Table 2.
Primary and secondary outcomes*
| Variable | Baseline (n=159) | 2 weeks (n=129) | 8 weeks (n=142) | 26 weeks (n=141) | 52 weeks (n=144) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPD | BD | IPD | BD | IPD | BD | IPD | BD | IPD | BD | |||||||||
| Primary outcome | OR (P value) | OR (P value) | OR (P value) | OR (P value) | ||||||||||||||
| % (95% CI) success ZCQ | NA | NA | 67 (55 to 77) | 57 (45 to 69) | 1.64 (0.18) | 63 (51 to 73) | 72 (60 to 81) | 0.73 (0.44) | 64 (53 to 74) | 63 (50 to 74) | 1.20 (0.64) | 66 (54 to 74) | 69 (57 to 78) | 0.90 (0.77) | ||||
| Secondary outcomes | MD | MD | MD | MD | ||||||||||||||
| Mean (95% CI) MRDQ (23 points) score | 13.0 (11.7 to 14.2) | 14.4 (13.3 to 15.5) | 10.4 (9.2 to 11.8) | 10.6 (9.3 to 12.0) | 0.1 | 7.5 (6.1 to 9.0) | 6.5 (5.3 to 7.8) | 1.0 | NA | NA | NA | 6.9 (5.4 to 8.5) | 8.1 (6.6 to 9.7) | 1.2 (0.28)† | ||||
| Mean (95% CI) VAS back pain (0-100 mm) | 60 (37 to 83) | 49 (44 to56) | 32 (27 to38) | 33 (28 to39) | 1 | 24 (19 to 30) | 23 (17 to 28) | 1 | NA | NA | NA | 23 (17 to 29) | 31 (24 to 37) | 8 (0.09)† | ||||
| Mean (95% CI) VAS leg pain (0-100 mm) | 52 (47 to 59) | 58 (52 to 64) | 23 (18 to 28) | 26 (20 to 32) | 3 | 26 (20 to 32) | 22 (18 to 27) | 4 | NA | NA | NA | 23 (17 to 30) | 26 (20 to 33) | 3 (0.54)† | ||||
| OR (P value) | OR (P value) | |||||||||||||||||
| Mean (95% CI) Likert % perceived success‡ | NA | NA | NA | NA | NA | 51 (40 to 63) | 53 (41 to 64) | 0.94 (0.85) | NA | NA | NA | 56 (45 to 67) | 49 (38 to 60) | 1.37 (0.37) | ||||
BD=bony decompression; IPD=interspinous process device; MD=mean difference; MRDQ=modified Roland Disability Questionnaire; NA=test not administered; OR=odds ratio; VAS=visual analogue scale; ZCQ=Zurich Claudication Questionnaire.
*Outcomes were analyzed with generalized estimating equations (GEE). Outcome was reported as odds ratio for better success rate when treated with IPD versus bony decompression and overall P value (based on GEE) of interaction between two groups based on continuous outcome scales with mean differences (MRDQ and VAS).
†Overall score in continuous outcome scales were not significant (MRDQ and VAS).
‡Likert global perceived recovery defined by 7 point scale from “worse” to “complete” recovery; score was dichotomized between good recovery (1-2) and bad recovery (3-7).