Letters

Intervention trial for late life depression defended

EDITORWe would like to respond to several criticisms of our study¹ raised in Haynes's editorial² and in Deeks and Juszczak's commentary.³

Haynes argues that our study is limited by patients refusing to participate or dropping out. In the context of studies in elderly people, we think that our refusal and follow up rates were reasonable. It is difficult to recruit elderly people into research studies⁴ and maintain their participation. Considerable drop out rates were expected because our sample consisted of frail older people, many of whom died or became too ill to participate.

Haynes considers that our difficulty in recruiting and retaining doctors and patients compromised our community based intervention. Although maximising participation is important, the practical difficulties of fully engaging a community of busy doctors and frail elderly people needs acknowledgment. The educational programme was attended by the general practitioners of 62% of the participants in the intervention group and by 35 of the 71 (49%) general practitioners caring for the 1036 residents interviewed. These attendance rates are above average for continuing medical education in general practice. Although 28% of participants in the intervention group attended exercise classes, remember that we sought to reach old, frail, depressed people unlikely to exercise. Furthermore, by design, we did not specifically target participants in the study.

Haynes is concerned that the intervention's effect was at the expense of additional resources. Although some additional resources were required, the intervention was not expensive to run. To enhance generalisability and sustainability we used existing resources whenever possible. Apart from having a full time project officer funded by a grant, the intervention project was encompassed within the other people's usual positions. As the educational materials have been devised, the intervention could be implemented at a similar site with reasonable healthcare resources for the cost of this position.

Deeks and Juszczak's commentary outlines four "deficiencies" that may have affected our findings but seems to conclude that only the lack of a concurrent control group is likely to have introduced bias. We considered these limitations at the outset, but there was no sensible, practical alternative design. Whereas the standard randomised controlled trial design is undeniably valuable, the value should be weighed against the methodological difficulties of health services research.

If undue weight is given to these criticisms our study's important clinical message may be lost. Late life depression is an important public health problem. The effect of intervention was modest but significant. Despite the methodological issues raised, this intervention is a promising way of addressing late life depression in residential care.

Robert H Llewellyn-Jones, lecturer. 
rljones{at}mail.usyd.edu.au

Karen A Baikie, senior research officer. 
Healthy Aging Research Unit, Hornsby Ku-ring-gai Hospital, Hornsby, New South Wales, 2077, Australia