
==== Front
Interv Pain Med
Interv Pain Med
Interventional Pain Medicine
2772-5944
Elsevier

S2772-5944(24)00027-X
10.1016/j.inpm.2024.100407
100407
Original Article
The association of payer type on genicular radiofrequency neurotomy treatment outcomes: Results of a cross-sectional study
Braun Samantha brauns@uthscsa.edu
a⁎
Mascoe Jason b
Caragea Marc b
Woodworth Tyler b
Curtis Tim b
Blatt Michael c
Cheney Cole e
Brown Todd c
Carson Daniel b
Kuo Keith c
Randall Dustin c
Huang Emily Y. f
Carefoot Andrea b
Teramoto Masaru b
Cooper Amanda b
Mills Megan d
Burnham Taylor b
Conger Aaron b
McCormick Zachary L. b
a Department of Physical Medicine and Rehabilitation, University of Texas San Antonio, San Antonio, TX, USA
b Department of Physical Medicine and Rehabilitation, University of Utah, Salt Lake City, UT, USA
c School of Medicine, University of Utah, Salt Lake City, UT, USA
d Department of Radiology, University of Utah, Salt Lake City, UT, USA
e Mayo Clinic Health System, Mankato, MN, USA
f Johns Hopkins University School of Medicine, Baltimore, MD, USA
⁎ Corresponding author. Department of Rehabilitation Medicine, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX, 78229, USA brauns@uthscsa.edu
05 4 2024
6 2024
05 4 2024
3 2 1004076 2 2024
15 3 2024
19 3 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by/4.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Background

Genicular radiofrequency neurotomy (GRFN) is an effective treatment for a subset of individuals with chronic knee pain. Previous studies demonstrate that Medicare and Medicaid beneficiaries report worse outcomes following various interventional procedures compared with commercially insured patients.

Objective

Evaluate the association of payer type on GRFN treatment outcomes.

Methods

Consecutive patients who underwent GRFN at a tertiary academic center were contacted for participation. Demographic, clinical, and procedural characteristics were collected from electronic medical records. Outcome data were collected by standardized telephone survey at 6–12 months, 12–24 months and ≥24 months. Treatment success was defined as ≥50% numerical pain rating scale (NPRS) score reduction from baseline. Data were analyzed using descriptive statistics for demographic, clinical, and procedural characteristics. Logistic and Poisson regression analyses were performed to examine the association of variables of interest and pain reduction.

Results

One hundred thirty-four patients treated with GRFN (mean 65.6 ± 12.7 years of age, 59.7% female) with a mean follow-up time of 23.3 ± 11.3 months were included. Payer type composition was 48.5% commercial (n = 65), 45.5% Medicare (n = 61), 3.7% Medicaid (n = 5), 1.5% government (n = 2), and 0.8% self-pay (n = 1). Overall, 47.8% of patients (n = 64) reported ≥50% NPRS score reduction after GRFN. After adjusting for age, follow-up duration, Kellgren-Lawrence osteoarthritis grade, baseline opioid use, antidepressant/antianxiety medication use, history of knee replacement, and number of RFN lesions placed, the logistic regression model showed no statically significant association between payer type and treatment outcome (OR = 2.11; 95% CI = 0.87, 5.11; p = 0.098).

Discussion/conclusion

In this study, after adjusting for demographic, clinical, and procedural characteristics, we found no association between payer type and treatment success following GRFN. This observation contrasts findings from other interventional studies reporting an association between payer category and treatment success.

Keywords

Knee pain
Genicular nerve
Radiofrequency ablation
Osteoarthritis
Payer type
==== Body
pmc1 Introduction

Osteoarthritis (OA) is the most common form of arthritis, involving degenerative changes to the protective cartilage and joint structures over time, leading to pain and functional impairments. The global prevalence of radiographically-confirmed symptomatic knee OA in 2010 was estimated to be 3.8% [1]. In 2008, nearly 27 million people were estimated to have clinically significant OA, which is one of the leading causes of disability in the US adult population [2]. OA has been shown to increase with age and additional risk factors such as obesity, specific occupations, previous injuries, and genetics [3]. It is anticipated that the prevalence of OA will increase globally as the population ages and obesity rates continue to rise [1].

Conservative treatment of painful knee OA includes exercise, self-management and education, weight loss, Tai Chi, bracing, physical therapy, topical and oral non-steroidal anti-inflammatory drugs (NSAIDS), and intra-articular (IA) corticosteroid injection [4]. In severe cases, total knee arthroplasty (TKA) is often performed. Risks associated with TKA include continued pain, infection, joint instability, periprosthetic fracture, and prosthetic joint failure [5]. Despite improvements in overall quality of life, approximately 20% of patients suffer from recalcitrant pain post-TKA [6].

Genicular radiofrequency neurotomy (GRFN) is a treatment for knee pain in patients who choose to avoid surgery, are not surgical candidates, or have post-TKA pain. GRFN was first studied in a prospective manner by Choi et al. [7]. The procedure involves partial sensory denervation of the anterior joint capsule through targeted delivery of radiofrequency energy to the genicular nerves [8]. Initially, GRFN protocols predominately focused on three nerves: the superior medial genicular nerve (SMGN) and inferior medial genicular nerve (IMGN), which innervate the medial aspect of the knee joint, along with the superior lateral genicular nerve (SLGN) that innervates the lateral aspect of the knee joint. However, recent anatomical studies have revealed a more extensive sensory innervation of the anterior knee joint capsule. Both Tran et al. and Fonkoue et al. identified 10 or more articular branches originating from the femoral, obturator, and sciatic nerves in independent dissection studies [9,10]. As more complete descriptions of knee joint neuroanatomy emerge, conventional GRFN protocols have been expanded to capture additional sensory targets and account for anatomic variation [11,12]. The majority of clinical outcome studies indicate that GRFN effectively reduces pain and improves function in patients with painful knee OA [[13], [14], [15]], with the most recent literature indicating that expanded GRFN protocols likely improve outcomes beyond that of the originally proposed technique [12,[16], [17], [18], [19]]. Although there are case reports of complications such as pes anserine tendon injury, septic arthritis, skin burns, and periarticular hematomas [[20], [21], [22], [23]], GRFN is generally regarded as a safe and well-tolerated procedure [24]. Positive outcomes have been associated with ≥80% pain relief with prognostic blocks [11,12,25], no opioid use, absence of coexisting mental health disorders, and lower baseline pain scores [12].

Patients in the United States can access various types of health insurance depending on their eligibility, including private (commercial) insurance and government-funded plans such as Medicaid and Medicare. There is a considerable body of literature investigating outcome discrepancies associated with payer type following various surgeries and interventional procedures. Badin et al. performed a systematic review of outcomes on elective spine surgery and found that Medicaid was associated with worse health outcomes, and increased emergency department utilization [26]. Additionally, a retrospective analysis of total hip arthroplasty (THA) by Shau et al. found Medicaid payer status was associated with increased resource usage and total cost following primary THA [27].

However, to our knowledge, the relationship between payer type and patient outcomes has not been evaluated for GRFN. As such, this study aimed to assess whether primary payer type was independently correlated with treatment success in patients who underwent GRFN for chronic knee pain.

2 Methods

2.1 Data collection

This study was conducted at a tertiary academic center. The protocol was approved by the University of Utah Institutional Review Board (IRB 00138414). Using CPT codes 64,624 and 64,640, electronic medical records (EMRs) of 226 consecutive patients who underwent GRFN from October 2015 to March 2021 were reviewed. Inclusion criteria were patients aged 18–80 years with painful knee OA or persistent post-TKA pain with at least one positive (≥50% pain relief) prognostic genicular nerve block prior to their GRFN procedure. Detailed descriptions of the genicular nerve block and GRFN procedural techniques were previously published [15]. Exclusion criteria included inability or refusal to participate in a phone call-based standardized outcome survey and missing or low (≤3 points) baseline Numeric Pain Rating Scale (NPRS) scores within two months before the index GRFN.

2.2 Procedures

Both prognostic genicular nerve blocks and GRFN procedures were performed as we have described previously [16].

2.2.1 Data collection

Data collected at baseline from the EMR included age, body mass index, duration of pain, gender, smoking status, workers’ compensation status, active opioid prescription for ≥6 months at the time of GRFN, current use of antidepressants or anxiolytic medications, presence and laterality of knee replacement and number of nerves targeted by prognostic genicular blocks and the index GRFN procedure, radiofrequency probe type, prognostic block response, GRFN laterality, the number of prior GRFN procedures, and payer type. Payer type was based on the primary coverage listed in the EMR and included commercial (private) insurance, Medicare, Medicaid, other government plan, and self-pay.

Outcome data were collected at a single follow-up assessment conducted via telephone call using a standardized survey at variable time points categorized between 6 and 12 months, 12–24 months, or ≥24 months post-procedure, depending on when the phone call occurred relative to the GRFN procedure. The standardized telephone survey included NPRS scores at the time of follow up and self-reported improvement as measured by the Patient Global Impression of Change (PGIC) at the time of follow up for the most recent GRFN. Data was input into REDCap, which is a secure web-based platform designed for validated data capture in research studies [28,29]. The primary outcomes were the proportion of participants with ≥50% NPRS reduction and ≥ 2-point NPRS reduction. Secondary outcomes included the proportion of participants who reported a PGIC score of ≥6 (consistent with an improvement rating of at least “much improved”), NPRS follow-up scores, and opioid cessation. Core outcome measures and minimally important differences were chosen in accordance with established recommendations for studies involving measurement of chronic pain [[30], [31], [32]].

2.3 Data analysis

Patient demographics and clinical characteristics were summarized using descriptive statistics, including mean, standard deviation (SD), median, and interquartile range (IQR) for continuous variables, as well as frequency and percentage for categorical variables. A 95% confidence interval (CI) was also calculated for select variables. Contingency table analyses were used to examine whether responder rates differed by payer types for the following categorical outcome variables: (1) ≥ 50% NPRS reduction from baseline, (2) ≥ 2-point NPRS reduction from baseline, and (3) ≥ 6 on PGIC indicating “much improved” or “very much improved”. A one-sample t-test was used to determine if NPRS follow-up scores had improved significantly from baseline (i.e., whether NPRS score reduction and percent change were significantly different from zero and 0%, respectively).

Multivariate logistic regression (LR) analysis was performed to explore the relationships between categorical outcome variables (≥50% NPRS reduction, ≥ 2-point NPRS reduction, and ≥6 on the PGIC) and payer type, while using select demographic and clinical characteristics as covariates. The included covariates were follow-up timeframe, worst compartment KL grade, opioid use at baseline, antidepressant/anxiolytic medication at baseline, history of knee replacement, number of nerves targeted by RFA, and age [11,12]. Odds ratios (ORs) and 95% CIs were calculated for each coefficient.

Lastly, a Poisson regression model was developed to explore the relationship between payer type and the continuous secondary outcome variable of NPRS score at follow-up. This model incorporated the same predictor variables as the LR models, with addition of baseline NPRS score as an extra covariate. Incidence rate ratios (IRRs) and 95% CIs were calculated for each coefficient. All analyses were conducted using Stata/MP 17.0 (StataCorp LLC, College Station, TX), with an α level of 0.05 for statistical significance.

3 Results

Of the 226 patients identified, a total of 134 participants were included in this study. Patient demographics and clinical characteristics, including payer types, are presented in Table 1a, Table 1ba and 1b. Commercial insurance (n = 65; 48.5%) and Medicare (n = 61; 45.5%) were the most common payer types. The remaining patients received primary coverage through Medicaid (n = 5; 3.7%) and other government insurance plans (n = 2; 1.5%), with one case (0.8%) of self-pay. For the majority of patients, follow-up occurred at 12–24 months (n = 60; 44.8%) and ≥24 months (n = 53; 39.5%) post-GRFN, whereas only 15.7% (n = 21) were contacted 6–12 months after the procedure. The average follow-up time was 23.4 ± 11.3 months.Table 1a Patient demographics and clinical characteristics (N = 134; continuous variables).

Table 1aContinuous Variable	N	Mean (SD)	Min, Max	
Age	134	65.6 (12.7)	34, 96	
Height (cm)	107	168.9 (10.5)	147.3, 190.5	
Weight (kg)	107	94.7 (26.3)	47.2, 161	
BMI (kg/m2)	134	33.4 (7.9)	20.1, 59	
Duration of pain (yr)	134	5.4 (5.8)	0.25, 49	
Follow-up months	112	23.3 (11.3)	5, 56	
Abbreviations: BMI = body mass index; Max = maximum value; Min = minimum value; SD = standard deviation.

Table 1b Patient demographics and clinical characteristics (N = 134; categorical variables).

Table 1bCategorical Variable	Frequency (%)	
Payer type	
 Commercial	65 (48.5)	
 Medicare	61 (45.5)	
 Medicaid	5 (3.7)	
 Other Government	2 (1.5)	
 Self-pay	1 (0.8)	
Follow-up interval	
 6–12 months	21 (15.7)	
 12–24 months	60 (44.8)	
 ≥ 24 months	53 (39.5)	
Gender	
 Male	54 (40.3)	
 Female	80 (59.7)	
Smoking	
 Never	100 (74.6)	
 Current	3 (2.2)	
 Former	31 (23.1)	
Opioid use at baseline	
 Yes	37 (27.6)	
 No	97 (72.4)	
Antidepressant/anxiolytic medication at time of GRFN	
 Yes	63 (47.0)	
 No	71 (53.0)	
Depression	
 Yes	66 (49.3)	
 No	68 (50.7)	
Anxiety	
 Yes	55 (41.0)	
 No	79 (59.0)	
Worker's compensation claim	
 Yes	4 (3.0)	
 No	130 (97.0)	
Worst compartment KL grade	
 0	16 (11.9)	
 1	6 (4.5)	
 2	18 (13.4)	
 3	25 (18.7)	
 4	69 (51.5)	
Prognostic block relief	
 > 50%	14 (10.5)	
 50–79%	15 (11.2)	
 80–99%	23 (17.2)	
 100%	82 (61.1)	
RFA probe type	
 Cooled	129 (96.3)	
 Conventional monopolar	5 (3.7)	
History of knee replacement in knee treated with GRFN	
 Yes	25 (18.7)	
 No	109 (81.3)	
Laterality of knee replacement	
 Left	12 (9.0)	
 Right	13 (9.7)	
Presence of lower extremity hardware	
 Yes	9 (6.7)	
 No	125 (93.3)	
GRFN laterality	
 Left	74 (55.2)	
 Right	60 (44.8)	
Number of previous GRFN procedures	
 0	117 (87.3)	
 1	16 (11.9)	
 2	1 (0.7)	
Number of nerves targeted with GRFN	
 3	13 (9.7)	
 4	39 (29.1)	
 5	67 (50.0)	
 6	4 (3.0)	
 7	7 (5.2)	
 8	4 (3.0)	
Abbreviations: GRFN = genicular radiofrequency neurotomy; KL = Kellgren and Lawrence; RFA = radiofrequency ablation.

Continuous NPRS scores at baseline and follow-up are summarized according to payer type in Table 2. The mean baseline NPRS score for all patients was 6.2 ± 1.7 (95% CI: 6.0, 6.5). NPRS score reductions at follow-up were significantly higher than zero (p < 0.05) for patients with commercial (2.5 ± 3.4 [95% CI: 1.6, 3.3]) and Medicare payer types (3.1 ± 3.7 [95% CI: 2.2, 4.0]), as the lower bounds of both 95% CIs were greater than zero. Likewise, percent NPRS reduction at follow-up was significantly greater than 0% (p < 0.05) for commercial and Medicare patients (36.4 ± 52.5% [95% CI: 23.5, 49.3] and 44.6 ± 54.4% [95% CI: 30.9, 58.4], respectively).Table 2 Continuous NPRS scores at baseline and follow-up.

Table 2Outcome	Payer Type	n	Mean (SD)	Median (IQR)	Min, Max	95% CI (mean)	
Baseline NPRS	Commercial	65	6.3 (1.5)	6 (5, 7)	4, 10	5.9, 6.7	
Medicare	61	6.1 (1.9)	6 (4, 7)	4, 10	5.6, 6.6	
Medicaid	5	6.6 (1.3)	6 (6, 8)	5, 8	5.4, 7.8	
Other Government	2	6.0 (1.4)	6 (5, 7)	5, 7	4.0, 8.0	
Self-pay	1	8.0 (N/A)	N/A		
Total	134	6.2 (1.7)	6 (5, 8)	4, 10	6.0, 6.5	
Follow-up NPRS	Commercial	65	3.9 (3.2)	4 (0, 7)	0, 10	3.1, 4.6	
Medicare	61	3.1 (2.9)	2 (0, 5)	0, 10	2.3, 3.8	
Medicaid	5	4.2 (1.9)	4 (3, 5)	2, 7	2.5, 5.9	
Other Government	2	3.0 (4.2)	3 (0, 6)	0, 6	−2.9, 8.9	
Self-pay	1	0.0 (N/A)	N/A		
Total	134	3.5 (3.0)	3 (0, 6)	0, 10	2.9, 4.0	
NPRS reduction (baseline NPRS minus follow-up NPRS)	Commercial*	65	2.5 (3.4)	2 (0, 5)	−5, 10	1.6, 3.3	
Medicare*	61	3.1 (3.7)	3 (0, 6)	−4, 10	2.2, 4.0	
Medicaid	5	2.4 (3.2)	2 (1, 5)	−2, 6	−0.4, 5.2	
Other Government	2	3.0 (5.7)	3 (−1, 7)	−1, 7	−4.9, 10.9	
Self-pay	1	8.0 (N/A)	N/A		
Total	134	2.8 (3.5)	2 (0, 6)	−5, 10	2.2, 3.4	
% NPRS change (from baseline to follow-up)	Commercial*	65	36.4 (52.5)	33 (0, 100)	−100, 100	23.5, 49.3	
Medicare*	61	44.6 (54.4)	50 (0, 100)	−100, 100	30.9, 58.4	
Medicaid	5	29.5 (45.2)	33 (16.7, 62.5)	−40, 75	−10.5, 69.5	
Other Government	2	40.0 (84.9)	40 (−20, 100)	−20, 100	−78.7, 158.7	
Self-pay	1	100.0 (N/A)	N/A		
Total	134	40.4 (53.1)	40 (0, 100)	−100, 100	31.3, 49.5	
* Significantly different from baseline according to one sample t-test (p < 0.05).

Abbreviations: CI = confidence interval; IQR = interquartile range; Max = maximum value; Min = minimum value; N/A = not applicable; NPRS = numeric pain rating scale; SD = standard deviation.

Responder rates for the three pre-defined categorical outcome variables are presented in Table 3a, Table 3b, Table 3c. Excluding the single patient with self-pay, NPRS responder rates were highest for the Medicare payer type group, with 57.4% (95% CI: 44.9, 69.0) and 63.9% (95% CI: 51.4, 74.8) of Medicare patients reporting ≥50% NPRS reduction and ≥ 2-point NPRS reduction, respectively. However, the 95% CIs for the proportion of responders in each group overlapped with one another, indicating that categorical improvements in NPRS pain scores did not differ significantly between payer types. Similarly, 60.0% of patients with commercial insurance (95% CI: 47.9, 71.0) and 55.7% of those with Medicaid (95% CI: 21.1, 88.2) reported a PGIC score of ≥6 points consistent with “improved” or “very much improved” at follow-up.Table 3a Proportion of participants with ≥50% NPRS reduction by payer type.

Table 3aPayer type	≥50% NPRS reduction	95% CI (yes)	
Yes	No	
Commercial	25 (38.5)	40 (61.5)	27.6, 50.6	
Medicare	35 (57.4)	26 (42.6)	44.9, 69.0	
Medicaid	2 (40.0)	3 (60.0)	11.8, 76.9	
Other Government	1 (50.0)	1 (50.0)	9.5, 90.5	
Self-pay	1 (100.0)	0 (0.0)	20.7, 100.0	
Total	64 (47.8)	70 (52.2)	39.5, 56.2	
Note: values are frequency (%).

Abbreviations: CI = confidence interval; NPRS = numeric pain rating scale.

Table 3b Proportion of participants with ≥2-point NPRS reduction by payer type.

Table 3bPayer type	≥2-point NPRS reduction	95% CI (yes)	
Yes	No	
Commercial	38 (58.5)	27 (41.5)	46.3, 69.6	
Medicare	39 (63.9)	22 (36.1)	51.4, 74.8	
Medicaid	3 (60.0)	2 (40.0)	23.1, 88.2	
Other Government	1 (50.0)	1 (50.0)	9.5, 90.5	
Self-pay	1 (100.0)	0 (0.0)	20.7, 100.0	
Total	82 (61.2)	52 (38.8)	52.7, 69.0	
Note: values are frequency (%).

Abbreviations: CI = confidence interval; NPRS = numeric pain rating scale.

Table 3c Proportion of participants with ≥6 on PGIC by payer type.

Table 3cPayer type	≥6 on PGIC	95% CI (yes)	
Yes	No	
Commercial	39 (60.0)	26 (40.0)	47.9, 71.0	
Medicare	34 (55.7)	27 (44.3)	43.3, 67.5	
Medicaid	3 (60.0)	2 (40.0)	23.1, 88.2	
Other Government	2 (100.0)	0 (0.0)	34.2, 100.0	
Self-pay	1 (100.0)	0 (0.0)	20.7, 100.0	
Total	79 (59.0)	55 (41.0)	50.5, 66.9	
Note: values are frequency (%).

Abbreviations: CI = confidence interval; PGIC = Patient Global Impression of Change (scores ≥6 indicate at least “much improved”).

According to results from the logistic regression (LR) analyses, payer type was not significantly associated with ≥50% NPRS reduction, ≥ 2-point NPRS reduction, or ≥ 6 on PGIC after adjusting for covariates (p > 0.05; Table 4a). However, two of the LR models contained statistically significant covariates: a follow-up time of ≥24 months was a significant predictor variable for ≥50% NPRS reduction (OR = 3.36 [95% CI: 1.04, 10.85]; p = 0.042), while a history of knee replacement in the knee treated with GRFN was associated with lower odds by 68% of achieving a PGIC score ≥6 points (OR = 0.32 [95% CI: 0.12, 0.88]; p = 0.027).Table 4a Logistic regression models on ≥ 50% NPRS reduction, ≥ 2-point NPRS reduction, and ≥6 PGIC.

Table 4aOutcome	Predictor	OR	95% CI	p	
≥50% NPRS reductiona	Payer type (vs. commercial)				
 Medicare	2.11	0.87, 5.11	0.098	
 Medicaid	1.42	0.20, 9.90	0.726	
 Other Government	2.86	0.14, 56.75	0.490	
 Self-pay*	–	–	–	
Follow-up (vs. 6–12 months)				
 12–24 months	2.11	0.68, 6.53	0.197	
 ≥24 months	3.36	1.04, 10.85	0.042	
Worst compartment KL grade (vs. 0–1)				
 2-3	3.50	0.97, 12.61	0.055	
 4	3.12	0.89, 10.94	0.076	
Opioid use at baseline (vs. no)				
 Yes	0.64	0.27, 1.52	0.316	
Antidepressant/anxiolytic medication at time of GRFN (vs. no)				
 Yes	0.69	0.31, 1.51	0.354	
History of knee replacement in knee treated with GRFN (vs. no)				
 Yes	0.59	0.2, 1.74	0.342	
Number of nerves targeted with GRFN (vs. 3)				
 >3	1.20	0.31, 4.68	0.796	
 Age	1.02	0.98, 1.05	0.376	
≥2-point NPRS reductionb	Payer type (vs. commercial)				
 Medicare	1.40	0.59, 3.34	0.450	
 Medicaid	1.26	0.19, 8.51	0.814	
 Other Government	0.57	0.03, 11.00	0.708	
 Self-pay*	–	–	–	
Follow-up (vs. 6–12 months)				
 12–24 months	1.37	0.48, 3.96	0.558	
 ≥24 months	2.17	0.71, 6.63	0.172	
Worst compartment KL grade (vs. 0–1)				
 2-3	1.69	0.54, 5.25	0.365	
 4	1.85	0.61, 5.62	0.278	
Opioid use at baseline (vs. no)				
 Yes	0.84	0.37, 1.95	0.693	
Antidepressant/anxiolytic medication at time of GRFN (vs. no)				
 Yes	0.51	0.24, 1.11	0.092	
History of knee replacement in knee treated with GRFN (vs. no)				
 Yes	0.74	0.28, 1.98	0.548	
Number of nerves targeted with GRFN (vs. 3)				
 >3	2.04	0.57, 7.29	0.272	
 Age	0.99	0.96, 1.02	0.573	
≥6 PGICc	Payer type (vs. commercial)				
 Medicare	0.80	0.34, 1.89	0.613	
 Medicaid	1.09	0.14, 8.60	0.934	
 Other Government*	–	–	–	
 Self-pay*	–	–	–	
Follow-up (vs. 6–12 months)				
 12–24 months	0.92	0.31, 2.77	0.886	
 ≥24 months	0.55	0.18, 1.69	0.300	
Worst compartment KL grade (vs. 0–1)				
 2-3	0.62	0.18, 2.16	0.457	
 4	0.51	0.15, 1.72	0.276	
Opioid use at baseline (vs. no)				
 Yes	1.04	0.45, 2.40	0.919	
Antidepressant/anxiolytic medication at time of GRFN (vs. no)				
 Yes	0.88	0.40, 1.91	0.742	
History of knee replacement in knee treated with GRFN (vs. no)				
 Yes	0.32	0.12, 0.88	0.027	
Number of nerves targeted with GRFN (vs. 3)				
 >3	3.23	0.85, 12.33	0.087	
 Age	1.00	0.97, 1.04	0.962	
Abbreviations: CI = confidence interval; OR = odds ratio; GRFN = genicular radiofrequency neurotomy; KL = Kellgren and Lawrence; NPRS = numeric pain rating scale; PGIC = Patient Global Impression of Change (scores ≥6 indicate at least “much improved”).

* Unable to reliably estimate regression coefficients due to insufficient number of cases.

As with the LR analyses, payer type was not a significant predictor variable in the Poisson regression model for NPRS follow-up score after accounting for the selected covariates, which included baseline NPRS score (p > 0.05; Table 4b). Compared to the LR models, the Poisson regression model contained a larger number of statistically significant predictor variables (p < 0.05), including worst compartment KL grade, baseline opioid use, and antidepressant/antianxiety medication at the time of GRFN. Patients with a worst compartment KL grade of 2–3 and 4 (vs. 0–1) were expected to have lower NPRS scores at follow-up by 32% (IRR = 0.68 [95% CI: 0.52, 0.88]; p = 0.004) and 35% (IRR = 0.65 [95% CI: 0.50, 0.84]; p = 0.001), respectively, after controlling for the other covariates. Conversely, baseline opioid use and antidepressant/anxiolytic medication at the time of GRFN were significantly associated with higher follow-up NPRS scores by 42% (IRR = 1.42 [95% CI: 1.17, 1.73]; p = 0.001) and 33% (IRR = 1.33 [95% CI: 1.09, 1.62]; p = 0.005), respectively. Although a follow-up time of ≥24 months (IRR = 0.75 [95% CI: 0.57, 0.99]) and >3 nerves (vs. 3) targeted by RFA (IRR = 0.75 [95% CI: 0.57, 0.99]) demonstrated statistical significance (p < 0.05), the 95% CIs of the IRRs were fairly close to 1.00, indicating that the effects of these predictor variables on follow-up NPRS score were likely minimal.Table 4b Poisson regression model on continuous NPRS score at follow-up.

Table 4bOutcome	Predictor	IRR	95% CI	p	
Follow-up NPRS	Baseline NPRS	0.95	0.90, 1.01	0.122	
Payer type (vs. commercial)				
 Medicare	0.89	0.71, 1.11	0.301	
 Medicaid	0.99	0.63, 1.55	0.958	
 Other Government	0.70	0.31, 1.62	0.411	
 Self-pay*	<0.001	–	0.982	
Follow-up (vs. 6–12 months)				
 12–24 months	0.86	0.66, 1.12	0.264	
 ≥24 months	0.75	0.57, 0.99	0.039	
Worst compartment KL grade (vs. 0–1)				
 2-3	0.68	0.52, 0.88	0.004	
 4	0.65	0.50, 0.84	0.001	
Opioid use at baseline (vs. no)				
 Yes	1.42	1.17, 1.73	0.001	
Antidepressant/anxiolytic medication at time of GRFN (vs. no)				
 Yes	1.33	1.09, 1.62	0.005	
History of knee replacement in knee treated with GRFN (vs. no)				
 Yes	1.06	0.83, 1.34	0.658	
Number of nerves targeted with GRFN (vs. 3)				
 >3	0.75	0.56, 0.99	0.042	
 Age	0.99	0.98, 1.00	0.094	
Abbreviations: CI = confidence interval; GRFN = genicular radiofrequency neurotomy; IRR = incidence rate ratio; KL = Kellgren and Lawrence; NPRS = numeric pain rating scale.

* 95% CI could not be calculated due to insufficient number of observations (only one case of self-pay).

4 Discussion

This study evaluated whether patient reported outcomes following treatment with GRFN varied by payer type. Initial analysis revealed that commercially-insured and Medicare patients experienced significantly greater reductions in NPRS scores compared to other payer types. However, after adjusting for demographic, clinical, and procedural characteristics, there was no significant association between payer type and treatment success, which was defined as NPRS score reductions of either ≥50% or ≥2 points from baseline. This finding was unexpected, as a large body of existing literature reports that Medicare and Medicaid beneficiaries have worse outcomes following various surgeries and interventional procedures compared to patients with commercial insurance.

Outcomes after interventional treatments for musculoskeletal pain may be influenced by payer type. Reisinger et al. evaluated ODI scores of 606 patients who underwent lumbosacral transforaminal epidural steroid injection for radicular pain [33]. Medicare and Medicaid payer types were found to be negative predictors of functional improvement after adjusting for other demographic, clinical, and procedural variables. Compared to those with commercial insurance, Medicare and Medicaid patients were significantly less likely to attain the minimal clinically important difference (MCID) of ≥15-point and ≥30% improvements in ODI scores.

When assessing outcomes of total knee replacement, Veltre et al. found that Medicare patients were at higher risk of mortality and post operative complications, including wound dehiscence, as compared to privately insured patients [34]. Similar outcomes were reported by Tirumala et al. when assessing major and minor complications following revisions of both total hip and knee joint arthroplasty [35]. In addition to higher cost burden and longer hospital stay, Medicaid beneficiary status was linked to a higher prevalence of wound dehiscence, hematoma, and postoperative in-hospital infection compared to non-Medicaid payer types [36].

Given the aging demographic of the United States and the anticipated increase in the prevalence of knee OA, GRFN presents itself as a compelling and viable treatment option for managing pain associated with knee OA. In a randomized clinical trial, Davis et al. found that, compared to IA corticosteroids, patients treated with cooled RFN experienced significantly greater pain reduction at six months as well as improvements in Oxford Knee Score (OKS), Global Perceived Effect, and analgesic use [13]. In a retrospective analysis of 31 patients who underwent GRFN, the procedure was shown on average to provide greater than 60% pain relief in patients for as long as 6 months [14]. In a systematic review, Fogarty et al. reported that 49–74% of participants achieved ≥50% pain relief at 6 months after undergoing GRFN treatment, which outperformed IA injections of steroid and hyaluronic acid, respectively [15]. Approximately half of patients in the present study reported a ≥50% reduction in NPRS score at an average follow-up time of 23.4 ± 11.3 months post-GRFN, suggesting that patients who undergo GRFN in “real-world” practice can be expected to report a similar rate of treatment success to that observed in clinical trials, particularly when expanded lesioning protocols are utilized. Importantly, the present observations were derived from a “real-world” patient cohort that included characteristics such as prior TKA, opioid usage, and significant depression, which were among the exclusion criteria for several clinical trials that have served to establish the effectiveness of GRFN [7,13,19].

We previously described negative and positive prognostic factors, as well as additional covariates associated with GRFN treatment success, for this study cohort in a separate publication [16].

Several important limitations to this study must be addressed. Of the 226 charts selected for review, only 134 patients completed follow-up telephone questionnaires and were thus included in this cross-sectional study. We found a higher probability of significant pain relief in those who were contacted after two years. It is possible that those who reported pain reduction at that distant time point might have had other treatment or those who did not benefit from the index procedure were lost to follow up. This sample size is smaller compared to similar studies evaluating the relationship between payer type and clinical outcomes across hundreds of thousands of patients and large studies will ultimately be required to confirm the present findings [34,36]. As a result of the reduced sample size and decreased power, potential smaller difference in outcomes based on payer status may have been overlooked. The generalizability of our results is also limited by the fact that all study participants were patients treated at a single academic medical center in Utah and future multicenter studies may be warranted. Additionally, Medicaid payer status (n = 5) was markedly underrepresented in our study population; only 5 participants reported it as their primary payer type compared to 65 patients with commercial insurance and 61 with Medicare. While the study did not include variables like household income and highest education level, it is worth considering that the socioeconomic diversity in this cohort might not have been extensive. This consideration is based on the fact that as of 2022, Utah had the lowest income inequality ranking among all 50 United States [37]. However, the precise extent of this influence remains uncertain. It would be prudent to conduct further research to ascertain further whether GRFN treatment outcomes differ according to payer type in more diverse populations, where a wider range of social and economic backgrounds are represented.

5 Conclusion

We present the results of a cross-sectional study evaluating whether GRFN treatment outcomes vary by patient payer type. In contrast to previous literature reports, this study found no significant associations between payer type and patient reported outcomes after adjusting for demographic, clinical, and procedural characteristics. Overall, GRFN was found to be an effective pain-relieving modality for symptomatic knee OA or persistent post-TKA pain. Further investigations, especially prospective studies collecting outcome data from large, diverse patient populations, are warranted to extend the clinical applications of the present findings.

Conflict of interest statement

Zachary L. McCormick, MD serves on the Board of Directors of the International Pain and Spine Intervention Society and has received research grants from Avanos Medical (paid directly to the 10.13039/100007747 University of Utah ). Taylor Burnham, DO, is a consultant for Avanos Medical and has received grant funding from DIROS Technology (paid directly to the 10.13039/100007747 University of Utah ). Aaron Conger, DO, has received research grants from Stratus Medical (paid directly to the 10.13039/100007747 University of Utah ). There are no other potential conflicts of interest to disclose on the part of any of the other authors.

Funding

This study was funded by research grants from (1) the Skaggs Foundation for Research and (2) the Mitchell and June Morris Foundation.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Zachary McCormick reports financial support was provided by Skaggs Foundation for Research. Zachary McCormick reports financial support was provided by Mitchell and June Morris Foundation. Zachary McCormick reports a relationship with International Pain and Spine Intervention Society that includes: board membership. Zachary McCormick reports a relationship with Avanos Medical Inc that includes: funding grants. Taylor Burnham reports a relationship with Avanos Medical Inc that includes: consulting or advisory. Taylor Burnham reports a relationship with Diros Technology Inc that includes: funding grants. Aaron Conger reports a relationship with Stratus that includes: funding grants. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
==== Refs
References

1 Cross M. Smith E. Hoy D. Nolte S. Ackerman I. Fransen M. Bridgett L. Williams S. Guillemin F. Hill C.L. Laslett L.L. The global burden of hip and knee osteoarthritis: estimates from the global burden of disease 2010 study Ann Rheum Dis 73 7 2014 1323 1330 10.1136/annrheumdis-2013-204763 24553908
2 Lawrence R.C. Felson D.T. Helmick C.G. Arnold L.M. Choi H. Deyo R.A. Gabriel S. Hirsch R. Hochberg M.C. Hunder G.G. Jordan J.M. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II Arthritis Rheum 58 1 2008 26 35 10.1002/art.23176 18163497
3 Brophy R.H. Fillingham Y.A. AAOS clinical practice guideline summary: management of osteoarthritis of the knee (nonarthroplasty), third edition J Am Acad Orthop Surg 30 9 2022 e721 e729 10.5435/JAAOS-D-21-01233 35383651
4 Kolasinski S.L. Neogi T. Hochberg M.C. Oatis C. Guyatt G. Block J. Callahan L. Copenhaver C. Dodge C. Felson D. Gellar K. Harvey W.F. Hawker G. Herzig E. Kwoh C.K. Nelson A.E. Samuels J. Scanzello C. White D. Wise B. Altman R.D. DiRenzo D. Fontanarosa J. Giradi G. Ishimori M. Misra D. Shah A.A. Shmagel A.K. Thoma L.M. Turgunbaev M. Turner A.S. Reston J. 2019 American College of rheumatology/arthritis foundation guideline for the management of osteoarthritis of the hand, hip, and knee Arthritis Care Res 72 2 2020 149 162 10.1002/acr.24131 PMID: 31908149
5 Ahmad S.S. Gantenbein B. Evangelopoulos D.S. Schär M.O. Schwienbacher S. Kohlhof H. Kohl S. Arthroplasty - current strategies for the management of knee osteoarthritis Swiss Med Wkly 145 708 2015 w14096 10.4414/smw.2015.14096
6 Wylde V. Beswick A. Bruce J. Blom A. Howells N. Gooberman-Hill R. Chronic pain after total knee arthroplasty EFORT Open Rev 3 8 2018 461 470 10.1302/2058-5241.3.180004 30237904
7 Choi W.J. Hwang S.J. Song J.G. Leem J.G. Kang Y.U. Park P.H. Shin J.W. Radiofrequency treatment relieves chronic knee osteoarthritis pain: a double-blind randomized controlled trial Pain 152 3 2011 481 487 10.1016/j.pain.2010.09.029 21055873
8 Conger A. Gililland J. Anderson L. Pelt C.E. Peters C. McCormick Z.L. Genicular nerve radiofrequency ablation for the treatment of painful knee osteoarthritis: current evidence and future directions Pain Med 22 Suppl 1 2021 S20 S23 34308957
9 Fonkoué L. Behets C. Kouassi J.É. Coyette M. Detrembleur C. Thienpont E. Cornu O. Distribution of sensory nerves supplying the knee joint capsule and implications for genicular blockade and radiofrequency ablation: an anatomical study Surg Radiol Anat 41 12 2019 1461 1471 10.1007/s00276-019-02291-y 31338537
10 Tran J. Peng P. Agur A. Evaluation of nerve capture using classical landmarks for genicular nerve radiofrequency ablation: 3D cadaveric study Reg Anesth Pain Med 45 2020 898 906 32928998
11 McCormick Z.L. Cohen S.P. Walega D.R. Kohan L. Technical considerations for genicular nerve radiofrequency ablation: optimizing outcomes Reg Anesth Pain Med 46 6 2021 518 523 33483425
12 Chen Y. Vu T.N. Chinchilli V.M. Farrag M. Roybal A.R. Huh A. Cohen Z.O. Becker A.B. Arvanaghi B. Agrawal M. Ogden J. Clinical and technical factors associated with knee radiofrequency ablation outcomes: a multicenter analysis Reg Anesth Pain Med 46 4 2021 298 304 33558282
13 Davis T. Loudermilk E. DePalma M. Hunter C. Lindley D. Patel N. Choi D. Soloman M. Gupta A. Desai M. Buvanendran A. Prospective, multicenter, randomized, crossover clinical trial comparing the safety and effectiveness of cooled radiofrequency ablation with corticosteroid injection in the management of knee pain from osteoarthritis Reg Anesth Pain Med 43 1 2018 84 91 10.1097/AAP.0000000000000690 29095245
14 Iannaccone F. Dixon S. Kaufman A. A review of long-term pain relief after genicular nerve radiofrequency ablation in chronic knee osteoarthritis Pain Physician 20 3 2017 E437 E444 28339444
15 Fogarty A.E. Burnham T. Kuo K. Tate Q. Sperry B.P. Cheney C. Walega D.R. Kohan L. Cohen S.P. Cushman D.M. McCormick Z.L. The effectiveness of fluoroscopically guided genicular nerve radiofrequency ablation for the treatment of chronic knee pain due to osteoarthritis: a systematic review Am J Phys Med Rehabil 101 5 2022 482 492 35006653
16 Caragea M. Woodworth T. Curtis T. Blatt M. Cheney C. Brown T. Carson D. Kuo K.T. Randall D. Huang E.Y. Carefoot A. Teramoto M. Mills M. Cooper A. Burnham T. Conger A. McCormick Z.L. Genicular nerve radiofrequency ablation for the treatment of chronic knee joint pain: a real-world cohort study with evaluation of prognostic factors Pain Med 24 12 2023 1332 1340 10.1093/pm/pnad095 PMID: 37428157 37428157
17 Koshi E. Cheney C.W. Sperry B.P. Conger A. McCormick Z.L. Genicular nerve radiofrequency ablation for chronic knee pain using a three-tined electrode: a technical description and case series Pain Med 21 12 2020 3344 3349 10.1093/pm/pnaa204 PMID: 32984887 32984887
18 Koshi E. Meiling J.B. Conger A.M. McCormick Z.L. Burnham T.R. Long-term clinical outcomes of genicular nerve radiofrequency ablation for chronic knee pain using a three-tined electrode for expanded nerve capture Interv Pain Med 1 1 2022 100003 10.1016/j.inpm.2021.100003
19 Guven Kose S. Kirac Unal Z. Kose H.C. Celikel F. Akkaya O.T. Ultrasound-guided genicular nerve radiofrequency treatment: prospective randomized comparative trial of a 3-nerve protocol versus a 5-nerve protocol Pain Med 24 7 2023 758 767 10.1093/pm/pnad025 PMID: 36869680 36869680
20 Conger A. McCormick Z.L. Henrie A.M. Pes anserine tendon injury resulting from cooled radiofrequency ablation of the inferior medial genicular nerve PM&R 11 11 2019 1244 1247 30859692
21 McCormick Z.L. Walega D.R. Third-degree skin burn from conventional radiofrequency ablation of the inferiomedial genicular nerve Pain Med 19 5 2018 1095 1097 29025030
22 Khan D. Nagpal G. Conger A. McCormick Z.L. Walega D.R. Clinically significant hematoma as a complication of cooled radiofrequency ablation of the genicular nerves; a case series Pain Med 21 7 2020 1513 1515 31880783
23 Khanna A. Knox N. Sekhri N. Septic arthritis following radiofrequency ablation of the genicular nerves Pain Med 20 7 2019 1454 1456 30698785
24 McCormick Z.L. Patel J. Conger A. Smith C.C. Spine Intervention Society's Patient Safety Committee. The safety of genicular nerve radiofrequency ablation Pain Med 22 2 2021 518 519 10.1093/pm/pnaa355 33517427
25 Reddy R.D. McCormick Z.L. Marshall B. Mattie R. Walega D.R. Cooled radiofrequency ablation of genicular nerves for knee osteoarthritis pain: a protocol for patient selection and case series Anesthesiol Pain Med 6 6 2016 e39696 10.5812/aapm.39696
26 Badin D. Ortiz-Babilonia C. Musharbash F.N. Jain A. Disparities in elective spine surgery for medicaid beneficiaries: a systematic review Global Spine J 13 2 2023 534 546 10.1177/21925682221103530 PMID: 35658589 35658589
27 Shau D. Shenvi N. Easley K. Smith M. Bradbury T. Guild G. 3rd Medicaid payer status is associated with increased 90-day morbidity and resource utilization following primary total hip arthroplasty: a propensity-score-matched analysis J Bone Joint Surg Am 100 23 2018 2041 2049 10.2106/JBJS.17.00834 30516627
28 Harris P.A. Taylor R. Thielke R. Payne J. Gonzalez N. Conde J.G. Research electronic data capture (REDCap) – a metadata-driven methodology and workflow process for providing translational research informatics support J Biomed Inf 42 2 2009 Apr 377 381
29 Harris P.A. Taylor R. Minor B.L. Elliott V. Fernandez M. O'Neal L. McLeod L. Delacqua G. Delacqua F. Kirby J. Duda S.N. The REDCap consortium: building an international community of software partners J Biomed Inf 95 2019 Jul 103208 10.1016/j.jbi.2019.103208 PMID: 31078660
30 Dworkin R.H. Turk D.C. Farrar J.T. Haythornthwaite J.A. Jensen M.P. Katz N.P. Kerns R.D. Stucki G. Allen R.R. Bellamy N. Carr D.B. Chandler J. Cowan P. Dionne R. Galer B.S. Hertz S. Jadad A.R. Kramer L.D. Manning D.C. Martin S. McCormick C.G. McDermott M.P. McGrath P. Quessy S. Rappaport B.A. Robbins W. Robinson J.P. Rothman M. Royal M.A. Simon L. Stauffer J.W. Stein W. Tollett J. Wernicke J. Witter J. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations Pain 113 1–2 2005 Jan 9 19 10.1016/j.pain.2004.09.012 PMID: 15621359 15621359
31 Farrar J.T. Young J.P. Jr. LaMoreaux L. Werth J.L. Poole M.R. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale Pain 94 2 2001 Nov 149 158 10.1016/S0304-3959(01)00349-9 PMID: 11690728 11690728
32 Salaffi F. Stancati A. Silvestri C.A. Ciapetti A. Grassi W. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale Eur J Pain 8 4 2004 Aug 283 291 10.1016/j.ejpain.2003.09.004 PMID: 15207508 15207508
33 Reisinger R. Caragea M. Teramoto M. Shipman H. Berry-Rieser D. Oli M. Kendall R. Burnham T. Conger A. McCormick Z. The association of payer type and opioid use on functional improvement at short-term follow-up after lumbosacral transforaminal epidural steroid injection: results of a large registry study Interv Pain Med 1 2 2022 100073
34 Veltre D.R. Yi P.H. Sing D.C. Curry E.J. Endo A. Smith E.L. Li X. Insurance status affects in-hospital complication rates after total knee arthroplasty Orthopedics 41 3 2018 e340 e347 10.3928/01477447-20180226-07 29494747
35 Tirumala V. Klemt C. Esposito J.G. Robinson M.G. Barghi A. Kwon Y.M. Insurance payer type affects outcomes after revision total joint arthroplasty: a matched cohort analysis Arch Bone Jt Surg 10 4 2022 328 338 10.22038/ABJS.2021.56165.2792 PMID: 35721591 35721591
36 Browne J.A. Novicoff W.M. D'Apuzzo M.R. Medicaid payer status is associated with in-hospital morbidity and resource utilization following primary total joint arthroplasty J Bone Joint Surg Am 96 21 2014 e180 10.2106/JBJS.N.00133
37 State Health Access Data Assistance Center (SHADAC) analysis of income inequality (Gini coefficient). State Health Compare, SHADAC, University of Minnesota https://statehealthcompare.shadac.org
