GAF scores were available for 396 patients. Demographics and baseline characteristics for these patients are shown in table 1.
Table 1: Demographics and baseline characteristics of patients with evaluable GAF scores.
|
Total population (n = 396)
|
Age (years), mean (SD)
|
38.7 (14.6)
|
Gender male, n (%)
|
238 (60.1)
|
Body Mass Index (kg/m²), mean (SD)
|
29.2 (6.9)
|
Age at diagnosis (years), mean (SD)
|
29.1 (11.6)
|
Duration of disease since diagnosis (years), mean (SD)
|
9.6 (10.2)
|
BPRS score at baseline, mean (SD)
|
48.1 (15.6)
|
CGI-S score at baseline, mean (SD)
|
4.47 (0.90)
|
GAF score at baseline, mean (SD)
|
47.7 (13.4)
|
BPRS, Brief Psychiatric Rating Scale; CGI-S, Clinical Global Impression – Severity; GAF, Global Assessment of Functioning; SD, standard deviation.
We compared the data of the original studies to see if they were similar enough to support pooling. Relevant GAF data are given in table 2.
Table 2: GAF data from the original studies.
Patients with baseline assessment
|
Patients from Canadian NIS (n = 169)
|
Patients from German NIS (n = 238)
|
p-value
|
GAF at baseline, mean (SD)
|
48.7 (12.6)
|
46.9 (13.9)
|
|
Patients with baseline and at least one post-baseline assessment
|
Patients from Canadian NIS (n = 168)
|
Patients from German NIS (n = 228)
|
|
GAF at baseline, mean (SD)
|
48.7 (12.6)
|
47.0 (13.9)
|
0.22a
|
GAF at month 6, mean (SD)
|
58.0 (13.8)
|
60.5 (17.1)
|
0.04a
|
Responders with improvement by at least 10 points, n (%)
|
75 (44.6)
|
129 (56.6)
|
0.02b
|
aWilcoxon Two-sample test; bFisher’s exact test. GAF, Global Assessment of Functioning; SD, standard deviation; NIS, non-interventional study.
Total population
In the analyzed population (n = 396), the mean GAF score at baseline was 47.7 (SD 13.4) [95% confidence interval [95% CI], 46.4-49.0] (figure 1). During six months of treatment with AOM, the mean GAF score increased to 59.4 (SD 15.8) [95% CI, 57.9-61.0]. Compared to baseline, improvements at month 3 and month 6 were statistically significant (p < 0.001).
Patients ≤35 years and >35 years
Patients ≤35 years started out with a mean GAF score of 49.0 (SD 12.7) [95% CI, 47.2-50.8] at baseline, while patients >35 years had a mean GAF score of 46.4 (SD 13.9) [95% CI, 44.5-48.3] (figure 2). Patients ≤35 years improved by 13.3 points (SD 16.4) [95% CI, 10.9-15.6] during 6 months of AOM treatment, reaching a mean score of 62.3 (SD 15.1) [95% CI, 60.1-64.4], compared to an improvement of 10.3 points (SD 13.9) [95% CI, 8.4-12.2] in patients >35 years, who reached a mean score of 56.7 (SD 16.1) [95% CI, 54.5-59.0].
Patient gender
At baseline, male patients had a mean GAF score of 48.0 (SD 13.4) [95% CI, 46.3-49.7], while female patients had a mean GAF score of 47.2 (SD 13.3) [95% CI, 45.1-49.3] (figure 3). During 6 months of treatment with AOM, male patients improved by 10.3 points (SD 14.8) [95% CI, 8.5-12.2] and reached a mean score of 58.4 (SD 15.8) [95% CI, 56.3-60.4], while female patients improved by 13.9 points (SD 15.8) [95% CI, 11.4-16.4] and reached a mean score of 61.1 (SD 15.8) [95% CI, 58.6-63.5].
Disease duration
Patients with a disease duration of ≤5 years had a mean GAF score of 47.9 (SD 13.5) [95% CI, 45.9-49.9] at baseline, while patients with a disease duration of >5 years had a mean GAF score of 47.6 (SD 13.3) [95% CI, 45.8-49.4] (figure 4). Patients with a disease duration of ≤5 years improved by 12.8 points (SD 16.3) [95% CI, 10.4-15.2] during 6 months of AOM treatment, reaching a mean score of 60.7 (SD 15.9) [95% CI, 58.4-63.0]. Patients with a disease duration of >5 years improved by 10.9 (SD 14.3) [95% CI, 9.0-12.8], reaching a mean of 58.5 (SD 15.6) [95% CI, 56.4-60.6] at month 6.
Baseline severity
Patients with greater disease severity at baseline had on average lower GAF scores than patients with less severe disease (figure 5). Patients with all levels of disease severity improved on the GAF during treatment, and intergroup differences decreased during treatment. Whereas higher CGI scores seem to correlate with greater improvements on the GAF scale, no significance was determined.
Responders
51.5% of the patients improved by at least 10 points, which corresponds to one descriptor interval on the GAF during 6 months, and were considered “responders” (figure 6). Among patients ≤35 years, 56.2% achieved a 10-point improvement, and among patients >35 years this was the case for 47.0%. Sensitivity analyses based on higher cut-offs revealed that 96 patients (24.2%) achieved an improvement of 20 points or more, 60 patients (15.2%) 30 points or more, and 28 patients (7.1%) 40 points or more.
For further sensitivity analysis, we looked at the proportion of patients in the total sample that moved to a higher descriptor interval during 6 months, irrespective of the actual number of points gained. This was the case for 250 patients (63.1%). 114 patients (28.8%) remained within the same interval during 6 months, and 32 patients (8.1%) moved to a lower interval. These changes did not depend on the baseline GAF score.
Composite responders
40.5% of the patients were considered “responders” on both the GAF scale and the Brief Psychiatric Rating Scale (BPRS) (figure 7), which means that they improved by at least 10 points on the GAF and by at least 20% on the BPRS. Among patients ≤35 years of age, 43.8% met the composite response criterion, which was true of only 37.3% of patients >35 years of age.
Functional remission
We analyzed the proportion of patients reaching >60, >70, or >80 points on the GAF at 6 months, representing possible criteria for functional remission. 179 patients (45.2%) reached >60 points, 82 patients (20.7%) reached >70 points, and 36 patients (9.1%) reached >80 points.