New drug prescribing by hospital doctors: The nature and meaning of knowledge
Introduction
In the UK, prescription drugs cost over £5 billion, or around 12% of total NHS costs. This increases annually by around 9%, due to a number of factors such as demographic changes (an ageing population with increased morbidity). A key factor accounting for over 60% of the rise is displacement of old drugs with new drugs at higher costs (Walley, Mrazek, & Mossialos, 2005). Some new drugs may expand therapeutics in areas of previously unmet clinical need, and their use in this situation is desirable. But many new drugs have only limited advantages over existing therapies, often at a substantial cost increase.
Most prescribing takes place in general practice—about 80% by cost. One of the key factors swaying a general practitioner to prescribe a particular drug is the advice received or the example set from hospital colleagues. The influence of secondary care on primary care prescribing is well recognized (deVries, van Diepen, Tromp, & Jong-van den Berg, 1996; Edgar & Girvin, 1999; Feely, Chan, McManus, & O’Shea, 1999; Pryce, Heatlie, & Chapman, 1996; Steffensen, Sorensen, & Olesen, 1999); the UK Audit Commission (1994) estimated that one prescription in secondary care resulted in seven or eight in primary care. In contrast to prescribing in UK general practice, which is closely monitored for managerial purposes and heavily studied, there is far less information available on hospital prescribing. It might be expected that the same basic principles apply in prescribing: that the doctor strives to prescribe effectively and safely for each patient. But key drivers of prescribing behaviour in general practice, personal clinical experience, social influence and local norms and practice (Greer, 1988; Prosser, Almond, & Walley, 2003) although still present, will differ. Clinical experience in hospitals is with a different group of more seriously ill patients. Often the clinical experience is not that of the individual doctor who may be relatively junior and inexperienced, but of a senior colleague. Managerial controls are often tighter in hospitals than in primary care, with formularies restricting the range of drugs available and controlled by Drugs and Therapeutics Committees (DTC). There is more informal scrutiny of prescribing by colleagues than in primary care. However, hospitals provide specialist led care. This brings with it a culture of expectation that hospital doctors will be up-to-date in their area of expertise and will to some extent try to push the boundaries of that specialty within the hospital, being keen to explore new diagnostic or therapeutic tools.
Research on what determines new drug prescribing in hospital is limited. Some research suggests the influence of peer reviewed scientific literature, clinical trial evidence, colleague contact and improved drug characteristics (Jones, Greenfield, Bradley, & Jowett, 2000; Jones, Greenfield, & Bradley, 2001; Peay & Peay, 1990). Other findings highlight the influence of social and experiential aspects (Allery, Owen, & Robling, 1997; Denig, Haaijer-Ruskamp, Wesseling, & Versluis, 1993). Further, while the former studies rated the pharmaceutical industry as influentially low despite frequent contact, a case-control study of drug requests for additions to a hospital formulary found significant association with how often the proposing doctors interacted with pharmaceutical representatives (Chren & Landefeld, 1994). These inconsistencies may be due to methodological confines limiting a comprehensive understanding of decision-making: surveys and interviews may simply capture normative responses and expressed attitudes about prescribing, rather than what actually occurs in practice. While the recent study by Jones et al. (2001) demonstrates the multi-factorial determinants of new drug prescribing in hospitals, how these factors become meaningful is not yet understood. Therefore, to try to understand prescribing in its context, we examined hospital doctors’ perspectives and explanations of actual prescribing events.
Section snippets
Method
We used the critical incident technique (Flanagan, 1954) to encourage doctors to give factual accounts of actual prescribing and to describe the factors that determined their initial prescribing from a list of new drugs. This method follows an inductive process of analysis, as it seeks to explain decision-making from the viewpoint of the prescriber.
Sampling was analytically driven from experience in pilot interviews and theoretically by previous research. This revealed that junior doctors are
Results
Twenty-three hospital doctors participated, eight consultants (two from each specialty except psychiatry where no consultant agreed to participate) and 15 specialist registrars (three from each specialty). Six doctors declined to participate: five were too busy and one declined because of a recent bereavement. A total of 153 critical prescribing incidents were identified. All doctors had prescribed some of the new drugs (range 3–14, median 6, mean 6.5).
Knowledge of the patient
When a new drug was first prescribed, it was initially considered individually for specific cases. The single most frequent reason to prescribe a new drug, cited in 65 incidents, was failure or unsuitability of the doctor's standard choice. This was assessed in relation to the doctor's knowledge of a specific patient. That a patient fell outside the norm legitimized the use of a new drug, and the new drug offered a possible solution to a treatment dilemma. For example, a new drug was proposed
Experiential knowledge
The fourth category of knowledge is based on the nature of subjective context provided by personal experience. This is essentially a reflective process grounded in the actual prescribing of a new drug that allows doctors to test its therapeutic outcomes and (re)interpret the evidence in the light of experience. Thus, actual prescribing was an experimental learning process with further prescribing contingent upon examination of use, experience and familiarity:
I use losartan because I have more
Discussion
Hospital doctors had a clear approach of the hierarchy of evidence, citing clinical trial data as the key evidence source for new drug uptake. However, it was seldom influential in isolation. In practice, the limits of scientific data meant that the ideal evidence threshold for new drug prescribing was often lowered and interrelated with other forms of knowledge that run counter to the formal scientific rationality of evidence-based medicine (EBM). A doctor's knowledge was eclectic, acquired
Acknowledgements
This study was funded by the Department of Health prescribing research initiative. We would like to thank all the hospital doctors who generously gave up their time to participate in this study. We would also like to thank the journal reviewers for their helpful comments in contributing to this paper.
References (42)
- et al.
Towards understanding treatment preferences of hospital physicians
Social Science & Medicine
(1993) - et al.
Patterns of preference for information sources in the adoption of new drugs by specialists
Social Science & Medicine
(1990) The social construction of premenstrual syndrome
Social Science & Medicine
(1992)- et al.
Achieving clinical behaviour change: A case of becoming indeterminate
Social Science & Medicine
(1998) - et al.
Why general practitioners and consultants change their clinical practice: A critical incident study
British Medical Journal
(1997) - et al.
Medical work, medical knowledge and health care: Themes and Perspectives
Clinical autonomy, individual and collective: The problem of changing doctors’ behaviour
Social Science & Medicine
(2001)Discourse, descriptions and diagnoses
Medical talk and medical work: The liturgy of the clinic
(1995)A prescription for improvement: Towards more rational prescribing in general practice
(1994)