Recent Advances: Posters for accident departments: simple method of sustaining reduction in x ray examinations
BMJ 1995; 310 doi: https://doi.org/10.1136/bmj.310.6980.640 (Published 11 March 1995) Cite this as: BMJ 1995;310:640
- Eugene McNally, senior registrar in diagnostic radiologya,
- Gerald de Lacey, consultant radiologista,
- Philippa Lovell, registrar in accident and emergencya,
- Theo Welch, consultant in charge of accident and emergency
- a Departments of Radiology and Accident and Emergency, Northwick Park Hospital, Harrow, Middlesex HA1 3UJ
- Correspondence to: Dr E McNally, Department of Radiology, Nuffield Orthopaedic Centre NHS Trust, Oxford OX3 7LD.
- Accepted 10 February 1995
Abstract
Objective: To assess whether a simple strategy would sustain a reduction in the number of unnecessary x ray examinations.
Design: Use of posters to display guidelines encouraging the more effective use of radiology in patients with head injuries, twisted ankles, neck injuries, and abdominal pain.
Setting: Accident department of a large metropolitan district general hospital.
Patients: 15875 patients attending the accident department over two years.
Main outcome measure: Proportion of patients having radiography.
Results: Referrals for skull radiography fell from 56% to 20% and those for abdominal radiography fell from 31% to 7%. Referral patterns for adults attending with twisted ankles and cervical spine injuries did not change. Reductions were sustained over two years.
Conclusion: Carefully designed posters provide a simple method of reducing unnecessary x ray examinations.
Key messages
Key messages
Good presentation is as important as the guidelines themselves
Simple methods are more likely to be suc- cessful in the long term than complex ones
Giant coloured posters displaying guidelines have reduced unnecessary abdominal and skull x ray examinations by over half
If applied elsewhere with equal success con- siderable savings in both costs and radiation dose would occur
Introduction
Concerns regarding the overuse of x ray examinations1 2 and increasing radiology workloads3 have led to attempts to reduce unnecessary radiography.4 5 6 7 8 9 10 Many studies have shown that, although guidelines and protocols can alter practice in the short term, the effect is difficult to sustain. We tested the effect of posters in reinforcing guidelines.
Patients and methods
The study was carried out in our accident department, which has 35000 new patient attendances each year. Guidelines were prepared indicating when to request a radiograph in patients with head injuries,11 12 13 cervical spine injuries,14 and abdominal pain15 and in adults with twisted ankles.16 A series of commercially designed 70x50 cm coloured posters (figs 1 and 2) containing the guidelines were displayed in the doctors' room in the accident department.
Guidelines for radiography of cervical spine
All patients presenting with any of the four target injuries between January 1989 and January 1991 were included in the analysis. Referral rates before the introduction of the guidelines were calculated by reviewing the accident department records during February, April, and July 1988. Non-consecutive months were chosen to avoid any bias from seasonal variation in the pattern of injuries presenting and to allow for changes in referral rates that might occur as accident doctors became more experienced.
All patients attending with a head or ankle injury during the first 12 months of the study who were not referred for radiography were contacted by post to ensure that no important abnormality had been missed. We did not contact patients who had been admitted or referred to a follow up clinic. A random sample of 38 of the non-responders was pursued more vigorously, first by letter and then by telephone if no reply was forthcoming.
Results
During January 1989 to January 1991, 6613 patients attended the department with head injury, 3833 with abdominal pain, 3973 adults with twisted ankles, and 1456 with cervical spine injuries. Figure 3 shows the referral rates expressed as the ratio of patients who had radiography to the total number of patients presenting with that particular injury. Referrals for skull radiography fell from 56% (399/709) to 20% (123/628) two years after the guidelines were applied. Abdominal radiography fell from 31% (83/270) to 7% (25/356). There was no significant change in the referral rates for ankle or cervical spine x ray examinations.
Referral rates for skull, cervical spine, ankle, and abdominal radiography after introduction of poster guidelines
Follow up letters were sent to 2259 patients with head or ankle injuries, and replies were received from 1175 (52%). Of the 38 non-responders followed up, 16 (42%) were untraceable, having either moved or given a false address.
Discussion
Many strategies using guidelines to alter practice have found that a sustained effect is difficult to achieve.4 13 This has led to the view that providing guidelines alone is insufficient, and that devising an effective method of re-enforcing guidelines is also important.17 The ideal reinforcement must be cheap, simple, and easily applied to other departments and hospitals. The colourful posters used in this study satisfy these criteria. Several alternative explanations for the marked reductions in referrals for skull radiographs were considered. Referral rates before the introduction of the guidelines were low (53/1000) compared with rates of 58-94/1000 in other published studies.12 13 Though we accept that comparison of referral rates in different hospitals is difficult because of the different populations served, these figures indicate that the large reductions in skull radiography were not due to excessively high rates before introducing the guidelines. Furthermore, the reductions in referral rates were sustained over several changes in accident staff. Our postal follow up detected no serious consequences resulting from the change in referral practice.
We thank Kodak Ltd (Health Sciences Division) for financial support, Shelley Briers for her computer skills, and Jo Bashford for word processing.