7A.3.5 Patients no longer meeting inclusion thresholds
of the trial
In order to be included in the trial, patients had to function
poorly in at least one of the areas of physical health, psychiatric status and
social functioning, according to specific threshold values on the corresponding
instruments in these domains. To investigate the extent to which patients improved
in such a way that they did not function poorly anymore according to these inclusion
thresholds, this measure of "poor functioning" was used as an alternative
dichotomous index of treatment outcome in the trial. Figure 16 shows the development
in percentage of patients who did not meet any of the inclusion thresholds for
"poor functioning" anymore in the course of the trial. As before,
these percentages were calculated for the intention-to-treat population, using
LOCF for each missing assessment during the trial.
Figure 16. Percentage of patients no longer meeting the
inclusion thresholds of the trial
Two
months after baseline, 18.4% of the patients in the heroin group had improved
above the level of "poor functioning", compared to 5.1% in the methadone
group. After 12 months of treatment, 32.9% of the patients in the heroin group,
and 13.3% of the patients in the methadone group, no longer met any of the inclusion
thresholds for "poor functioning".
Similar to the analyses conducted earlier for "response", the difference
of 19.6% between the treatment groups at month 12 was tested for significance
in a logistic regression model, with "poor functioning" as the dependent
variable, and treatment site as the only covariate. With non-significant relationships
between "poor functioning" and both treatment site, and treatment-by-site
interaction, and with a good fit of the data to the logistic regression model
(X²=0.80; df=7; p=0.99), the observed difference of 19.6% corresponded with an
adjusted Odds-Ratio of 3.27 (95%-CI: 1.52-7.03; p=0.0024).
- The observed effect in terms of percentage responders
in the co-prescribed injectable heroin group did not go at the expense
of an increased percentage of non-responders who did not improve and
deteriorated.
- In the group with co-prescribed heroin, a considerable
response rate could be observed two months after the start of this treatment.
This response rate gradually increased during the last six months of
the trial. In the methadone alone treatment group, on the other hand,
the early response rate of 20-25% remained largely unchanged until the
end of the trial.
- Response in the co-prescribed heroin group was
based on considerable contributions from each of the outcome domains
of physical health, psychiatric status, and social functioning. The
effect of heroin treatment could not be attributed to a dominating influence
of decreased illegal activities by the participants. The effect of methadone
alone treatment, in contrast, was largely accounted for by contributions
from the physical health domain alone, psychiatric domain alone, or
social domain alone.
- The co-prescribed injectable heroin treatment did
not only result in a higher percentage of responders than the methadone
alone treatment, but also yielded a significantly higher percentage
of multi-domain responders.
- Co-prescribed injectable heroin treatment resulted
in a significantly higher percentage of stable, sustained responders
than methadone alone treatment.
- Co-prescribed injectable heroin treatment
resulted in significantly more participants who no longer met the inclusion
thresholds for the trial, indicative of "poor functioning",
than methadone alone treatment.
|