Basic HTML Version
SELF-HYPNOSIS FOR DEPRESSION IN PRIMARY CARE
307
Benchmarking
We chose three studies to benchmark against (Bedi et al., 2000;
Proudfoot et al., 2004; Ward et al., 2000). All benchmarked studies
were also conducted in NHS primary care and recruited referrals from
GPs of newly depressed patients seen in their everyday consulting ses-
sions. Two of the studies were PRPTs randomizing or offering a choice
between therapist-led psychological treatment or antidepressants/
treatment as usual (Bedi et al.; Ward et al.), and we benchmarked
against the randomized arms of these studies. The third study was an
RCT between psychological treatment plus treatment as usual (which
may or may not have included antidepressants) or treatment as usual
alone (Proudfoot et al.). We modeled our benchmarking on that of
Merrill et al. (2003).
R
ESULTS
Twelve out of the 14 practices within the LHCC agreed to take part
resulting in the potential participation of 72 GPs (out of 80). Thirty-
three GPs (46%) subsequently made 85 referrals to the trial. Of these,
6 patients had contact problems, and 79 patients agreed to take part in
the trial. The total number of patients accepted into the trial was 58,
and 21 were not included for the following reasons: didn’t fulfill crite-
ria (
n
= 11), subsequently declined (
n
= 7), other reasons (
n
= 3). Figure 2
is a flowchart showing the pathways of patients in the trial.
The patient preference pathway as detailed in Figure 2 shows that of
the 50 patients who elected to receive self-hypnosis, 9 were not avail-
able at follow-up because 2 stopped all treatment completely and 7 com-
pleted treatment but did not complete forms. Of the 4 patients who
elected to receive antidepressant drugs, all were available at follow-up.
Of those randomized, 2 patients were randomized to self-hypnosis
and 2 patients to antidepressants. These groups were too small for
statistical analysis; we have not included a flow chart on these
patients.
Table 1 shows the comparison of demographic data between this
study’s preference arms and the benchmarking studies’ randomized
data. Tables 2 and 3 show the evaluation measures, BDI and BSI scores,
at intake and 12-weeks follow-up, respectively, and compare these
measures with the benchmarking studies. The results show a marked
similarity between demographic data and pre- and posttreatment
results. Figure 3 shows the BDI comparison scores with confidence
intervals for this study’s preference arms and the benchmarked studies
randomized data.
Results from our study show that the proportion of patients choos-
ing self-hypnosis was significantly greater than chance,
z
= 6.26,
p
<
.001. The 95% confidence limits for mean difference in the 12-week
Downloaded by [ ] at 04:44 26 March 2012