Margit Kooijman

90 Appendix B. Characteristics of studies Author Study quality (%) Study population Duration follow- up/ outcome measures Prognostic factor(s) Strength of association (95% CI) Primary care Brox and Brevik (1996) 78 125 patients; 6 months Multivariate analysis, adjusted for age, gender, symptom duration, baseline Neer score Rotator tendinosis Referred by general practitioner Neer shoulder score (0–100) Succes: ≥80 points Not on sick leave OR = 4.4 (1.6, 12.1) Not on regular medication OR = 4.2 (1.5, 11.1) Active treatment (ref = not active) OR = 4.8 (1.7, 13.6) Years of education, overhead work activity, comorbidity, isometric strength endurance, locus of control beliefs, emotional distress n.s. Croft et al. (1996) 50 166 patients 6 months Shoulder pain General practice Validated 22-item disability questionnaire Baseline disability score >10, symptom duration (>1 month), injection at baseline, previous episodes of shoulder pain, severely restricted passive elevation (<101°). Poorer outcome (P < 0.05) (Beta’s not presented) Kennedy et al. (2006) 89 361 patients max 12 weeks Multivariate analysis (higher log 1 + DASH at discharge) Shoulder complaints Physiotherapy practice DASH at discharge (higher DASH is greater disability) Age (older) β by decade = 0.13 (0.06, 0.19), p = 0.0004 Gender (female) β = -0.24 (-0.45, -0.04), p = 0.0219 Function/disability (baseline DASH) β = 0.02 (0.01, 0.03), p < 0.0001 Workers'compensation claim β = 0.55 (0.18, 0.92), p = 0.0036 Therapist predicts more restricton of patient return to usual activity β = 0.42 (0.21, 0.63), p < 0.0001 Onset, duration of pretreatment symptoms, familiair with shoulder complaints, pain (intensity), pain (global rating injury/pain), pretreatment ROM (restriction), pretreatment weakness (muscle strength), general health (SF 36 physical), additional medical problems (co-morbidity), work status, surgery, GP treatment (prescribed medication), psychological factors (SF 36 mental), patient outcome expectancy (patient prediction recovery) n.s.