Table 3 Analysis characteristics of included studies
Analysis characteristics | Rare (n = 13) | Non-rare (n = 61) | Total (n = 74) |
|---|---|---|---|
Definition of treatment response, n (%) | |||
P value < 0.05 | 4 (30.8) | 11 (18.0) | 15 (20.2) |
P value < 0.1 | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Statistical significance (p value unspecified) | 0 (0.0) | 3 (4.9) | 3 (4.1) |
Other statistical definition | 1 (7.7) | 2 (3.3) | 3 (4.1) |
Specified numeric change in outcome | 1 (7.7) | 10 (16.4) | 11 (14.9) |
Favourable response in all cycles | 0 (0.0) | 4 (6.6) | 4 (5.4) |
Favourable response in set number of cycles | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Specified symptom change | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Not given | 7 (53.8) | 28 (45.9) | 35 (47.3) |
Analysis approach, n (%)* | |||
Bayesian model | 4 (30.8) | 7 (11.5) | 11 (14.9) |
Non-parametric test | 1 (7.7) | 8 (13.1) | 9 (12.2) |
t-test | 1 (7.7) | 16 (26.2) | 17 (23.0) |
Graph or visual examination | 0 (0.0) | 4 (6.6) | 4 (5.4) |
Regression model | 4 (30.8) | 13 (21.3) | 17 (23.0) |
No formal statistical analysis | 2 (15.4) | 8 (13.1) | 10 (13.5) |
Linear mixed model | 0 (0.0) | 3 (4.9) | 3 (4.1) |
Random subjects effects model | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Randomisation test | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Time series analysis | 0 (0.0) | 1 (1.6) | 1 (1.4) |
Not given | 1 (7.7) | 7 (11.5) | 8 (10.8) |
Individual data pooled, n (%) | |||
Yes | 6 (46.2) | 26 (42.6) | 32 (43.2) |
No | 2 (15.4) | 12 (19.7) | 14 (18.9) |
Planned | 4 (30.8) | 12 (19.7) | 16 (21.6) |
Not applicable (single patient) | 1 (7.7) | 11 (18.0) | 12 (16.2) |
Numerical results reported, n (%) | |||
Yes | 8 (61.5) | 31 (50.8) | 39 (52.7) |
No | 0 (0.0) | 16 (26.2) | 16 (21.6) |
Not applicable (protocol only) | 5 (38.5) | 14 (23.0) | 19 (25.7) |
P values reported, n (%) | |||
Yes | 3 (23.1) | 24 (39.3) | 27 (36.5) |
No | 5 (38.5) | 23 (37.7) | 28 (37.8) |
Not applicable (protocol only) | 5 (38.5) | 14 (23.0) | 19 (25.7) |