internationally. Indeed keeping such information confidential seems to take a higher priority than finding a way to use it to prevent similar accidents from taking place. [4]

There are a large number of other serious incidents that have been identified under the Freedom of Information Act and published both in the press [26] and on websites [30]. For the most part, these have not been fed back to other departments in an effective way.

Hospitals fear the public disclosure of reports that may damage their reputation [31]. However, the media usually learn about embarrassing accidents from other sources [31], and this occurred in the Leeds case when the error was reported the in local press [4]; the article is available on the internet [30]. Others have wrestled more effectively with the dilemma posed for a hospital by the public management of medical error [19]: at Sturdy Memorial Hospital mistakes in prostate biopsy reporting were discovered, patients were notified and a review of 6000 specimens was commissioned. There was little support from professional bodies and a flurry of media attention, but overall the experience was judged to have been positive and there was no impact on the hospital's workload or finances. Openness reaffirmed the hospital's reputation for putting patients first. Patients accepted the inevitability of human error and were impressed that something was being done about it [19].

	International reporting systems

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
Lessons learned from the investigation of therapy misadministration events have been reported by the United States Nuclear Regulatory Commission (NRC) [32]. They reviewed seven events, involving radioisotopes, investigated in 1991–92, and emphasised the importance of a comprehensive safety culture. The International Atomic Energy Agency (IAEA) has published a report on "Lessons learned from accidental exposures in radiotherapy", reviewing 92 anonymized radiotherapy incidents together with a classification of causes and recommendations on prevention [8]. The International Commission on Radiological Protection (ICRP) has published "Prevention of accidental exposures to patients undergoing radiation therapy", reviewing seven case histories of major large scale accidental exposures [9]. Their main recommendations were for quality assurance systems and a programme of in vivo dosimetry within departments [9]. A taxonomy of radiation incidents has been proposed to support risk analysis and organizational learning: 437 incidents from databases held by the NRC, the IAEA and the Radiation Oncology Safety Information System (ROSIS) were reviewed and classified into systemic and sporadic, and by the process domain in which they occurred [33].

The reporting of adverse events in healthcare has been reviewed by Leape [31]. He argues that to be effective it is necessary to mimic the aviation safety system so that reporting is simple, safe (for the reporter) and worthwhile. Successful reporting systems are non-punitive, confidential, independent, timely, responsive and subject to expert analysis [31]. Recommendations should focus on changes in systems, processes or products rather than on individual performance [31]. Both the IAEA and ICRP reports draw attention to the virtues of the open reporting systems in aviation [8, 9] and the desirability of anonymous reporting and dissemination to the wider community [8]. Nevertheless, there is no such comprehensive system in the UK or elsewhere. The nearest approximation is ROSIS, which was established in 2001 with pump-priming funds from the European Society for Therapeutic Radiology and Oncology (ESTRO). It is a voluntary reporting system that now holds on its website over 700 incidents from 37 departments across 19 countries [34]. The data are anonymized, made fully searchable and are in the public domain. Details of incidents and their causes are given in an abbreviated form, and it is not possible to analyse them in detail.

In response to reports of harm to patients caused by radiotherapy, a new initiative has been launched by the World Health Organization World Alliance for Patient Safety [35, 36]. There are two primary issues of concern. First, can a set of standardized safety interventions be developed to reduce harm to patients? Second, can lessons from errors be translated into international learning? A timely and worldwide response to safety incidents has been achieved in aviation but not in medicine [37]. One of the aims of the World Alliance for Patient Safety is that the learning from the inadvertent death of a patient in a hospital in one country could save the lives of many others around the world [37].

	Recommendations

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 
First, learning from the past should be consolidated by setting up a publicly available website to hold copies of the reports of previous radiotherapy incident enquiries. This should help to ensure that the learning from them is not forgotten. In the UK alone, there have been a number of public enquiries about radiotherapy incidents that affected a series of patients. These include those at Exeter [6] and North Staffordshire [2]. These are in the public domain, but the incident at Glasgow showed us that very similar errors can still occur [1].

Second, dissemination of learning from near misses and non-reportable incidents should be improved. This could be achieved by changing the way in which the NPSA database works and by providing effective analysis and feedback [18]. At present, radiotherapy incidents are submitted internally and then sent to a central server where they are anonymized [22]; no analysis has been published. Lack of feedback was one of several criticisms made by the National Audit Office [29], and the Department of Health has responded with a plan to remedy the deficiencies [38]. The NPSA system as it applies to radiotherapy could be improved if a classification of radiotherapy incidents could be agreed and if reports included:

• Agreed key words
  
• A description of how the error occurred
  
• Identification of how the error was detected
  
• Patient outcome and corrective action taken
  
• Lessons to be learnt
  
• Free text for further comments.
  

The Royal College of Radiologists has established a multidisciplinary working party whose remit includes these issues; it will also make recommendations on error prevention and detection. The report will be completed by the end of 2007 and should provide the basis for developing an open reporting system for radiotherapy incidents. We hope to work with the NPSA to achieve this objective. In addition, it is hoped that full analysis of these complex incidents could be assisted by support services provided by the Health Protection Agency, working with the professions involved in radiotherapy delivery to provide specialist expertise. Detailed investigation of selected near misses and other incidents that are not statutorily reportable would maximize learning. The results must then be linked to an effective system of dissemination that would reassure staff that reporting is worthwhile.

Third, the development of a more open culture for reports under the IR(ME)R is essential. Since 1 November 2006, the Healthcare Commission has been the responsible authority; this change may provide the opportunity for greater openness. The Scottish Executive has already shown that reports initiated under these regulations can be published. The Healthcare Commission's vision is "to make a difference to the delivery and quality of healthcare by inspecting, informing and improving". To fulfil their informing role, they should routinely publish the results of enquires under the IR(ME)R. This would be consistent with their policy on the reports of other investigations and would help to ensure that errors are not repeated elsewhere. This could be facilitated by establishing a website to host anonymized reports of enquiries, or at the very least their main findings and recommendations.

In conclusion, if these objectives could be achieved, then the UK could become a world leader in reporting and learning from radiotherapy incidents. This should improve the safety of our patients and would take us a step nearer to passing the "orange wire test", whereby a safety incident results in a rapid worldwide response as in aviation [37].

Received for publication January 13, 2007. Revision received February 26, 2007. Accepted for publication March 9, 2007.

	References

Top Abstract Introduction Open reporting and the... Internal reporting of... Reports under the Ionising... Reports of inquiries into... International reporting systems Recommendations References

 

1. Scottish Executive Publications. Report into unintended overexposure of Lisa Norris at Beatson Oncology Centre, Glasgow in January 2006. Report of an investigation by the inspector appointed by the Scottish Ministers under Ionising Radiation (Medical Exposures) Regulations - IRMER. October 2006. Available from: http://www.scotland.gov.uk/Publications/2006/10/27084909/0
2. Baldwin Report. Report of an investigation into the conduct of isocentric radiotherapy at the North Staffordshire Royal Infirmary between 1982 and 1991. Birmingham, West Midlands Health Authority. First report, August 1992. Second report, March 1994
3. Toft B, Mascie-Taylor H. Involuntary automaticity: a work system induced risk to safe healthcare. Health Serv Manage Res 2005;18:211–16.
4. Independent review of the circumstances surrounding a serious adverse incident that occurred in the Cookridge Hospital. [Redacted copy obtained under the Freedom of Information Act.]
5. Department of Health. DH Alert 4181. Safe delivery of radiotherapy treatment. 19 November 2004. Available from: http://www.dh.gov.uk/en/index.htm
6. Report of the Committee of Inquiry into the incident in the radiotherapy department. Exeter, UK: Exeter Health Authority, 1988
7. BBC News. Hospital admits cancer mistakes. Plymouth Hospital NHS Trust, 7 January 2002. Available from: http://news.bbc.co.uk/1/hi/england/1746774.stm
8. International Atomic Energy Agency. Lessons learned from accidental exposures in radiotherapy. Safety reports series no. 17. Vienna, Austria: International Atomic Energy Agency, 2000
9. Valentin J, editor. International Commission on Radiological Protection. Prevention of accidental exposures to patients undergoing radiation therapy. ICRP publication 86. Oxford, UK: Elsevier Science, 2001
10. Helmreich RL. On error management: lessons from aviation. Br Med J 2000;320:781–5.[Free Full Text]
11. Reason J. Human error: models and managements. Br Med J 2000;320:768–70.[Free Full Text]
12. Department of Health. An organisation with a memory. Report of an expert group on learning from adverse events in the NHS chaired by the Chief Medical Officer. London, UK: Stationery Office, 2000. Available from: http://www.dh.gov.uk/assetRoot/04/08/89/48/04088948.pdf
13. Department of Health. Building a safer NHS for patients – implementing an organisation with a memory. London, UK: Stationery Office, 2001. Available from: http://www.dh.gov.uk/assetRoot/04/05/80/94/04058094.pdf
14. Vincent C. Patient safety. London, UK: Elsevier, 2006
15. Williams MV. Radiotherapy near misses, incidents and errors: radiotherapy incident at Glasgow. Clin Oncol 2007;90:1–3.
16. Barach P, Small SD. Reporting and preventing medical mishaps: lessons from non-medical near miss reporting systems. Br Med J 2000;320:759–63.[Free Full Text]
17. Gaba M. Anaesthesiology as a model for patient safety in health care. Br Med J 2000;320:785–8.[Free Full Text]
18. Cohen MR. Why error reporting systems should be voluntary. They provide better information for reducing errors. Br Med J 2000;320:728–9.[Free Full Text]
19. Pietro DA, Shyavitz LJ, Smith RA, Auerbach BS. Detecting and reporting medical errors: why the dilemma? Br Med J 2000;320:794–6.[Free Full Text]
20. Reinertsen JL. Let's talk about error. Br Med J 2000;320:730[Free Full Text]
21. Bleehen Report. Quality assurance in radiotherapy: a quality management system for radiotherapy. PL/CMO (94).7. London, UK: Department of Health, 1994
22. National Patient Safety Agency. Building a memory: preventing harm, reducing risks and improving patient safety. The first report of the National Reporting and Learning System and the Patient Safety Observatory. NPSA, July 2005. Available from: http://www.npsa.nhs.uk/site/media/documents/1269_PSO_Report_FINAL.pdf
23. Department of Health, The Ionising Radiation (Medical Exposure) Regulations 2000. London, UK: Department of Health, 2001. Available from: http://www.opsi.gov.uk/si/si2000/20001059.htm
24. Council Directive 97/43/Euratom of 30 June 1997 on health protection of individuals against the dangers of ionizing radiation in relation to medical exposure, and repealing directive 84/466/Euratom. Available from: http://www.bnsa.bas.bg/eurolex/31997L0043_en.pdf
25. Clewer PR, Jackson PC. Reporting overexposures and unintended exposures in diagnostic procedures. Br J Radiol 2006;79:866–9.[Abstract/Free Full Text]
26. Lois Rogers. Over 200 hurt or killed by botched radiation. Sunday Times. London, UK: Times Newspapers Ltd. 30 April 2006
27. Tylko K, Blennerhassett M. How the NHS could better protect the safety of radiotherapy patients. Health Care Risk Rep 2006;12:18–19.
28. Act on patient safety in the Danish health care system. Act No. 429 of 10/6/2003 (current). Available from: http://www.patientsikkerhed.dk/admin/media/pdf/133907d0940e4d5f751852ec8f6b1795.pdf
29. National Audit Office. A safer place for patients: learning to improve patient safety. London, The Stationery Office, 2005. Available from: http://www.nao.org.uk/publications/nao_reports/05-06/0506456.pdf
30. "Radiotherapy Damaged Me" website. http://www.radiotherapydamaged.me.uk/newspapers.htm
31. Leape LL. Reporting of adverse events. N Engl J Med 2002;347:1633–8.[Free Full Text]
32. Ostrom LT, Rathbun P, Cumberlin R, Horton J, Gastorf R, Leahy TJ. Lessons learned from investigations of therapy misadministration events. Int J Radiat Oncol Biol Phys 1996;34:227–34.[CrossRef][Medline]
33. Ekaette EU, Lee RC, Cooke DL, Kelly K-L, Dunscombe PB. Risk analysis in radiation treatment: application of a new taxonomic structure. Radiother Oncol 2006;80:282–7.[CrossRef][Medline]
34. Radiation Oncology Safety Information System (ROSIS). Available from: http://www.clin.radfys.lu.se/default.asp
35. Donaldson L. Reducing harm from radiotherapy. Br Med J 2007;334:272[Free Full Text]
36. World Health Organization. Patient Safety: harm to patients caused by radiotherapy. Available from: http://www.who.int/patientsafety/activities/technical/radiotherapy/en/index.html
37. Donaldson L. When will health care pass the orange-wire test? Lancet 2004;364:1567–8.[CrossRef][Medline]
38. Department of Health. Safety first. A report for patients, clinicians and healthcare managers. London, UK: The Stationery Office, 2006. Available from: http://www.dh.gov.uk/assetRoot/04/14/14/41/04141441.pdf

This article has been cited by other articles:

				BJR review of the year -- 2007 Br. J. Radiol., April 1, 2008; 81(964): 265 - 269. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Williams, M V Search for Related Content PubMed PubMed Citation Articles by Williams, M V