Background: In 2005, guidance on how to prevent wrong site surgery in the form of a national safety alert was issued to all NHS hospital trusts in England and Wales by the National Patient Safety Agency.
Objective: To investigate the response to the alert among clinicians in England and Wales 12–15 months after it had been issued.
Methods: A before–after study, using telephone/face-to-face interviews with consultant surgeons and senior nurses in ophthalmology, orthopaedics and urology in 11 NHS hospitals in England & Wales in the year prior to the alert and 12–15 months after. The interviews were coded and analysed thematically.
Results: The study revealed marked heterogeneity in organisational processes in response to a national alert. There was a significant change in surgeons’ self-reported practice, with only 48% of surgeons routinely marking patients prior to the alert and 85% after (p<0.001). However, inter-specialty differences remained and change in practice was not always matched by change in attitude. Compliance with the detailed recommendations about how marking should be carried out was inconsistent. There were unintended consequences in terms of greater bureaucracy and concerns about diffusion of responsibility and hastily performed marking to enable release of patients from wards.
Conclusion: The alert was effective in promoting presurgical marking and encouraging awareness of safety issues in relation to correct site surgery. However, care should be taken to monitor unintended consequences and whether change is sustained. Greater flexibility for local adaptation coupled with better design and early testing of safety alerts prior to national dissemination may facilitate more sustainable changes in practice.
Statistics from Altmetric.com
Wrong site surgery is one of the most serious, yet neglected, medical errors.1 In the UK, a study by the National Patient Safety Agency (NPSA) found ambiguity and uncertainty among hospital staff relating to routine policies for preventing wrong site surgery, particularly marking.2 As a result, guidance on how to prevent wrong site surgery in the form of a national safety alert (box 1) was issued in March 2005 to all National Health Service (NHS) hospital trusts in England and Wales, with the expectation that the recommendations would be fully implemented in all surgical specialties within a year.3 A patient safety alert requires prompt action by a given date to address an important safety problem4 and is distributed electronically to the NHS via the Safety Alert Broadcast System for cascade within hospital and other NHS trusts. The NPSA had issued 18 patient safety alerts and advice notices between April 2001 and October 2008.4 This paper reports on the results of a study to assess responses to this intervention by hospitals.
Box 1: Correct Site Surgery Alert recommendations
1 How to mark
An indelible marker pen should be used. The mark should be an arrow that extends to, or near to, the incision site and remain visible after the application of skin preparation. It is desirable that the mark should also remain visible after the application of theatre drapes
2 Where to mark
Surgical operations involving side (laterality) should be marked at, or near, the intended incision. For digits on the hand and foot the mark should extend to the correct specific digit. Ascertain intended surgical site from reliable documentation and images
3 Who marks
Marking should be undertaken by the operating surgeon, or nominated deputy, who will be present in the operating theatre at the time of the patient’s procedure
4 With whom
The process of preoperative marking of the intended site should involve the patient and/or family members/significant others wherever possible
5 Time and place
The surgical site should, ideally, be marked on the ward or day care area prior to patient transfer to the operating theatre. Marking should take place before premedication
6 Verify the mark
The surgical site mark should subsequently be checked against reliable documentation to confirm it is (a) correctly located, and (b) still legible. This checking should occur at each transfer of the patient’s care and end with a final verification prior to commencement of surgery. All team members should be involved in checking the mark
We used a before–after study design. Baseline data on surgical marking practices among surgeons in England and Wales were collected using a semi-structured questionnaire administered in face-to-face or telephone interviews with medical and nursing leaders in three surgical specialties (urology, ophthalmology and orthopaedics) at 11 NHS hospitals in England and Wales during 2004/5.5 A purposive sample was selected comprising: five hospitals selected randomly from NHS trusts in England and Wales, four hospitals that were part of an existing clinical epidemiology research network (for ease of access), and two hospitals that were part of the Health Foundation’s Safer Patients Initiative6 (with a view to finding possible exemplar sites). All interviews were tape-recorded.
Follow-up interviews were conducted with the same sample 12–15 months after the issue of the alert. The interviews covered: awareness of the alert and its contents; handling of the guidance at both hospital and departmental levels; changes in marking practice; views about marking; documentation changes; and verification processes (table 1).
The qualitative data were transcribed and downloaded into QSR Nud*ist 6 (QSR International, ©March 2002) for coding. For each phase, six transcripts were coded independently by the three researchers who conducted the interviews and, from this exercise, a coding frame was developed, which was then applied to each of the transcripts.7 Emergent themes and subthemes were discussed within the team. The surgeons’ accounts were corroborated by those of the nurses interviewed. Where there were differences, these have been noted in the text.
A total of 66 nurses and surgeons were approached to participate and over 90% responded (table 2). We describe, first, participants’ awareness of the alert and how it was handled at both hospital and departmental levels and, second, reported changes in marking practice.
Dissemination and implementation of the alert by hospitals
The majority of surgeons (71%, 22/31) claimed to have heard about the alert through some route (eg, hospital route, colleagues, professional body, professional meetings), although many were unsure of its contents and also from where it had been sent. Some (29%, 9/31) claimed to have neither seen the alert nor been aware that it had been issued. There were often separate paths of dissemination for nursing and medical staff and separate fora for discussion. It was common for nurses to be unaware of how medical staff had been informed and, similarly, doctors to be unaware of the channels of communication for nurses. Nurses were more likely to be aware of local change without necessarily knowing about the national alert. Staff (both medical and nursing) in the two Safer Patient Initiative hospitals were often unable to differentiate the alert from other safety initiatives. We identified three approaches to implementation and these are discussed below.
No dissemination or implementation
In one hospital, only one person interviewed had heard about the alert (an ophthalmology nurse who had heard about it from an outside source) and there had therefore been no changes in practice in line with alert recommendations in any of the three specialties sampled.
Dissemination and passive implementation
The alert was circulated to clinical leads and other staff and they were required to assume responsibility for implementation within their own departments.
In those hospitals where decisions about and responsibility for implementation were delegated to individual specialties, departmental responses were variable. Three patterns were identified:
surgical leads simply noted the guidance, decided that it was already being implemented and made no further effort to inform colleagues;
guidance was circulated to colleagues and individuals were left to decide their own practice;
the recommendations were discussed at audit and clinical governance meetings.
Some senior surgeons, in deference to individual clinical autonomy, had simply passed on the guidance to their colleagues for them to decide their own practice. Others had interpreted the alert in terms of a mandate and tried to ensure (at least partial) implementation within their departments, whereas others had responded to it purely as guidance and as an opportunity to review existing procedures. One ophthalmology consultant, for example, commented:
“I think it has raised awareness of wrong site surgery as an issue and it’s caused us to look at our own systems to see whether or not we do have a robust system, so it did raise the issue and the questions.”
Active dissemination and implementation
This involved integration of the alert’s recommendations (often with modifications) into central hospital policies, with which all relevant departments were expected to comply. Hospitals where the guidance had been adopted as trust policy and disseminated as such were likely to have set up a working group to examine the alert, consult with staff and devise a local policy and necessary supporting documentation. However, with insufficient preparation and consultation, a top-down approach could generate antagonism and hostility. Criticisms voiced about the handling of the alert at the hospital level included:
lack of planning and preparation;
poor coordination of documentary and practice changes;
lack of consultation with clinicians;
lack of discussion with, or preparation of, nursing staff;
treatment of the alert as a managerial as opposed to a clinical issue;
an overly dogmatic approach;
a rushed timetable to meet deadlines.
In one hospital visited, all staff interviewed (both nursing and medical) had been critical of the implementation process. The following comments were typical:
“I think in this organisation it has been rushed through. There has been a sense that we have to get something in place for a deadline… A lot of people didn’t know what was expected of them, and that is not just the surgeons because it involves the ward nurses and theatre nurses… If we had been given the time to work through the policy, perhaps we would have all bought into it.” (Ophthalmology consultant)
“It caused uproar, because there was no communication about it. It just appeared and everybody said, ‘What the heck is this? Where has this come from? Who did this?’ The staff in recovery were not happy about it and rang up the manager and they ended up having one big massive row over it. I walked in on it on the Monday morning and [my colleague] said, ‘Have you seen this?’ and I said, ‘I don’t know what you are talking about.’ We didn’t actually know [about it].” (Urology nurse)
Hospitals with an active dissemination and implementation policy modified the recommendations to their local context. Modifications included:
dispensing with the requirement of doctors to sign;
requiring the operating surgeon, rather than the doctor who made the mark, to sign to indicate that the correct site had been marked;
making provision for surgeons who chose not to mark.
In four hospitals, some surgical specialties (eg, ophthalmology, urology and spinal surgery) had argued successfully for exemption from the new policy. A consultant ophthalmologist explained what happened in his department:
“[The alert] was disseminated through the risk management team…. The trust asked us to either implement it or explain that we had some kind of robust alternative. We said that we had a robust alternative…. We discussed it at consultant level, particularly to do with the marking, and decided that we wouldn’t introduce the marking and therefore that our current checks that we have in place were sufficient that we didn’t need to use the guidance…. Well, we in ophthalmology have a problem with marking and we feel that marking beforehand actually reduces the accuracy of the checking process. In so much as our operating is done under local anaesthetic, we can still make a final check at the end by asking the patient.”
Elsewhere, provision had been made for staff who refused to comply with the requirement to mark, as an ophthalmology consultant explained.
“I think we are required to have a policy. A local decision had to be made, because I know that a certain group [urologists and spinal surgeons] do not agree with it. So, if you don’t mark, then you have to sign for it.”
In at least three hospitals, the requirement to mark was introduced prior to tackling the issue of signing to indicate that the patient had been marked. In one hospital, for example, the forms for medical staff to sign had not been introduced in order to avoid delay in implementation of the less controversial aspects of the guidance. As the medical director explained:
“It’s not that the practice isn’t occurring; it’s the documentation of the practice…. That’s our stumbling block at the moment and that has gone to our consent committee. The clinicians on the consent committee are, you know, saying enough is enough. That doesn’t mean to say that it won’t happen in the future but, at the present time, there’s a final link in the chain that we have not got…. The final bit of the documentation, that is checking in the anaesthetic room, is the bit that we’re stumbling over.”
A variety of measures had been instituted to increase compliance, including:
audits of compliance with the alert recommendations;
personal letters from the medical director requiring an explanation for failures to comply;
introduction of a veto on releasing unmarked patients from the wards and/or admitting them to theatre.
Impact of the alert
There was a statistically significant change in surgeons’ self-reported practice between baseline and follow-up (McNemar test, p<0.001). Prior to the Correct Site Surgery Alert, only 48% of the surgeons interviewed said they marked patients routinely. At the follow-up, 85% said they were marking routinely, with 71% (12/17) of those who had previously never marked, or marked only for certain non-routine procedures, having changed their practice to one of marking (table 3). In the words of one ophthalmic surgeon, “We didn’t want to stand out like a sore thumb.”
Changes in practice had been initiated in some settings without staff necessarily having been aware of what had prompted the changes. In hospitals which had adopted the alert recommendations as hospital policy, staff in the departments sampled were aware of the new hospital policy but not necessarily of the NPSA alert.
Inter-specialty differences noted at baseline remained.5 Prior to the alert, all but one orthopaedic surgeon marked all patients before surgery, either personally or using a deputy, whereas only a minority of urologists and ophthalmologists did so (table 4). At the follow-up, all orthopaedic surgeons and a majority of urologists and ophthalmologists were marking. However, 29% (5/17) continued not to mark or to mark only for certain non-routine procedures, including two people who had adopted a policy of not marking in the wake of a previous incident/near miss of wrong site surgery involving a patient who had been wrongly marked.
Views about marking
Prior to the alert, views about the value of marking as a contribution to patient safety varied. Even among those who marked routinely, some held no strong views either way and two people, given a choice, would have preferred not to mark. When asked to identify the main causes of wrong site/side, wrong procedure, wrong patient errors, marking was mentioned on only two occasions and, in each case, involved patients wrongly marked. Both nurses and doctors were more likely to refer to the context in which the surgery took place and issues such as increased volume of work, rapid throughput of patients, time pressures, targets to reduce waiting lists, multi-site working, high staff turnover, and language and communication barriers between staff and between staff and patients. Those surgeons who rarely marked believed that marking did not improve patient safety and that the practice was itself inherently dangerous. Many also described a number of practical factors which militated against marking, including structural issues, such as the distance between wards and theatres, and lack of privacy for patients resulting from the practice of admitting patients before beds are available for them, and the use of “arrivals lounges”.5 As one urology consultant explained:
“Patients won’t come into a bed before the operation; they come into the so-called ‘arrivals lounge’ and it’s a nightmare there. How on earth in an arrivals lounge, in this sort of set up, are you going to sort of strip the patient to put the mark on and privacy and so on?”
At follow-up, surgeons’ practical concerns and reservations about the value of marking remained. At least half of the 12 who had changed their practice to one of marking retained their negative views. They believed, for example, that marking was inherently unsafe, especially if carried out by someone other than the surgeon conducting the procedure, and that it could induce a dangerous sense of complacency and diffusion of responsibility. In other words, they had been coerced into compliance against their clinical judgement, through the imposition of the alert as an organisational mandate.
Objections to the recommendations were more likely to be expressed by doctors (14/31) than nurses (2/31) (p<0.005) and generally fell into two categories: objections to marking (from staff in departments of ophthalmology and urology), and objections to the additional paperwork, specifically the requirement to sign to indicate that the checks had been completed (voiced across all three specialties).
“All that piece of paper does is bounce the responsibility back on the surgeon, which is where it was anyway. You know, great strides are achieved here!” (Urologist)
“I now have to sign a piece of paper in two or three places to say that I have decided that marking is not appropriate, so it’s an extra piece of paper. It’s very reassuring for everyone!” (Urologist)
Who marks the patient, when and where
The alert stipulates that marking should be undertaken by the surgeon, or nominated deputy, who will be present at the time of the patient’s operation (see box 1). However, at the follow-up, delegation to junior doctors or nurses who would rarely assist in theatre continued to be routine practice in some departments, especially orthopaedics. Some informants were unaware of this specific recommendation; others cited the impracticality of following it.
In all but one of the departments where patients were marked, the person who obtained the patient’s consent also carried out the marking. Since the baseline interviews, there had been a general strengthening of the consent process, initiated in large part by the professional bodies of the different specialties (Royal College of Ophthalmologists, Royal College of Surgeons, British Association of Urological Surgeons), focusing on details of what should be included and when and by whom a patient’s consent should be taken. Tension between a model of good practice which emphasised the importance of the person obtaining the consent having adequate knowledge of the intended surgical procedure and the alert requirement of subsequent presence in theatre was resolved in favour of adequate knowledge in some departments and in favour of presence in theatre in others.
Patients were usually marked on the ward prior to leaving for theatre—only in exceptional cases would they be marked in the anaesthetic room or even in theatre. In some departments, nurses were refusing to release unmarked patients from wards and/or admit them to theatre and, in two hospitals, this practice had been introduced as a hospital-wide policy. However, although this strategy may have encouraged doctors to mark, it may also have introduced a potential element of risk through delays that would put additional pressure on theatre staff and increase the risk of operations being cancelled. Under these circumstances, there was a danger that marking could be carried out in haste by a doctor who was not assisting in theatre and who did not take the patient’s consent, as happened in the following incident recounted by a urology consultant:
“A hip replacement patient was received in theatre, wasn’t marked, an SHO [senior house surgeon] was called from the ward to mark them, he went down, put an arrow on the patient, who happened to be a urology patient rather than a hip replacement patient…. I can assume that they sent for the patient, (when the patient) came to reception, somebody or other realised that the side wasn’t marked, rang up whoever was on call, he rushed along, realised he was holding everybody up and dashed in and saw somebody… [As far as the patient was concerned] some strange man rushed in, spoke something about hips and drew an arrow on his leg.
Visibility of the mark
The alert states that “it is desirable that the mark should remain visible after the application of theatre drapes”. However, nurses, in particular, noted that this was often not the case.
Verification of marking
Subsequent to the alert, most hospitals had produced (or were in the process of producing) new ward and theatre checklists that included a check of whether or not the patient had been marked. One hospital had adapted the documentation so that the operating surgeon (rather than the person who marked the patient) signed to indicate that the patient had been correctly marked; another had dispensed with the requirement for doctors to sign. Nurses observed that doctors’ compliance with the requirement to sign at the time of marking was often poor. Failure to reach an agreement over documentation had delayed implementation in at least three hospitals. An orthopaedic nurse from one hospital explained:
“We’re looking at it, we haven’t as yet. At the moment it has gone to our consultant team, because they have a few issues with having to sign and sign again.”
Elsewhere, an uneasy compromise had been reached, where nurses alone signed to indicate completion of the checks. When asked whether doctors were signing, another orthopaedic nurse commented:
“They don’t do anything, we do that. They do all the talking, we do the writing…. They’ve implemented the practice, but they’re not actually documenting it.”
By contrast, nurses expressed little opposition to the documentary changes and reported general compliance with the new system of checks.
Least likely to have been fully implemented of the four recommended checks (table 5) was the final check, the pause for verbal briefing prior to commencement of surgery. Most of those questioned had been unaware that it had been one of the alert recommendations and were not implementing it as a formal procedure that had to be documented, although at least two hospitals were attempting to introduce the pause as trust policy. However, staff were generally implementing the pause for verbal briefing (or “stop” as it was known locally) in the two Safer Patient Initiative hospitals, where it had been introduced as a separate initiative.
In the Safer Patient Initiative hospitals marking tended to be seen simply as one of a range of patient safety measures. Those interviewed were more likely to identify the pre-procedural briefing (“huddle”) and pause for verbal briefing (“stop”) as the most significant changes in practice in relation to correct site surgery. However, even in the these hospitals, this final check was not necessarily signed for and documented.
The study found that 12–15 months after the issue of the NPSA alert on correct site surgery there had been a significant increase in self-reported routine marking practice. However, compliance with the detail of the recommendations about how marking should be carried out was variable. Inter-specialty differences in marking practice were reduced but not eliminated and surgeons continued to hold reservations about the value and safety of marking.
This study provides a valuable insight into the reception and impact of a major national intervention to reduce one of the most serious medical errors. The study covered a broad sample of three surgical specialties in 11 hospitals across the UK. The hospitals chosen may not be representative of all hospitals in the UK; however, they include a range of district general and teaching hospitals, serving a mix of rural, urban and ethnically diverse populations. It is possible that the combination of face-to-face and telephone interviews may have encouraged participants to respond in different ways; however, no systematic differences were discernible from the analysis. The study relied on participants’ accounts of their practice rather than direct observation. There is a tendency for self-report to stress the positive aspects of performance and to downplay the negative. For example, there has been extensive documentation of self-reporting of errors representing a much lower error rate than those recorded through direct observation.1–12 Participants’ accounts may have emphasised compliance with the requirements of the alert. The study findings may therefore overestimate levels of compliance and reflect awareness as much as compliance. However, the willingness of participants to discuss areas of non-compliance suggests that such glossing may have been minimal.
Dissemination not enough
Box 2: An integrated model of change management
Multiple channels of communication
Interdisciplinary communication, especially between nursing and medical staff
Local consultation with all staff likely to be affected by the guidelines
Adaptation to the local context
Coordination of documentary with practice changes
Audit of the changes
In responding to a patient safety alert, it is not enough for trusts simply to disseminate the information to relevant staff.1314 Successful implementation is likely to involve a more integrated model of change management (box 2).13–16 An important finding of the study was the insularity of nursing and medical staff from each other and the lack of awareness of each other’s channels of communication, fora for discussion and different work practices. Despite the emphasis in the literature on promoting inter-disciplinary discussions and collaboration,15 there was often little evidence of this in practice.17
The factors conducive to implementation included the adoption of the recommendations of the alert as trust-wide policy, in other words the imposition of a central organisational will. Given the persistence of often deeply felt and sincerely held reservations, it is unlikely that many surgeons would have been persuaded to change their practice without this element of central authority. As one urologist explained, when asked what had prompted the change in practice:
“It, perhaps, wasn’t directly that [the alert] but it was the fact that it came through the hospital council and the clinical governance mechanism in the hospital, and so it has been rolled out across the trust.”
Resistance to the documentary changes was greatest from doctors and, in some hospitals, was responsible for delays in implementation. In a culture of targets and deadlines, doctors are likely to resent anything which may be perceived as inessential.
Although the alert was generally successful in promoting marking practice, it seems to have had less effect on underlying attitudes to patient safety improvement. Inter-specialty variation persisted along with cynicism about the alert being about obligatory paperwork rather than an effective intervention to reduce error. The alert may well improve safety culture over time by raising awareness and prompting changes in marking practice; however, there is a risk that practice changes may not be sustained in the absence of change in core attitudes.15 In the absence of actual experiences of wrong site surgery, and given its relative infrequency, it remains to be seen how individuals and organisations will be able to maintain adherence to risk reducing processes without visible rewards.
Reification of the mark
The alert was about more than simply marking practice. However, reported changes in marking practice were not always accompanied by compliance with the detail of the guidance (for example, who should carry out the marking, its visibility once the patient is draped, documentary verification) and its underlying spirit of patient safety improvement.
There is a danger that an undue emphasis is placed on the presence of the mark per se as opposed to the circumstances in which the marking takes place. For example, in an attempt to ensure doctors’ compliance with the requirement to mark, nurses in some hospitals have introduced practices (such as refusing to release patients who have not been marked from the wards or refusing to admit them to theatre) which could themselves compromise patient safety. Similarly, orthopaedic surgeons, although most likely to mark patients, were among those least likely to adhere to the detailed recommendations of the alert. Unlike ophthalmologists and urologists who had no general culture of marking, orthopaedic surgeons had been inculcated with the need to mark during their training. As a result, these surgeons had sometimes paid little attention to the guidance, in the belief that they were already compliant, and were therefore unaware of much of the accompanying detail about how and by whom the marking should be carried out.
The central focus of the alert was preoperative marking and how, when, where, by whom and with whom it should be done. As a consequence, the significance of other important recommendations, such as the pause for verbal briefing, may have been diluted.
Implications for policy
This study provides evidence to inform the increasing volume of nationally issued patient safety guidance. Previous evaluations of national guidance have demonstrated variation in adoption and implementation.1618 Frustration over the ineffectiveness of similar top-down approaches to quality improvement has been a driving force for continuous quality improvement methods such as total quality management (TQM), kaizen, communities of practice or Institute for Health Improvement (IHI) methodologies. In contrast to the managed cascade of knowledge from official channels, ideas for improvement can emerge and be adapted from existing knowledge at a local level by empowered practitioners, coupled with an iterative cycle of measurement, analysis, testing and action.19
Future national guidance should take more account of the environment in which the recommendations will be introduced, encourage greater flexibility for adaptation to local contexts, and address changes in attitudes as well as practice. It should concentrate not simply on what should be done but the reasons why it should be done, and keep documentary requirements to a minimum. Where documentation is required, the reasoning behind this should be addressed more explicitly. In particular, better design and early testing of safety alerts prior to national dissemination may facilitate sustainable changes in practice.
Various strategies have been promoted for reducing wrong site/side, wrong procedure, wrong patient errors, including: surgical smart chips, bar coded patient wrist bands, computer-based patient identification systems, paper checks, site marking, patient involvement (including patients marking themselves), preoperative confirmation of laterality and procedure using documentary, imaging and other evidence and pre-procedure time-out.1 None on its own is sufficient to prevent errors and, even with combined strategies, such as the US Joint Commission of Accreditation of Healthcare Organizations’ universal protocol and the NPSA’s correct site surgery alert, errors and near misses continue to occur.120 As one investigation of reported errors and site verification protocols concluded, “No protocol will prevent all cases”.19 Protocols on their own may be of little value without an environment that is conducive to their full implementation and which fosters commitment to a culture of safety consciousness.
The national patient safety alert increased the practice of presurgical marking and promoted greater awareness of safety issues in relation to correct site surgery. However, there were also unintended consequences in terms of greater bureaucracy and concerns from surgical staff that the alert could have adverse effects due to diffusion of responsibility and hastily performed marking to enable release of patients from wards.
The authors wish to acknowledge the NHS Patient Safety Research Programme for funding this project. We thank the hospitals which took part and their staff for giving so generously of their time to help organise and take part in the interviews. We also thank the two anonymous referees for useful comments.
Competing interests: None.
Funding: The authors wish to acknowledge the Patient Safety Research Programme for funding this project.
See Commentary, p 396
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.