| PMC full text: | Spine J. Author manuscript; available in PMC 2011 Feb 1. Published in final edited form as: Spine J. 2010 Feb; 10(2): 117. doi: 10.1016/j.spinee.2009.09.002 |
Table 4
Patients achieving a 50% reduction in symptoms and adjusted odds ratios (95% CI)*
| 8 treatment sessions n (%) | 16 treatment sessions n (%) | Intervention Main Effect: SMT / LM OR (95% CI) | Dose Main Effect: 16 / 8 sessions OR (95% CI) | Interaction Effect: SMT & 16 sessions OR (95% CI) | ||
|---|---|---|---|---|---|---|
| CGH pain scale † | ||||||
| profile | SMT | 10 (53%) | 9 (53%) | 3.6‡ | 1.1 | 1.9 |
| LM | 9 (47%) | 7 (35%) | (1.6, 8.1) | (0.5, 2.4) | (0.4, 9.9) | |
| 12 wk | SMT | 6 (38%) | 6 (35%) | 3.1 | 0.5 | 4.0 |
| LM | 6 (35%) | 1 ( 6%) | (0.9, 9.8) | (0.2, 1.5) | (0.4, 41.2) | |
| 24 wk | SMT | 5 (28%) | 8 (47%) | 3.1 | 1.2 | 4.0 |
| LM | 5 (28%) | 3 (16%) | (0.9, 10.3) | (0.4, 3.6) | (0.4, 37.9) | |
| CGH number (in last 4 wk) | ||||||
| profile | SMT | 16 (84%) | 15 (88%) | 2.1‡ | 1.0 | 1.2 |
| LM | 13 (68%) | 14 (70%) | (1.0, 4.5) | (0.5, 2.0) | (0.3, 4.4) | |
| 12 wk | SMT | 12 (75%) | 10 (59%) | 2.6 | 0.5 | 1.0 |
| LM | 8 (47%) | 6 (38%) | (0.9, 7.4) | (0.2, 1.5) | (0.1, 7.4) | |
| 24 wk | SMT | 10 (56%) | 11 (65%) | 1.8 | 0.8 | 2.7 |
| LM | 10 (56%) | 9 (47%) | (0.7, 5.1) | (0.3, 2.2) | (0.4, 19.5) | |
| Neck pain† | ||||||
| profile | SMT | 7 (37%) | 8 (47%) | 2.5 | 1.2 | 2.7 |
| LM | 7 (37%) | 5 (25%) | (1.0, 6.4) | (0.4, 3.3) | (0.4, 18.4) | |
| 12 wk | SMT | 5 (31%) | 6 (35%) | 2.5 | 1.6 | 1.3 |
| LM | 3 (18%) | 4 (25%) | (0.7, 8.3) | (0.5, 5.1) | (0.1, 13.2) | |
| 24 wk | SMT | 6 (33%) | 8 (47%) | 2.9 | 1.0 | 6.0 |
| LM | 6 (33%) | 4 (21%) | (0.9, 9.3) | (0.3, 3.1) | (0.6, 61.3) | |
SMT - spinal manipulative therapy; LM - light massage; CGH - cervicogenic headache; OR - adjusted odds ratio.
*Outcomes are presented for the 12- and 24-week follow-ups, as well as for the full profile across all follow-ups (4 to 24 weeks). Original data, n (%) are included for the 4 study groups. The main effects are adjusted odds ratios (n = 40 per comparison group). The odds ratios were adjusted in the analysis for baseline differences between groups as in Table 2. The effects for the individual time points were computed using logistic regression. The profile effects were computed across all follow-up time points using logistic regression with generalized estimating equations. Missing data were imputed except for five patients with no follow-up data. A main effect with OR > 1 favors SMT over LM and favors 16 treatment sessions over 8 treatment sessions.
†Modified Von Korff scale (scored from 0 to 100 points with lower score preferable). The pain score is the primary outcome.
‡P < .05.