C5 Subcutaneous Fat Thickness Ratio Predicts Surgical Site Infection After Cervical Laminoplasty: A Propensity Score-Matched Study

Abstract

Study Design

Retrospective case-control study.

Objective

Surgical site infection(SSI) is a common complication following cervical spine surgery. Previous studies have used C5-level neck parameters to predict SSI risk. This study employed propensity score matching(PSM) to control for body mass index(BMI), aiming to compare the predictive validity of C5-level parameters against average neck parameters at C5, C6, and C7, while also exploring risk factors of SSI.

Methods

Patients who underwent posterior single-door laminoplasty between November 2012 and April 2025 and developed postoperative SSI were selected as the case group(n = 42). Controls were patients who underwent the same procedure on the same day or within one day before or after those in the case group(n = 69). Data collected included baseline characteristics, surgical indicators, perioperative data, as well as C5, C6 and C7 neck parameters. PSM was conducted using a 1:1 nearest neighbor match with a caliper value set at 20%, matching variables including gender, age, height, weight, and BMI.

Results

After PSM, both groups comprised 38 patients each with no significant differences in baseline characteristics pre- or post-PSM. LASSO regression identified 2 variables: C5 subcutaneous fat thickness ratio and average subcutaneous fat thickness ratio. Multivariate logistic regression indicated that the C5 subcutaneous fat thickness ratio(P = 0.002, OR = 1.068, 95%CI:1.025-1.112) is an independent risk factor for SSI following cervical laminoplasty.

Conclusions

The C5 subcutaneous fat thickness ratio demonstrates a modest predictive advantage for SSI compared to average subcutaneous fat thickness ratio, supporting the feasibility of using preoperative MRI-based C5-level indicators to predict SSI after cervical spine surgery.

Keywords

surgical site infection (SSI)propensity score matching (PSM)subcutaneous fat thickness posterior single-door laminoplasty

Introduction

The posterior single-door laminoplasty, a commonly employed surgical technique for the treatment of myelopathic cervical spondylosis and ossification of the posterior longitudinal ligament (OPLL), effectively alleviates intradural pressure and mitigates patient symptoms while preserving the integrity of the posterior neck muscles.¹ However, surgical site infection (SSI) remains a common complication after posterior cervical procedures, with reported incidence rates ranging from 1.3% to 4.93%.^2,3 SSI results in prolonged hospital stays, increased healthcare costs, and negatively impact patient outcomes; furthermore, it is one of the most prevalent causes for readmission following posterior cervical surgeries.⁴ In previous studies, advanced age,⁵ high body mass index (BMI),⁶ diabetes mellitus (DM),^7,8 prolonged postoperative drainage² and frailty⁹ have been identified as risk factors for SSI after cervical surgeries.

Notably, neck thickness and subcutaneous fat thickness are increasingly recognized as significant factors influencing the occurrence of SSI after posterior cervical surgery in clinical practice. However, there is currently a limited amount of literature addressing the relationship between neck thickness and SSI. In 2013, Mehta et al,¹⁰ through a retrospective study involving 22 patients with SSI, observed that increased subcutaneous fat thickness correlated with a higher incidence of SSI after cervical fusion surgery. Subsequent research by Porche et al¹¹ and Donnally et al¹² in 2021 and 2022 respectively—each including 20 cases of SSI—also demonstrated that neck thickness and subcutaneous fat thickness were closely related to SSI. In 2023, Wang et al² conducted a retrospective study comprising 20 patients with SSI and indicated that subcutaneous fat thickness served as a risk factor for SSIs following cervical laminoplasty.

Previous studies have indicated that patients with a high BMI tend to exhibit increased neck subcutaneous fat thickness.¹⁰ None of the aforementioned 4 studies strictly controlled for differences in BMI between the SSI group and the non-SSI group. This study included 42 patients with SSI following cervical laminoplasty and employed propensity score matching (PSM) to control for variables encompassing age, gender, height, weight and particularly BMI, in order to investigate the relationship between subcutaneous fat thickness and postoperative SSI. Furthermore, while 4 previous studies measured neck thickness and fat thickness at the C5 vertebral level, this study additionally assessed muscle and fat thickness at C5, C6, and C7 levels along with their average measurements to identify an effective predictive tool for SSI following cervical laminoplasty.

Materials and Methods

Patient Population

The medical ethics committee of our hospital approved the study protocol, and the approval number was 2024PHB157-001. A retrospective analysis of medical records from patients following posterior single-door laminoplasty at our center between November 2012 and April 2025 was conducted. A total of 42 postoperative patients with SSI were included in the case group (n = 42). The control group (n = 69) comprised patients who underwent the same procedure on the same day or within one day before or after those in the case group. This temporal proximity matching^13,14 was employed primarily to control for potential confounding by time-varying institutional factors, such as variations in operating room protocols, nursing staff, or seasonal infection rates, thereby isolating the association between patient-specific anatomical factors and SSI risk.

The diagnosis of SSI follows the definition from the Centers for Disease Control and Prevention.^15,16 SSI is classified into superficial incisional SSI, deep incisional SSI, and organ/space SSI. If no implant is present, infection may occur within 30 days post-surgery; if an implant is in place, the observation period extends to one year post-surgery.¹⁷

Exclusion criteria: (1) Patients with a history of spinal surgery; (2) Patients with incomplete medical records; (3) Patients experienced prolonged hospitalization due to non-surgical reasons postoperatively.

Data Collection

The baseline characteristics of patients included age, gender, height, weight, BMI, penicillin allergy, hypertension, coronary artery disease, DM, smoking history, chronic obstructive pulmonary disease (COPD), and long-term use of corticosteroids. Patients who had quit smoking for less than 4 weeks preoperatively were defined as having a smoking history, whereas those who were non-smokers or had quit smoking for ≥4 weeks preoperatively were defined as having no smoking history.¹⁸ The surgical-related indicators involved number of surgical segments, surgical duration and intraoperative blood loss. Perioperative data encompassed the American Society of Anesthesiologists’ (ASA) physical status classification, preoperative hemoglobin (Hb), postoperative 1-day Hb, ΔHb (preoperative Hb-postoperative 1-day Hb), preoperative white blood cell (WBC) count, postoperative 1-day WBC count, ΔWBC (postoperative 1-day WBC-preoperative WBC), preoperative albumin, types of antibiotics administered within 48 h post-surgery, volume of wound drainage after surgery, duration with drains postoperatively, maximum temperature on postoperative day 1, preoperative maximum C-reactive protein (CRP) within 3 weeks, preoperative maximum procalcitonin (PCT) within 3 weeks and allogeneic transfusion post-surgery.

The measurement of neck parameters in patients was conducted using preoperative magnetic resonance imaging (MRI), which include neck thickness at C5, C6, C7 levels and average neck thickness; muscle thickness at C5, C6, C7 levels as well as average muscle thickness and the ratio of muscle thickness; subcutaneous fat thickness at C5, C6, C7 levels as well as average subcutaneous fat thickness and the ratio of subcutaneous fat thickness. The neck parameters were measured at the upper endplate levels of C5, C6, and C7.

Average neck thickness is defined as the average of C5, C6, and C7 neck thickness. Average muscle thickness is defined as the average of C5, C6, and C7 muscle thickness. The ratio of muscle thickness is defined as muscle thickness divided by neck thickness. Average subcutaneous fat thickness is defined as the average of C5, C6, and C7 subcutaneous fat thickness. The ratio of subcutaneous fat thickness is defined as subcutaneous fat thickness divided by neck thickness (Figure 1).

Figure 1.

Neck parameter diagram. (A) C5 muscle thickness; (B) C5 subcutaneous fat thickness; (ab) C5 neck thickness; (D) C6 muscle thickness; (E) C6 subcutaneous fat thickness; (de) C6 neck thickness; (F) C7 muscle thickness; (H) C7 subcutaneous fat thickness; (fh) C7 neck thickness.

Perioperative Procedures

All patients underwent posterior single-door laminoplasty. Following the administration of general anesthesia, patients were positioned in a prone orientation on the operating table. Prophylactic antibiotics were administered 30 to 60 min prior to the surgical procedure and continued for a duration of 48 h postoperatively. For patients with known penicillin allergies, clindamycin or fluoroquinolone antibiotics were utilized as an alternative prophylactic measure.

The incision was closed using a multilayer suture technique. Prior to wound closure, extensive irrigation of the incision was conducted using normal saline; no antibiotic solutions or powders, including vancomycin powder, were applied at this stage. Bilateral drainage tubes were placed intraoperatively; if postoperative drainage was less than 50 mL, the corresponding drain was removed and dressing changed accordingly. Typically, dressings are changed every 3 days following surgery, with sutures being removed on postoperative day seven.

Statistical Analysis

PSM was utilized for 1:1 matching, with the matching variables including age, gender, height, weight and BMI. Using SPSS (version 26; IBM), an independent samples T-test was applied to continuous variables that adhered to a normal distribution; the Mann-Whitney U test was employed for continuous variables that adhered to an abnormal distribution; and the chi-square test was used for categorical variables. These tests were utilized to compare the case group with the control group in order to identify risk factors associated with SSI following cervical laminoplasty. Subsequently, LASSO regression analysis in R version 4.3.3 further refined the selection of SSI risk factors among matched patients. Finally, the identified risk factors were incorporated into a multivariate logistic regression model to ascertain independent risk factors and the receiver operating characteristic (ROC) curve was employed to identify the optimal cutoff value.

Results

The Baseline Characteristics

A total of 42 patients with SSI following cervical laminoplasty were included in the case group, while 69 patients undergoing cervical laminoplasty during the same period comprised the control group. PSM was utilized to perform a 1:1 nearest neighbor match between the 2 groups, with a caliper value established at 20%. The matched variables included gender, age, height, weight and BMI. After matching, both the case and control groups consisted of 38 patients each. Prior to and following matching, no significant differences were observed between the 2 groups concerning gender, age, height, weight, BMI, penicillin allergy, hypertension, coronary artery disease, DM, smoking history, COPD or long-term use of corticosteroids (Table 1).

Table 1.

Univariate Analysis of Baseline Characteristics for SSI After Cervical Laminoplasty

Variables	Before PSM					After PSM
Variables	Case group (SSI, n = 42)	Control group (non-SSI, n = 69)	T value/χ²	P value	SMD	Case group (SSI, n = 38)	Control group (non-SSI, n = 38)	T value/χ²	P value	SMD
Gender^a			1.175	0.278	0.197			0.070	0.791	0.056
Male	29 (69.0%)	54 (78.3%)				28 (73.7%)	29 (76.3%)
Female	13 (31.0%)	15 (21.7%)				10 (26.3%)	9 (23.7%)
Age^a (years)	62.520 ± 10.850	61.300 ± 7.307	−0.645	0.521	0.112	62.080 ± 10.901	63.320 ± 6.439	0.602	0.549	−0.114
Height^a (cm)	166.600 ± 7.924	167.190 ± 6.399	0.410	0.683	−0.075	166.790 ± 8.121	167.110 ± 5.550	0.198	0.844	−0.040
Weight^a (kg)	74.202 ± 10.007	72.341 ± 9.537	−0.979	0.330	0.186	72.829 ± 9.355	72.408 ± 8.3633	−0.207	0.837	0.042
BMI^a (kg/m²)	26.721 ± 3.065	25.853 ± 2.913	−1.492	0.139	0.283	26.148 ± 2.555	25.936 ± 2.799	−0.344	0.732	0.069
Penicillin allergy	6 (14.3%)	3 (4.3%)	2.255	0.133	-	6 (15.8%)	3 (7.9%)	0.504	0.478	-
Hypertension	21 (50.0%)	36 (52.2%)	0.049	0.824	-	19 (50.0%)	21 (55.3%)	0.211	0.646	-
Coronary artery disease	7 (16.7%)	11 (15.9%)	0.010	0.920	-	5 (13.2%)	6 (15.8%)	0.106	0.744	-
DM	17 (40.5%)	17 (24.6%)	3.082	0.079	-	16 (42.1%)	9 (23.7%)	2.921	0.087	-
Smoking history	26 (61.9%)	31(44.9%)	3.012	0.083	-	24 (63.2%)	18 (47.4%)	1.916	0.166	-
COPD	2 (4.8%)	0	1.196	0.274	-	2 (5.3%)	0	0.514	0.474	-
Long-term use of corticosteroids	2 (4.8%)	2 (2.9%)	<0.001	>0.999	-	2 (5.3%)	1 (2.6%)	<0.001	>0.999	-

BMI, Body mass index; DM, Diabetes mellitus; COPD, Chronic obstructive pulmonary disease; PSM, Propensity score matching; SMD, Standardized mean difference.

^aMatched variables.

The overall standardized mean difference (SMD) decreased from 0.3269 before PSM to 0.0250 after PSM, indicating an improvement in overall balance of 92.35%. Notably, the SMD for BMI reduced from 0.2830 before PSM to 0.0690 after PSM, reflecting an enhancement in balance of 89.92%. The Jitter plot and Histogram (Figure 2) demonstrate that the propensity scores for both the case group and control group transitioned from an unbalanced state prior to matching to a state of equilibrium post-matching, illustrating effective matching results.

Figure 2.

Jitter plot (A) and Histogram (B) of propensity score distribution before and after PSM. The distribution of propensity scores between the case and control groups after PSM is comparable, suggesting effective matching

The Surgical-Related Indicators and Perioperative Data

The surgical-related indicators of the SSI group and the non-SSI group revealed no significant differences both prior to and following PSM (Table 2). In the univariate analysis of perioperative data, before PSM, preoperative albumin in the SSI group were significantly lower than those observed in the non-SSI group (P = 0.001). Furthermore, a smaller proportion of patients within the SSI group received prophylactic cephalosporin antibiotics compared to their counterparts in the non-SSI group (P = 0.032). However, after PSM, there were no significant differences between the 2 groups concerning preoperative albumin or types of prophylactic antibiotics administered within 48 h (Table 2).

Table 2.

Univariate Analysis of Surgical-Related Indicators and Perioperative Data for SSI After Cervical Laminoplasty

Variables	Before PSM				After PSM
Variables	Case group (SSI, n = 42)	Control group (non-SSI, n = 69)	T value/Z value/χ²	P value	Case group (SSI, n = 38)	Control group (non-SSI, n = 38)	T value/Z value/χ²	P value
Surgical-related indicators
Number of surgical segments			2.424	0.298			5.198	0.074
3	3 (7.1%)	4 (5.8%)			3 (7.9%)	1 (2.6%)
4	11 (26.2%)	28 (40.6%)			9 (23.7%)	18 (47.4%)
5	28 (66.7%)	37 (53.6%)			26 (68.4%)	19 (50.0%)
Surgical duration (hours)	2.290 (2.000, 2.808)	2.000 (1.920, 2.420)	−1.682	0.092	2.170 (1.980, 2.663)	2.080 (1.958, 2.500)	−0.360	0.718
Intraoperative blood loss (ml)	100.000 (100.000, 200.000)	120.000 (100.000, 200.000)	−0.435	0.664	100.00 (100.00, 200.00)	100.00 (95.00, 200.00)	−0.348	0.728
Perioperative data
ASA classification			1.072	0.585			1.985	0.371
I	2 (4.8%)	4 (5.8%)			2 (5.3%)	3 (7.9%)
II	35 (83.3%)	52 (75.4%)			32 (84.2%)	27 (71.1%)
III	5 (11.9%)	13 (18.8%)			4 (10.5%)	8 (21.1%)
Preoperative Hb (g/L)	138.500 (133.500, 150.000)	143.000 (132.000, 150.500)	−0.684	0.494	139.500 (134.000, 150.250)	139.000 (128.000, 151.750)	−0.140	0.888
Postoperative 1-day Hb (g/L)	128.140 ± 15.425	130.610 ± 13.496	0.884	0.379	129.000 ± 15.093	130.500 ± 14.057	0.448	0.655
ΔHb (g/L)	10.500 ± 7.794	10.620 ± 8.079	0.079	0.937	10.340 ± 8.072	9.870 ± 9.265	−0.238	0.813
Preoperative WBC (×10⁹/L)	6.350 (5.100, 7.525)	6.000 (4.950, 6.920)	−0.557	0.578	6.350 (5.100, 7.185)	6.100 (4.775, 7.400)	−0.094	0.925
Postoperative 1-day WBC (×10⁹/L)	11.713 ± 3.527	11.425 ± 2.838	−0.472	0.638	11.706 ± 3.526	11.070 ± 2.754	−0.876	0.384
ΔWBC (×10⁹/L)	5.375 (3.363, 6.440)	5.300 (3.450, 7.300)	−0.164	0.870	5.375 (3.513, 6.575)	4.650 (2.770, 7.225)	−0.753	0.451
Preoperative albumin (g/L)	40.686 ± 3.858	42.857 ± 3.081	3.268	0.001	40.903 ± 3.773	42.318 ± 3.225	1.758	0.083
Postoperative drainage volume (ml)
Surgery day	205.000 (170.000, 260.000)	220.000 (190.000, 280.000)	−0.821	0.412	205.000 (160.000, 252.500)	200.000 (186.250, 260.000)	−0.094	0.925
Post-op 1 day	145.000 (100.000, 181.250)	140.000 (100.000, 170.000)	−0.621	0.534	140.000 (97.500, 181.250)	140.000 (100.000, 170.000)	−0.505	0.613
Post-op 2 day	95.000 (68.750, 125.000)	80.000 (60.000, 120.000)	−0.513	0.608	100.000 (70.000, 127.500)	82.500 (58.750, 112.500)	−0.914	0.361
Post-op 3 day	70.000 (50.000, 90.000)	60.000 (30.000, 100.000)	−0.739	0.460	70.000 (50.000, 90.000)	60.000 (26.250, 100.000)	−0.705	0.481
Post-op total	425.000 (311.250, 592.500)	480.000 (359.500, 575.000)	−0.730	0.465	425.000 (311.250, 600.000)	465.000 (348.750, 552.500)	−0.125	0.901
Postoperative duration with drains (days)	3.000 (2.000, 4.000)	3.000(3.000, 4.000)	−1.160	0.246	3.000 (2.000, 4.000)	3.000 (3.000, 4.000)	−0.844	0.399
Allogeneic transfusion	3 (7.1%)	2 (2.9%)	0.329	0.566	3 (7.9%)	1 (2.6%)	0.264	0.607
Max temperature on post-op 1 day (°C)	37.142 ± 0.434	37.020 ± 0.438	−1.420	0.158	37.128 ± 0.431	36.984 ± 0.427	−1.458	0.149
Postoperative 48-h antibiotics			8.800	0.032			4.120	0.249
2nd-generation cephalosporins	8 (19.0%)	23 (33.3%)			5 (13.2%)	9 (23.7%)
3rd-generation cephalosporins	26 (61.9%)	40 (58.0%)			25 (65.8%)	26 (68.4%)
Clindamycin	7 (16.7%)	2 (2.9%)			7 (18.4)	2 (5.3%)
Quinolones	1 (2.4%)	4 (5.8%)			1 (2.6%)	1 (2.6%)

PSM, Propensity score matching; ASA classification, the American Society of Anesthesiologists’ (ASA) physical status classification; Hb, Hemoglobin; WBC, White blood cell.

ΔHb = preoperative Hb-postoperative 1-day Hb; ΔWBC = postoperative 1-day WBC-preoperative WBC.

Neck Parameters

The C5 neck thickness and average neck thickness in the case group demonstrated significant differences when compared to the control group prior to PSM. However, after matching age, gender, height, weight and particularly BMI, no significant differences were observed in C5, C6, C7 or average neck thickness between the 2 groups.

Regarding the muscle thickness, before PSM, the significant difference was only evidenced at C5 level between the 2 groups, while no significant differences were observed at C6, C7 levels or average muscle thickness. After PSM, there were no significant differences at C5, C6, C7 levels or average muscle thickness. Concerning the ratios of muscle thickness, a notable difference was identified at C7 level between the 2 groups before PSM; however, this difference disappeared after PSM. The C5, C6 and average muscle thickness ratio remained significantly lower in the case group compared to the control group both before and after PSM.

Prior to PSM, there was a notable difference in C7 subcutaneous fat thickness between 2 groups; however, this discrepancy disappeared following PSM. In contrast, C5 and C6 subcutaneous fat thickness, C7 subcutaneous fat thickness ratio, as well as average subcutaneous fat thickness and its ratio remained significantly higher in the case group compared to the control group both before and after PSM (Table 3).

Table 3.

Univariate Analysis of Neck Thickness, Muscle Thickness and Subcutaneous Fat Thickness for SSI After Cervical Laminoplasty

Variables	Before PSM				After PSM
Variables	Case group (SSI, n = 42)	Control group (non-SSI, n = 69)	T value/χ²	P value	Case group (SSI, n = 38)	Control group (non-SSI, n = 38)	T value/χ²	P value
Neck thickness (cm)
C5	6.560 ± 1.194	5.924 ± 0.982	−3.047	0.003	6.546 ± 1.211	6.123 ± 1.022	−1.645	0.104
C6	6.760 ± 1.244	6.365 ± 1.087	−1.759	0.081	6.731 ± 1.266	6.496 ± 1.083	−0.871	0.386
C7	6.992 ± 1.204	6.669 ± 1.041	−1.491	0.139	6.950±1.223	6.770±0.994	−0.705	0.483
Average	6.771±1.177	6.319±0.967	−2.194	0.030	6.742±1.194	6.463±0.983	−1.114	0.269
Muscle thickness (cm)
C5	3.415 ± 0.658	3.711 ± 0.538	2.580	0.011	3.443 ± 0.658	3.689 ± 0.533	1.786	0.078
C6	3.422 ± 0.622	3.613 ± 0.424	1.755	0.084	3.470 ± 0.626	3.639 ± 0.411	1.391	0.169
C7	3.689 ± 0.594	3.718 ± 0.376	0.283	0.778	3.732 ± 0.594	3.709 ± 0.379	−0.198	0.844
Average	3.509 ± 0.571	3.680 ± 0.382	1.729	0.089	3.548 ± 0.570	3.679 ± 0.362	1.192	0.238
Muscle thickness ratio (%)
C5	52.883 ± 9.779	63.897 ± 11.453	5.185	<0.01	53.427 ± 9.706	61.562 ±11.769	3.288	0.002
C6	51.491 ± 8.930	57.875 ± 8.965	3.644	<0.01	52.412 ± 8.698	57.019 ± 8.288	2.364	0.021
C7	53.398 ± 7.201	56.645 ± 7.654	2.216	0.029	54.330 ± 6.827	55.606 ± 7.598	0.770	0.444
Average	52.591 ± 7.575	59.472 ± 7.967	4.495	<0.01	53.390 ± 7.293	58.063 ± 8.034	2.655	0.010
Subcutaneous fat thickness (cm)
C5	2.706 ± 1.070	1.710 ± 0.940	−5.134	<0.01	2.653 ± 1.061	1.901±1.015	−3.155	0.002
C6	2.963 ± 1.006	2.241 ± 0.918	−3.877	<0.01	2.886 ± 0.992	2.350 ± 0.906	−2.459	0.016
C7	2.921 ± 0.915	2.375 ± 0.845	−3.200	0.002	2.833 ± 0.892	2.474 ± 0.859	−1.787	0.078
Average	2.863 ± 0.934	2.109 ± 0.792	−4.368	<0.01	2.790 ± 0.917	2.242 ± 0 .847	−2.709	0.008
Subcutaneous fat thickness ratio (%)
C5	40.303 ± 11.621	27.622 ± 12.179	−5.412	<0.01	39.562 ± 11.567	29.772 ± 12.705	−3.512	0.001
C6	42.804 ± 9.891	34.068 ± 9.603	−4.596	<0.01	41.827 ± 9.750	35.134 ± 9.207	−3.077	0.003
C7	41.005 ± 8.412	34.727 ± 8.029	−3.924	<0.01	39.984 ± 8.057	35.692 ± 8.312	−2.286	0.025
Average	41.371 ± 9.040	32.139 ± 8.474	−5.427	<0.01	40.458 ± 8.820	33.533 ± 8.859	−3.415	0.001

The LASSO Regression, Multivariate Logistic Regression and ROC Curve

The preoperative albumin, types of prophylactic antibiotics administered within 48 hours as well as neck thickness, muscle thickness, subcutaneous fat thickness and their ratio of the patients after PSM were incorporated into the LASSO regression analysis. The path diagram and cross-validation curve for LASSO regression are illustrated in Figure 3. LASSO regression identified 2 variables with non-zero regression coefficients: C5 subcutaneous fat thickness ratio and average subcutaneous fat thickness ratio. The regression coefficients are presented in Table 4.

Figure 3.

Variable selection by the LASSO model. (A) Path diagram for LASSO regression; (B) Cross-validation curve for LASSO regression: the lambda (λ) value corresponding to the minimum mean squared error (indicated by the rightmost dashed line) is 0.1081495

Table 4.

The LASSO Regression, Multivariate Logistic Regression and ROC Curve for SSI After Cervical Laminoplasty

Variable	LASSO regression	Multivariate logistic regression			ROC curve
Variable	Coefficient	P	OR	95%CI	AUC	95%CI	Optimal cutoff value (%)	Sensitivity	Specificity	Youden’s index
Average subcutaneous fat thickness ratio	0.005006851	0.555	1.037	0.918-1.171	0.715	0.598-0.833	36.053	0.789	0.632	0.421
C5 subcutaneous fat thickness ratio	0.018768583	0.002	1.068	1.025-1.112	0.723	0.608-0.838	37.136	0.711	0.737	0.447

The 2 variables selected by LASSO regression after matching patients were included in a multivariate logistic regression analysis. The results indicated that C5 subcutaneous fat thickness ratio (P = 0.002, OR = 1.068, 95% CI: 1.025-1.112) is an independent risk factor for SSI following cervical single-door laminoplasty (Table 4).

Utilizing ROC curve, we compared the AUC, specificity, and sensitivity of C5 subcutaneous fat thickness ratio against the average subcutaneous fat thickness ratio as follows: The AUC for predicting SSI using the C5 subcutaneous fat thickness ratio was calculated to be 0.723, with a sensitivity of 0.711 and specificity of 0.737; thus yielding a Youden index of 0.447 and an optimal cutoff value of 37.136%. In contrast, the AUC for predicting SSI based on the average subcutaneous fat thickness ratio was determined to be 0.715, with a sensitivity of 0.789 and specificity of 0.632; resulting in a Youden index of 0.421 and an optimal cutoff value of 36.053% (Table 4).

Based on the LASSO regression coefficients, multivariate logistic regression P-values, and AUC, C5 subcutaneous fat thickness ratio emerges as a superior predictor of SSI following cervical laminoplasty compared to average subcutaneous fat thickness ratio (Figure 4).

Figure 4.

The ROC curve of C5 and average subcutaneous fat thickness ratio

Outcomes of SSI Patients After Cervical Laminoplasty

The median postoperative length of stay for patients with SSI was significantly longer at 14.50 days (IQR: 7.75 to 21.25) compared to non-SSI patients, whose median postoperative length of stay was only 8.00 days (IQR: 7.00 to 12.00). The Z value was −3.929 and the P value was <0.001. Among the patients with SSI, wound culture results were positive in 12 cases: Staphylococcus aureus was identified in 3, coagulase-negative staphylococci in 3, Enterococcus faecalis in 2, as well as 1 case each of Corynebacterium striatum, Klebsiella pneumoniae, Streptococcus pyogenes, and Enterobacter aerogenes.

In a cohort of 42 patients following posterior single-door laminoplasty and subsequently developed SSI, 20 patients were diagnosed with SSI during their hospital stay, while 22 patients received an SSI diagnosis post-discharge and required readmission. The median time to readmission was 27.00 days (IQR: 19.00 to 39.00) after surgery. A total of 16 patients experienced wound dehiscence; notably, the risk of wound dehiscence among patients with SSI-related readmission (54.5%) was significantly higher than that observed in those diagnosed with SSI during hospitalization (20%). Furthermore, 28 patients underwent debridement; similarly, the probability of requiring debridement for SSI-related readmissions (95.5%) was significantly higher than that for SSI patients during hospitalization (35.0%) (Table 5).

Table 5.

Outcomes of SSI Patients After Cervical Laminoplasty

SSI	Positive wound culture	Hospital stay duration (days)	Wound dehiscence		Wound management measures
SSI	Positive wound culture	Hospital stay duration (days)	N (%)	P	Debridement	Regular dressing change	P
SSI during hospitalization (n = 20)	4 (20.0%)	21.00 (8.25, 26.50)	4 (20.0%)	0.047	7 (35.0%)	13 (65.0%)	<0.001
SSI-related readmission (n = 22)	8 (36.4%)	12.00 (7.00, 15.00)	12 (54.5%)	0.047	21 (95.5%)	1 (4.5%)	<0.001
Overall (n = 42)	12 (28.6%)	14.50 (7.75, 21.25)	16 (38.1%)	-	28 (66.7%)	14 (33.3%)	-

Discussion

C5 Subcutaneous Fat Thickness Compared to C6, C7, and Average Subcutaneous Fat Thickness

SSI following cervical spine surgery is a prevalent complication that can result in extended hospitalization, increased medical expenses,¹⁹ patient readmission,²⁰ and may even lead to suboptimal recovery of neurological function after surgery.²¹ Previous studies have identified several risk factors associated with postoperative SSI, including smoking,⁴ intraoperative cerebrospinal fluid leaks,²² poorly controlled DM,²³ and metabolic syndrome.²⁴

It is important to note that neck thickness and subcutaneous fat thickness have been identified as significant factors influencing SSI following cervical spine surgery, according to both researchers and practitioners. Mehta et al,¹⁰ Porche et al,¹¹ Donnally et al,¹² and Wang et al.² have all corroborated through retrospective studies that C5 subcutaneous fat thickness serves as a risk factor for SSI after posterior cervical surgery. However, considering that posterior cervical surgeries are frequently multilevel procedures, this raises the question of whether evaluating postoperative SSI risk solely based on neck and fat thickness at the C5 level is sufficiently robust.

The data concerning subcutaneous fat across multiple segments of the lumbar spine has been employed to predict SSI following lumbar surgery. Shaw et al²⁵ proposed measuring the ratio of subcutaneous fat thickness to paravertebral muscle thickness at the deepest segment of the subcutaneous fat layer, which they defined as the Subcutaneous Lumbar Spine Index (SLSI), which serves as a predictive factor for postoperative SSI in lumbar surgeries. Subsequently, Shen et al²⁶ refined the measurement method for SLSI by utilizing the ratio of average subcutaneous fat thickness to average paravertebral muscle thickness at the surgical segment; this modified SLSI exhibited excellent predictive efficacy. In a similar vein, when considering predictions for SSI after cervical spine surgery, is it more compelling to base indicators on multi-segmental neck and subcutaneous fat thickness rather than relying solely on measurements taken at the C5 level?

In this study, we identified the C5 subcutaneous fat thickness ratio and the average subcutaneous fat ratio at the C5, C6 and C7 levels as potential risk factors for SSI through LASSO regression analysis. Subsequently, through a multivariate logistic regression, we found that the C5 subcutaneous fat thickness ratio (P = 0.002, OR = 1.068, 95%CI:1.025-1.112), rather than the average subcutaneous fat ratio (P = 0.555, OR = 1.037, 95% CI: 0.918-1.171), is an independent risk factor for SSI.

From a surgical perspective, the C5 vertebra, a key vertebra influencing cervical lordosis,²⁷ often lies at the apex of the cervical lordosis and is frequently the central level exposed during posterior cervical laminoplasty. The surgical incision and maximal soft tissue retraction typically center around this region, making the local tissue environment at C5 most susceptible to the surgical trauma that predisposes to SSI. Although the fat distribution at C5-C7 is important, the C5 subcutaneous fat thickness ratio contains the most specific and non-redundant risk information related to the surgical center. While conceptually reasonable, averaging measurements from different levels may dilute this localized risk signal. Furthermore, considering that measuring single-level neck thickness and fat thickness at C5 is more straightforward compared to calculating an average from measurements taken at the C5-7 levels, we propose that utilizing the C5 neck parameters as predictive factors for postoperative SSI in cervical spine surgery is both reasonable and clinically applicable.

The Impact of BMI on SSI

Mehta et al,¹⁰ Porche et al,¹¹ Donnally et al,¹² and Wang et al² conducted studies that didn’t rigorously control for inter-group BMI, leading to the conclusion that subcutaneous fat thickness is a risk factor for SSI. In contrast, our study meticulously controlled for BMI and established that the ratio of subcutaneous fat thickness—not merely subcutaneous fat thickness itself—serves as an independent risk factor for SSI. This discrepancy may be attributed to the positive correlation between BMI and subcutaneous fat thickness in both the waist and neck regions.^10,26 Therefore, when investigating the impact of subcutaneous fat thickness on postoperative SSI, it is essential to control for patients’ BMI in order to mitigate its influence on outcomes. After controlling for BMI, simple measures of fat thickness become less reliable; rather, ratios of fat thickness present a more compelling argument.

Previous studies have demonstrated that a high BMI is associated with an increased risk of SSI following cervical spine surgery. Specifically, for each unit increase in BMI, the likelihood of developing SSI after posterior cervical instrumented fusion increases by a factor of 1.048.²⁸ In a retrospective study involving 5441 patients, Sebastian et al²⁹ identified BMI >35 kg/m² (P = 0.003, OR = 1.780) as an independent risk factor for postoperative SSI following posterior cervical surgery. Furthermore, elevated BMI may indirectly contribute to SSI by prolonging surgical duration and increasing the risk of hematoma formation.³⁰

This study effectively controlled for the differences in BMI between the case group and the control group through PSM. In this research, the SMD of BMI prior to PSM was 0.2830, which decreased to 0.0690 following PSM. The significant reduction in the difference in BMI between these 2 groups post-matching indicates an improvement of 89.92% in balance concerning BMI. Furthermore, after PSM, the case group comprised 38 patients, whose sample size is notably larger compared to those reported by Mehta et al¹⁰ (22 SSI patients), Porche et al¹¹ (20 SSI patients), Donnally et al¹² (20 SSI patients), and Wang et al² (20 SSI patients).

Although the AUC of 0.723 for the C5 subcutaneous fat thickness ratio indicates a moderate level of predictive performance, its clinical value lies in providing specific information beyond traditional systemic indicators such as BMI and preoperative albumin. By strictly controlling for BMI through PSM, this study found that the C5 subcutaneous fat thickness ratio remains an independent risk factor for SSI, suggesting that local fat distribution is a pathogenic element independent of overall obesity. Compared to albumin, which requires laboratory testing and showed no significant difference after matching, the C5 subcutaneous fat thickness ratio can be obtained non-invasively from preoperative MRI already planned for surgery, offering surgeons immediate anatomical risk assessment, especially during the initial outpatient visit when patients often lack laboratory test results. Furthermore, compared to absolute neck thickness, the subcutaneous fat thickness ratio normalizes for individual variations in body size (such as neck circumference) and overall obesity (BMI), more directly reflecting the local fat distribution characteristics in the patient’s neck. This assists surgeons in preoperatively identifying a subset of patients who, despite having a normal BMI, exhibit a higher local fat proportion in the neck and thus may be at increased risk for SSI. Therefore, we propose that the C5 subcutaneous fat thickness ratio should be regarded as a valuable supplement to—rather than a replacement for—existing risk assessment models (including BMI, albumin, etc.), particularly aiding in the identification of patients with normal BMI but high local anatomical risk.

The Impact of Muscle Thickness and Subcutaneous Fat Thickness on SSI

In this study, we found that neck muscle thickness and its ratio were negatively correlated with SSI, while subcutaneous fat thickness and its ratio exhibited a positive correlation and C5 subcutaneous fat thickness ratio was identified as an independent risk factor for SSI. The increased susceptibility of the cervical subcutaneous fat layer to SSI can be attributed to a confluence of anatomical, physiological, and surgical factors. (1) Poor vascularity and perfusion. Adipose tissue is inherently hypovascular compared to muscle, leading to poor perfusion which compromises local oxygen delivery, immune cell trafficking, and the effective concentration of systemic prophylactic antibiotics at the wound site.^11,31,32 (2) Increased skin tension and dead space formation. A thicker subcutaneous fat layer necessitates more extensive dissection and retraction to achieve adequate surgical exposure to the deeper cervical spine. This often results in larger tissue flaps and a more pronounced separation between the skin and the underlying fascia. Upon closure, the overlying skin edges are under greater tension to approximate, which can further compromise microcirculation at the incision line. More critically, the potential space created within the avascular or poorly vascularized fat layer acts as a dead space, prone to accumulating serous fluid, blood (hematoma), or liquefied necrotic fat.^11,33,34 This collection serves as an ideal culture medium for bacteria, significantly elevating infection risk. The placement of subcutaneous drains aims to mitigate this specific risk by evacuating these fluids.³⁵ (3) Prolonged retraction and tissue trauma. The need to retract thicker subcutaneous and muscular layers for a prolonged duration to maintain a deep surgical field may lead to direct mechanical trauma and ischemia-reperfusion injury to the fat and overlying skin edges.³⁰ Extended retraction time is often correlated with longer operative duration, which itself is a documented risk factor for SSI.

Therefore, a high C5 subcutaneous fat thickness ratio is not merely an anatomic measurement; it encapsulates a high-risk wound environment characterized by poor vascularity, increased closure tension, and greater surgical trauma. Our finding that this ratio remains an independent predictor even after controlling for BMI underscores its direct mechanistic relevance to SSI pathogenesis in cervical laminoplasty.

The Impact of Preoperative Serum Albumin and Prophylactic Antibiotic on SSI

Previous studies have indicated that preoperative albumin <3.5 g/dL, defined as hypoalbuminemia, is an independent risk factor for SSI following spinal surgery.^36,37 Chaker et al³⁸ further noted that not only does an albumin level <3.5 g/dL increase the risk of SSI, but levels <4.0 g/dL also significantly elevate this risk. In our study, prior to PSM, the preoperative albumin in the SSI group were significantly lower than those in the non-SSI group. However, after controlling for confounding factors such as age, gender, height, weight and BMI, no significant difference in preoperative albumin was observed between the 2 groups. Similarly, before PSM, a significantly higher proportion of patients in the non-SSI group received prophylactic cephalosporins compared to SSI group; however, this difference disappeared post-PSM. These findings suggest that while preoperative albumin and prophylactic use of non-cephalosporin antibiotics may contribute to an increased incidence of SSI to some extent following cervical spine surgery, they are not decisive factors.

The Impact of DM and Smoking History on SSI

Both DM and smoking history are well-established, strong systemic risk factors for SSI following spine surgery, including cervical procedures.^4,8 DM, particularly when poorly controlled, can impair immune function, microvascular circulation, and tissue healing, creating a permissive environment for infection.²³ Similarly, smoking induces vasoconstriction, reduces tissue oxygenation, and hampers neutrophil activity, all of which compromise the surgical wound’s ability to resist infection.⁴

Interestingly, in our matched cohort where baseline characteristics like BMI were balanced, neither DM (P = 0.087) nor smoking history (P = 0.166) showed a statistically significant difference between the case group and the control group. This finding does not negate the established importance of DM and smoking as general risk factors. Rather, it may be interpreted in several ways within the scope of our study: First, our PSM design, which actively controlled for BMI—a factor often correlated with metabolic health—might have partially accounted for the risk profile associated with DM. Second, our study focused on a homogenous population undergoing posterior single-door laminoplasty, and the sample size, though larger than previous similar anatomical studies, may still limit the power to detect the effect of these systemic factors compared to large national database studies. Most importantly, this result underscores the primary aim of our study: to isolate and identify local anatomical risk factors that contribute to SSI risk independent of and potentially concurrent with these powerful systemic comorbidities. The persistence of the C5 subcutaneous fat thickness ratio as an independent predictor even after accounting for DM and smoking highlights its unique and complementary role in risk stratification. In clinical practice, optimizing glycemic control and encouraging smoking cessation remain important. However, for patients undergoing cervical laminoplasty, assessing the C5 subcutaneous fat thickness ratio on preoperative MRI may provide an additional, immediate anatomical risk assessment that is particularly relevant to the local surgical site environment.

Given this complementary role, the C5 subcutaneous fat thickness ratio provides a practical, imaging-based tool for pre-operative SSI risk stratification. For patients exceeding the 37.136% cutoff, we propose considering them at elevated “anatomical risk”. This finding should be integrated with traditional systemic risk factors (eg, DM, smoking history). In such cases, a bundle of enhanced perioperative measures may be justified, including stringent glycemic control, weight-based antibiotic dosing, and meticulous surgical handling of the subcutaneous layer to reduce dead space. Postoperatively, closer wound surveillance and patient education on infection signs are advisable. This ratio aids in personalizing risk communication and focusing preventive resources, particularly for patients with unremarkable systemic profiles but high local risk.

Limitations

The present study has several limitations. Firstly, while our case group (SSI group) increased to 42 patients compared to previous studies—Mehta et al¹⁰ (22 SSI patients), Porche et al¹¹ (20 SSI patients), Donnally et al¹² (20 SSI patients), and Wang et al² (20 SSI patients)—only 38 matched cases remained after PSM; the stability of variable selection via LASSO regression may be limited by the sample size in our matched cohort. To address this concern and bolster the robustness of our findings, the variables identified by LASSO were subsequently validated using confirmatory multivariate logistic regression. Secondly, in our study, no patient received the topical application of vancomycin powder before or during wound closure; therefore, we did not investigate the impact of local vancomycin powder use on SSI following posterior cervical surgery. In the future, we plan to conduct a prospective study to further explore its effect on SSI. Finally, our case-control design utilized temporal proximity (same day or adjacent days) for control selection. While this approach helps control for non-patient, time-related confounders, it carries inherent limitations. If SSI cases were clustered in time due to unmeasured factors (eg, a specific surgical team or instrument batch), this could introduce selection bias. Furthermore, the potential overlap of controls between multiple case patients introduces non-independence in the dataset, which may affect the precision of estimates and the generalizability of findings. Although we employed PSM to balance measured patient characteristics between groups, this design choice remains a notable limitation of our retrospective study.

Conclusions

The C5 subcutaneous fat thickness ratio serves as an independent risk factor for surgical site infection (SSI) following cervical laminoplasty. This study demonstrates that, compared to the average subcutaneous fat thickness ratio, the C5 subcutaneous fat thickness ratio exhibits a slight advantage in predictive performance and offers a more convenient measurement method in clinical practice. Furthermore, our findings provide a foundational basis for clinical applications employing preoperative magnetic resonance imaging (MRI) based on indicators at the C5 level to forecast postoperative SSI in cervical spine surgery.

Footnotes

Acknowledgments

We thank all members of the Department of Spinal Surgery, Peking University People’s Hospital, Beijing, China, for their assistance of this research.

ORCID iDs

Yonghao Wu

Shuai Xu

Hongzhen Li

Jiahao Zhang

Kaifeng Wang

Consent to Participate

Informed consent was exempt from the study because it was a retrospective study.

Author Contributions

Conceptualization, Y.W. and K.W.; Formal analysis, C.Z. and Y.L.; Funding acquisition, K.W., and H.Liu; Investigation, Y.W., S.X. and H.Z.; Methodology, K.W.; Project ad-ministration, J.Zhang.; Software, J.Zhu. and H. Li; Supervision, C.Z., H.L. and K.W.; Visualization, S.X.; Writing – original draft, Y.W.; Writing – review & editing, Y.W. and K.W.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by National Key Research and Development Program of China (2022YFB4703004, 2022YFB4703005), and Beijing Natural Science Foundation-Haidian Original Innovation Joint Fund: Research on Intelligent Micro Nano Robot System (grant number: L252173).

Declaration of Conflicting Interests

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

Data Availability Statement

The datasets used and analyzed during the current study are available from the corresponding author on reasonable request. The data are not publicly available because of privacy or ethical restrictions.*

Clinical Trials Registration

Not applicable. Because this study is a retrospective observational study.

Institutional Review Board Statement

The Medical Ethics Committee of Peking University People’s Hospital approved the study protocol, and the approval number was 2024PHB157-001. The study was conducted in accordance with the ethical standards of the Helsinki Declaration.

References

Liang

Liu

Zhong

. Quantitatively biomechanical response analysis of posterior musculature reconstruction in cervical single-door laminoplasty. Comput Methods Progr Biomed. 2023;233:107479.

Wang

Chang

Suo

Huang

Liu

. Incidence and risk factors of surgical site infection following cervical laminoplasty: a retrospective clinical study. Int Wound J. 2023;21(2):e14450.

Ogihara

Yamazaki

Shiibashi

, et al. Risk factors for deep surgical site infection after posterior cervical spine surgery in adults: a multicentre observational cohort study. Sci Rep. 2021;11(1):7519.

Badiee

Chan

Rivera

, et al. Smoking is an independent risk factor for 90-day readmission and reoperation following posterior cervical decompression and fusion. Neurosurgery. 2021;88(6):1088-1094.

Takasawa

Iizuka

Mieda

, et al. Trends in cervical laminoplasty and 30-day postoperative complications: 10-year results from a retrospective, multi-institutional study of 1095 patients. Eur Spine J. 2023;32(10):3575-3582.

Sridharan

Malik

Kim

Khan

. Does increasing body mass index correlate with adverse outcomes following posterior cervical fusions? World Neurosurg. 2020;133:e789-e795.

Ban

Minei

Laronga

, et al. American college of surgeons and surgical infection society: surgical site infection guidelines, 2016 update. J Am Coll Surg. 2017;224(1):59-74.

Wang

Lin

Yao

Zhang

Gao

Feng

. Surgical site infection in spinal surgery: a bibliometric analysis. J Orthop Surg Res. 2023;18(1):337.

Cheng

, et al. Impact of frailty on short- and long-term outcomes following cervical laminoplasty for ossification of the posterior longitudinal ligament: a retrospective cohort study. Eur Spine J. 2025;34(8):3345-3355.

10.

Mehta

Babu

Sharma

, et al. Thickness of subcutaneous fat as a risk factor for infection in cervical spine fusion surgery. J Bone Joint Surg Am. 2013;95(4):323-328.

11.

Porche

Lockney

Gooldy

Kubilis

Murad

. Nuchal thickness and increased risk of surgical site infection in posterior cervical operations. Clin Neurol Neurosurg. 2021;205:106653.

12.

Donnally

Henstenburg

Pezzulo

, et al. Increased surgical site subcutaneous fat thickness is associated with infection after posterior cervical fusion. Surg Infect (Larchmt). 2022;23(4):364-371.

13.

Orabi

Naples

Brundidge

, et al. Postoperative respiratory failure after elective abdominal surgery: a case-control study. J Surg Res. 2022;274:160-168.

14.

Cuihong

Fan

Yingmo

. Lichtenstein repair for giant inguinoscrotal hernia: a retrospective case-control study. Hernia. 2024;29(1):48.

15.

Mangram

Horan

Pearson

Silver

Jarvis

. Guideline for prevention of surgical site infection, 1999. Centers for disease control and prevention (CDC) hospital infection control practices advisory committee. Am J Infect Control. 1999;27(2):97-132. quiz 133-134; discussion 196.

16.

BerríOS-Torres

Umscheid

Bratzler

, et al. Centers for disease control and prevention guideline for the prevention of surgical site infection, 2017. JAMA Surg. 2017;152(8):784-791.

17.

Horan

Andrus

Dudeck

. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36(5):309-332.

18.

Nolan

Martin

Thompson

Schroeder

Hanson

Warner

. Association between smoking status, preoperative exhaled carbon monoxide levels, and postoperative surgical site infection in patients undergoing elective surgery. JAMA Surg. 2017;152(5):476-483.

19.

Spatenkova

Bradac

Mareckova

, et al. Incidence of surgical site infections after cervical spine surgery: results of a single-center cohort study adhering to multimodal preventive wound control protocol. Eur J Orthop Surg Traumatol. 2023;33(5):1997-2004.

20.

Shlobin

Cloney

Hopkins

, et al. Surgical site infection, readmission, and reoperation after posterior long segment fusion. Spine. 2021;46(9):624-629.

21.

Ahorukomeye

Gwiazdon

Liu

Stephenson

Pittman

Gandhi

. Salmonella surgical site infection after cervical spine surgery: a case report. JBJS Case Connect 2025;15(2).

22.

Oshima

Nakamoto

Doi

, et al. Impact of incidental dural tears on postoperative outcomes in patients undergoing cervical spine surgery: a multicenter retrospective cohort study. Spine J. 2025;25(1):91-98.

23.

Roth

Chanbour

Gupta

, et al. Optimal hemoglobin A1C target in diabetics undergoing elective cervical spine surgery. Spine J. 2022;22(7):1149-1159.

24.

Shahbandi

Ghamasaee

Yassin Kassab

, et al. The impact of body mass index and metabolic syndrome on perioperative outcomes following cervical laminoplasty. Glob Spine J 2026;16(1):483-495.

25.

Shaw

Chen

Sheppard

, et al. Use of the subcutaneous lumbar spine (SLS) index as a predictor for surgical complications in lumbar spine surgery. Spine J. 2018;18(12):2181-2186.

26.

Shen

Gao

Zhang

, et al. Subcutaneous Lumbar Spine Index (SLSI) as a risk factor for surgical site infection after lumbar fusion surgery: a retrospective matched case-control study. Glob Spine J. 2024;14(5):1498-1503.

27.

Lee

Mcgregor

Bull

Wragg

. Dynamic response of the cervical spine to posteroanterior mobilisation. Clin Biomech. 2005;20(2):228-231.

28.

Cheng

Cizik

Dagal

AHC

, et al. Body mass index and the risk of deep surgical site infection following posterior cervical instrumented fusion. Spine J. 2019;19(4):602-609.

29.

Sebastian

Huddleston

3rd Kakar

Habermann

Wagie

Nassr

. Risk factors for surgical site infection after posterior cervical spine surgery: an analysis of 5,441 patients from the ACS NSQIP 2005-2012. Spine J. 2016;16(4):504-509.

30.

Miao

Liang

Sun

. Treatment of hematomas after anterior cervical spine surgery: a retrospective study of 15 cases. Neurochirurgie. 2018;64(3):166-170.

31.

Edmiston

Krepel

Kelly

, et al.

Perioperative antibiotic prophylaxis in the gastric bypass patient: do we achieve therapeutic levels?

Surgery. 2004;136(4):738-747.

32.

Zhang

Chen

Zhu

, et al. Construct validation of machine learning for accurately predicting the risk of postoperative surgical site infection following spine surgery. J Hosp Infect. 2024;146:232-241.

33.

Vaidya

Carp

Bartol

Ouellette

Lee

Sethi

. Lumbar spine fusion in obese and morbidly obese patients. Spine. 2009;34(5):495-500.

34.

Abdallah

Jadaan

Mccabe

. Body mass index and risk of surgical site infection following spine surgery: a meta-analysis. Eur Spine J. 2013;22(12):2800-2809.

35.

Sucato

Richards

. Risk factors for the development of delayed infections following posterior spinal fusion and instrumentation in adolescent idiopathic scoliosis patients. Spine. 2007;32(20):2272-2277.

36.

Konishi

Sano

Kawano

, et al. Factors related to surgical site infection in spinal instrumentation surgery: a retrospective study in Japan. Asian Spine J. 2024;18(6):822-828.

37.

Ramesh

Parel

Cui

, et al. Data-driven preoperative albumin level predicts increased risk of 30-day surgical site infection following anterior cervical discectomy and fusion. Spine Surg Relat Res. 2025;9(4):392-397.

38.

Chaker

Rademacher

Easton

, et al. The impact of serum albumin levels on postoperative complications in lumbar and cervical spine surgery: an analysis of the Michigan spine surgery improvement collaborative registry. J Neurosurg Spine. 2024;41(6):792-802.