Multi-Level Posterior Cervical Foraminotomy Associated With Increased Post-operative Infection Rates and Overall Re-Operation Relative to Anterior Cervical Discectomy With Fusion or Cervical Disc Arthroplasty

Abstract

Study Design

Retrospective Cohort Study

Objective

Cervical radiculopathy meeting operative criteria has traditionally been managed using anterior cervical discectomy and fusion (ACDF). However, cervical disc arthroplasty (CDA) and posterior cervical foraminotomy (PCF) are also reasonable options. This study aimed to assess differences in postoperative outcomes among patients undergoing multi-level ACDF, CDA, or PCF comparing medical/surgical complications and healthcare utilization parameters.

Methods

Patients who underwent multi-level ACDF, CDA, or PCF between 2012 and 2019 were identified from the American College of Surgeons National Surgical Quality Improvement Project (ACS-NSQIP) database. Patients were stratified based on procedure type and propensity score matched to resolve baseline differences. ANOVA was performed to identify differences in medical complications, surgical complications, and healthcare utilization metrics.

Results

A total of 31 344 patients who underwent an eligible procedure were identified (ACDF: n = 28 089, CDA: n = 1748, PCF: n = 1507), and 684 patients remained in each group following propensity score matching. Patients undergoing multi-level PCF were found to experience longer lengths of hospital stay (PCF: 1.67 ± 1.61 days, ACDF: 1.50 ± 1.32 days, CDA: 1.27 ± 1.05 days, P < .001), higher rates of reoperation (PCF: 3.2%, ACDF: 1.0%, CDA: .4%, P = .020), superficial infection (PCF: 1.3%, ACDF: .3%, CDA: .1%, P = .008) and deep infection (PCF: 1.2%, ACDF: 0%, CDA: 0%, P < .001). There were no outcome differences between multi-level ACDF and CDA.

Conclusions

Patients undergoing multi-level PCF were at increased risk for longer hospital stay, re-operation, and infection relative to those undergoing ACDF and CDA. Future research should aim to uncover the precise mechanisms underlying these complications, as well as analyze long term outcomes.

Level of Evidence

III

Keywords

anterior cervical discectomy and fusion posterior cervical foraminotomy cervical disc arthroplasty radiculopathy

Introduction

Cervical radiculopathy is a debilitating condition characterized by upper arm pain accompanied by sensory deficits in about 85%, reflex deficits in about 71%, and motor deficits in 68% of patients.^1,2 The most common causes of such presentations are cervical disc herniations leading to posterolateral or intra-foraminal impingement, and diminished disc height with degenerative changes to the uncovertebral joints secondary to cervical spondylosis.^3,4

For ∼75% of patients, cervical radiculopathy is self-limiting and the symptoms resolve over time without intervention. ^5,6 However, non-operative management is often trialed in an effort to provide pain relief and alleviate patient symptoms.⁷ Conservative treatment modalities that have demonstrated some efficacy include pharmacologic therapies such as nonsteroidal anti-inflammatory drugs, neck immobilization using cervical collar, mechanical traction, physical therapy, and steroid injections.^8‐12 Despite these measures, up to one third of patients continue to experience symptoms after 6 weeks of treatment and are considered for referral to a spine surgeon.¹³

There are several surgical procedures available to address cervical radiculopathy, which can be broadly categorized into anterior and posterior approaches. The anterior surgeries include anterior cervical discectomy and fusion (ACDF) and cervical disc arthroplasty (CDA), while the posterior procedure is a posterior cervical foraminotomy (PCF).¹⁴ ACDF is one of the most commonly performed spine surgeries and is considered to be the standard surgical treatment for cervical disc disease¹⁵. Although it offers excellent outcomes, ACDF is associated with perioperative complications such as dysphagia and hematoma formation, decreased range of motion, nonunion and delayed complications such as adjacent segment disease.¹⁶ In contrast, CDA is a relatively novel procedure that gained popularity in the U.S. market within the last 2 decades, with the purported benefits of maintaining physiologic motion and minimizing adjacent segment degeneration.¹⁷ The surgical technique is essentially identical to ACDF with the principal exception being the implantation of a disc arthroplasty device instead of an intervertebral graft following discectomy.¹⁸ While typically asymptomatic, CDA can result in complications such as heterotopic ossification and osteolysis, increasing the risk for spinal segment fusion and device subsidence, respectively.¹⁹ Lastly, PCF involves enlarging the neuro-foramen by removing a portion of the involved facet joint with/without disc removal in order to decompress the nerve root causing the patient’s symptoms.²⁰ Similar to the aforementioned surgical approaches, PCF also offers favorable outcomes for the treatment of cervical radiculopathy, reducing common risks associated with the anterior approach such as implant-related complications and dysphagia.²¹ However, it also requires posterior muscle dissection which can lead to increased intra-operative bleeding, pain, and wound/nerve complications.

While outcomes from ACDF, CDA, and PCF have been studied for single-level procedures,²² there is a paucity of literature comparing them in the context of multi-level procedures. To offer patients, the most optimal treatment option for multi-segment cervical radiculopathy, it is important to identify any significant disparities among these surgical approaches. To this end, this study aims to determine differences in postoperative outcomes among patients undergoing multi-level ACDF, CDA, and PCF by evaluating: (1) 30-day postoperative medical complications; (2) surgical complications and (3) healthcare utilization parameters (reoperation, readmission, and length of stay).

Methods

Study Design and Data Sources

Cervical radiculopathy patients who underwent multi-level ACDF, CDA, or PCF (defined as >1 level) between January 1st, 2012 and December 31st, 2019 were retrospectively identified from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database. Created and validated by surgeons, ACS-NSQIP currently serves as a leading national quality improvement database that provides incredible utility to researchers due to its data collection from patients who undergo major surgeries in over 500 centers in North America.²³ The ACS-NSQIP database encompasses risk-adjusted and case-mixed adjusted information for important patient variables, including patient demographics (sex, age, race, and body mass index (BMI)), American Society of Anesthesiologists (ASA) physical status class, existing comorbidities, and complications spanning up to 30 days post-operation. ACS-NSQIP captures major complications including any readmissions or reoperations, superficial surgical site infections (SSI) that involve only the skin or subcutaneous tissue of the incision, deep SSI (involving fascial and muscle layers), organ/space infections (infections involving any anatomy that was opened or manipulated during surgery apart from incision), sepsis (systemic inflammatory response syndrome with documented infection) and septic shock (sepsis with documented organ and/or circulatory dysfunction), renal complications, cardiac dysfunction, and wound dehiscence.²⁴ Standard definitions of complications are established by the central NSQIP to ensure uniform adherence among cooperating surgical centers. This study did not necessitate Institutional Review Board (IRB) approval nor individual consent as any patient data was presented in a de-identified manner through a publicly available dataset.

Study Population

Patients were identified as having undergone multi-level ACDF, CDA or PCF by the current procedural terminology (CPT) codes. Multi-level ACDF cases were identified with CPT code 22 551 with the addition of a single or multiple 22 552 codes and with CPT code 22 554 with the addition of a single or multiple 22 585 codes. Multi-level CDA cases were identified with CPT code 22 856 with the addition of a single or multiple 22 858 codes. Multi-level PCF cases were identified with CPT code 63 020 with the addition of a single or multiple 63 035 codes. Following the identification of patients who underwent multi-level ACDF, CDA, or PCF, exclusion criteria were applied. Additional procedure CPT codes identified patients who underwent only a singular level of fusion, those who underwent fusion from a posterior approach, and those who received additional concomitant spinal procedures, such as vertebral corpectomy, decompression, or removal of an artificial disc. These patients were subsequently excluded.

A total of 31 344 cervical radiculopathy patients undergoing surgery were identified and included in the study (ACDF: n = 28 089, CDA: n = 1,748, PCF: n = 1507). (See Table 1). Of the included patients, a slight female predominance was detected in the multi-level ACDF (50.6%) and multi-level CDA (50.1%) cohorts. A more skewed distribution of sex existed in the multi-level PCF cohort, which exhibited a male predominance (60.6%). Most patients in each cohort presented with a BMI >30 (multi-level ACDF: 50.9%; multi-level CDA: 45.6%; multi-level PCF: 48.7%; P < .001). The multi-level ACDF cohort demonstrated the highest proportion of diabetics (16.4%; P < .001), smokers (28.0%; P < .001), and patients with dyspnea, on exertion or resting (5.2%; P < .001), patients with COPD (4.5%; P < .001), patients with hypertension (45.6%; P < .001), and patients with partially or totally dependent functional status (1.6%; P < .001). The highest proportion of patients who were identified as ASA class ≥ 3, a classification system that captures the burden of patient comorbidities, was found in the multi-level ACDF cohort (multi-level ACDF: 42.8%; multi-level CDA: 23.1%; multi-level PCF: 36.6%; P < .001).

Table 1.

Before Matching Demographics of Cervical Radiculopathy Patients Who Underwent 2+ Levels ACDF, CDA, OR PCF

	ACDF (n = 28 089)	CDA (n = 1748)	PCF (n = 1507)	P
Age	54.21 +/− 11.66	45.54 +/− 10.39	54.22 +/− 12.72	<.001
Sex				<.001
1. Female	14 200 (50.6%)	875 (50.1%)	594 (39.4%)
2. Male	13 889 (49.4%)	873 (49.9%)	913 (60.6%)
Body Mass Index	30.58 +/− 6.75	29.76 +/− 6.29	30.20 +/− 6.36	<.001
1. <18.5	217 (.8%)	7 (.4%)	14 (.9%)
2. 18.5 - 24.9	4525 (16.1%)	332 (19.0%)	239 (15.9%)
3. 25.0 - 29.9	9035 (32.2%)	611 (35.0%)	520 (34.5%)
4. >29.9	14 312 (50.9%)	798 (45.6%)	734 (48.7%)
Diabetes				<.001
1. No	23 489 (83.6%)	1596 (91.3%)	1289 (85.5%)
2. Non-insulin	2924 (10.4%)	107 (6.1%)	143 (9.5%)
3. Insulin	1676 (6.0%)	45 (2.6%)	75 (5.0%)
Smoker				<.001
1. No	20 233 (72.0%)	1365 (78.1%)	1125 (74.7%)
2. Yes	7856 (28.0%)	383 (21.9%)	382 (25.3%)
Dyspnea				<.001
1. No	26 618 (94.8%)	1716 (98.2%)	1463 (97.1%)
2. Exertion	1392 (4.9%)	29 (1.6%)	40 (2.6%)
3. Resting	79 (.3%)	3 (.2%)	4 (.3%)
COPD				<.001
1. No	26 816 (95.5%)	1728 (98.9%)	1441 (95.6%)
2. Yes	1273 (4.5%)	20 (1.1%)	66 (4.4%)
Hypertension				<.001
1. No	15 276 (54.4%)	1333 (76.3%)	876 (58.1%)
2. Yes	12 813 (45.6%)	415 (23.7%)	631 (41.9%)
Steroids				.1364
1. No	27 175 (96.7%)	1706 (97.6%)	1456 (96.6%)
2. Yes	914 (3.3%)	42 (2.4%)	31 (3.4%)
Functional Status				<.001
1. Independent	27 634 (98.4%)	1742 (99.7%)	1487 (98.7%)
2. Partially Dependent	419 (1.5%)	6 (.3%)	19 (1.2%)
3. Totally Dependent	36 (.1%)	0 (.0%)	1 (.1%)
ASA Classification				<.001
1	959 (3.4%)	164 (9.4%)	87 (5.8%)
2	15 114 (53.8%)	1180 (67.5%)	868 (57.6%)
3	11 483 (40.9%)	399 (22.8%)	527 (35.0%)
4	530 (1.9%)	5 (.3%)	25 (1.6%)
5	3 (.0)	0	0

Outcomes of Interest

Primary outcomes of interest were: operative time, length of stay, 30-day readmission and 30-day re-operation. Secondary outcomes of interest were 30-day postoperative complications, which included superficial infection, deep infection, organ/space infection, wound dehiscence, as well as renal complications (urinary tract infections, progressive renal insufficiency, acute renal failure), cardiopulmonary complications (use of mechanical ventilation >48 hours postoperatively, pneumonia, unplanned intubation, cardiac arrest, myocardial infarction), thromboembolic complications (postoperative transfusions, pulmonary embolism (PE), deep vein thrombosis (DVT), cerebrovascular accident (CVA) and septic complications (sepsis, septic shock).

Statistical Analysis

Univariate analysis identified the distribution of patient demographics, existing comorbidities and patient outcomes among patients undergoing multi-level ACDF, CDA, or PCF. Statistically significant differences in baseline patient demographics, existing comorbidities, steroid use, functional status and ASA classification were resolved using propensity score matching (1:1:1). The propensity score-matching process included the following variables: age, sex, BMI category, diabetes, smoking status, and the presence of chronic obstructive pulmonary disease (COPD), hypertension, overall functional status, and ASA classification. The propensity score-matching process was then followed by a comparison of outcomes. As appropriate, Fisher’s exact test and Chi-square tests were utilized to analyze differences in the counts and percentages of categorical variables, such as readmissions, reoperations, and medical/surgical complications. Analysis of variance (ANOVA) was utilized to analyze differences in the means ± standard deviations (SDs) of continuous variables, such as operative time and length of stay. All tests were 2 tailed with an alpha level of .05 (significance level at P < .05).

Results

A total of 31 344 cervical radiculopathy patients undergoing multi-level ACDF, CDA or PCD were identified (See Table 1). Of these, 28 089 patients were identified to have undergone multi-level ACDF, with a mean (±SD) age of 54.21 (±11.66) years, a BMI of 30.58 (±6.75), and a slight female majority (50.6%). 1748 patients were identified to have undergone multi-level CDA, with a mean (±SD) age of 45.54 (±10.39) years, a BMI of 29.76 (±6.29), and a slight female majority (50.1%). 1507 patients were identified to have undergone multi-level PCF, with a mean (±SD) age of 54.22 (±12.72) years, a BMI of 30.20 (±6.36), and a male majority (60.6%). Univariate analysis showed significant differences among the age, sex, and BMI of the cohorts, as well as among existing comorbidities, including diabetes (P < .001), smoking status (P < .001), dyspnea (P < .001), COPD (P < .001), hypertension (P < .001), functional status (P < .001), and ASA classification (P < .001), but not steroid use (P = .1364).

Propensity score matching allowed for proper adjusting of the statistically significant differences established through univariate analysis (See Table 2), which included 684 patients with cervical radiculopathy who received either multi-level ACDF, CDA or PCF. All statistically significant discrepancies in patient demographics and existing comorbidities were removed, which included age (P = .106), dyspnea (P = .9941), sex, BMI, diabetes, smoking status, COPD, hypertension, steroid use, functional status, and ASA classification.

Table 2.

After Matching Demographics of Cervical Radiculopathy Patients Who Underwent 2+ Levels ACDF, CDA, or PCF

	ACDF (n = 684)	CDA (n = 684)	PCF (n = 684)	P
Age	49.17 +/− 10.68	48.43 +/− 10.00	49.61 +/− 10.45	.106
Sex				1
1. Female	295 (43.1%)	295 (43.1%)	295 (43.1%)
2. Male	389 (56.9%)	389 (56.9%)	389 (56.9%)
Body Mass Index	30.63 +/− 6.90	30.63 +/− 6.90	30.63 +/− 6.90	1
1. <18.5	1 (.1%)	1 (.1%)	1 (.1%)
2. 18.5 - 24.9	119 (17.4%)	119 (17.4%)	119 (17.4%)
3. 25.0 - 29.9	275 (40.2%)	275 (40.2%)	275 (40.2%)
4. >29.9	289 (42.3%)	289 (42.3%)	289 (42.3%)
Diabetes				1
1. No	657 (96.1%)	657 (96.1%)	657 (96.1%)
2. Non-insulin	18 (2.6%)	18 (2.6%)	18 (2.6%)
3. Insulin	9 (1.3%)	9 (1.3%)	9 (1.3%)
Smoker				1
1. No	559 (81.7%)	559 (81.7%)	559 (81.7%)
2. Yes	125 (18.3%)	125 (18.3%)	125 (18.3%)
Dyspnea				.9941
1. No	673 (98.4%)	674 (98.6%)	675 (98.7%)
2. Exertion	10 (1.5%)	9 (1.3%)	8 (1.2%)
3. Resting	1 (.1%)	1 (.1%)	1 (.1%)
COPD				1
1. No	683 (99.9%)	683 (99.9%)	683 (99.9%)
2. Yes	1 (.1%)	1 (.1%)	1 (.1%)
Hypertension				1
1. No	522 (76.3%)	522 (76.3%)	522 (76.3%)
2. Yes	162 (23.7%)	162 (23.7%)	162 (23.7%)
Steroids				1
1. No	681 (99.6%)	681 (99.6%)	681 (99.6%)
2. Yes	3 (.4%)	3 (.4%)	3 (.4%)
Functional Status				1
1. Independent	684 (100%)	684 (100%)	684 (100%)
2. Partially Dependent	0 (.0%)	0 (.0%)	0 (.0%)
3. Totally Dependent	0 (.0%)	0 (.0%)	0 (.0%)
ASA Classification				1
1	43 (6.3%)	43 (6.3%)	43 (6.3%)
2	518 (75.7%)	518 (75.7%)	518 (75.7%)
3	123 (18.0%)	123 (18.0%)	123 (18.0%)
4	0 (.0%)	0 (.0%)	0 (.0%)
5	0 (.0%)	0 (.0%)	0 (.0%)

Before propensity score matching, statistically significant discrepancies for postoperative outcomes existed in operation time (P < .001), length of stay (P < .001), any readmission (P < .001), any reoperation (P = .029), superficial infection (P < .001), deep infection (P < .001), wound dehiscence (P < .001), transfusions (P < .001), mechanical ventilation (P = .015), pneumonia (P = .001), unplanned intubation (P = .016), and sepsis (P = .001) among patients with cervical radiculopathy who received multi-level ACDF, CDA, or PCF treatment (See Table 3). No statistically significant differences among the cohorts existed in organ/space infection, urinary tract infection, progressive renal insufficiency, acute renal failure, pulmonary embolism, deep vein thrombosis, cardiac arrest, myocardial infarction, cerebrovascular accident, and septic shock (See Table 3).

Table 3.

Before Matching Operative and Post-Operative Data for Cervical Radiculopathy Patients Who Underwent 2 or More Levels ACDF, CDA, OR PCF

	ACDF (n = 28 089)	CDA (n = 1748)	PCF (n = 1507)	P
Operation Time (min)	116.66 +/− 49.64	109.58 +/− 39.80	108.96 +/− 55.21	<.001
Length of Stay (days)	1.77 +/− 2.17	1.30 +/− 1.00	1.89 +/− 1.94	<.001
Any Readmission	1052 (3.7%)	24 (1.4%)	55 (3.6%)	<.001
Any Reoperation	421 (1.5%)	14 (.8%)	28 (1.9%)	.029
Superficial Infection	81 (.3%)	4 (.2%)	21 (1.4%)	<.001
Deep Infection	35 (.1%)	1 (.1%)	13 (.9%)	<.001
Organ/Space Infection	37 (.1%)	0 (.0%)	3 (.2%)	.238
Wound Dehiscence	6 (.0%)	0 (.0%)	5 (.3%)	<.001
Transfusions	52 (.2%)	2 (.1%)	14 (.9%)	<.001
Urinary Tract Infection	153 (.5%)	3 (.2%)	9 (.6%)	.104
Progressive Renal Insufficiency	8 (.0%)	0 (.0%)	0 (.0%)	.629
Acute Renal Failure	15 (.1%)	1 (.1%)	1 (.1%)	.977
Mechanical Ventilation	89 (.3%)	0 (.0%)	1 (.1%)	.015
Pneumonia	177 (.6%)	0 (.0%)	4 (.3%)	.001
Unplanned Intubation	119 (.4%)	2 (.1%)	1 (.1%)	.016
Pulmonary Embolism	65 (.2%)	0 (.0%)	1 (.1%)	.056
Deep Vein Thrombosis	82 (.3%)	1 (.1%)	5 (.3%)	.184
Cardiac Arrest	33 (.1%)	0 (.0%)	1 (.1%)	.308
Myocardial Infarction	39 (.1%)	1 (.1%)	4 (.3%)	.280
Cerebrovascular Accident	16 (.1%)	1 (.1%)	1 (.1%)	.989
Sepsis	64 (.2%)	1 (.1%)	10 (.7%)	.001
Septic Shock	21 (.1%)	0 (.0%)	2 (.1%)	.365

After 1:1:1 propensity score matching, there were no differences in patient demographics or comorbidities of ACDF, CDA, or PCF groups (P > .05). After propensity score matching, several post-operative outcomes maintained statistically significant differences (See Table 4). There were statistically significant discrepancies in length of stay of multi-level ACDF (1.50 ± 1.32 days), multi-level CDA (1.27±1.05 days), and multi-level PCF (1.67±1.61 days) (P < .001). Patients undergoing multi-level ACDF experienced the same rates of sepsis events (0%) relative to those who underwent multi-level CDA (0%) or multi-level PCF (0%) (P = .050). Namely, cervical radiculopathy patients receiving multi-level PCF experienced higher rates of reoperation (3.2%) vs those receiving multi-level ACDF (1.0%) or multi-level CDA (.4%) (P = .020). Patients who received multi-level PCF also experienced higher rates of superficial infection (1.3%) compared to those receiving multi-level ACDF (.3%) and those receiving multi-level CDA (.1%) (P = .008). Correspondingly, patients who received multi-level PCF also experienced higher rates of deep infection (1.2%) compared to those who received multi-level ACDF (0%) or multi-level CDA (0%) (P < .001). There were no other significant differences in medical or surgical complications among the ACDF, CDA, or PCF patients, including any readmission, organ/space infection, wound dehiscence, transfusions, urinary tract infection, acute renal failure, pneumonia, unplanned intubation, pulmonary embolism, deep vein thrombosis, and myocardial infarction (P > .05, each).

Table 4.

After Matching Operative and Post-Operative Data for Cervical Radiculopathy Patients Who Underwent 2 OR More Levels ACDF, CDA, OR PCF

	ACDF (n = 684)	CDA (n = 684)	PCF (n = 684)	P
Operation Time (min)	114.51 +/− 48.95	110.10 +/− 39.69	108.26 +/− 56.89	.054
Length of Stay (days)	1.50 +/− 1.32	1.27 +/− 1.05	1.67 +/− 1.61	<.001
Any Readmission	20 (2.9%)	11 (1.6%)	22 (3.2%)	.136
Any Reoperation	7 (1.0%)	3 (.4%)	14 (2.0%)	.020
Superficial Infection	1 (.1%)	2 (.3%)	9 (1.3%)	.008
Deep Infection	0 (.0%)	0 (.0%)	8 (1.2%)	<.001
Organ/Space Infection	2 (.3%)	0 (.0%)	1 (.1%)	.367
Wound Dehiscence	0 (.0%)	0 (.0%)	2 (.3%)	.135
Transfusions	0 (.0%)	0 (.0%)	2 (.3%)	.135
Urinary Tract Infection	2 (.3%)	0 (.0%)	3 (.4%)	.246
Progressive Renal Insufficiency	0 (.0%)	0 (.0%)	0 (.0%)	N/A
Acute Renal Failure	0 (.0%)	1 (.0%)	0 (.0%)	.368
Mechanical Ventilation	0 (.0%)	0 (.0%)	0 (.0%)	N/A
Pneumonia	2 (.3%)	0 (.0%)	1 (.1%)	.367
Unplanned Intubation	0 (.0%)	2 (.3%)	0 (.0%)	.135
Pulmonary Embolism	2 (.3%)	0 (.0%)	0 (.0%)	.135
Deep Vein Thrombosis	2 (.3%)	0 (.0%)	0 (.0%)	.135
Cardiac Arrest	0 (.0%)	0 (.0%)	0 (.0%)	N/A
Myocardial Infarction	1 (.1%)	1 (.1%)	0 (.0%)	.606
Cerebrovascular Accident	0 (.0%)	0 (.0%)	0 (.0%)	N/A
Sepsis	0 (.0%)	0 (.0%)	0 (.0%)	.050
Septic Shock	0 (.0%)	0 (.0%)	0 (.0%)	N/A

Discussion

To our knowledge, this is the first study investigating and comparing outcomes among the 3 most common multi-level surgeries performed for symptomatic cervical radiculopathy. Our initial hypothesis was that there would be no difference in postoperative outcomes among the ACDF, CDA, and PCF cohorts, as corroborated by a meta-analysis on the topic.²⁵ However, the analysis conducted in this retrospective cohort study demonstrated that compared to ACDF and CDA, multi-segment PCF is associated with significantly prolonged length of hospital stay and increased rates of reoperation and infection including superficial skin infection and deep infection. This is consistent with the findings reported by Ng et al., who detected similar trends in reoperation and infection following single-level PCF.²² Contrary to their additional discovery of differences in operative time based on procedure type, however, we found no other significant differences in outcomes.

The limitations of the present study are primarily 2-fold: stemming from its design as a retrospective cohort study and from its use of the data available through the ACS-NSQIP database. First, our study results rely on data available in the NSQIP database, identifying patients through Current Procedural Terminology codes regardless of their diagnosis. To this end, coding errors can be present, although NSQIP has been well-established for its robust and consistent auditing resulting in <1.8% inter-rater variability. Second, the CPT codes available were binary in nature (single-level and multi-level) and were unable to distinguish between different multi-level procedures. This is important to consider when ACDF involving multiple levels have high rates of pseudarthrosis and increased nonunion rates, and varies based on the number of fused segments (eg 0-10% for single-level and up to 50% for 3-level constructs) ¹⁵. To this end, single-institution studies may allow for more precise discernment of multi-level procedures, although they would likely lack the statistical power of this study. In addition, as a retrospective cohort study, the study lacks completely randomized data, and is therefore subject to selection bias. In other words, without randomization, this study has poor control over pre-existing covariates and confounders, which makes it a possibility that patient demographics among the multi-level ACDF, CDA, and PCF groups had differed at baseline. To address this limitation, propensity score matching identified patients from each of the 3 groups with similar baseline characteristics. This was confirmed with a P-value greater than .05 in patient demographics or comorbidities among the patients in each group, suggesting there were no statistically significant differences in baseline characteristics. Secondly, the study’s use of the ACS-NSQIP database allowed analysis of outcomes within a 30-day period after operation but lacked outcomes that were followed over a longer term. Furthermore, the study was limited only to the complications defined and made available by NSQIP, which lacked information on complications more specific to surgery on the cervical spine, as detailed earlier, such as dysphagia and adjacent segment disease.¹⁶ Future research efforts may aim to analyze longer-term complications as well as complications specific to cervical spine surgery.

In recent years, various studies have begun exploring outcomes following PCF. One such study by Witiw et al. utilized a propensity score matching methodology like this paper to compare PCF and ACDF and also found a higher infection rate in the PCF cohort.²⁶ Another retrospective cohort study found posterior approaches in general to be a risk factor for superficial and deep surgical site infections, as consistent with the findings in this paper. The authors surmised that this might be due to the relatively decreased vascularity and lymphatic drainage of the posterior spine compared to the anterior spine, which may diminish the clearance of incidental bacterial contamination.²⁷ This elevated incidence can be attributed to a variety of characteristics, including procedural factors such as instrumented surgeries, multi-level surgeries, open exposures, lack of local antibiotic usage, increased estimated blood loss, and longer surgeries. One study suggested that the extensive muscle dissection required for posterior approaches might be to blame, which can de-vascularize paraspinal muscles and predispose the patient to a deep infection. Retractors used in posterior spinal surgery can also compress the paraspinal musculature and cause ischemia-induced infections.²⁸ Some patient characteristics such as diabetes, history of postoperative wound infection, obesity, smoking, radiation exposure, and cancer have also been found to be associated with increased infection rates, although the precise mechanism and contribution of these factors have yet to be uncovered.^27‐30 While many of these risk factors are not modifiable, taking judicious measures such as preparation of the surgical site and drapes with alcohol foam, and/or considering insertion of an additional suprafascial drain in obese patients or placement of powdered antibiotic such as vancomycin into the wound in addition to administering standard perioperative intravenous antibiotics may decrease the incidence in patients predisposed to infections.³¹

Contrary to our findings on the other hand, 2 studies comparing patients who underwent ACDF or PCF found no differences in reoperation rates at the index level within 2 years following the initial surgery.^32,33 While intraoperative complications such as hematomas, infections, and cerebrospinal fluid leaks secondary to dural puncture can necessitate a revision procedure, the most common indications are recurrent disc herniation in PCF and inadequate decompression.²⁹ Patients could also be counseled to reduce or avoid spinal loading in the immediate postoperative period in order to minimize the risk of disc reherniation.²⁹

Fortunately, there has been an increasing amount of literature published that has delved into the comparison between PCF and other procedure types, as well as between open and minimally invasive approaches. Although these have been exclusive to single level spinal surgeries, they offer novel perspectives that surgeons can benefit from when risk-stratifying their patients. For instance, the advent and increasing utilization of minimally invasive and endoscopic approaches, especially in the context of PCF, can reduce the incidence of both infection and reoperation in the future by minimizing tissue exposure, contamination, and disruption.^25,35 In addition, open posterior exposures are associated with neck and shoulder pain/spasm,³⁶ and using minimally invasive approaches can therefore minimize these undesirable outcomes.³⁷ On the other hand, it has quite a few limitations and potential complications that must be considered prior to patient selection. When performed without arthrodesis for example, there is concern for iatrogenic cervical spine instability following facetectomy.³⁹ Similar to the findings in the current study, Lin et al. also found a significantly increased rate of reoperation in their PCF cohort relative to the ACDF and CDA cohorts.⁴⁰ Lastly, a paper published shortly prior to the submission of the current study reported shorter operative time for minimally invasive PCF and greater estimated blood loss with no postoperative improvement in arm pain for open PCF patients compared to ACDF and CDA groups.⁴¹ Thus, all things considered including the pre-existing literature as well as the findings from this study, a cautious risk-benefit analysis should be carried out prior to choosing PCF for patients, and when possible, minimally invasive approaches should be utilized.

Conclusion

In conclusion, the present study suggests that patients with cervical radiculopathy who underwent multi-level PCF experience longer length of stay, higher rates of reoperation as well as higher rates of superficial and deep infection 30 days after operation than those who underwent treatment via multi-level ACDF or multi-level CDA. As this study now adds to the literature that suggests an association between the posterior approach and higher rates of infection, future efforts should be made to elucidate the precise mechanism behind this phenomenon. Additionally, future studies should analyze postoperative complications on a longer time scale as well as complications specific to those most associated with surgery of the cervical spine.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Mitchell K. Ng

Joydeep Baidya

References

Woods

Hilibrand

. Cervical Radiculopathy. Journal of Spinal Disorders & Techniques. 2015;28(5):E251-E259. doi:10.1097/BSD.0000000000000284

Abbed

Coumans

JVCE

. Cervical radiculopathy: pathophysiology, presentation, and clinical evaluation. Neurosurgery. 2007;60(1 Supp1 1):S28-34. doi:10.1227/01.NEU.0000249223.51871.C2

Rhee

Yoon

Riew

. Cervical radiculopathy. J Am Acad Orthop Surg. 2007;15(8):486-494. doi:10.5435/00124635-200708000-00005

Corey

Comeau

. Cervical radiculopathy. Med Clin North Am. 2014;98(4):791-799, xii. doi:10.1016/j.mcna.2014.04.001

Shelerud

Paynter

. Rarer causes of radiculopathy: Spinal tumors, infections, and other unusual causes. Physical Medicine and Rehabilitation Clinics of North America. 2002;13(3):645-696. doi:10.1016/S1047-9651(02)00012-8

Bono

Ghiselli

Gilbert

, et al. An evidence-based clinical guideline for the diagnosis and treatment of cervical radiculopathy from degenerative disorders. Spine J. 2011;11(1):64-72. doi:10.1016/j.spinee.2010.10.023

Childress

Becker

. Nonoperative management of cervical radiculopathy. Am Fam Physician. 2016;93(9):746-754.

Machado

Maher

Ferreira

Day

Pinheiro

Ferreira

. Non-steroidal anti-inflammatory drugs for spinal pain: a systematic review and meta-analysis. Ann Rheum Dis. 2017;76(7):1269-1278. doi:10.1136/annrheumdis-2016-210597

Levine

Albert

Smith . Cervical radiculopathy: Diagnosis and nonoperative management. J Am Acad Orthop Surg. 1996;4(6):305-316. doi:10.5435/00124635-199611000-00003

10.

Graham

Gross

Goldsmith

, et al. Mechanical traction for neck pain with or without radiculopathy. Cochrane Database of Systematic Reviews. Published online July 16, 2008. doi:10.1002/14651858.CD006408.pub2

11.

Boyles

Toy

Mellon

Hayes

Hammer

. Effectiveness of manual physical therapy in the treatment of cervical radiculopathy: A systematic review. J Man Manip Ther. 2011;19(3):135-142. doi:10.1179/2042618611Y.0000000011

12.

Diwan

Manchikanti

Benyamin

, et al. Effectiveness of cervical epidural injections in the management of chronic neck and upper extremity pain. Pain Physician. 2012;15(4):E405-E434.

13.

Eubanks

. Cervical radiculopathy: nonoperative management of neck pain and radicular symptoms. Am Fam Physician. 2010;81(1):33-40.

14.

Leveque

JCA

Marong-Ceesay

Cooper

Howe

. Diagnosis and treatment of cervical radiculopathy and myelopathy. Phys Med Rehabil Clin N Am. 2015;26(3):491-511. doi:10.1016/j.pmr.2015.04.008

15.

Hauerberg

Kosteljanetz

Bøge-Rasmussen

, et al. Anterior cervical discectomy with or without fusion with ray titanium cage. Spine (Phila Pa 1976). 2008;33(5):458-464. doi:10.1097/BRS.0b013e3181657dac

16.

Tasiou

Giannis

Brotis

, et al. Anterior cervical spine surgery-associated complications in a retrospective case-control study. J Spine Surg. 2017;3(3):444-459. doi:10.21037/jss.2017.08.03

17.

Shin

Kim

Son

, et al. Cervical disc arthroplasty: What we know in 2020 and a literature review. Journal of Orthopaedic Surgery. 2021;29(1_suppl):230949902110069. doi:10.1177/23094990211006934

18.

Nunley

Schouwen

KF van

Stone

. Cervical total disc replacement: Indications and technique. Neurosurg Clin N Am. 2021;32(4):419-424. doi:10.1016/j.nec.2021.05.001

19.

Price

Coric

Ray

. Cervical total disc replacement: Complications and complication avoidance. Neurosurg Clin N Am. 2021;32(4):473-481. doi:10.1016/j.nec.2021.05.006

20.

Dodwad

SJM

Dodwad

SNM

Prasarn

Savage

Patel

Hsu

. Posterior cervical foraminotomy. Clinical Spine Surgery: A Spine Publication 2016;29(5):177-185. doi:10.1097/BSD.0000000000000384

21.

Clarke

Ecker

Krauss

McClelland

Dekutoski

. Same-segment and adjacent-segment disease following posterior cervical foraminotomy. Journal of Neurosurgery: Spine. 2007;6(1):5-9. doi:10.3171/spi.2007.6.1.2

22.

Emara

Rajan

P v.

, et al. Single-level posterior cervical foraminotomy associated with increased incidence of early postoperative wound infection rates relative to anterior cervical discectomy with fusion and cervical disc arthroplasty. Journal of Spine Surgery. 2021;0(0):0-0. doi:10.21037/jss-21-39

23.

Bilimoria

Liu

Paruch

, et al. Development and evaluation of the universal ACS NSQIP surgical risk calculator: A decision aid and informed consent tool for patients and surgeons. J Am Coll Surg. 2013;217(5):833-842.e1-3. doi:10.1016/j.jamcollsurg.2013.07.385

24.

Molina

Thakore

R v.

Blumer

Obremskey

Sethi

. Use of the national surgical quality improvement program in orthopaedic surgery. Clinical Orthopaedics & Related Research. 2015;473(5):1574-1581. doi:10.1007/s11999-014-3597-7

25.

Platt

Fessler

Traynelis

O’Toole

. Minimally invasive posterior cervical foraminotomy versus anterior cervical fusion and arthroplasty: Systematic review and meta-analysis. Global Spine Journal. Published online December 8, 2021. doi:10.1177/21925682211055094

26.

Witiw

Smieliauskas

O’Toole

Fehlings

Fessler

. Comparison of anterior cervical discectomy and fusion to posterior cervical foraminotomy for cervical radiculopathy: Utilization, costs, and adverse events 2003 to 2014. Neurosurgery. 2019;84(2):413-420. doi:10.1093/neuros/nyy051

27.

Pull ter Gunne

Cohen

. Incidence, prevalence, and analysis of risk factors for surgical site infection following adult spinal surgery. Spine (Phila Pa 1976). 2009;34(13):1422-1428. doi:10.1097/BRS.0b013e3181a03013

28.

Levi

Dickman

Sonntag

. Management of postoperative infections after spinal instrumentation. J Neurosurg. 1997;86(6):975-980. doi:10.3171/jns.1997.86.6.0975

29.

Stadler

Wong

Graham

Liu

. Complications associated with posterior approaches in minimally invasive spine decompression. Neurosurg Clin N Am. 2014;25(2):233-245. doi:10.1016/j.nec.2013.12.003

30.

Zhou

Wang

Huo

Xiong

Kang

Xue

. Incidence of surgical site infection after spine surgery. Spine (Phila Pa 1976). 2020;45(3):208-216. doi:10.1097/BRS.0000000000003218

31.

Pahys

Cho

, et al. Methods to decrease postoperative infections following posterior cervical spine surgery. The Journal of Bone and Joint Surgery. 2013;95(6):549-554. doi:10.2106/JBJS.K.00756

32.

Lubelski

Healy

Silverstein

, et al. Reoperation rates after anterior cervical discectomy and fusion versus posterior cervical foraminotomy: a propensity-matched analysis. Spine J. 2015;15(6):1277-1283. doi:10.1016/j.spinee.2015.02.026

33.

Ruetten

Komp

Merk

Godolias

. Full-Endoscopic cervical posterior foraminotomy for the operation of lateral disc herniations using 5.9-mm endoscopes. Spine (Phila Pa 1976). 2008;33(9):940-948. doi:10.1097/BRS.0b013e31816c8b67

34.

McGirt

Ambrossi

GLG

Datoo

, et al. Recurrent disc herniation and long‐term back pain after primary lumbar discectomy. Neurosurgery. 2009;64(2):338-345. doi:10.1227/01.NEU.0000337574.58662.E2

35.

Platt

Gerard

O’Toole

. Comparison of outcomes following minimally invasive and open posterior cervical foraminotomy: description of minimally invasive technique and review of literature. J Spine Surg. 2020;6(1):243-251. doi:10.21037/jss.2020.01.08

36.

Hosono

Yonenobu

Ono

. Neck and shoulder pain after laminoplasty. A noticeable complication. Spine (Phila Pa 1976). 1996;21(17):1969-1973. doi:10.1097/00007632-199609010-00005

37.

Song

Zhang

Hao

, et al. Microsurgery or open cervical foraminotomy for cervical radiculopathy? A systematic review. Int Orthop. 2016;40(6):1335-1343. doi:10.1007/s00264-016-3193-4

38.

Rossi

Adamson

. Cervical spine surgery. Neurosurgery Clinics of North America. 2021;32(4):483-492. doi:10.1016/j.nec.2021.05.005

39.

Mazas

Benzakour

Castelain

Damade

Ghailane

Gille

. Cervical disc herniation: which surgery? International Orthopaedics. 2019;43(4):761-766. doi:10.1007/s00264-018-4221-3

40.

Lin

Rui

Sharma

Kotheeranurak

Suen

Kim

. Does the neck pain, function, or range of motion differ after anterior cervical fusion, cervical disc replacement, and posterior cervical foraminotomy? World Neurosurgery. 2019;129:e485-e493. doi:10.1016/j.wneu.2019.05.188

41.

Padhye

Shultz

Alcala

, et al. Surgical treatment of single level cervical radiculopathy. Clinical Spine Surgery: A Spine Publication. 2022;35(4):149-154. doi:10.1097/BSD.0000000000001316