Judging the use of clinical protocols by fellow professionals
Introduction
A range of methods is available to organisations for controlling the behaviour of employees (Hopwood, 1974). One such method, administrative control, typically involves the development of written rules specifying correct and desirable behaviour. While rules designed to govern behaviour in the UK NHS are nothing new, initially clinical practice was rarely targeted. However, in recent years several disparate factors have brought the management and control of health professionals’ behaviour close to the top of the agenda in the British healthcare system. First, the financial risks associated with complaints and litigation now lie with hospital Trusts themselves. Ensuring quality in standards of care now has direct financial benefits via reduced insurance premiums payable to the Clinical Negligence Scheme for Trusts established in 1995. At the same time, the Patient’s Charter has raised patient awareness of quality issues and complaints are on the increase. These pressures resulted in the 1998 White Paper A First Class Service: Quality in the New NHS, in which the UK government announced a commitment to the principles of clinical governance as a way of ensuring quality of patient care (Scally & Donaldson, 1998).
Clinical governance, a term imported from the US, is a management system aiming, among other things, to ‘tackle the wide variations in quality of care throughout the country’ (Gulland, 1999), by spreading best practice and giving Trusts a statutory duty to seek quality improvement. Thus, the legal responsibility for ensuring quality of care will rest with Trust Chief Executives (Black, 1998). The guidelines on which best practice will be based, will be produced by the National Insititute for Clinical Excellence, which began work in April 1999. Compliance with the aims of clinical governance will be monitored by a new Commission for Health Improvement.
The development of guidelines reflects a drive towards evidence based medicine, and is designed to achieve a reduction in practice variation, and some degree of standardisation of clinical practice for the benefit of patients. This standardisation of health care professionals’ behaviour around agreed best practice should reduce error, which is acknowledged to be a problem in medicine (Leape et al., 1993, Gopher et al., 1989). Indeed, The UK’s Chief Medical Officer, Professor Liam Donaldson, includes well trained staff and clear procedures as aspects of his conceptualization of an integrated approach to clinical governance (Scally & Donaldson, 1998). Although medicine is a complex and hazardous business, Leape (1994) points out that an error rate of 1%, as found in Gopher et al.’s (1989) study of an intensive care unit, would be unacceptable in other high hazard industries. Moreover, recent high visibility cases of error have brought the issue to the attention of the public, and there is growing concern about the accountability of health care professionals.
All of these factors mean that there is a growing interest in getting health care professionals, and especially doctors, to follow newly introduced clinical protocols (e.g., Grimshaw and Russell, 1993, Robertson et al., 1996). In the future there are likely to be increasing numbers of written protocols developed both nationally and locally in the British NHS. This process will require careful management, as the development and introduction of protocols can be problematic. Problems are especially likely among doctors whose training emphasises self-reliance and error-free practice. Some evidence of disquiet has already surfaced, for example in a response to Scally and Donaldson, which characterised the disintegration of clinical governance, due to, in part, lack of evidence and conflicting guidelines (Goodman, 1998).
However, in attempting to facilitate the introduction and uptake of procedures, it may be possible to learn some lessons from research in other industries, and there has already been some interest in extending psychological work on the human factors involved in adverse events in complex sociotechnical systems to a consideration of the National Health Service (NHS). For example Vincent, Taylor-Adams and Stanhope (1998) have considered the relevance of Reason, 1990, Reason, 1995) model of organisational accidents to medicine. They specifically note that Reason’s model indicates that issues surrounding the availability and use of protocols require attention in ensuring a safe organisation. Simply having protocols in place is not sufficient to improve safety (Lawton, 1998, Reason, 1997). Indeed, in the area of healthcare it has been noted that the availability of clinical guidelines has not noticeably promoted changes in clinical practice (Goldberg et al., 1998). Steps must also be taken to ensure that the staff meant to use the protocols accept them and are willing to comply with them.
One reason why the use of protocols may be particularly difficult to manage in medicine concerns the professional independence fostered by the culture of medicine. In other industries one of the main methods for implementing protocols involves monitoring and enforcement. In the British NHS this is not a viable option, at least at a senior level. The work of a consultant is largely unsupervised, and even junior medical staff frequently work on their own initiative. However, if adherence to clinical protocols is to be improved, it will be necessary to foster a climate where the reporting of rule violations is not only encouraged, but seen as a duty. This will require quite a change of culture, as doctors’ reluctance to criticise each other is well documented (Donaldson, 1994, Rosenthal, 1995). In fact it could be argued that the professional culture of doctors appears to promote several of the violation-producing conditions noted by Reason (1995), i.e., a culture that encourages taking risks, the belief that a bad outcome will not happen, and professional attitudes hostile towards procedures (or in the case of medicine, protocols and guidelines).
The fact that the practice of healthcare professionals is largely self-regulated may explain why much of the medical research on clinical protocols to date, which concentrates on compliance rates, has shown that compliance appears to be low (Grilli and Lomas, 1994, Weingarten et al., 1998, Yoong et al., 1992). This finding of low compliance is not restricted to UK physicians. Indeed, of US physicians Caper (1988, p.1535) writes “as fast as regulations and review protocols are written physicians learn to circumvent them….”. Lomas, Anderson, Domnick-Pierre, Vayda, Eakin and Hannah (1989) in a study of Canadian obstetricians found that practice guidelines had little effect on actual clinical behaviour. There is also a suggestion that non-compliance is more prevalent among senior staff than their juniors (Lloyd, 1995). In our own research we have shown that doctors in Britain feel that given the nature of their work, protocols are often inappropriate. In a focus group study, the need to use clinical judgement to deal with each case as it arose and the individualised approach to patient consultation were often mentioned by doctors, who perceived protocols to have a straight-jacketing effect on clinical practice (Lawton & Parker, 1998).
Given that the number of clinical protocols is set to increase, it is important to understand healthcare professionals’ attitudes to their use. The study reported here involves an empirical investigation of the judgements of a range of healthcare professionals with respect to rule-related behaviour. As such, it extends previous research in an industrial context (Reason, Parker & Lawton, 1998) to a hospital setting.
Section snippets
Method
A large-scale questionnaire survey was carried out among healthcare professionals (nurses, doctors and midwives) of all grades, asking for their opinions and judgements with respect to a range of hypothetical scenarios describing the behaviour of a fellow health care professional. The questionnaire was constructed in this way to avoid asking respondents to report on their own clinical behaviour. In the incident scenarios two factors were orthogonally manipulated. The first factor, which was
Results
Multivariate analyses of variance (MANOVA) were computed. The first included judgements of appropriateness as the dependent variable. Likelihood of reporting was the dependent variable in the second analysis. In both cases the two repeated measure variables were behaviour type (compliance, improvisation and violation) and outcome (good, bad and poor). Professional group (nurse, doctor and midwife) was included as a between subjects variable.
Appropriateness
As might be expected, there was an enormous difference in the judgement of different behaviour types, in that the violation of a protocol was judged as less appropriate behaviour than either improvisation or compliance (F(2,261)=762.42, p<0.001). However there were interesting differences between the professional groups. While all professional groups made distinctions between behaviour types, those made by midwives were significantly more striking than those made by doctors and nurses (F
Likelihood of reporting
When the reporting behaviour of the three professional groups was considered, there was a large and significant main effect for behaviour type (F(2,520)=198.77, p<0.001) which supported judgements of appropriateness. Violations were more likely to be reported than improvisations and compliance was the least likely to be reported. There was also a large and significant main effect for outcome (F(2,520)=82.01, p<0.001). Unsurprisingly, there were linear trends showing that when the behaviour lead
Discussion
The results of this study show that when judging how appropriate the behaviour of a fellow professional is, and when considering whether to report the behaviour of a colleague, British healthcare professionals take into account both the outcome of the behaviour and whether or not it follows a protocol. Behaviour reflecting compliance with a protocol is likely to be judged appropriate, while the violation of a protocol is thought to be inappropriate by all professional groups. However, midwives
Acknowledgements
Research supported by ESRC grant L211252050.
References (28)
- et al.
A randomized control trial of CQI teams and academic detailing: Can they alter compliance with guidelines?
Joint Commission Journal on Quality Improvement
(1998) Not working to rule: Understanding procedural violations at work
Safety Science
(1998)- et al.
Preventing medical injury
Quality Review Bulletin
(1993) - et al.
Can practice guidelines safely reduce hospital length of stay? Results from a multicenter intervention study
American Journal of Medicine
(1998) Clinical governance: Fine words or action
British Medical Journal
(1998)Solving the medical dilemma
New England Journal of Medicine
(1988)Doctors with problems in the NHS workforce
British Medical Journal
(1994)British Medical Journal
(1998)- et al.
The nature and causes of human errors in a medical intensive care unit
- et al.
Evaluating the message: The relationship between compliance rate and the subject of a practice guideline
Medical Care
(1994)