Introduction
The incidence of first-time traumatic anterior shoulder dislocation (FTASD) is around 23 per 100 000 person-years1 with increased dislocation rate in contact athletes.2 Recurrent instability following such injuries ranges from 26%3 to 92%.4 This wide variation may be explained by the heterogeneous populations in these studies.5 Risk factors for recurrent instability following an FTASD have been described in recent systematic reviews and meta-analyses.6–8 These risk factors may be categorised as either modifiable or non-modifiable. Modifiable risk factors include manual occupations,9 occupations where the upper limb is used above shoulder height,10 immobilisation following the dislocation,11 involvement in collision sport12 and time to return to sport.12 Psychosocial factors such as higher levels of pain13 and fear of reinjury14 15 may also be modifiable risk factors. Non-modifiable risk factors include greater tuberosity fractures,3 9 11 12 16 Hill-Sachs lesions,11 16 17 bony Bankart lesions,9 11 17 18 axillary nerve palsy,12 18 age,3 9 11 16–18 gender3 9 11 12 and hypermobility.12 17
While knowledge of individual risk factors is important in clinical practice, prognostic models use multiple predictive factors to calculate risk of recurrent instability in individual patients.19 Personal, customised care is a goal of clinical practice.20 Customising healthcare carries the costs of additional communication, cognition, coordination and capability.20 21 However, these costs can be minimised through the use of decision-making aids which enable patients and clinicians to make informed treatment choices.
Decision-making tools and algorithms have been developed to assist clinicians and to enable patients with shoulder instability to make informed choices about their healthcare.22–30 Of these, only two5 27 tools have been developed to predict outcomes or management for people with an FTASD. These tools5 27 primarily use age and gender to predict recurrent instability. Incorporation of other known significant risk factors into a prognostic model would improve the accuracy of identifying those more likely to have recurrent instability and enhance clinical decision-making for people following an FTASD.
There are very few high-quality prospective studies examining risk factors for recurrent instability following an FTASD.6 7 We aimed to develop a multivariate prognostic tool which can be used to predict recurrent instability, based on known risk factors. We hypothesised that key variables, other than the established variables age and gender,12 would predict recurrent instability following an FTASD.