Crisis in London's mental health servicesBMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7089.1278b (Published 26 April 1997) Cite this as: BMJ 1997;314:1278
Empirical evidence was lacking in study
- Scott Weich, Senior lecturera
- a University Department of Psychiatry, Royal Free Hospital School of Medicine, London %NW3 2PF
- b Section of Disability, United Medical and Dental Schools of Guy's and St Thomas's Hospitals, Guy's Hospital, London SEI 9RT
- c Research Section, Sainsbury Centre for Mental Health, London SE1 1LB
- d Institute of Psychiatry, de Crespigny Park, London SE5 8AF
- e Tower Hamlets Healthcare NHS Trust, The Royal London Hospital, London E3 4LL
- f Westminster Association for Health, London W9 3QF
Editor—Given the attention finally being paid to the funding of mental health services, the article by G Shepherd and colleagues came as a great disappointment.1 Indeed, it is hard to say which was more surprising, the serious methodological limitations of the study or the omission of any reference to these in the discussion. More disturbingly, the authors' conclusions were almost entirely unsupported by empirical evidence.
I should like to draw attention to some of the more glaring problems with this article.
Firstly, the authors fail to acknowledge the possibility of non-response bias. Data on bed occupancy were based on the responses of just 26 out of 42 providers (62%) and data on alternative mental health resources on just 28 respondents. Similarly, neither the size of the target patient population nor the completeness of patient data from responder units was stated. For example, staff may conceivably have been more motivated to record information on patients they believed to be inappropriately placed on acute wards.
Secondly, all of the data were collected by representatives of the provider units. We do not know how many data collectors were recruited in this way or (more importantly) whether any checks were carried out to assess the reliability or validity of ratings obtained in this way. In the absence of independent ratings the possibility of bias in the reporting of, for example, bed occupancy levels cannot be excluded.
I could go on. Instead of qualifying their findings in the light of such limitations, the authors devote their discussion instead to supposition about the optimal provision of mental health services. Quite how they arrive at the conclusion that the relocation of all patients who are inappropriately placed on acute wards would solve the problem of overoccupancy is difficult to comprehend, since no data are presented on …