Intended for healthcare professionals


Studying only admissions is a source of potential bias

BMJ 1995; 311 doi: (Published 19 August 1995) Cite this as: BMJ 1995;311:509
  1. Clive Weston,
  2. Peter Donnelly
  1. Senior lecturer Department of Cardiology, Pinderfields Hospital, Wakefield WF1 4DG
  2. Senior lecturer Centre for Applied Public Health Medicine, University of Wales College of Medicine, Cardiff CF1 3NW

    EDITOR,--U M Guly and colleagues state that their study, which claims to show that “paramedics and technicians are equally successful at managing cardiac arrest outside hospital,” does not “diminish the role of paramedics.”1 Yet the paragraph about their paper in This week in BMJ concludes that such patients “are best treated” by technicians and calls into question the requirement of having a paramedic in every emergency ambulance. We do not believe that such a conclusion can be safely drawn from the data presented.

    The methodology gives rise to several sources of bias. Information is presented for those patients taken to the emergency department and not for all patients sustaining cardiac arrests in the community. In our series, based on telephone interviews with ambulance staff, 30% of all patients were certified dead at the scene, and for every three cases in which resuscitation was attempted there were two cases in which it was not; paramedics were more likely to start resuscitation.2 Moreover, if ambulance controllers base their decision to dispatch technicians or paramedics on clinical information, random allocation of crew is unlikely. Thus the two types of crew may not resuscitate patients with the same likelihood of success before the intervention.

    Furthermore, a comparison of times spent at the scene and outcome may be distorted by the inclusion of patients attended first by technicians and then by paramedics in the group treated by paramedics. Our data (table) show that these patients spend the longest times at the scene of the arrest. In our community based study of arrests due to all causes, paramedics, who (unlike those in Guly and colleagues' study) were able to give drugs, compared favourably with other crews. Therefore, while we agree that it is most important to provide rapid defibrillation, giving drugs (according to the European Resuscitation Council's guidelines) may be important.

    Outcome of cardiorespiratory arrest outside hospital managed by South Glamorgan Ambulance Service (figures are numbers (percentages))

    View this table:


    View Abstract