Intra-articular sacroiliac joint steroid injection in patients with risk factors may have greater effect

Abstract

Background

Low back pain (LBP) is highly prevalent, with the sacroiliac joint (SIJ) implicated in up to ∼30 % of cases. Intra-articular steroid injections are an established treatment for SIJ pain. Several risk factors—including lumbosacral fusion, total hip arthroplasty, pelvic fracture, inflammatory bowel disease, and seronegative spondyloarthropathy—are known to predispose patients to SIJ pain. However, it remains unclear whether these preconditions predict treatment response.

Objective

To determine whether the presence of known SIJ pain risk factors correlates with improved patient-reported outcomes following SIJ intra-articular steroid injection.

Methods

A retrospective cohort study was conducted at the University of Rochester using electronic medical records from September 1, 2013 to February 1, 2023. Patients aged 18–95 years with SIJ pain who underwent fluoroscopic or ultrasound-guided SIJ injection and completed Patient Reported Outcome Measurement Information System (PROMIS) Pain Interference (PI) and/or Physical Function (PF) surveys before and after injection were included. Outcomes were compared between patients with and without preconditions. Statistical significance was set at p < 0.05. Cohen's d test was preformed to determine the effect of having a precondition.

Results

Of 4780 patients identified, 91 completed PROMIS PI and 168 completed PROMIS PF surveys within the required timeframe. Among these, 27 (PI) and 40 (PF) had a precondition. The mean PI score change was 0.854 in the precondition group and 0.509 in the non-precondition group; PF score change was 1.30 and −0.378, respectively. Neither within-group changes nor between-group differences reached statistical significance (PI p = 0.739; PF p = 0.114). The effect size, measured by Cohen's d, of preconditions versus no preconditions, was d = 0.08 in the PI group, indicating little to no effect. The effect size, measured by Cohen's d, of preconditions versus no preconditions, was d = 0.28 in the PF group, indicating a small effect.

Conclusion

The presence of established SIJ pain risk factors did not predict improved response to intra-articular steroid injection when using t-test for comparison. When using Cohen's d test however, a small effect size of (d = 0.28) was shown for increased PF following SIJ injections in patients with preconditions compared to patients without preconditions. This finding may indicate that patients with the preconditions investigated in this study can be expected to have a marginally better physical function following SIJ injection compared to their peers without preconditions. These findings suggest that preconditions such as fusion, arthroplasty, or inflammatory disease may influence injection efficacy. Larger multicenter studies are warranted to validate these results.

1. Background

Low back pain (LBP) is highly prevalent in the United States where the lifetime prevalence is currently 40 % and set to rise substantially [1]. The sacroiliac joint (SIJ) is implicated in ∼31 % of patients with LBP [2]. It has been established that intra-articular steroid injections in the sacroiliac joint are an effective treatment for sacroiliac joint pain [[3], [4], [5]]. The Patient Reported Outcomes Measurement Information System (PROMIS) was developed by the National Institute of Health in 2004 to monitor the effectiveness of medical interventions in a person-centered matter. Kristoff et al. have shown a significant improvement in PROMIS scores at 3-months following SIJ injections [4].

There are certain well established risk factors for developing sacroiliac joint pain including spondyloarthropathies, inflammatory bowel disease, trauma, lumbosacral fusion (LSF) surgery, and hip replacement surgery (THA) [[5], [6], [7], [8], [9], [10], [11], [12], [13]]. The interplay between the lumbar spine, SIJ, and the hip joint has been well established in recent literature, likely attributed to increased motion at the SIJ following fusion and THA [9,12]. A 2023 systemic review of 12 studies found a pooled SIJ pain incidence of 15.8 after lumbar fusion, increasing to 32.9 % when the fusion was extended to the sacrum [10]. Interestingly SIJ fusion has been shown to nearly double the risk of THA dislocation further suggesting bidirectional influence of surgery between the SIJ and hip joint [[11], [12], [13]]. These structures all articulate with the pelvic ring and thus trauma to this structure has also been linked to increased risk for SIJ pain [3]. Unlike the mechanical alterations thought to cause pain in post-operative and post-traumatic patients, the inherent increased systemic inflammation is thought to drive increased SIJ pain in patients with seronegative spondyloarthopathies and IBS [5].

Despite these known risk factors, there remains a paucity of studies evaluating the responsiveness to intra-articular steroid injection based on the presence of one of these known risk factors utilizing patient reported outcomes.

2. Objective

The purpose of this study was to investigate if the presence of certain preconditions would correlate with patient outcomes following SIJ injection compared to the absence of these conditions.

3. Design

This study was a retrospective cohort study of patient charts via the electronic medical record who received a fluoroscopically and/or ultrasound guided sacroiliac joint steroid injection at the University of Rochester between the dates of September 1, 2013 and February 1, 2023 in the Rheumatology, Physical Medicine & Rehabilitation, and Anesthesiology Departments. This study was approved by the Institutional Review Board at the University of Rochester. All procedures were performed in compliance with relevant laws and institutional guidelines including obtaining informed consent from patients prior to procedure. Potential subjects were aged 18–95 years old at time of injection and had a body mass index (BMI) less than 50 were identified. A very high BMI cutoff was included due to the presumed medical comorbidities, substantial biomechanical alterations, and evidence of reduced intraarticular steroid effectiveness with increased BMI [14]. Patients also required a diagnosis of sacroiliac joint pain (ICD.10 code M53.3). Patients were only included if PROMIS Pain Interference (PI) and Physical Function (PF) questionnaires were completed at most 60 days prior to SIJ injection and 90 days following the SIJ injection. Medical records were then reviewed for the presence of precondition defined as a history of lumbosacral fusion (LSF), total hip arthroscopy (THA), pelvic fracture, irritable bowel syndrome (IBS), inflammatory bowel disease (IBD), and seronegative spondyloarthropathy. These conditions were chosen as they are encountered frequent enough in the authors’ non-operative interventional spine clinic in patients with SIJ pain to generate curiosity for study and evidence which shows an increased risk of SIJ pain in patients with some of these preconditions [[5], [6], [7], [8], [9], [10], [11], [12], [13]]. Patients who underwent LSF, THA, and pelvic fracture following SIJ injection were not included. A two tailed equal variance t-test was then preformed comparing the difference between pre and post injection PROMIS PI and PF scores between a group with the presence of a precondition and those without a specified precondition. Statistical significance was attributed for a p value less than 0.05.

4. Results

A total of 4780 patients were identified to have undergone SIJ injection within the given timeframe. Of that sample total of 1175 patients were identified as having a qualifying precondition. When filtering for completed PROMIS PI and PF within the timeframe designated above, 91 patients had completed PROMIS PI and 168 had completed PROMIS PF questionnaires. Within the PI group 27 had the presence of a precondition, while 40 patients had the presence of a precondition in the PF group. A flow diagram depicts this reduction of initial study subjects to the final sample size in Fig. 1. The PI group was 75.8 % female and the PF group was 74.4 % female. The mean difference in PI scores was 0.854 (95 % CI: 0.597, 2.30) in the precondition group and 0.509 (95 % CI: −0.655, 1.67) in the non-precondition group. The mean difference in PF scores was 1.30 (95 % CI: 0.600, 3.20) in the precondition group and −0.378 (95 % CI: −1.37, 0.618) in the non-precondition group. There was not a statistically significant change in pre vs post PI and PF scores. There was not a statistical difference in the difference between pre and post injection PROMIS PI (P = 0.739) and PF (P = 0.114) in patients who had a precondition versus those without the specified preconditions. The results of this comparison are outlined in Table 1.

Fig. 1.

Patient selection flowchart. Diagram illustrating the screening of 4,647 patients and the final inclusion of participants meeting precondition criteria and available outcome metrics for pain interference and physical function.

Table 1.

Comparison of average change in Pain Interference (PI) and Physical Function (PF) scores between patients with and without preconditions.

		N	% Female	Average Change
PI	Preconditions	27	75.8	0.854 (95 % CI: 0.597, 2.30)	P = 0.739
	No Preconditions	64		0.509 (95 % CI: −0.655, 1.67)
PF	Preconditions	40	74.4	1.30 (95 % CI: 0.600, 3.20)	P = 0.114
	No Preconditions	128		−0.378 (95 % CI: −1.37, 0.618)

1: Average change represents the mean difference in pre vs post PI/PF scores. P values represent the comparison of average change between PI/PF in those with preconditions and those without.

The effect size, measured by Cohen's d, of preconditions versus no preconditions, was d = 0.08 in the PI group, indicating no effect. The effect size, measured by Cohen's d, of preconditions versus no preconditions, was d = 0.28 in the PF group, indicating a small effect.

4. Discussion

This retrospective cohort study sought to determine if the presence of certain risk factors for developing SIJ pain would be more responsive to SIJ intraarticular steroid injection compared to the absence of these known risk factors using PROMIS PI and PF scores. This study did not find a statistically significant difference in either PI or PF between patients with and patients without the preconditions studied in this paper using t-test for analysis. This study also found no statistically significant change in PROMIS PI or PF treated with SIJ injection regardless of the presence or absence of a precondition. This finding is in contrast to other studies likely due to decreased sample compared to other higher power studies [4].

One notable finding this study did uncover is a small effect size of (d = 0.28) for increased PF following SIJ injections in patients with preconditions compared to patients without preconditions. This finding may indicate that patients with the preconditions investigated in this study can be expected to have a marginally better physical function following SIJ injection compared to their peers without preconditions.

This study has notable limitations including a sample size limited by loss of participants from the initial subject number of 4780 to 259 analyzed, representing a 95 % sample loss. This sample loss is primarily due to lack of completed PROMIS PI and PF questionaries within a reasonable timeframe. This limited sample size confounds the ability to interpret the results as a small sample size increases the possibility of making a Type II error (failing to reject the null-hypothesis) especially when using t-tests. This study is also limited by the inherent limitations of retrospective study including but not limited to loss to follow-up, limited control of confounding variables, and single center data set. The female predominance in this study can be partially explained by the increased incidence of SIJ pain in women, thought to be due morphological and biomechanical differences [15,16]. It may also be explained by the notion that IBS, one of the preconditions included in this study, is more likely to occur in women [17]. Furthermore, other studies which have used PROMIS as a measure for studying SIJ injections report similar female predominance [4]. Ultimately further investigation is warranted and would likely require the participation of multiple large tertiary treatment centers to obtain a sufficient population to meet the inclusion criteria. Future investigations could be aided by improved collection of PROMIS PI and PF by implementing systems which increase collection of data prior to and within 3 months of injection.