Oxford Shoulder Instability Score

Original Editor - Heba El Saeid

Top Contributors - Heba El Saeid and Kim Jackson  

Objective

A patient-reported measure (questionnaire) of shoulder dislocation/subluxation events, pain, and function domains combined in one composite scale for assessing outcomes of shoulder surgery for conditions associated with shoulder instability, including recurrent dislocation/subluxation.

Intended Population

The Oxford Shoulder Instability Score (OSIS) is a short, 12-item, condition-specific, patient reported outcome (PRO) measure developed and validated for measuring surgical and non-surgical therapeutic outcomes of patients presenting with unidirectional or multi-directional instability of the shoulder.

Method of Use

The Oxford shoulder instability score (OSIS) consists of 12 questionnaire items with 5 ordinal response options each. Recall period “during the last 6 months” for 1 item; “during the last 3 months” for six items, and “during the last 4 weeks” for five items. The scoring of each item is 0–4 (4 = best/least problems).

All item scores are summed to produce a single composite scale of 0–48 (48 = best/least problems).

Reliability

Internal consistency: Cronbach’s alpha was 0.91 at the pre-intervention assessment (n = 92) and 0.92 at the 6-month follow-up (n = 64). All items correlated with the total score of >0.5. Test-retest reliability is assessed in the developmental study on n = 43 with 24 h between assessments. The estimated mean (0.26) of score differences was not significantly different from 0. The coefficient of reliability was calculated as 5.7 using the Bland and Altman (19861) method, and 95 % of score differences fell between 0 and 5.7. Overall, 88 % of score differences lay between 0 and 4 points. The intraclass correlation coefficient (ICC) has since been calculated as 0.90 (95 % C.I. 0.74, 0.96).

Validity

Evidence of convergent validity included large correlations between the OSIS and the clinically assessed Rowe standard rating scale (pre-op, r = 0.51, p < 0.01; post-op, r = 0.84, p < 0.01) and the Constant shoulder score (pre-op, r = 0.56, p < 0.01; post-op, r = 0.76, p < 0.01). There were also moderate to high correlations (r > 0.69 to r = 0.71, p < 0.01) between the preoperative OSIS questionnaire and relevant domains of the SF-36 Health Survey (particularly physical function, role physical, and pain domains).

Responsiveness

Patients reported substantial improvement at the 6-month follow-up assessment. The effect size (0.8) was large for the OSIS questionnaire, although the effect size for the clinically assessed, shoulder instability-specific Rowe was even larger at 1.2. The constant clinical score attained only a small effect size: 0.2. The OSIS effect size was much larger than any attained by individual subscales of the SF-36 questionnaire, where the largest effect size was 0.5 (for the role physical domain). The change scores for the OSIS were significantly greater (p < 0.01) for patients who rated their treatment as “very successful” compared with all other patient responses to that item. Change scores on all other measures (either patient or clinician assessed) did not detect this difference. The change scores for the OSIS were also significantly greater (p < 0.001) for patients who reported the most improvement in their shoulder problem using a Transition questions or items compared with lesser ratings. Here, neither clinical assessment distinguished these differences, while the SF-36 health perceptions and role physical domains did produce significant differences (p < 0.05) in their change scores when comparing these groups.