Appropriateness of referrals to hospital

BMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6963.1233a (Published 05 November 1994) Cite this as: BMJ 1994;309:1233
  1. D T Lipman
  1. Medical Centre, Westerhope, Newcastle upon Tyne NE5 2LH

    EDITOR, - In their study of the appropriateness of referrals Glyn Jones Elwyn and Nigel CH Stott used the concept of avoidable referrals.1 They found that 38% of the referrals studied over one year had been avoidable, whereas in a larger study Fertig et al found 15.9% of referrals to have been possibly inappropriate when judged against locally determined guidelines.2 Thirty two of the avoidable referrals in the authors' study were considered to have been due to lack of resources, and it is perverse to judge the decision on referral in these cases as inappropriate. If these cases were included in the appropriate group the proportion of inappropriate or avoidable referrals would fall to 15.5%, almost exactly the same as Fertig et al's finding and well within their 95% confidence interval (11.8% to 20%).2

    Since inappropriate referrals do not seem to explain variation in referral rates the authors' approach, which classifies the reason for referral, may offer a way forward; perhaps the proportions of the types of referral vary between doctors with high and low referral rates.1,2 Unfortunately, it is not clear how significant was the agreement between the referrer's assessment and that of his collaborator, and it would be interesting to know what they disagreed about (nor are we told the referral rate per 100 consultations in this study). It would have been helpful if at least two other general practitioners had assessed the referrals so that we could know both how closely they agreed with the published assessments and whether the referrer's assessments were significantly different from those of the others; the degree of agreement can be measured by calculating K for each pair.3

    Two important points made by this study are the extent to which referrals in general practice are necessitated by lack of resources and inadequate communication with hospitals. The second point clearly needs to be addressed, but the first is not so clear cut as may be assumed. For example, if general practitioners were able to carry out upper gastrointestinal endoscopy in their surgeries they would not have to refer patients for this investigation; the costs in terms of equipment and time, however, are so great that it is almost certainly more economical for endoscopy to continue to be done in the secondary sector. We should be wary of the suggestion that a reduction in referral to hospital is necessarily desirable.


    1. 1.
    2. 2.
    3. 3.
    View Abstract

    Sign in

    Log in through your institution

    Free trial

    Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
    Sign up for a free trial