How much does an MRI change over a period of up to 2 years in patients with chronic low back pain? Is a repeated MRI really necessary in the follow-up of patients with chronic low back pain?

Purpose: Clinical practices vary between healthcare providers when it comes to asking for a Magnetic Resonance Imaging (MRI) during follow-up for chronic low-back pain (LBP). The association between progressive changes on the MRI and the clinical relevance of these findings is not clearly defined. The objective of our study is to investigate to what extent do MRI findings change during a period less than or equal to two years in patients with chronic LBP. We question the efficacy of its routinary use as a tool for follow-up and we also study the correlation between new changes on MRI and modifications in therapeutic attitude.

Methods: Data was collected from 468 lumbar spine MRIs from 209 patients undergoing two or more MRIs between January 2015 and December 2019 with a mean of 2.24 MRIs per patient. The evaluated data included diagnosis, reason for request, MRI findings and treatment offered post-MRI. MRIs were assessed according to a standardized scoring system from 0 to 14 points according to the severity in findings (modified Babinska Score). Radiological changes were defined as increased severity of findings in the most affected segment.

Results: 51.06% of MRI requests had no documented reason to be asked for. The average score of the findings on the first MRI was 5,733 (SD 2,462) and 6,131 (SD 2,376) on the second, not reaching a statistically significant difference (p = 0.062). There was no difference on the findings between the first and the second MRI in 40, 15% (n = 104) and up to 89, 96% with only mild changes (-1/ + 2 points over 14 possibles). After repeating the MRI, no modification to the treatment plan was made in 44, 79% of patients (n = 116) and only in 11.58% (n = 30) was surgical treatment indicated.

Conclusion: The rate of lumbar MRI has risen to an alarming pace without evidence of consequent improvements in patient outcomes. A significant number of repeated MRIs did not show radiological changes, nor did they give rise to further surgical treatment after obtaining these images. This study should help to review the real applications of clinical guides on the appropriate use for image tests.