Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Computerised decision support and reflection in action
- Toby Lipman (10 February 2003)
-
Similar findings
- Michael A Krall MD MS, Dean F. Sittig, PhD (10 February 2003)
-
User friendly guidelines - is the GP software to blame or are GPs just prejudiced?
- Kenneth M Holton (13 February 2003)
-
Interactive Decision Support is alive and well and essential for 21st Century Primary Care
- Clive B Prince, Bewdley DY12 2EG (14 February 2003)
|
|
|||
|
Toby Lipman, General Practitioner Westerhope Medical Group, Newcastle upon Tyne, NE5 2LH
Send response to journal:
|
Eccles et al have done us a service in demonstrating not only the lack of effectiveness of a computerised decision support system based on evidence-based guidelines (1), but also some of the reasons why general practitioners (GPs) found it unusable in everyday clinical practice (2). Essentially, they found it inflexible and obstructive to the normal process of consultation, even though they were in general supportive of the guidelines themselves. As the authors found, GPs valued on-demand information. However, this not only needs to be available, it needs GPs to be able to recognise the need for information and to possess the skills to obtain and interpret it at the point of use. These skills (usually taught as evidence-based medicine (3)) are difficult and not widespread (4), so it is understandable that efforts to implement the findings of research should have concentrated on implementing the recommendations of expertly produced evidence-based guidelines in a linear, top-down process. The use of research evidence is thus simplified and codified as a checklist of tasks in response to pre-defined prompts. The issue with this approach is not whether its computerised form is user friendly, but whether it is a practical or desirable way of making clinical decisions about individuals. Schön (5) points out that most of the time experienced clinicians make decisions in an apparently intuitive way - "reflection in action" - rather than proceeding through a rational technical process such as that exemplified by computerised guidelines. In a setting like general practice, where GPs are attempting to address patients' agendas in psychological and social as well as biological terms, the intrusion of a checklist (whether computerised or not) may disrupt the smooth flow of a consultation and lead to the worst of all worlds - neither to implementation of a guideline nor to tackling the patient's own concerns. On-demand information is now available on-line through the National Electronic Library for Health to all GPs. All NICE guidelines, the Cochrane Library, Clinical Evidence and much else may be accessed while the patient is present. The limitation here is the lack of confidence in IT and EBM skills among GPs and other clinicians. More emphasis in research and training should be placed on developing these skills in clinicians and evaluating the effects of their use rather than on seeking to achieve "technical fixes" to circumvent their lack. 1. Eccles M, McColl E, Steen N, Rousseau N, Grimshaw J, Parkin D, et al. Effect of computerised evidence based guidelines on management of asthma and angina in adults in primary care: cluster randomised controlled trial. BMJ 2002;325(7370):941-. 2. Rousseau N, McColl E, Newton J, Grimshaw J, Eccles M. Practice based, longitudinal, qualitative interview study of computerised evidence based guidelines in primary care. BMJ 2003;326(7384):314-. 3. Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine. How to Practice and Teach EBM. London: Churchill, Livingstone; 1997. 4. McColl A, Smith H, White P, Field J. General practitioner's perceptions of the route to evidence based medicine: a questionnaire survey. Bmj 1998;316(7128):361-5. 5. Schon DA. From technical rationality to reflection-in-action. In: Dowie J, Elstein A, editors. Professional Judgement: a reader in clinical decision making. Cambridge: Cambridge University Press; 1983. p. 60-77. Competing interests: I am co-organiser of an evidence-based practice workshop |
|||
|
|
|||
|
Michael A Krall MD MS, Family Physician Kaiser Permanente 2850 NW Nicolai Street, Portland, OR USA 97210, Dean F. Sittig, PhD
Send response to journal:
|
We very much appreciated the article by Rousseau et al (BMJ 2003;326:314-318) and believe it has contributed to our understanding of usability requirements for computerized decision support. Their findings related to timing of the guideline, ease of use of the system, and helpfulness of the content, as well as many of their other observations are very similar to, and extend the findings noted in our work(1). We performed a qualitative study of 16 primary care clinician users of a comprehensive electronic medical record system to discern the elements that determine whether alerts and reminders will be seen as a help or a hindrance. During analysis, five themes emerged: Efficiency, Usefulness, Information Content, User Interface, and Workflow. Specific usability and usefulness requirements were identified from within the themes and were described. A key difference between the type of information presented in our studies appears to be the comprehensive guidelines in the Rousseau study compared to the "just in time", brief elements extracted from reference sources and generally presented in the order entry pathway or at the point of decision making, in our study. Where our system failed to present useful information at the right time, our users also were dissatisfied, and in fact, this concept of the right time and place within the workflow emerged as the number one requirement, following speed and efficiency. Our users, while offering numerous suggestions for improvement, generally endorsed the alerts and reminders. There is no doubt that attempting to provide electronic quidance for chronic disease management is complex. Taken together, our studies suggest that breaking these guidelines into "bite sized" chunks and presenting them at an optimal point in the workflow, or at the user's discretion, may be the most acceptable approach. (1) Krall MA, Sittig DF. Clinician's Assessments of Outpatient Electronic Medical Record Alert and Reminder Usability and Usefulness Requirements. Proc AMIA Symp (United States), 2002, p400-4 Competing interests: None declared |
|||
|
|
|||
|
Kenneth M Holton, General Medical Practitioner Holbrooks Health Team, 75-77 Wheelwright Lane, Coventry, CV6 4HN
Send response to journal:
|
Rousseau et al fascinating study highlights problems with use of guidelines that are more to do with the software environment than the guidelines themselves. It is not surprising that a system presenting itself at the wrong time or constraining you to follow a fixed route will be unpopular in the context of a GP consultation. Other GP software not included in the study unobstrusively announces the existence of guidelines appropriate to an individual patient and allows the user to open them with one click and navigate just to those components desired at the time. A guideline can also include links to external resources, thus improving the utility of the guideline itself and reducing the time overhead in accessing an external resource. My partners and I have this at our disposal during consultation, but still tend not to use guidelines. The partners generally feel that guidelines present users with information and decisions they do not want at that time, or that have already been discarded for sound clinical reasons, and this frictional overhead detracts from the overall consultation quality. To be acceptable, guidelines have to be much less obtrusive into the consultation process. Competing interests: None declared |
|||
|
|
|||
|
Clive B Prince, General practitioner. PCT tutor Bewdley Medical Centre,Dog Lane, Bewdley DY12 2EG
Send response to journal:
|
Dear Editor, I am worried that the paper "Practice based, longitudinal, qualitative interview study of computerised evidence based guidelines in primary care" may inhibit the development of primary care in the 21st century. We have been using interactive decision support as an important facet of modern general practice for several years. We were rewarded with "Doctor of the Year Award" a couple of years ago, but would not have been able to demonstrate the high quality general practice without the locally created doctor and nurse chronic and acute disease decision support software that we use. Most chronic illness including hypertension, diabetes and heart failure and acute care of illnesses such as back pain, asthma and mental illness are successfully managed within the confines of routine general practice. The authors of this week’s BMJ paper should make it quite clear what software has been used to allow primary care to distinguish this project from decision support software which is working well in every day practice. The analogy of ploughing a field may be useful. If a formula one racing car had been used to pull the plough, the readers would not be surprised if it was not up to the job! This could then have been reported as “internal combustion engine is less effective than the horse”. We must not tar all decision support with the same brush -we couldn't work successfully without locally written Synergy Sophies decision support in our practice. Competing interests: Bewdley Medical Centre has been a Beacon Practice for cardiovascular disease.Our Practice Manager is on the Torex user group |
|||