Abstract
Objectives:
There is a paucity of evidence quantifying the utility and value of pre-referral imaging tests among patients with atraumatic shoulder pain. The primary aim of this study was to assess the prevalence and relative influence of radiographs and MRIs obtained prior to referral to an orthopedic specialist among patients with atraumatic shoulder pain. We hypothesized that there are no patient factors independently associated with undergoing a pre-referral MRI or radiograph. Similarly, there are no factors associated with change of care based on imaging findings, nor do patient factors influence post-referral imaging.
Methods:
A total of 665 consecutive new adult patients age 35 and older presenting to a single specialty care clinic were included in the study cohort. PROMIS Global Health, PROMIS Mental Health, Patient Health Questionnaire (symptoms of depression), and Generalized Anxiety Disorder questionnaire were completed at the first specialty care visit. Pre-referral shoulder imaging studies were obtained in 462 of 665 (69%) patients; 49% (n = 324) had radiographs and 21% (n = 138) had MRI. We used logistic regression to identify factors independently associated with prereferral imaging and post-referral imaging. We also recorded whether or not pre-referral imaging changed treatment.
Results:
Pre-referral MRI was associated with worse overall health (P=0.012), greater symptoms of depression (P=0.049), and longer duration of symptoms (P=0.0072). Pre-referral radiographs were associated with worse PROMIS Global Physical Health (P<0.001), worse PROMIS Global Mental Health (p=0.012), greater symptoms of anxiety (p=0.055), greater symptoms of depression (p=0.026), and longer duration of symptoms (p<0.005). There was no relationship between diagnosis and pre-referral MRI or radiographs (p=0.34, p= 0.17). There were no factors associated with post-referral MRI. Post-referral radiograph was associated with worse PROMIS Global Mental Health (p=0.013) and duration of symptoms (p=0.014). Eight of the pre-referral radiographs (2.5%) helped distinguish arthritis from adhesive capsulitis, and six (4%) of the pre-referral MRIs were useful for planning surgery. Ultimately, no images were necessary prior to referral, all could have waited for an order from the specialist.
Conclusions:
The finding that patient factors, including symptoms of depression or anxiety and poorer overall health correlated with pre-referral imaging suggests that non-specialists may feel pressed to order imaging when patients are more incapable, worried and despairing, perhaps to demonstrate compassion, and possibly in an attempt to prioritize specialty care. Specialists were less influenced by patient factors in ordering imaging tests, perhaps due to a more selective and discerning use of tests largely for planning surgery on discrete pathology. Given that pre-referral imaging does little to influence specialist management, a preferred approach may be immediate specialist referral or electronic consultation without imaging when typical supportive treatments such as a simple set of rotator cuff strengthening exercises are unsatisfying to the patient.