Return to PJ Online Home Page
The Pharmaceutical Journal Vol 265 No 7114 September 16, 2000
Pharmacy Practice Research
Papers presented at the British Pharmaceutical Conference, Birmingham, September 10 to 13, 2000 pR17

Prescribing errors in hospital inpatients: why do they occur?

By Bryony Dean, Nick Barber, Mike Schachter* and Charles Vincent†

Introduction Studies carried out in United States hospitals suggest that prescribing errors cause harm in about 1 per cent of all inpatients.1 A United Kingdom pilot study suggests similar findings.2 However, no studies have explored the reasons why prescribing errors occur from a prescriber's perspective. According to the theories of human error,3 errors can result from latent failures in management decisions, personal and environmental performance influencing factors, and active failures at the level of the individual.
The objectives of this study were to find out whether human error theory can explain why prescribing errors occur, to identify factors contributing to their occurrence and to suggest ways in which they can be prevented.

Method The study took place at a London teaching hospital and had ethics committee approval.
Pharmacists were asked to give the researcher details of any potentially serious prescribing errors identified during an eight-week period. Prescribers were then contacted and invited to participate in the study. It was emphasised that the study was non-disciplinary in nature.
Semi-structured interviews were conducted with each participant, exploring the reasons for the error's occurrence. A questionnaire exploring potential performance influencing factors was also administered. Some interviews were taped and transcribed; detailed notes were taken for the remainder.
Content analysis was used to analyse the interviews using human error theory and to explore other themes within the data.

Focal points

  • Human error theory suggests that errors result from (i) latent failures in management decisions, (ii) performance influencing factors, (iii) active failures at the level of the individual
  • This is a useful model for the study of prescribing errors
  • Latent failures identified included gaps in medical school training, staffing policies and misunderstandings about responsibility
  • Performance influencing factors included fatigue, interruptions and covering other doctors' patients with whom the prescriber is unfamiliar
  • Active failures include slips in concentration, lapses in memory and mistakes resulting from lack of knowledge

Results Eighty-eight potentially serious prescribing errors were reported. For 35 of these, it was not possible to identify the doctor concerned. Of the 53 prescribers contacted, 44 (83 per cent) agreed to participate; these represented a wide range of grades, clinical specialties and types of error.
Latent failures identified include gaps in medical school training (interviewees reported no training in drug dosing or in the practicalities of prescribing), staff shortages and misunderstandings about responsibility when one doctor asks another to write a prescription. Common performance influencing factors were fatigue, interruptions, time pressures and covering other doctors' patients with whom the prescriber was unfamiliar. Active failures included slips in concentration, lapses in memory and mistakes resulting from lack of knowledge.
Several interviewees indicated that writing discharge prescriptions was regarded as transcription rather than as prescribing, and that little thought was given to this process. Many interviewees made unprompted comments that hospital pharmacists had an important role in preventing prescribing errors, although some stated that they might prescribe more carefully if there were no pharmacy check.
Doctors varied in terms of their attitude to error. Some felt that errors were inevitable and were relatively unconcerned. Others were very upset to have made a prescribing error and were keen to reflect on and change their practice accordingly.

Discussion A high proportion of the doctors contacted were willing to participate in the study, and human error theory proved to be a useful model for studying prescribing errors. Many examples of latent failures, performance influencing factors and active failures were identified.
It is suggested that clinical governance and error reduction strategies should be focused on medical school training, teamwork and hospital culture rather than at the level of the individual. The impact on prescribing errors of other models of prescribing should also be explored.

School of Pharmacy, London WC1N 1AX; *Imperial College school of medicine, London; †clinical risk unit, University College London

References

1. Bates DW et al. Incidence of adverse drug events and potential adverse drug events. JAMA 1995;274:29-34.
2. Study into medical errors planned for the UK. BMJ 1999;319:1091.
3. Reason J. Human error. Cambridge: University of Cambridge; 1990.