Quality of life as an outcome measure

Page 392
Black blue
392 HEALTH PSYCHOLOGY
following an assessment of the Cochrane Controlled Trials Register from 1980 to 1997,
Sanders et al. (1998) reported that although the frequency of reporting quality of life data
had increased from 0.63 per cent to 4.2 per cent for trials from all disciplines, from
1.5 per cent to 8.2 per cent for cancer trials and from 0.34 per cent to 3.6 per cent for
cardiovascular trials, less than 5 per cent of all trials reported data on quality of life.
Furthermore, they showed that this proportion was below 10 per cent even for cancer
trials. In addition, they indicated that whilst 72 per cent of the trials used established
measures of quality of life, 22 per cent used measures developed by the authors themselves. Therefore, it would seem that although quality of life is in vogue and is a required
part of outcome research, it still remains underused. For those trials that do include a
measure of quality of life, it is used mainly as an outcome variable and the data are
analysed to assess whether the intervention has an impact on the individual’s health
status, including their quality of life.
Quality of life as an outcome measure
Research has examined how a range of interventions influence an individual’s quality of
life using a repeated measures design. For example, a trial of breast reduction surgery
compared women’s quality of life before and after the operation (Klassen et al. 1996).
The study involved 166 women who were referred for plastic surgery, mainly for physical
reasons and their health status was assessed using the SF36 to assess general quality of
life, the 28-item GHQ to assess mood and Rosenberg’s self-esteem scale. The results
showed that the women reported significantly lower quality of life both before and after
the operation than a control group of women in the general population and further, that
the operation resulted in a reduction in the women’s physical, social and psychological
functioning including their levels of ‘caseness’ for psychiatric morbidity. Accordingly,
the authors concluded that breast reduction surgery is beneficial for quality of life and
should be included in NHS purchasing contracts.
Quality of life has also been included as an outcome variable for disease-specific
randomized controlled trials. For example, Grunfeld et al. (1996) examined the relative
impact of providing either hospital (routine care) or primary care follow-ups for women
with breast cancer. The study included 296 women with breast cancer who were in
remission and randomly allocated them to receive follow-up care either in hospital or by
their general practitioner. Quality of life was assessed using some of the dimensions from
the SF36 and the HAD scale. The results showed that general practice care was not
associated with any deterioration in quality of life. In addition, it was not related to
an increased time to diagnose any recurrence of the cancer. Therefore, the authors
concluded that general practice care of women in remission from breast cancer is as good
as hospital care.
Other studies have explored the impact of an intervention for a range of illnesses.
For example, the DAFNE study group (2002) explored the impact of teaching diabetic
patients flexible intensive treatment which combines dietary freedom and insulin
adjustment (dose adjustment for normal eating – DAFNE). The results showed that
this approach to self-management improved both the patients’ glycaemic control and
their quality of life at follow-up. Shepperd et al. (1998) also used quality of life as an
Page 392
Black blue