Basic HTML Version
216
MARK P. JENSEN ET AL.
complete data did not experience a meaningful decrease in pain
following treatment, these findings indicate that at least 30% of all of
the participants recruited into the study experienced a clinically mean-
ingful change in pain, pre- to posttreatment. One participant (3%)
reported an increase in pain of 30% or greater, pretreatment to post-
treatment.
As would be expected, a larger treatment effect was found in the
efficacy analysis than in the more conservative intent-to-treat analysis
(Table 1). These participants who both completed all 10 treatment
sessions and who did not report developing a new medical condition
(that might impact pain) during the study unrelated to treatment
reported a decrease of 0.94 (21% change) in pain intensity, pretreat-
ment to posttreatment, on average. This represents an effect size of .53,
which is considered a “medium” effect size (Cohen, 1988). Among the
26 efficacy analysis participants (and also including the single partici-
pant who did not provide complete data but who did complete all 10
sessions and who did not develop a medical problem during the study
period), 10 (37%) reported a clinically meaningful change in pain, and
only 1 reported a meaningful increase in pain, pretreatment to post-
treatment.
Secondary Outcome Variables: Pain Unpleasantness, Pain Interference,
Depressive Symptoms, and Perceived Control Over Pain
The results of the ANOVAs examining changes in the secondary
outcome variables from the initial baseline period through the 3-month
follow-up assessment are presented in Table 2. The results reported
above with pain intensity as the dependent variable were replicated
for pain unpleasantness. Significant time effects were found in both the
intent-to-treat and efficacy analysis samples, with no significant change
during the baseline period, a significant decrease in pain unpleasant-
ness from pre- to posttreatment, and maintenance of gains through the
3-month follow-up.
However, no evidence was found for a treatment effect on pain
interference or depressive symptoms in either sample. Although a sig-
nificant time effect was found for pain interference in both the intent-
to-treat and efficacy analyses, this effect was due to a significant
decrease in pain interference during the baseline period (pre- to post-
baseline) for the efficacy analysis participants, and from baseline to
posttreatment and follow-up, but not pre- to posttreatment, for the
intent-to-treat analysis participants. No significant time effect for
depressive symptoms was found.
The results of the analyses concerning perceived control over pain
were inconsistent across the two analysis groups. No statistically
significant time effect was found in the intent-to-treat analyses, but a
significant time effect was found in the efficacy analyses. This later