Intended for healthcare professionals

Editorials

Patients' demands for prescriptions in primary care

BMJ 1995; 310 doi: https://doi.org/10.1136/bmj.310.6987.1084 (Published 29 April 1995) Cite this as: BMJ 1995;310:1084
  1. Nicky Britten
  1. Lecturer in medical sociology Department of General Practice, United Medical and Dental Schools of Guy's and St Thomas's Hospitals, London SE11 6SP

    Patients cannot take all the blame for overprescribing

    The Audit Commission's recent report on prescribing in general practice in England and Wales estimated that up to pounds sterling275m could be saved from the NHS drugs bill if overprescribing was reduced.1 The report lists several overprescribed drugs, including antibiotics and non-steroidal anti-inflammatory drugs. The authors clearly believe that patients' expectations of treatment are partly responsible for the problem, as do many general practitioners.2 Perhaps the most memorable view came from Marinker, who said, “We may see the doctor as helpless in the face of a population of patients who have an overwhelming need to alter chemically their experiences of the world in which they live.”3 He compared a general practitioner in a consulting room to a barmaid in a gin shop, implying that not only do patients know exactly what they want but that they usually get it. But what evidence is there that patients' demands for prescriptions have any effect on doctors' prescribing habits other than prompting repeat prescribing?

    Much of the evidence is equivocal because researchers have not directly defined or measured demand for prescriptions. Instead, studies have focused on doctors' perceptions of patients' demands and doctors' statements that patients' expectations influenced real or hypothetical decisions about prescribing. Little attempt has been made to measure either patients' overt requests for prescriptions in the consultation or their expectations beforehand. About 5-7% of prescriptions, however, are not dispensed,4 and many drugs are dispensed but not consumed. This suggests that prescribing levels actually exceed patients' expectations. It seems that demand (either real or perceived) is greater than need.

    When general practitioners are surveyed they describe high levels of demand,5 but objective evidence consistently suggests that doctors overestimate patients' expectations.6 Reanalysis of published data shows that about a fifth of patients leave general practice consultations with prescriptions they did not expect.7

    In contrast, studies that have considered whether demand from patients influences prescribing habits have yielded inconsistent results. Two studies showed that demand had no influence on prescribing,8 9 one that perceived demand resulted in lower prescribing,10 and five that demand was associated with higher rates of prescribing.7 11 12 13 14 None, however, was able to look directly or reliably at the effect of such demand. The two studies that showed demand to have no influence were based on hypothetical consultations and may not accurately have reflected real prescribing behaviour: they were also based in North America and may not apply to general practice in Britain. The studies with negative findings covered a limited range of drugs. Finally, four of the five studies with positive results looked at the whole range of prescribing behaviour and found that severity of disease, type of drug, and whether the patient had an appointment also influenced prescribing. On balance, demand from patients is probably only one of many factors that lead to overprescribing by general practitioners.

    Less superficial evidence about the complex reasons for overprescribing will probably come only from qualitative research. Qualitative interview techniques could find out what patients and doctors really think about prescribing. Firstly, detailed and patient-centred investigation should explore patients' ideas and expectations before consultations, perceptions of symptoms and illnesses, reasons for consulting the doctors, ideas and preferences about treatment, attitudes to drugs, previous experiences with doctors, and what patients expect consultations to achieve. Researchers could ask also whether doctors educate their patients by word or deed to expect prescriptions (or no prescriptions) and whether patients choose doctors whose prescribing habits they like. Careful questioning should disentangle patients' ideal expectations for prescriptions (what they hope for or want) from their actual expectations (what they think will really happen).

    Secondly, a similar approach is needed to sort out what general practitioners think and do about prescribing. Perhaps some doctors justify their poor prescribing habits by blaming patients instead of recognising that they sometimes misuse prescribing—for example, to close a difficult consultation. Qualitative studies could investigate doctors' perceptions of patients' preferences for prescriptions, previous experience with and knowledge of individual patients, beliefs about drugs, sense of time and other pressures during consultations, clinical assessments of patients' conditions, expectations of consultations, and other factors influencing the decisions to prescribe. Linking the two kinds of approach on a case by case basis would verify the accuracy of doctors' perceptions.

    Finally, we still need to know more about the process of clinical consultations: how and how often patients make their expectations known, how doctors assess patients' expectations, and how well the two parties communicate on this subject. One way of taking this further would be to show a videotape of the consultation separately to each party afterwards and to interview them about what they were thinking at each point (I Cromarty, unpublished findings).

    If doctors' perceptions do not correspond with patients' preferences poor or inappropriate prescribing, wastage of drugs, and unsatisfactory doctor-patient relationships may result. Reassuringly, recent research on patients' ideas about drugs suggests that more patient centred practice would not necessarily lead to higher rates of prescribing.15

    References

    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.