Clinical audit is enjoying a high profile life style at present. It is impossible to pick up any journal or newspaper concerned with medical management without seeing an article on audit.
Everyone knows what it is all about -or believes that they do. The fundamental principles of clinical audit are that it should:
(a) be professionally led
(b) be seen as educational
(c) form part of routine clinical practice
(d) be based on setting of standards
(e) generate results that can be used to improve outcome of quality care
(f) involve management in both the process and the outcome of audit
(g) be confidential at the individual patient/clinician level
(h) be informed by the views of patients / clients
These are wonderful platitudes which must satisfy the dreams of our masters, the healthcare administrators, as a method of underpinning their pronouncements that patients are put first. Like many other good questions, "what is wrong with medical audit?" can be interpreted in several ways. At the conceptual level it invites the answer "nothing". Indeed, one can argue strongly that a key to. improvement in clinical performance and to the optimal use of resources lies in accurate, appropriate and relevant audit of clinicians and their activities.
Nevertheless, at the implementational level, we have to face up to the fact that audit has so often failed to deliver what was expected or could reasonably be expected. So, the question "what is wrong with medical audit" also invites a different answer, for it invites a search, not so much destructively for the ways in which audit has failed to deliver, but constructively in the sense of learning lessons so that we can progress more rapidly towards the fulfilling of the potential of medical audit.
This paper will consider the sort of problems that have risen in practice and some ways in which they can be overcome. In particular, the audit process has to be looked at critically but constructively, to try and find out how we can benefit from the unsatisfactory state of affairs in other areas and thereby avoid making the same mistakes, so that the audit process may ultimately be improved.
The Audit Cycle
We are all familiar with the concept of the audit cycle which comprises definition, observation and analysis of performance, and result in remedial action being taken -leading to improvement - and the whole cycle starts allover again.
Unfortunately, those familiar with auditing practice will know very well that things are not as straightforward. (Figure 1) shows the audit cycle as it often is, resulting in no improvement -and this is so because there are problems at each stage of the cycle. It follows that if we are to achieve benefits from audit, we must discuss these problems critically and propose solutions.
Problems With Audit: Phase 1, The Setting Up
In the first phase of setting up audit and defining "performance" there are a number of obvious problems, both technical and "people problems" which are addressed later. If we want audit to succeed then:
(a) we must define its aims
(b) we must define what we mean by clinical audit and clinical performance
(c) we must set out (in advance) clear criteria by which success or failure are to be recognized
d) we must define a terminology so that reproducible valid scientific comparisons can be made
Problems With Audit: Phase 2, The Observation
Audit depends upon observation. If inappropriate observations are made, audit will fail. Here again, there are problems:
(a) what data should be collected?
(b) how are these data to be defined?
(c) what quality control measures are to be adopted?
(d) what remedial staff education is available?
Once again, these fundamental problems have been poorly addressed. If audit is to succeed, then the data which are to be collected must be defined in advance. They should allow the aims of Phase 1 of the audit cycle to be achieved, but should be as few as possible to maximize user compliance. It is important that each term to be used in the audit should be defined and those definitions should be circulated beforehand so that everyone knows what they are. There should be continuing education programme so that all staff are aware of the definitions and terminology that are being used and there should be a continuing quality control to ensure the maximum reproducibility of the data collected.
Problems with Audit: Phase 3, The Analysis
Once the aims have been defined and the data collected for an audit project, they must be analyzed. Here again, there are problems:
(a) what is "normal" and what is "acceptable"?
(b) how is the analysis to be performed?
(c) who is to hear about it?
Again, these questions are fundamental. For example, the "normal" range and the "acceptable" range are by no means the same thing. A further problem of analysis that has bedeviled audit is "who is told what?". One of the major problems with audit, which often precludes clinical acceptance, is the nagging worry that an uncaring, unfeeling, unknowledgeable health administrator will draw inappropriate conclusions from the data. It is therefore essential that before doing an audit the reference range should be discussed, the acceptable level should be set widely in the first instance, and it must be made clear who sees the analysis and under what conditions.
Problems with Audit: Phase 4, Remedial Action
This is perhaps the most difficult phase of all. There are numerous problems here, even when the audit has proceded successfully this far:
(a) does the audit process have "teeth"?
(b) if guidelines are provided, what type are they?
(c) how is remedial education to be provided?
(d) what evidence of remedy is acceptable?
(e) what changes are envisaged with the time?
It is quite clear that without some form of remedial education audit will clearly fail to achieve its potential other than of a "stick" with which to beat overworked and increasingly resentful staff. Without remedy, the staff simply lose interest. Yet, the consequent question "what sort of remedial education?" is almost never raised. Remedial education must be focused, it must be specific and it must be seen to be effective.
These conditions are almost totally lacking in some institutions. All too often some facile statement is made as such "guidelines will be provided". Unfortunately, most guidelines are inappropriate or incomplete; often they remain nothing more than a personal opinion. If audit is to achieve its potential, failure to fall within an acceptable range of performance must be tied specifically and carefully to pre-agreed focused remedial education measures and there must be agreed evidence as to whether these guidelines themselves are succeeding or failing as the audit cycle revolves.
Audit in this decade - in Perspective
In the previous paragraphs some specific continuing problems with audit have been discussed. There are, of course, other syndromes encountered in the audit process relating to "people problems". They will be familiar to most readers:
(a) the "Hilton" syndrome. Ten people meet for a pleasant weekend (all paid for) in a nice hotel. To justify such a pleasant occasion they produce a set of guidelines for the junior staff, who were not invited but upon whom (together with the audit staff) the guidelines are imposed (rigorously, to justify the weekend)
(b) "Pink Floyd" syndrome - as in the opening line of the well known song by this group : "We don\'t need no education"

The clinical problems of resource allocation remain, indeed, they are worse than before. The problems of resource wastage remain, indeed, they are worse than before. The clinical course for medical student is finite: the faculty teaching time and number has been reduced or eroded, and the amount to learn has exploded. Evaluation of audit has been a disaster! Given all the problems, it is not surprising that audit has run into difficulties. I would, however, be distressed if these thoughts were to be construed as an attack on the principles of audit. They are not. The question "what is wrong with audit?" still invites the answer "nothing". Nor is it an attack on audit people. All of the mistakes and problems outlined have been made, and continue to be made, in many healthcare institutions. We have to face two problems and face them squarely. There is far too much information in the institutions. The problem with institutional decision making is not lack of data, it is the lack of right data. Much of the information sadly is invalid and some of it is junk. All pragmatists belief that facing problems squarely enables us to overcome them more easily. In the present instance, if we are to face the problems that have been outlined and can adopt appropriate solutions (Figure 2) then audit will be seen for the beneficial process which we intrinsically believe it to be.