Genicular nerve block with corticosteroid for chronic knee Pain: Patient-Reported outcomes across 9-years of practice

Abstract

Background

Genicular nerve blocks (GNB) for knee pain secondary to osteoarthritis (OA) are often performed with an anesthetic only, but corticosteroid can be used as an adjuvant with the goal of extending therapeutic relief. The efficacy of GNB with corticosteroid has been established in small, randomized trials, but its effectiveness in practice remains largely unknown.

Objectives

Evaluate the therapeutic effect of GNB with corticosteroid using patient-reported outcomes (PROs) in a large patient sample.

Methods

Data from consecutive GNBs with corticosteroid performed at an academic medical center from 2015 to 2024 was collected retrospectively. We included all patients aged ≥18 who received a GNB with corticosteroid for chronic knee pain due to OA, and excluded diagnostic GNBs with anesthetic alone and procedures performed using non-classical/modified techniques. Comparisons between pre-versus post-procedure PROs were by mixed-effect regression models with multiple testing corrections.

Results

A total of 123 GNBs (96 patients) with corticosteroid were identified. Post-procedurally, the mean NRS reductions were 3.56 (2–30 days), 2.14 (1–3 months), 1.92 (3–6 months), and 1.23 (6–12 months). 50 % of the patients maintained a pain reduction greater than or equal to the NRS MCID of −2.0 for at least 6.9 months. Post-procedure improvements in PROMIS-GH and PHQ9 were not clinically significant. Mean time to total knee arthroplasty was 16.5 months in 13.5 % of patients (13/96). For active opioid users, opioid usage was significantly decreased from the baseline 28.01 ± 13.20 MME/day to 13.58 ± 16.23 MME/day (p < 0.002) in the first 6 months post-GNB, and this opioid reduction was maintained at 15.59 ± 12.51 MME/day (p < 0.05) after the initial 6-month follow-up period.

Conclusion

For the majority of patients who received GNB with corticosteroid, we observed statistically and clinically significant pain reduction for 3 months or more in this cohort. While more comparative studies are required to evaluate its effectiveness, GNB with corticosteroid has promising potential as a safe and effective treatment for chronic knee pain.

1. Introduction

Knee osteoarthritis (OA) accounts for nearly 80 % of the osteoarthritis disease burden worldwide, and its progression results in significant debility, impaired knee function and gait, and reduced quality of life due to chronic knee pain [1,2]. In the spectrum of non-surgical interventions available for knee OA, there have been new developments in recent decades that can help delay the time to a total knee arthroplasty (TKA) and provide options for patients who are not TKA candidates, or for those who prefer non-surgical options. One set of promising non-surgical options is the genicular nerve block (GNB) and geniculate nerve radiofrequency ablation (GNRFA), which target genicular nerves that provide the sensory innervation to the knee joint capsule [3]. A GNB involves injecting a local anesthetic with or without a corticosteroid under image guidance [4]. Classically, genicular nerve interventions are performed using a 3-needle technique to target the superior medial (SMGN), superior lateral (SLGN), and inferior medial genicular nerves (IMGN) [3,5]. Of note, recent clinical advancement suggests that expanded genicular nerve intervention protocols (classical targets as well as nerves to vastus medialis, lateralis, and intermedius) could more completely denervate the knee joint and result in significant long term pain relief [6].

In practice, GNB with a local anesthetic alone is often used as a prognostic tool to predict the success of a GNRFA [7,8]. Pharmacologically, the pain relief from the local anesthetic is intended to be rapid-onset but short-acting, usually wearing off after several hours to a day [9,10]. Although the precise pharmacologic mechanism of the adjuvant corticosteroid is unknown [11], it is known that corticosteroids can signal through local glucocorticoid receptors to induce vasoconstriction, thereby decreasing systemic absorption of the local anesthetic [[12], [13], [14]], and potentially suppress pain transmission via unmyelinated C‐fibers to reduce neural discharge [15]. In the past decade, several randomized-controlled trials demonstrated that knee OA patients may receive pain relief for up to 12 weeks when treated using GNB with an adjuvant corticosteroid, such as dexamethasone, betamethasone, or triamcinolone [[16], [17], [18], [19]]. For instance, a 2018 RCT found that pain scores were significantly lower in the patient group that received lidocaine plus triamcinolone compared to lidocaine alone at 2 and 4 weeks after a GNB [19]. Moreover, a 2022 placebo-controlled RCT also found meaningful pain reduction following GNB using bupivacaine and celestone, with pain score reductions ranging 2.4–3.5 points at 2–8 weeks after intervention [16]. Importantly, the group who received GNB with corticosteroid has significantly lower pain scores compared to the saline placebo group. Although results from the small randomized-controlled trials supported the efficacy of GNB with corticosteroid, there is limited long-term data from large samples in the literature about the effectiveness of this procedure since its initial introduction into clinical practice.

Currently in the United States, the adoption of GNB with corticosteroid adjuvant is subject to variations in provider and geographic practice patterns, and its coverage differs by insurance company. Active debate remains over the addition of a corticosteroid adjuvant to peripheral nerve block procedures such as the GNB. Unlike a diagnostic GNB that is intended to provide short-lasting relief with a local anesthetic alone, the potential advantage of GNB with a corticosteroid adjuvant is the therapeutic potential to provide extended pain relief beyond the immediate post-treatment period.

Given the paucity of large sample data about the effectiveness of GNB with corticosteroid, the primary objective of this study is to evaluate the therapeutic effect of GNB with adjuvant corticosteroid across multiple years of clinical practice and in a larger patient cohort with long-term follow-up. The main outcome measures used to assess treatment effect are patient-reported pain scores and patient-reported quality of life questionnaires collected at post-procedural follow-up visits. Secondary outcome measures include assessing patient-reported percent relief and duration, changes in prescription opioid usage, and time to TKA.

2. Methods

2.1. Overview and study design

A retrospective review of the electronic medical records was conducted to identify all consecutive genicular nerve blocks with corticosteroid for treatment of knee pain from January 2015 to January 2024 at a large academic tertiary care center under a single fellowship-trained physician. Institutional review board approval was obtained before the start of this study.

2.2. Study population

Inclusion criteria were all patients aged 18 years or older diagnosed with knee pain who underwent a genicular nerve block (local anesthetic plus corticosteroid) using the classical 3-needle technique (SMGN, SLGN, IMGN) (Fig. 1). We excluded patients with a Kellgren-Lawrence (KL) grade of <2 to include only patients with definitive knee OA [20,21]. KL grading was performed on weight-bearing knee radiographs by two reviewers (including a musculoskeletal radiologist) who were blinded to the patient-reported outcomes.

Fig. 1.

Knee radiographs with AP views (left top & bottom) and lateral views (right top & bottom) showing the classical 3-needle technique for genicular nerve block targeting the superior medial, superior lateral, and inferior medial genicular nerves.

We excluded patients who received GNBs without corticosteroid (diagnostic GNBs) to focus our investigation on GNBs with anesthetic and corticosteroid. In addition, we excluded GNBs with needle placement beyond the classical targets due to an inadequate sample size of GNBs performed using the non-classical technique, also known as expanded protocols [22]. Manual review of patient charts, radiographic images, radiology reports, and procedural images was performed to ensure accurate adherence to the eligibility criteria.

2.3. Data collection

Following eligibility selection, we collected data on the following pre-procedural variables: basic demographic information, knee OA characteristics (laterality, KL grade at diagnosis), numerical rating scale (NRS) pain score at diagnosis, duration of knee pain prior to procedure, history of knee OA treatments (physical therapy, intra-articular injections, lower extremity surgeries), and comorbidities including diabetes, smoking status, alcohol abuse, fibromyalgia, chronic pain syndrome, gout, meniscal tear (confirmed on MRI), chronic oral corticosteroid use, depression, anxiety disorder, and prescription opioid usage (quantified by morphine milligram equivalents, MME). We also collected data on the following periprocedural variables: type of procedure, laterality, KL grade at procedure, American Society of Anesthesiologists (ASA) classification, type of sedation, anesthetic and corticosteroid type, and pre-procedural NRS pain level and patient-reported outcomes. We also collected data on the following post-procedural variables at follow-up visits: complications, patient-reported outcome measures, adjuvant treatment utilization (prescription opioid usage at follow-up, number of knee OA physical therapy sessions, and other knee injections), and progression to total knee replacement/arthroplasty. All patient-reported outcomes were collected from existing validated and standardized questionnaires, including the numeric rating scale (NRS) for knee pain levels, the patient health questionnaire-9 (PHQ9), and the Patient Reported Outcomes Measurement Information System (PROMIS) Global Health 10-item questionnaire [23]. The PROMIS-Global Health forms include assessments for both Physical Health (PH) and Mental Health (MH). Initially designed to alleviate the burden of questionnaires and enhance overall efficiency, PROMIS-GH forms have been found to be effective at measuring changes in physical health among chronic knee pain patients and individuals undergoing TKA [24,25]. Only pre-procedure PRO scores within 3-months of the procedure were used in analyses because inclusion of older scores may not reflect an accurate baseline comparison.

2.4. Statistical analysis

For GNB outcomes, we grouped all PRO scores (NRS, PROMIS, PHQ-9) into the following post-procedure follow-up periods: 2–30 days, 31–90 days (1–3 months), 91–182 days (3–6 months), 183–365 days (6–12 months), 366+ days. The comparison of paired measurements between pre-versus post-procedure PRO scores (NRS, PROMIS, PHQ-9) and opioid usage was by Wilcoxon signed-rank test. To compare the PRO score changes in each post-procedure period and account for repeat measures, we fitted a linear mixed-effects model in which the post-procedure PRO score change was the dependent variable while post-procedural follow-up period and potential confounding variables (age, sex, BMI, opioid use) were the independent variables. The significance of the fixed effects was evaluated by Wald test.

We used the minimal clinically important difference (MCID) of −2.0 for knee pain NRS scores, as previously reported in literature [[26], [27], [28]]. A non-response subgroup and a corresponding positive response subgroup were selected to compare potential factors associated with limited treatment response. A lack of clinically robust response (non-response) was defined as < 50 % NRS score reduction [28], or < 75 % patient-reported pain relief in the initial 3-month follow-up period. Factors suspected to be associated with no response to treatment were evaluated by Welch's t-test assuming unequal variance. All tests were two-sided. When applicable, multiple comparisons were corrected by the Bonferroni approach. Available-case analysis (pairwise deletion) was used to handle all missing data. Data analyses were performed in R version 4.3.1.

3. Results

3.1. Demographic and procedure characteristics

A total of 123 GNBs (163 knees) from 96 individual patients were eligible for analysis. Of the GNB procedures, 78.0 % (96/123) were first-time procedures and 22.0 % (27/123) were repeat procedures.

The average age in this study was 66.1 ± 13.4 years, with a large portion of patients being female (69.8 %) (Table 1). The racial background of study subjects was diverse with 61.1 % White, 21.1 % Black, 14.6 % multiracial or other, 2.1 % Asian, and 2.1 % Native American. Additionally, 16.7 % of patients were Hispanic while the rest were non-Hispanic. Before their initial GNB, a majority of patients had a documented intra-articular corticosteroid knee injection (65.6 %, 63/96) while some patients also received visco-supplementation (38.5 %, 37/96). Prior to their first GNB, a large proportion of patients in this cohort had a history of anxiety (45.3 %) and depression (58.3 %), and the average duration of knee pain was 5.75 years. More than a third of patients (36.5 %, 35/96) had a history of meniscal tear in the affected knee. Pre-procedurally, 62.1 % of patients underwent physical therapy for an average of 5 sessions within a year prior to their GNB. Of note, 20.8 % of subjects reported active prescription opioid use prior to their first GNB, with an average daily morphine milligram equivalent (MME) of 28.01 ± 13.20.

Table 1.

Demographic and baseline clinical characteristics of patients.

n	96
Sex (%)
Female	67 (69.79)
Male	29 (30.21)
Age (mean (SD))	66.07 (13.35)
Body Mass Index (mean (SD))	34.27 (9.23)
Race (%)
White	58 (61.05)
Black	20 (21.05)
Other/Multi-racial	14 (14.58)
Asian/Pacific Islander	2 (2.11)
Native American	2 (2.11)
Ethnicity (%)
Non-Hispanic	80 (83.33)
Hispanic	16 (16.67)
Diabetic (%)	32 (33.33)
Smoking Status (%)
Current Smoker	10 (10.42)
Ever Smoking (quit smoking for at least 1 Year)	38 (39.58)
Never Smoking	48 (50.00)
History of Meniscal Tear (%)
None	53 (55.21)
Ipsilateral	30 (31.25)
Bilateral	5 (5.21)
Contralateral	8 (8.33)
History of Fibromyalgia (%)	16 (16.67)
History of Chronic Pain Syndrome (%)	20 (20.83)
History of Anxiety Disorder (%)	43 (45.26)
History of Depression (%)	56 (58.33)
History of Alcohol Abuse (%)	8 (8.33)
History of Gout or Pseudogout (%)	15 (15.63)
History of Chronic Corticosteroid Use (%)	2 (2.08)
Prescription Opioid Use (%)
Current	20 (20.83)
Daily Morphine Milligram Equivalents (mean (SD))	28.75 (12.47)
Prior Opioid Use and Discontinue	14 (14.58)
Never	62 (64.58)
Laterality of Knee OA (%)
Bilateral	64 (66.67)
Left	9 (9.38)
Right	23 (23.96)
Duration of Knee Pain, months (mean (SD))	68.78 (47.67)
Physical Therapy Use Before Procedure (%)	59 (62.11)
Number of PT Sessions within 1-year (mean (SD))	5.04 (5.08)
Number of Intra-articular Knee Steroid Injections (%)
None or Unknown	33 (34.40)
1	30 (31.25)
2	19 (19.79)
≥3	14 (14.58)
Number of Knee Visco-supplement Treatments (%)
0	59 (61.46)
1	32 (33.33)
2	4 (4.17)
≥3	1 (1.04)
History of Prior Knee Surgery (%)	52 (54.17)
Type of Prior Knee Surgery
Knee Arthroscopy with meniscectomy or meniscal repair	33 (34.38)
Knee Arthroplasty	23 (23.96)
Ligament Repair Reconstruction	2 (2.08)
Other	1 (1.04)
History of Prior Hip Surgery (%)	15 (15.62)
History of Prior Ankle Foot Surgery (%)	5 (5.21)
Referral Specialty (%)
Orthopaedic Surgery/Sports Medicine	38 (39.58)
Spine Medicine/Physical Medicine & Rehabilitation	24 (25.00)
Primary Care	7 (7.29)
Rheumatology	3 (3.13)
Other	17 (17.71)
No Referral	7 (7.29)

Pre-procedurally, the majority of GNB (59.0 %) patients had severe knee OA (KL grade 4). There were 23 patients (24.0 %) with a history of TKA prior to their first GNB, and the subgroup analysis did not reveal any statistically significant baseline or outcome differences in post-TKA versus native knees (P-value > 0.05). The majority of GNB patients received bupivacaine (97.6 %) as the anesthetic and triamcinolone 40 mg (94.3 %) as the corticosteroid adjuvant (Table 2). The average follow-up duration was 9.4 months. We did not find any documented complications following any GNB procedure. Regarding physical therapy adherence, approximately one-fifth (23 %) of patients were adherent to their recommended post-intervention physical therapy.

Table 2.

Procedure characteristics of genicular nerve blocks (GNB) with corticosteroid.

	GNB
n	123
Procedure Laterality (%)
Bilateral	40 (32.52)
Unilateral
Left	30 (24.39)
Right	53 (43.09)
ASA Physical Status Classification (%)
ASA II	44 (35.77)
ASA III	74 (60.16)
ASA IV	5 (4.07)
Sedation Performed (%)	87 (70.73)
Sedative Type, may be a combination (%)
PO Clonazepam	1 (0.81)
PO Diphenhydramine	1 (0.81)
IV Midazolam	85 (69.11)
IV Fentanyl	0 (0.00)
None	36 (29.27)
Nerve Block Anesthetic (%)a
Bupivacaine (0.75 %)	120 (97.56)
Lidocaine (1 %)	3 (2.44)
Steroid Type (%)a
Triamcinolone 40 mg (%)	116 (94.31)
Triamcinolone 80 mg (%)	1 (0.81)
Methylprednisolone 40 mg (%)	2 (1.63)
Methylprednisolone 80 mg (%)	3 (2.44)
Betamethasone	1 (0.81)
Kellgren-Lawrence Grade at Procedure (%)
Grade 2 (minimal)	19 (19.00)
Grade 3 (moderate)	22 (22.00)
Grade 4 (severe)	59 (59.00)
NRS Pain Score Pre-Procedure, mean (SD)	6.50 (1.99)
Immediate Pain Relief, mean percentage (SD)	79.60 (20.56)
Amount of Physical Therapy After Procedure
0 Sessions	95 (77.24)
1–5 Sessions	17 (13.82)
6–10 Sessions	8 (6.50)
≥11 Sessionsx	3 (2.44)

^aAt each injection site: 0.5 mL of total medication and 0.2 mL contrast agent.

3.2. Short-term and long-term PROs for GNB with corticosteroid

In paired comparisons to pre-procedure baseline, the mean NRS reductions after GNBs were 3.36 (2–30 days; P-value <0.00001), 2.24 (1–3 months; P-value <0.00001), 1.77 (3–6 months; P-value <0.001), 1.11 (6–12 months; P-value <0.01), and not significantly different than baseline after 1-year post-procedure (Table 3; Fig. 2). Despite the trend of diminishing mean NRS reductions over time, we observed higher mean NRS reductions in every follow-up period after excluding the GNB non-responders. For patients who experienced a positive response to GNB (n = 97 GNBs; ≥50 % NRS score reduction or ≥ 75 % patient-reported pain relief), the mean NRS reductions were 4.46 (2–30 days; P-value <0.0001), 3.22 (1–3 months; P-value <0.00001), 2.34 (3–6 months; P-value <0.0001), 1.84 (6–12 months; P-value <0.02), and non-significant after 1-year post-procedure. By contrast, none of the NRS reductions for GNB non-responders were statistically significant (P-values >0.05) in any follow-up period and the changes were lower in magnitude: 1.34 (2–30 days), 0.23 (1–3 months), 0.21 (3–6 months), 0.71 (6–12 months), 0.79 (366+ days). Interestingly, a survival curve of the NRS reduction revealed that 50 % of patients maintained a NRS reduction greater than or equal to the MCID of −2.0 for at least 6.9 months (95 % CI = 5.52–9.86) after their GNB with corticosteroid (Fig. 3). Approximately a third of patients (30.9 %, 95 % CI = 23.05–39.95) achieved a >50 % NRS pain score reduction within 3 months following their procedure. Subjectively, the overall patient-reported percent relief after GNBs was 79.6 % for an average of 3.55 months (SD 5.00 months; Range: 1 day–2.50 years).

Table 3.

Overview of the changes in patient-reported outcome measures (post-procedure minus pre-procedure scores) for NRS, PROMIS-PH, PROMIS-MH, and PHQ9 for patients who underwent a genicular nerve block (GNB) with corticosteroid.

	GNB
	n	Baseline	2–30 days	31–90 days	91–182 days	183–365 days	366+ days
Δ NRS, mean (SD)	123	6.50 (1.99)	−3.36 (2.54)	−2.24 (2.60)	−1.77 (2.81)	−1.11 (2.50)	−0.27 (2.25)
Δ PROMIS-PH, mean (SD)	123	38.68 (6.58)	−2.08 (3.53)	−0.05 (4.36)	−0.95 (4.55)	−0.65 (3.56)	−0.29 (3.17)
Δ PROMIS-MH, mean (SD)	123	44.50 (9.01)	+1.11 (4.51)	−0.14 (4.67)	−0.49 (4.18)	−0.87 (5.01)	+1.49 (5.82)
Δ PHQ-9, mean (SD)	123	7.29 (5.92)	−0.81 (7.53)	−2.02 (7.74)	−0.94 (8.92)	−0.58 (8.30)	−2.81 (4.14)

Fig. 2.

Paired comparisons of baseline versus post-procedure numerical rating scale (NRS) scores from patients who underwent a genicular nerve block (GNB), stratified by post-procedure follow-up period. P-values are by Wilcoxon signed-rank test.

Fig. 3.

Survival curve portraying the number of months that patients sustained the MCID of −2.0 for NRS reduction after a genicular nerve block (GNB) with corticosteroid.

For active opioid users, their opioid usage was significantly decreased from the baseline 28.01 ± 13.20 MME/day to 13.58 ± 16.23 MME/day (P-value = 0.0019) in the first 6 months post-GNB, and this opioid reduction was maintained at 15.59 ± 12.51 MME/day (P-value = 0.0436) after the initial 6-month follow-up period (Fig. 4).

Fig. 4.

Comparison of daily opioid usage (daily morphine milligram equivalents, MME/day) among active prescription opioid users before and after a genicular nerve block (GNB). P-values are by Wilcoxon signed-rank test.

Of the 13 patients (13/73, 17.8 %) who received a TKA after their genicular nerve interventions, the average time to TKA was 16.5 months. Of the patients with native knees before procedure, we found that those with a history of meniscal tear were significantly more likely to progress to an TKA after genicular nerve intervention (OR 2.61, 95 % CI = 1.08–6.61, P-value = 0.034) compared to patients without meniscal tears.

3.3. Risk factors of treatment failure

KL grade was identified as an independent risk factor associated with non-response to GNB (P-value = 0.01952), with a significantly greater proportion of KL grade 4 patients in the non-response group. In contrast, the following factors were not significantly associated with treatment failure (P-values > 0.05): obesity, prescription opioid use, history of depression, anxiety, chronic pain syndrome, fibromyalgia, history of TKA, and prior intra-articular knee injections with corticosteroid or visco-supplementation.

4. Discussion

Given the lack of large sample data on the topic, our study offers a broader perspective on the long-term therapeutic effect of GNB with corticosteroid as seen across nine years of clinical practice serving a diverse patient population. Overall, our results supported the existing evidence from smaller RCTs that GNB with corticosteroid can provide therapeutic pain relief for an extended period beyond the effect of the anesthetic alone. Notably, 100 % of GNBs with corticosteroid provided greater than 24 h of pain relief; in other words, no patient in this study reported less than one day of relief. This implies that steroid placement with anesthetic has a therapeutic effect locally. Although corticosteroids have systemic effects, this cohort's duration of benefit of >3 months suggests a therapeutic effect beyond the maximum systemic response of 2–4 weeks reported in literature [[29], [30], [31]].

We found both a statistically and clinically significant reduction in NRS pain scores in the 2–30 day and 1–3 month follow-up period. Also, the NRS pain score reduction was more prominent in the positive response subgroup of patients (n = 97/123), with significant mean reductions of 4.46 (2–30 days), 3.22 (1–3 months), and 2.34 (3–6 months), while diminishing relief over time. In this study, the outcomes after GNB with corticosteroid was consistent with data from existing RCTs, which reported NRS reductions ranging from 3.1 to 6.2 and 1.7–4.1 at the 1-month and 3-month follow-up respectively [[16], [17], [18]]. Importantly, patients who underwent GNB with corticosteroid reported ∼80 % overall subjective relief for average of 3.6 months.

The magnitude of NRS reduction after GNB with corticosteroid from this study was comparable to several prior GNRFA studies at the 1-month and 3-month follow-ups, with diminishing pain relief after 6-months post-procedurally and beyond [7,32,33]. It is possible that the corticosteroid adjuvant extended the pain relief duration of the GNB beyond any temporary anesthetic effects, although we did not find a clear mechanism in literature that can fully explain this finding. Potential mechanisms include decreasing the systemic absorption of the local anesthetic [[12], [13], [14]], suppressing pain transmission via unmyelinated C‐fibers to reduce neural discharge, and by down-regulating genetic expression of pro-inflammatory proteins [34].

Despite the intention of delaying knee arthroplasty, GNB with corticosteroid can still be an effective therapeutic option to consider for persistent knee pain after knee replacement, given we did not observe any significant baseline or post-procedural outcomes difference in TKA versus native knees.

As a 2021 randomized-controlled trial by Rambhia et al. revealed that GNB with corticosteroid may reduce opioid use in the post-operative period [17], our results demonstrated that GNB with corticosteroid can be effective at significantly lowering daily MME among active opioid users in both the short-term (≤6-months) and long-term (>6-months).

In terms of factors contributing to treatment non-response, we identified that higher KL grade was associated with a lack of clinically significant treatment response. Given that GNB is intended to relieve pain rather than alter the underlying knee pathology, it is likely that patients with more severe knee OA may be less responsive to GNB with corticosteroid, and more appropriate for GNRFA [35,36]. Our cohort reflects patients who have limited options as many have already been treated and referred by an orthopaedic surgeon or sports medicine practitioner. Treating patients earlier in their disease course of OA (lower KL grade) may result in more sustained therapeutic relief from GNBs with corticosteroid. Lastly, it is important to consider the potential impact of patient complexity on treatment response. An interventional spine practice often treats complex knee OA patients such as our patient cohort, with anxiety, depression, chronic pain syndrome, fibromyalgia, or sometimes concomitant spine pathology. By contrast, a pre-dominantly orthopaedic or sports medicine practice may see a larger proportion of patients with knee pathology alone and a smaller proportion of patients with concomitant spine pathology, chronic pain syndrome, and fibromyalgia.

4.1. Strengths and limitations

The main strength of this study is its large sample size, extended study period, and long-term post-treatment data achieved through the manual chart review. Importantly, we obtained a patient sample that was largely reflective of the demographic composition within the United States general population [37]. Although we gathered long-term follow-up information for many patients, a sizable portion was lost to follow-up after the 6-month post-procedural period. Since data collection was limited to our institutional health records, we could have missed potentially relevant medical and surgical history as well as follow-up information from other health systems. Additionally, our electronic records did not include specific knee outcomes measures like Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and Knee injury and Osteoarthritis Outcome Score (KOOS). Given intrinsic limitations of retrospective studies, our findings should be interpreted within the limitations of this study design, such as non-standardized follow-up time points, incomplete data, and variation in follow-up documentation. Because we did not include GNBs with anesthetic alone as a comparison/control group, we cannot make definite conclusions about the impact of the corticosteroid adjuvant on GNB outcomes due to inability to control for potential placebo effect. Additionally, the changes in PROs may be confounded by other factors related to the patients’ medical management, including psychological support, physical therapy, and home exercises, and over-the-counter oral analgesics. We also cannot exclude the possibility that the local anesthetic by itself may be contributing to a portion of the long-term analgesic effects reported by this cohort. Another limitation for this study compared to other academic and private practices is our referral source for GNBs. Lastly, because most patients from this sample lived in a metropolitan area, our findings may be more generalizable to other metropolitan populations but less applicable to certain rural populations.

5. Conclusions

GNB with corticosteroid adjuvant has shown promise as a therapeutic option in the management of chronic knee pain, with evidence in support of its potential to provide extended pain relief for symptomatic knee OA. For the majority of patients who received GNB with corticosteroid, we observed statistically and clinically significant pain reduction for 3 months or more in this cohort. While more comparative studies are required to evaluate its effectiveness, GNB with corticosteroid has promising potential as a safe and effective treatment for chronic knee pain.

Disclosures

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Chen, Tosun, Thompson, Goyal.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.