Basic HTML Version
218
MARK P. JENSEN ET AL.
finding was due to a significant and medium effect size (
d
= .62)
increase in perceived control over pain from pre- to posttreatment in
the efficacy analysis sample. This improvement was maintained
through the 3-month follow-up.
Predicting Treatment Outcome
The two-level repeated measures ANOVA in the efficacy analysis
sample, with average pain intensity as the dependent variable and
time and diagnostic group as the independent variables, did not yield
a significant diagnostic group by time interaction,
F
(2, 22) = 1.82,
p
= .185,
suggesting that the treatment was similarly effective for each diagnostic
group. However, there was limited power in these analyses for detect-
ing an interaction, given the small number of participants in each of
the three diagnostic groups examined (SCI = 11, amputation = 5, MS = 9).
An examination of the pre- to posttreatment percent change in average
pain intensity for each group separately (SCI = 17%, amputation =
43%, MS = 10%) and the percent of participants in each diagnostic
group who obtained a 30% or greater pre- to posttreatment decrease in
average pain (27%, 60%, and 33% for participants with SCI, amputa-
tion, and MS, respectively), suggests a larger and more frequent
response to treatment among participants with amputation than
among participants with SCI or MS.
Most of the other predictors, including number of treatment days
(
r
= .20,
p
=
ns
), change in pain intensity in the first treatment session
(
r
= .09,
p
=
ns
), hypnotizability (
r
= .16,
p
=
ns
), and treatment-outcome
expectancy rated before the first session (
r
= .10,
p
=
ns
), were not sig-
nificantly associated with pre- to posttreatment change in average pain
intensity. However, patient treatment-outcome expectancy rated just
after the first treatment session did show a moderate (
r
= .39,
p
< .05)
and statistically significant association with pre- to posttreatment change
in pain.
D
ISCUSSION
Change in Pain with Hypnotic-Analgesia Treatment
The standardized hypnotic-analgesia protocol used in this study
reduced average pain intensity, relative to a no-treatment baseline
period, in patients with chronic pain from a wide variety of disabilities.
Although the average change in pain intensity from pre- to posttreat-
ment was not large for the group as a whole and represented only a
small to medium effect size in the intent-to-treat sample (all partici-
pants for whom complete data were available), responder analyses
indicated that about one third of the study sample reported a clinically
meaningful (30% or greater) decrease in pain with treatment. Only 1
patient of 30 reported a clinically meaningful increase in pain. Moreover,