Background
Specific objectives
Methods
Inclusion criteria
-
Design: prospective cohort study;
-
Population: patients consulting with spinal pain (back or neck), shoulder pain, or knee pain in primary care;
-
Prognostic factor measurement: at least two assessments of pain: at baseline and short-term (within 4–6 weeks of baseline assessment);
-
Outcome ascertainment: Longer-term follow-up (at least 6 months after baseline assessment).
Statistical Analysis
Objective 1: Comparison of repeat assessment with baseline value
Objective 2: Comparison of repeat assessment with multivariable baseline prognostic model
Objective 3: Scenario combining optimal baseline prediction with brief repeated assessment
Results
Descriptive Results
Reference | Setting and study design | Sample | Follow-up | Outcome measures | Absolute scores (mean (SD)) on disability outcome | N(%) of participants meeting 30% improvement cut-off at long-term follow-up |
---|---|---|---|---|---|---|
Dunn & Croft 2006 [22] | Primary care; Prospective cohort | 430 patients with LBP (mean age 44.7 years; 47% male) | Four weeks and 12 months | Disability (RMDQ) – score of between 0–24 with a higher score indicating higher disability | Baseline: 9.02 (6.63) 4 weeks: 7.48 (6.48) 12 months: 6.61 (6.74) | 228 (24) at 12 months |
Kuijpers et al. 2006 [20] | Primary care; prospective cohort | 512 patients with a new episode of shoulder pain (mean age 51.49 years; 49.7% male) | Six weeks and six months | Recovery: Two categories (1: 0 = Not recovered and 1 = Recovered; 2: Global perceived recovery based on a 7-point Likert scale (6 = very much improved to 0 = very much deteriorated)) Disability (SDQ – score of between 0–100 – higher score indicating higher disability) | Baseline: 59.89 (24.21) 6 weeks: 40.32 (30.28) 6 months: 27.30 (31.26) | 354 (60) at six months |
Swinkels-Meewisse et al. 2006 [23] | Primary care; Prospective inception cohort | 300 patients with acute LBP (less than four weeks duration) (mean age 43.01 years; 58.2% male) | Six weeks and six months | Disability (RMDQ, see Dunn) | Baseline: 13.15 (5.02) 6 weeks: 6.15 (5.45) 6 months: 4.42 (5.17) | 303 (55) at six months |
van der Windt et al. 1996 [19] | Primary care; Prospective cohort | 300 patients with a new episode of shoulder pain (mean age 49.52 years; 44.4% male) | Four weeks and 12 months | Recovery: Two categories (1: 0 = Persisting symptoms and 1 = Full recovery/much improved; 2: Global perceived recovery based on a 5-point Likert scale (5 = Much improved to 0 = Much deteriorated)) Disability (SDQ, see Kuijpers) | Baseline: 66.6 (23.4) 4 weeks: 47.1 (31.3) 12 months: 29.0 (30.3) | 214 (64) at 12 months |
Objective 1
Study Author | Prediction model | OR (95% CI) | Goodness-of-fit statistics | c-statistic (95% CI) | Comparison of c-statistics (p-value) | |
---|---|---|---|---|---|---|
Hosmer & Lemeshow test | Nagelkerke pseudo R square | |||||
Dunn & Croft [22] LBP, RMDQ outcome | Baseline Pain Score (0–10) (n = 430) | 0.94 (0.88 to 1.01) |
X
2
(7) = 8.18, p = 0.32 | 0.01 | 0.55 (0.49 to 0.60) | Baseline vs. 4w pain, p < 0.001* Baseline vs. Change, p = 0.24 Change vs. 4w pain, p = 0.06 |
4w** Change in Pain (n = 332) | 1.21 (1.10 to 1.33)* |
X
2
(6) = 4.85, p = 0.56 | 0.07 | 0.62 (0.56 to 0.68) | ||
4w Pain Score (n = 334) | 0.78 (0.72 to 0.85)* |
X
2
(7) = 13.48, p = 0.06 | 0.13 | 0.68 (0.63 to 0.74) | ||
Kuijpers et al. [20] Shoulder pain, SDQ outcome | Baseline Pain Score (0–10) (n = 586) | 0.92 (0.85 to 0.98)* |
X
2
(6) = 6.10, p = 0.41 | 0.01 | 0.56 (0.51 to 0.61) | Baseline vs. 6w pain, p <0.001* Baseline vs. Change, p = 0.30 Change vs. 6w pain, p = 0.02* |
6w Change in Pain (n = 477) | 1.17 (1.09 to 1.26)* |
X
2
(7) = 5.09, p = 0.65 | 0.05 | 0.61 (0.56 to 0.67) | ||
6w Pain Score (n = 478) | 0.78 (0.72 to 0.85)* |
X
2
(6) = 3.68, p = 0.72 | 0.11 | 0.67 (0.62 to 0.72) | ||
Swinkels-Meewisse et al. [23] LBP, RMDQ outcome | Baseline Pain Score (0–10) (n = 300) | 0.99 (0.98 to 1.01) |
X
2
(8) = 6.99, p = 0.54 | 0.01 | 0.56 (0.47 to 0.64) | Baseline vs. 6w pain, p <0.001* Baseline vs. Change, p = 0.03* Change vs. 6w pain, p = 0.09 |
6w Change in Pain (n = 279) | 1.03 (1.02 to 1.04)* |
X
2
(8) = 11.20, p = 0.19 | 0.13 | 0.71 (0.63 to 0.79) | ||
6w Pain Score (n = 281) | 0.97 (0.95 to 0.98)* |
X
2
(8) = 6.25, p = 0.62 | 0.20 | 0.77 (0.70 to 0.84) | ||
van der Windt et al. [19] Shoulder pain, SDQ outcome | Baseline Pain Score (0–10) (n = 293) | 0.83 (0.74 to 0.94)* |
X
2
(5) = 5.24, p = 0.39 | 0.05 | 0.62 (0.55 to 0.69) | Baseline vs. 4w pain, p = 0.44 Baseline vs. Change, p = 0.22 Change vs. 4w pain, p <0.001* |
4w Change in Pain (n = 280) | 1.07 (0.97 to 1.16) |
X
2
(7) = 7.04, p = 0.43 | 0.01 | 0.55 (0.48 to 0.63) | ||
4w Pain Score (n = 287) | 0.83 (0.75 to 0.91)* |
X
2
(6) = 3.50, p = 0.74 | 0.08 | 0.65 (0.58 to 0.72) |
Objective 2
Study author | Prediction model | OR (95% CI) | Goodness-of-fit statistics | c-statistic (95% CI) | Comparison of c-statistics (p-value) | |
---|---|---|---|---|---|---|
Hosmer & Lemeshow test | Nagelkerke pseudo R square | |||||
Kuijpers |
Full model (n = 441)
Baseline pain (0–10) Duration of complaint: - less than 5 weeks - 6–11 weeks - more than 3 m Concomitant LBP (yes/no) Shoulder pain score at physical examination (0–14) Gradual onset (yes/no) | ᅟ 0.89 (0.82 to 0.98)* ᅟ 1 0.68 (0.40 to 1.18) 0.36 (0.22 to 0.60)* 0.52 (0.32 to 0.84)* 0.96 (0.91 to 1.01) ᅟ 0.60 (0.39 to 0.95)* | ᅟ
X
2
(8) = 10.45, p = 0.24 | ᅟ 0.19 | ᅟ 0.71 (0.67 to 0.76) | |
Full model plus 6w Pain Score (n = 441) | 0.72 (0.65 to 0.80)* |
X
2
(8) = 10.89, p = 0.21 | 0.29 | 0.78 (0.74 to 0.83) | Full model vs. full model plus 6w pain score, p <0.001* | |
Van der Windt |
Full model (n = 282)
Baseline Pain (0–10) Co-existing neck pain (yes/no) Preceding trauma (yes/no) Diagnosis Acute Bursitis (yes/no) | ᅟ 0.80 (0.69 to 0.91)* 0.47 (0.27 to 0.82)* 7.44 (1.69 to 32.85)* 2.32 (1.03 to 5.20)* | ᅟ
X
2
(8) = 9.05, p = 0.34 | ᅟ 0.17 | ᅟ 0.72 (0.66 to 0.78) | |
Full model plus 4w Pain Score (n = 270) | 0.85 (0.76 to 0.95)* |
X
2
(8) = 15.07, p = 0.06 | 0.22 | 0.75 (0.69 to 0.82) | Full model vs. full model plus 4w pain score, p = 0.10 |
Objective 3
Hypothetical scenario | Classification | Frequency n(%) | Perceived recovery at six months | |
---|---|---|---|---|
Not recovered N = 209 n(%) | Recovered N = 232 n(%) | |||
Stage 1 Full prediction model at baseline (n = 441) | 1 – Low probability of recovery | 87 (19.7) | 64 (73.6) | 23 (26.4) |
2 – Intermediate probability of recovery | 229 (51.9) | 116 (50.7) | 113 (49.3) | |
3 – High probability of recovery | 125 (28.3) | 29 (23.2) | 96 (76.8) | |
Stage 1&2 Full model plus re-assessment of pain at six weeks – Intermediate group only (n = 229) | 1 – Low probability of recovery | 52 (22.7) | 40 (76.9) | 12 (23.1) |
2 – Intermediate probability of recovery | 121 (52.8) | 67 (55.4) | 54 (44.6) | |
3 – High probability of recovery | 56 (24.5) | 9 (16.1) | 47 (83.9) | |
Model Stage 1&2 Full model plus re-assessment of pain at six weeks (n = 441) | 1 – Low probability of recovery | 87 + 52 = 139 (31.5) | 64 + 40 = 104 (74.8) | 23 + 12 = 35 (25.2) |
2 – Intermediate probability of recovery | 121(27.4) | 67 (55.4) | 54 (44.6) | |
3 – High probability of recovery | 125 + 56 = 181 (41.0) | 29 + 9 = 38 (21.0) | 96 + 47 = 143 (79.0) |