Increased actual sling wear is associated with better early patient-reported and image-based outcomes after shoulder
Tristan B Weir1, Blessing Enobun1, Michael G Livesey1
1Department of Orthopaedics, University of Maryland School of Medicine, Baltimore, MD, USA.
Background:
Prior studies have failed to show differences in functional outcomes for patient-reported sling use after rotator cuff repair. Temperature-sensing devices are used to more accurately measure brace adherence. The purposes of this study were to quantify actual sling adherence and its predictors and to establish whether increased sling adherence is associated with improved functional and image-based outcomes.
Methods:
We performed a prospective cohort study of 65 patients undergoing shoulder surgery requiring ≥4 weeks of postoperative sling use. Temperature-sensing devices were implanted in the slings to monitor sling adherence. Patient-reported sling adherence was determined from a questionnaire. Patients were considered 80% adherent if they wore the sling 16 h/d (112 h/week) when 20 h/d was prescribed. The primary outcomes were patient-reported and actual sling adherence, patient-reported outcomes (American Shoulder and Elbow Surgeons score and visual analog scale pain score) within 12 months postoperatively, and image-based failure based on ultrasound or radiography at 6 weeks and 1 year postoperatively.
Results:
Patient-reported sling adherence was highly sensitive (82.8%), was poorly specific (28.6%), had low accuracy (53.1%), and was weakly correlated with actual sling adherence (r = 0.32, P = .009). On multivariable logistic regression analysis, male patients were 91% less likely than female patients to be adherent with sling use (odds ratio, 0.09; 95% confidence interval [CI], 0.02-0.42; P = .002). Additionally, obese and morbidly obese patients were 88% (95% CI, 0.02-0.84; P = .033) and 98% (95% CI, 0.002-0.27; P = .003), respectively, less likely than non-obese patients to adhere to sling wear postoperatively. After we controlled for surgical procedure, visual analog scale pain scores were significantly better at 6 weeks (β = -1.47; 95% CI, -2.88 to -0.05; P = .04) and 3 months (β = -1.68; 95% CI, -3.28 to -0.08; P = .04) if patients adhered to sling wear. A receiver operating characteristic curve showed that 13.6 hours and 15.4 hours of daily sling wear optimized image-based outcomes at 6 weeks (failure rate, 0% vs. 16%; P = .01) and 1 year (failure rate, 3% vs. 28%; P = .008) postoperatively, respectively.
Conclusion:
The results of this study demonstrate that patient-reported sling adherence is unreliable, adherence can be predicted by female sex and lower body mass index, and increased sling adherence is associated with improved early pain scores and image-based outcomes. These data can help inform future studies using postoperative sling protocols as patient-reported sling adherence is not an accurate method to assess sling use.
