How many of us have heard patients tell us they'll be relieved if they could just get an MRI to "tell them what's wrong."? I'll be the first to admit I too jumped on this bandwagon early in my career. I was very eager to see an individuals MRI report to correlate their clinical symptoms. You've heard me quote a wise old PT friend of mine who said "We don't treat MRIs". Well RV, this one goes out to you. ;)
This fascination with "seeing" what's wrong has lost a bit of its luster as I've read some very interesting reports on the lack of agreement between MR and clinical presentations. We've seen this phenomenon in low back and shoulder pain where imaging studies don't necessarily correlate with the patients clinical presentation.
It'd been a while since I've seen a good study on this topic so I figured I'd look up what the American Journal of Roentgenology had to say on the issue. The article takes a whopping 1079 consecutive patients referred for shoulder MRI. The subjects were asked to complete validated shoulder questionnaires regarding pain and disability. These were correlated with the radiologists' MR findings.
What did the authors conclude?
- MRI is highly accurate at detecting the presence of a partial or full-thickness rotator cuff lesion.
- There was no relationship between pain or disability with the size or location of the rotator cuff tear.
- Rotator cuff lesions may be thought of as a natural correlate of aging
- It is not clearly identified why some rotator cuff lesions are symptomatic while others are "silent"
- Factors such as bursitis, capsuloligamentous lesions, or cartilage lesions may confound the findings of MRI as they pertain to the rotator cuff
- There may be "no relationship between rotator cuff tear size and the inflammatory reaction responsible for the pain and disability, like low back pain is unrelated to the size of disc herniations."
What are the author's recommendations?
- "Despite the absence of correlation between the size of the rotator cuff tears and the level of disability,MRI provides important data that may affect the management of rotator cuff lesions and should be performed before rehabilitation or surgery."
Really AJR? Let me make sure I understand. There is little to no correlation between imaging and symptoms, yet folks should go ahead and get the MRI anyway? Even before rehab? That's a hard one to swallow. In his defense of the study's author, the investigation took place in France within a socialized health care system where utilization would doubtless be far different from ours.
Regardless of the author's curious conclusion, the study is an honest representation of how limited the value of MRI can be for common musculoskeletal complaints. It also underscores the importance for physical therapists not to get too caught up in hounding the referring physician for the MRI report.
Treat the patient, not the report!
Krief, O.P. (2006). Shoulder Pain and Disability: Comparison with MR Findings. American Journal of Roentgenology, 186(5), 1234-1239. DOI: 10.2214/AJR.04.1766