Pressure to prescribe

Involves a complex interplay of factors>

Two thirds of consultations with general practitioners end with the issuing of a prescription.1 The decision to prescribe is influenced by many factors, to do with the doctor, the patient, the doctor-patient interaction, and the wider social context, including the effects of advertising and the financial incentives and disincentives for all parties.2 3 4 5 6 Hardline advocates of rational drug use do not look kindly on variations in prescribing patterns that cannot be explained by purely clinical factors.1 The prescriber who allows the “Friday night penicillin” phenomenon to sway his or her clinical judgment tends to do so surreptitiously and with a guilty conscience.

But such behaviour is the rule rather than the exception. Several studies have shown that the prescribing behaviour of doctors is heavily influenced by their perceptions of the social background, beliefs, attitudes, and expectations of the patient,2 as well as the uncertainty of the diagnosis.5 7

Bradley identified several patient factors associated with doctors' discomfort when prescribing (or refusing to prescribe) drugs: extremes of age and of social class, non-white ethnic group (because of perceived differences in expectations), patients with a medical or paramedical background, and patients whose history the doctor either did not know or knew only too well (frequent attenders, “heartsink,” and “fat file” patients).5

Although the powerful placebo effects of drugs prescribed in such situations are well documented,8 so are the maladaptive behaviours that follow. Patients who receive a prescription for a self limiting condition are more likely to expect (and receive) one if the same symptoms recur.9 The profession is regularly called on to acknowledge its vulnerability to the allegedly intense pressure to issue a prescription when none is needed,5 yet hard data on the extent of this pressure remain sparse. Patients originating from the Asian subcontinent in particular have been accused of attending their general practitioners expecting prescriptions for “trivial complaints” and “ill defined conditions,”10 and patients of Pakistani or Indian origin are more likely to be given a prescription than those from white and West Indian ethnic groups.7

Two recent studies in the BMJ,11 12 one of them published this week, address the gap between the prescriptions patients actually expect and what their doctors assume that they expect. An Australian study confirmed previous findings that about half of all patients have clear expectations for a prescription.11 After controlling for presenting condition and patients' actual expectations, doctors' perceptions of these expectations also independently influenced the decision to prescribe. A British study by Britten and Ukoumunne (p 000) found that doctors' perceptions of patients' expectations for a prescription were significantly related to patients' hopes and educational level, and to the broad category of diagnosis, as well as to characteristics of the individual doctor.12 The decision to prescribe was strongly related to the doctor's perception of the patient's expectations, and, overall, doctors classified 21% of their own prescriptions as “not strictly necessary.”

Cockburn and Pitt speculate that failure to ascertain patients' expectations is a major reason why practitioners prescribe more drugs in total than patients expect.11 But although these two studies suggest that a doctor's assessment of a patient's expectation is wrong in about a quarter11 and a sixth12 of cases, neither study provides direct evidence that this misconstruction leads to substantial overprescribing. Of the 255 patients in Cockburn and Pitt's study who expressed their expectations as anything other than “Don't know,” 71% were correctly classified by their doctor as either expecting or not expecting a prescription.11 A further 21% were incorrectly classified as not expecting a prescription; despite this, half of them received one. Only 8% (20) of the 255 patients did not expect a prescription when their doctor thought they did, and 16 of these received one.

In Britten and Ukoumunne's study, doctors admitted feeling pressure to prescribe in 66 out of 540 encounters (12%), and were more likely to prescribe for this group of patients.12 But the number of excess prescriptions given to this group (that is, the number beyond what would have been expected if there had been no perceived pressure to prescribe) amounted to 16 in 540 encounters and 5% of all prescriptions issued.

These two most recent studies concur with previous findings that patients who expect a prescription are many times more likely to receive one than those who do not. This evidence is compatible with the stereotype of demanding and manipulative patients repeatedly forcing the hands of their reluctant doctors. But it is also compatible with the fact that patients may have more insight into their medical needs than their doctors give them credit for, and that both doctor and patient may legitimately take account of non-medical factors in deciding whether a particular drug is necessary at a particular point in time.

The act of issuing a prescription is the culmination of a complex chain of decisions.2 It is open to biomedical, historical, psychosocial, and commercial influences, no aspect of which can be singled out as the “cause” of non-rational prescribing. The search should continue for methods to measure the interplay of these disparate factors on the decision to prescribe.