Risk stratification of surgical site infection in high-risk spine surgery under prophylactic ciNPWT

Abstract

Background

This retrospective observational single-center cohort study evaluated real-world outcomes of prophylactic closed-incision negative pressure wound therapy (ciNPWT) in high-risk spine surgery patients and explored the distribution of surgical site infection (SSI) across cumulative risk burden (RiskN) strata within this selectively treated cohort.

Methods

We reviewed consecutive spine surgery patients who received prophylactic single-use PICO 7 (Smith & Nephew) ciNPWT immediately after wound closure between April 2023 and October 2025. High-risk status was defined according to Japanese reimbursable criteria. Cumulative risk burden was quantified as RiskN (1, 2, or ≥ 3). SSI was defined based on CDC criteria and reported with Wilson 95% confidence intervals (CIs).

Results

A total of 104 patients were analyzed (mean age 70.2 ± 12.7 years; BMI 24.1 ± 4.5 kg/m²). The mean RiskN was 1.50 ± 0.78 (RiskN 1: 68 patients; RiskN 2: 22; RiskN 3–4: 14). The most frequent risk factor was hypoalbuminemia (57/104, 54.8%). Overall SSI occurred in 4 of 104 patients (3.8%; 95% CI, 1.5–9.5%). SSI rates were 2.9% (2/68) in RiskN 1, 0% (0/22) in RiskN 2, and 14.3% (2/14) in RiskN ≥ 3. Surgical procedures were mainly lumbar posterior (56/104, 53.8%) and thoracolumbar posterior (21/104, 20.2%); SSIs occurred after thoracolumbar posterior surgery (2/21), cervical posterior surgery (1/14), and lumbar posterior surgery (1/56). The most commonly used dressing size was 15 × 15 cm (51/104, 49.0%).

Conclusions

In this selectively treated, reimbursement-linked high-risk spine surgery cohort managed with prophylactic ciNPWT, SSI was uncommon overall but remained observed in patients with higher cumulative risk burden. RiskN was used to contextualize residual SSI risk within this pragmatically treated population. The observed SSI rate should be interpreted as a residual (“breakthrough”) infection rate within a high-risk population. As a descriptive study without a control group, this study does not directly demonstrate the treatment effect of ciNPWT. Further comparative studies are needed to determine its incremental benefit, cost-effectiveness, and appropriate indications.

Introduction

Surgical site infection (SSI) after spine surgery remains one of the most consequential complications, even when the absolute incidence is relatively low. SSI can necessitate surgical debridement and prolonged antimicrobial therapy, delay rehabilitation, and compromise clinical outcomes after fusion procedures [1]. In addition to its clinical impact, SSI imposes a substantial economic burden; a systematic review of spine surgery complications identified infection as the most commonly reported complication and reported high direct treatment costs [2]. Contemporary meta-analyses suggest that the pooled incidence of SSI following spine surgery is approximately 3% [3]. However, SSI risk varies markedly by procedure type and complexity, and the incidence after instrumented spinal operations has been reported to range from 2% to 20% [4]. Accordingly, SSI prevention in spine surgery requires a bundled approach spanning pre-, intra-, and postoperative phases, while postoperative wound management remains a practical target for risk-adapted escalation. In a spine-specific review, negative pressure wound therapy (NPWT) is described as a promising strategy not only for SSI management but also for prophylaxis in high-risk populations, although further spine-focused evidence is still needed [5]. Collectively, these considerations underscore SSI prevention as a central priority in modern spine care.

Current prevention strategies extend beyond standard perioperative measures (e.g., appropriate antimicrobial prophylaxis and meticulous soft-tissue handling) toward risk-adapted interventions. The CDC guideline for SSI prevention and the WHO global guidelines emphasize implementation of evidence-based measures within comprehensive quality-improvement programs, and both include recommendations across the pre-, intra-, and postoperative periods [6, 7]. In high-risk surgical wounds, prophylactic closed-incision negative pressure wound therapy (ciNPWT) has gained attention as an adjunct strategy intended to reduce incisional complications, including SSI.

The biologic rationale for negative pressure wound therapy is supported by foundational experimental and clinical work demonstrating that controlled subatmospheric pressure can improve the wound microenvironment and facilitate healing [8, 9]. Applied to closed incisions, ciNPWT is proposed to stabilize the incision, reduce lateral tension, manage exudate, and limit edema—mechanisms that may be particularly relevant in patients with impaired wound healing capacity.

In spine surgery, growing evidence suggests a potential benefit of ciNPWT for SSI prevention, although reported effects vary by population and procedure. Recent meta-analyses focusing on spinal fusion/spine surgery have suggested that prophylactic NPWT/ciNPWT can reduce SSI risk, while effects on other wound outcomes (e.g., dehiscence, readmission, reoperation) have been less consistent [10, 11]. Cohort data in complex spine surgery, including long-segment thoracolumbar fusion, have also reported lower rates of wound infection and dehiscence with incisional NPWT [12]. Nevertheless, the spine-specific literature remains heterogeneous with respect to patient selection, device type, surgical approach, and outcome definitions, and real-world data reflecting pragmatic “high-risk–only” implementation are still limited.

In Japan, reimbursement conditions for ciNPWT in closed incisions were revised in association with the national fee schedule revision, facilitating broader implementation within defined indications. Against this background, we report our initial institutional experience using a single-use ciNPWT system in reimbursable high-risk spine surgery patients, representing reimbursement-linked, high-risk-only implementation. The objective of this study was to describe the patient characteristics and surgical features under this real-world implementation strategy and to report SSI occurrence. In addition, we explored whether RiskN, defined as the simple sum of reimbursable risk factors, could serve as an indicator to capture differences in residual SSI risk within this high-risk cohort.

Methods

Patient selection

We reviewed all patients who underwent spine surgery at our institution during the study period and received prophylactic ciNPWT using a single-use PICO 7 system (Smith & Nephew, London, UK) immediately after wound closure [13]. Inclusion criteria were: (1) spine surgery performed at our institution and (2) prophylactic ciNPWT (PICO 7) applied to the closed incision. We excluded cases with missing documentation of the primary outcome (SSI status).

Indications for ciNPWT (high-risk criteria)

ciNPWT was applied selectively to patients deemed at high risk for SSI/wound complications according to institutional practice consistent with Japanese reimbursement criteria for single-use ciNPWT on closed incisions (PICO 7) [14]. In Japan, reimbursement is applicable only when both conditions are met: (1) the patient is managed under designated “special inpatient care” fee categories (e.g., ICU- or stroke care unit–related inpatient management fees), and (2) ciNPWT is used to reduce the risk of postoperative incisional SSI in patients meeting at least one of the predefined high-risk conditions (obesity [BMI ≥ 30 kg/m²], poorly controlled diabetes [HbA1c ≥ 7.0% NGSP, equivalent to ≥ 6.6% JDS], ongoing systemic steroid therapy, chronic maintenance dialysis, immunodeficiency state, malnutrition/undernutrition, skin disease or impaired cutaneous perfusion associated with delayed wound healing, and revision surgery at the same anatomical site).

To enhance reproducibility and reflect real-world reimbursement-linked risk stratification, “malnutrition/undernutrition” was operationally captured as hypoalbuminemia, defined as a preoperative serum albumin level < 4.1 g/dL (the lower limit of our institutional laboratory reference range [4.1–5.1 g/dL]), measured before surgery as part of routine preoperative assessment. Because this threshold corresponds to the lower limit of our institutional laboratory reference range, it may identify a relatively broad subgroup and influence the overall distribution of RiskN. An immunodeficiency state was defined operationally as patients with malignancy who were under ongoing clinical management at the time of surgery (e.g., active treatment and/or regular follow-up), consistent with our institutional risk screening practice. Skin disease or impaired cutaneous perfusion was defined as active inflammatory skin disease involving or adjacent to the planned incision area and/or locally compromised skin integrity (e.g., fragile skin) or prior local irradiation to the operative field.

When PICO 7 was used outside these reimbursement conditions, device-related costs were considered bundled within the procedure fee, according to the reimbursement notice.

For each patient, the number of risk factors (RiskN) was calculated as the count of applicable criteria. RiskN was used for exploratory stratified analyses within this high-risk cohort. RiskN represents a pragmatic, reimbursement-linked composite indicator rather than a validated risk prediction model.

Surgical approach categories

Surgical approach was categorized using the following institutional coding scheme:

thoracolumbar posterior, cervical posterior, thoracic posterior, lumbar posterior, cervical anterior, and combined lumbar anterior–posterior approach.

Perioperative management and wound closure

All patients received standard perioperative infection prophylaxis. Cefazolin was administered intravenously 30 min prior to incision, re-dosed every 3 h intraoperatively, and continued postoperatively at 12-hour intervals for up to 24 h, unless contraindicated or modified by clinical judgment [15]. Skin preparation was performed using either chlorhexidine gluconate in alcohol or povidone-iodine. Vancomycin powder was not used in this cohort. Intraoperatively, the wound was irrigated with either normal saline or diluted povidone-iodine solution; antibiotic irrigation was not performed. Fascial closure was performed using 1 − 0 polyglactin or barbed suture (STRATAFIX; Ethicon, Johnson and Johnson, Somerville, New Jersey) with interrupted or running technique based on surgeon preference. Subcutaneous tissue was closed with 3 − 0 polyglactin or barbed suture, and skin was closed with staples. A closed-suction drain was placed in all patients, and it was generally removed when the 24-hour drainage volume decreased to ≤ 100 mL.

ciNPWT application and postoperative management

After skin closure, the PICO 7 dressing was applied to cover the incision with an appropriate pad size, ensuring an airtight seal. The dressing was reinforced with fixation strips/occlusive tape as needed to minimize air leaks, particularly around anatomical contours and high-mobility areas. The therapy unit was then connected to the dressing via an extension tube and connector, and negative pressure therapy was initiated by pressing the start button (Fig. 1). Continuous negative pressure was applied at − 80 mmHg according to the manufacturer’s instructions [16]. PICO 7 consists of a compact therapy unit and a disposable dressing connected through tubing, enabling a simple setup at the bedside; multiple dressing sizes are available to accommodate various incision lengths and shapes. ciNPWT was continued for 7 days postoperatively, after which the dressing was replaced with standard wound dressings.

Fig. 1.

PICO7 ciNPWT application procedure (representative photographs). Representative intraoperative/postoperative photographs illustrating the PICO7 setup. After wound closure, the dressing is applied to the incision and reinforced with fixation strips/occlusive tape to prevent air leakage. The therapy unit is connected to the dressing via an extension tube and connector, and negative pressure therapy is initiated (A) Closed incision after skin closure (B) Dressing placement and reinforcement to maintain an airtight seal (C) Connection of the therapy unit and initiation of therapy (D) Example of the device in situ postoperatively

Data collection

Data were extracted from electronic medical records and operative/anesthesia records. Variables included:

• Demographics: age, sex, height, weight, BMI

• Surgical factors: approach category, operative time (min), estimated blood loss (mL)

• Laboratory values: white blood cell count and C-reactive protein (CRP) (recorded in the perioperative period; CRP at postoperative day 7 was used when available in routine care)

• Hospitalization: Hospital length of stay (days)

• ciNPWT parameters: dressing size (pad width and length; area calculated as width × length for descriptive analyses)

• Risk factors: individual high-risk criteria and RiskN

Outcome definition

The primary outcome was SSI. SSI was defined according to CDC criteria [17], classified as: superficial incisional SSI occurring within 30 days after surgery; and deep incisional SSI occurring within 30–90 days when an implant was involved [18].

Because this was a pragmatic early experience study focused on clinically meaningful events, SSI was primarily captured as infections requiring active treatment during the index hospitalization and/or leading to surgical management, based on clinical documentation and operative findings. All patients were routinely followed at our outpatient clinic until at least 90 days after surgery (follow-up rate, 100%). SSI events were identified during postoperative clinic visits and through electronic medical record review for wound-related symptoms/signs, additional antibiotic treatment, unplanned visits, readmissions, and/or reoperations occurring within 90 days postoperatively.

Statistical analysis

Patient characteristics and perioperative variables were summarized using mean ± standard deviation (SD) for continuous variables and counts (percentages) for categorical variables. SSI incidence was reported as proportions with 95% confidence intervals (CIs) (Wilson method). Exploratory analyses stratified patients by RiskN (1, 2, and ≥ 3). To enhance interpretability, graphical visualization of the RiskN–SSI relationship was performed using forest plots with Wilson confidence intervals, along with a cumulative clinical risk curve to visualize the relationship between cumulative risk burden and SSI occurrence. Between-group comparisons were performed using Fisher’s exact test for categorical variables and Student’s t-test or Mann–Whitney U test for continuous variables, as appropriate. Given the low number of SSI events, multivariable modeling was not prespecified; any regression-based analyses were considered exploratory only. A sensitivity analysis excluding the single pediatric patient (age 14 years) was also performed to assess the robustness of SSI incidence estimates and the RiskN-stratified pattern. Statistical significance was set at p < 0.05. All analyses were conducted using Python (version 3.14.3) with standard statistical libraries.

Results

A total of 104 patients who underwent spine surgery and received prophylactic ciNPWT using PICO 7 were analyzed. Baseline demographics and perioperative characteristics are summarized in Table 1. The mean age was 70.2 ± 12.7 years and the mean BMI was 24.1 ± 4.5 kg/m² (both n = 104). There were 56 men (53.8%) and 48 women (46.2%). The mean operative time was 170.8 ± 79.9 min, mean estimated blood loss was 200.6 ± 195.8 mL, and the mean length of hospital stay was 24.5 ± 16.1 days. Perioperative laboratory data were available in patients, with a mean WBC of 7056 ± 2170/µL and a mean CRP of 3.81 ± 3.90 mg/dL (Table 1).

Table 1.

Patient demographics and perioperative characteristics (n=104)

Variable	Overall (n=104)
Age, years	70.2 ± 12.7
Sex, male, n (%)	56 (53.8%)
Sex, female, n (%)	48 (46.2%)
Height, cm	157.9 ± 9.8
Body weight, kg	60.4 ± 13.8
BMI, kg/m²	24.1 ± 4.5
Operative time, min	170.8 ± 79.9
Estimated blood loss, mL	200.6 ± 195.8
WBC at POD7, /µL	7056 ± 2170
CRP at POD7, mg/dL	3.81 ± 3.90
Hospital length of stay, days	24.5 ± 16.1
HbA1c, %	6.25 ± 1.16
Serum albumin, g/dL	3.94 ± 0.47
PICO dressing area (width × length), cm²	254.6 ± 72.6
SSI, n (%)	4 (3.8%)

Data are presented for 104 consecutive spine surgery patients who received prophylactic closed-incision negative pressure wound therapy (ciNPWT) using PICO7. Continuous variables are reported as mean ± standard deviation (SD) and categorical variables as number (percentage). For variables with missing data, the number of available observations is shown as (n). PICO dressing area was calculated as pad width × length. SSI indicates surgical site infection as defined by CDC criteria within the surveillance period

Abbreviations: BMI Body mass index, WBC White blood cell count, CRP C-reactive protein, HbA1c Hemoglobin A1c, SSI Surgical site infection, ciNPWT Closed-incision negative pressure wound therapy, POD Postoperative day

High-risk indications for ciNPWT and the cumulative risk-factor burden are summarized in Table 2. The mean number of predefined risk factors (RiskN) was 1.50 ± 0.78 (n = 104). As shown in Fig. 2, RiskN was distributed as 1 in 68 patients (65.4%), 2 in 22 (21.2%), 3 in 12 (11.5%), and 4 in 2 (1.9%). Among individual risk factors, malnutrition (hypoalbuminemia; Alb < 4.1 g/dL) was the most common (57/104, 54.8%), followed by revision surgery at the same level (22/104, 21.2%), immunodeficiency (20/104, 19.2%), and poorly controlled diabetes (HbA1c ≥ 7.0% NGSP) (17/104, 16.3%) (Table 2). Skin disease/impaired perfusion was present in 15/104 (14.4%), systemic steroid use in 11/104 (10.6%), obesity (BMI ≥ 30 kg/m²) in 9/104 (8.7%), and maintenance dialysis in 5/104 (4.8%) (Table 2).

Table 2.

High-risk factors for ciNPWT and distribution of cumulative risk burden (RiskN) (n = 104)

High-risk factor	n (%)
Hypoalbuminemia (Alb <4.1 g/dL)	57 (54.8)
Revision surgery at the same level	22 (21.2)
Immunodeficiency	20 (19.2)
Poorly controlled diabetes (HbA1c ≥7.0% NGSP)	17 (16.3)
Skin disease / impaired perfusion	15 (14.4)
Systemic steroid therapy	11 (10.6)
Obesity (BMI ≥30 kg/m²)	9 (8.7)
Maintenance dialysis	5 (4.8)

RiskN	n (%)
RiskN, mean ± SD	1.50 ± 0.78
RiskN = 1	68 (65.4)
RiskN = 2	22 (21.2)
RiskN = 3	12 (11.5)
RiskN = 4	2 (1.9)

High-risk factors were predefined according to reimbursable indications in Japan and recorded as binary variables. RiskN represents the cumulative number of applicable high-risk factors per patient and was calculated as the sum of these binary indicators. Categorical variables are presented as number/total (percentage), and RiskN is additionally summarized as mean ± standard deviation (SD)

Fig. 2.

Distribution of RiskN in the ciNPWT cohort. Histogram showing the number of pre-defined surgical site infection (SSI) risk factors (RiskN) per case among patients who received closed-incision negative pressure wound therapy (ciNPWT). Numbers above bars indicate case counts

Sensitivity analysis (albumin-based definition): When malnutrition was alternatively defined as serum albumin < 3.5 g/dL, RiskN was redistributed as 0 in 19 patients, 1 in 66, 2 in 14, and ≥ 3 in 5, indicating that the operational albumin threshold materially influenced the distribution of the composite RiskN variable.

Overall, SSI occurred in 3.8% of patients (4/104; Wilson 95% CI, 1.5–9.5%). Clinical characteristics of the four SSI cases are summarized in Table 3. RiskN-stratified SSI rates were 2.9% (95% CI, 0.8–10.1%) for RiskN = 1 (2/68), 0% (95% CI, 0–14.9%) for RiskN = 2 (0/22), and 14.3% (95% CI, 4.0–39.9%) for RiskN ≥ 3 (2/14). When dichotomized as RiskN ≥ 3 versus RiskN ≤ 2, SSI occurred in 2/14 (14.3%) versus 2/90 (2.2%), respectively; however, given the very small number of events, this comparison should be interpreted cautiously and as descriptive only. Excluding the single pediatric SSI case did not materially change the overall SSI rate (2.9%, 3/103) or the high-risk pattern (RiskN ≥ 3, 14.3%, 2/14). The relationship between cumulative risk burden and SSI occurrence is illustrated in Fig. 3, which includes both stratified risk estimates (forest plot) and cumulative clinical risk visualization.

Table 3.

Characteristics of surgical site infection (SSI) cases in the prophylactic PICO7 ciNPWT cohort

Age	Sex	BMI	Approach	Op time (min)	EBL (mL)	RiskN	Risk factors (predefined)	PICO 7 dressing (cm)	LOS (days)
14	F	19.8	Thoracolumbar posterior	191	293	1	Skin	10×40	12
69	F	22.9	Thoracolumbar posterior	289	103	3	Steroid, Immunodef, Malnutrition	10×40	78
69	M	25.7	Lumbar  posterior	101	177	4	DM≥7.0,  Dialysis, Immunodef, Malnutrition	15×15	25
55	F	34.0	Cervical  posterior	292	490	1	Obesity	10×30	15

This table summarizes the four patients who developed SSI among 104 high-risk spine surgery cases managed with prophylactic closed-incision negative pressure wound therapy (ciNPWT) using a single-use PICO 7 system. Variables include demographics, surgical approach category, operative time, estimated blood loss, cumulative risk-factor burden (RiskN), predefined high-risk factors, dressing size, and length of stay. High-risk factors were defined according to institutional criteria aligned with reimbursable indications in Japan, and RiskN represents the number of applicable predefined risk factors per patient

Abbreviations: ciNPWT closed-incision negative pressure wound therapy, SSI Surgical site infection, EBL Estimated blood loss, LOS Length of stay, RiskN Number of predefined risk factors

Fig. 3.

Relationship between cumulative risk burden and SSI. A Forest plot showing SSI incidence across RiskN strata (B) Cumulative clinical risk curve illustrating the distribution of SSI occurrence across cumulative risk strata. These figures are intended for descriptive visualization only and should not be interpreted as evidence of statistical trends or inference SSI, surgical site infection

In the albumin-based sensitivity analysis (albumin < 3.5 g/dL), SSI rates were 0% for RiskN = 0 (0/19), 3.0% for RiskN = 1 (2/66), 7.1% for RiskN = 2 (1/14), and 20.0% for RiskN ≥ 3 (1/5), although the small number of events limits precision.

All surgeries were performed with spinal fusion instrumentation (instrumented surgery, 100%). Regarding surgical approach, lumbar posterior procedures accounted for the largest proportion (56/104, 53.8%), followed by thoracolumbar posterior (21/104, 20.2%), cervical posterior (14/104, 13.5%), and thoracic posterior (10/104, 9.6%). Cervical anterior surgery was performed in 2/104 (1.9%), and a combined lumbar anterior–posterior approach in 1/104 (1.0%). The approach distribution and SSI counts are summarized in Fig. 4. SSI events were observed after thoracolumbar posterior surgery (2/21, 9.5%), cervical posterior surgery (1/14, 7.1%), and lumbar posterior surgery (1/56, 1.8%), whereas no SSI events were recorded in the remaining approach categories.

Fig. 4.

Surgical approach and SSI counts. Stacked horizontal bar chart showing the distribution of cases across surgical approach categories (thoracolumbar posterior, cervical posterior, thoracic posterior, lumbar posterior, cervical anterior, and combined lumbar anterior–posterior). Blue bars indicate cases without SSI and orange bars indicate SSI events. “n=” denotes the total number of cases in each category

Finally, PICO 7 dressing size selection reflected incision coverage requirements. The most frequently used dressing was 15 × 15 cm (51/104, 49.0%), followed by 10 × 20 cm (28/104, 26.9%), 10 × 40 cm (18/104, 17.3%), 10 × 30 cm (6/104, 5.8%), and 20 × 20 cm (1/104, 1.0%). The mean dressing area (width × length) was 254.6 ± 72.6 cm² (n = 104) (Table 1). Dressing area by RiskN group is shown in Fig. 5 (RiskN = 1, n = 68; RiskN = 2, n = 22; RiskN ≥ 3, n = 14), with a median area of 225 cm² across groups and comparable mean areas (257.4, 260.2, and 232.1 cm², respectively).

Fig. 5.

PICO dressing area by RiskN group. Box-and-whisker plot showing the distribution of PICO7 ciNPWT dressing area across RiskN strata (1, 2, and ≥ 3). Dressing area was calculated as pad width × length (cm²). The box represents the interquartile range (IQR) with the median shown as a horizontal line; whiskers extend to 1.5× IQR, and points indicate outliers. The green triangle denotes the mean

Representative clinical photographs demonstrate typical postoperative incision coverage with 10 × 40 cm and 10 × 20 cm PICO 7 dressings immediately after surgery and the wound appearance after suture removal (Fig. 6).

Fig. 6.

Representative clinical photographs of incisions managed with PICO7 ciNPWT dressings of different sizes. Left panels: A patient treated with a 10 × 40 cm PICO7 dressing immediately after surgery (A) and the wound appearance after suture removal (B) Right panels: A patient treated with a 10 × 20 cm PICO7 dressing immediately after surgery C) and the wound appearance after suture removal (D)

Discussion

In this retrospective cohort of 104 high-risk spine surgery patients managed with prophylactic ciNPWT using PICO 7, this study provides real-world evidence from reimbursement-linked, high-risk-only implementation. Within this selectively treated cohort, simple RiskN stratification (a count of reimbursable risk factors) was used to contextualize residual SSI risk. Three observations are notable. First, the overall SSI rate was 3.8%, providing a pragmatic benchmark for real-world performance in a selectively treated, comorbidity-burdened population. Second, even within this high-risk cohort, SSI was numerically more frequent in patients with greater cumulative risk burden (2.9% in RiskN = 1 vs. 14.3% in RiskN ≥ 3); however, this pattern should be interpreted cautiously because confidence intervals were wide and the total number of SSI events was very small. Third, device utilization was feasible across a broad range of spine procedures and incision coverage requirements, with a limited set of dressing sizes accounting for most cases. Collectively, these findings describe the feasibility of ciNPWT use in reimbursable high-risk spine surgery patients and provide a pragmatic benchmark of residual SSI occurrence within this selectively treated population, rather than evidence supporting treatment effectiveness. This real-world SSI rate under selectively applied PICO 7 ciNPWT should be interpreted within the context of spine-specific perioperative prevention frameworks, where NPWT has been proposed as an adjunct for prevention in high-risk patients and for management of established SSI [5]. Given the very small number of events, a dose–response relationship cannot be assessed, and the findings should be interpreted strictly as descriptive observations. The figures are intended to visualize these descriptive patterns and do not represent statistical inference. Accordingly, the primary value of this study lies in describing the selective real-world implementation of ciNPWT in reimbursement-defined high-risk spine surgery patients, rather than in evaluating treatment effectiveness or elucidating underlying mechanisms.

SSI after spine surgery remains clinically and economically consequential, particularly in instrumented cases [1, 2]. Prior studies and meta-analyses have consistently identified diabetes/poor glycemic control, obesity, prolonged operative time, revision surgery, and increasing procedural complexity as important risk factors for SSI and wound complications [3, 19, 20], and these factors overlap substantially with the reimbursable high-risk criteria that guided ciNPWT use in our practice.

The pattern observed with RiskN stratification should be interpreted with caution; however, it may provide useful descriptive insights into the characteristics of a cohort in which ciNPWT was selectively applied to high-risk patients. In many institutions, “high risk” is often treated as a binary category, but our findings suggest that heterogeneity may still exist within this group. Patients with higher RiskN showed numerically higher SSI rates despite receiving ciNPWT, although this observation is based on very small numbers and should not be overinterpreted. This pattern is clinically intuitive and broadly consistent with prior literature showing that cumulative comorbidity burden and impaired healing capacity are associated with greater susceptibility to SSI [19, 20]. From an implementation standpoint, RiskN may provide a simple descriptive framework to organize cumulative risk burden within this selectively treated cohort, but should not be interpreted as a predictive or validated stratification tool. Any potential use for perioperative decision support remains hypothetical and requires validation in comparative studies. The descriptive pattern illustrated in Fig. 3 may suggest heterogeneity in residual SSI occurrence within this high-risk cohort, although no inferential conclusion can be drawn from these data.

When stratified by surgical approach, most procedures were posterior lumbar or thoracolumbar operations, and SSI events were observed in thoracolumbar posterior, cervical posterior, and lumbar posterior procedures. Although interpretation should be cautious given the small number of cases and events, this pattern is broadly consistent with previous reports indicating that posterior approaches and extensive soft-tissue exposure—particularly in deformity correction and long-segment constructs—are associated with a higher risk of wound complications [19, 20]. In this setting, a low-profile ciNPWT system may offer practical advantages. Dressing selection with PICO 7 was straightforward, and the selected pad sizes were able to cover the incisions, supporting the feasibility of routine use in high-risk patients. Single-use ciNPWT systems typically deliver lower continuous pressures (e.g., − 80 mmHg), which may still support exudate control and reduction of lateral tension at the incision. Recent meta-analyses in spine surgery have suggested that prophylactic NPWT/ciNPWT is associated with a lower SSI risk, although effects on other wound outcomes have been less consistent [11, 21]. Similarly, an RCT-only systematic review and meta-analysis across orthopedic and trauma surgery reported fewer SSIs with prophylactic NPWT compared with conventional dressings (pooled OR 0.64, 95% CI 0.52–0.78), along with fewer wound dehiscence events (pooled OR 0.39, 95% CI 0.26–0.60). Although procedure-specific evidence varied and the overall certainty was rated moderate, these randomized data provide complementary support for prophylactic NPWT as part of a risk-adapted SSI-prevention bundle. Notably, spine-specific RCT evidence is still sparse, with only one spine trial included [22]. Moreover, cohort data in complex spine surgery, including long-segment thoracolumbar fusion, have reported lower rates of postoperative wound infection and dehiscence with incisional NPWT [12]. Beyond prophylaxis, NPWT has also been used as an adjunct in the treatment of postoperative spinal wound infections. A systematic review summarized multiple series in which NPWT facilitated infection control and often enabled retention of instrumentation, and it concluded that NPWT may be a valuable adjuvant therapy for SSI after spine surgery, although higher-quality studies are needed and special situations such as concomitant CSF leak remain debated [23].

The present study should be interpreted in light of several limitations. First, this was a retrospective, single-center cohort in which ciNPWT was selectively applied to reimbursable high-risk patients, without a contemporaneous control group; therefore, no causal inference regarding the effectiveness of ciNPWT can be made. The observed SSI incidence of 3.8% should be interpreted as a residual (“breakthrough”) SSI rate within a selectively treated, comorbidity-burdened high-risk cohort, rather than as evidence of comparative benefit. Because the baseline risk in this population is expected to be higher than that of unselected spine surgery patients, direct comparison with historical or pooled SSI rates is not appropriate. In addition, a directly comparable historical cohort from our institution, with consistent outcome definitions, surveillance methods, and risk profiles, was not available for rigorous comparison; therefore, we elected not to include historical SSI rates to avoid potentially misleading contextual comparisons. Second, the overall number of SSI events was very low (n = 4), which substantially limits interpretability and precludes robust statistical inference or multivariable modeling. Therefore, the RiskN stratification should be considered exploratory, underpowered, and hypothesis-generating. In addition, RiskN is a simplified construct derived from the same reimbursable criteria that guided ciNPWT use in this cohort, and therefore may be subject to circularity. Furthermore, RiskN assigns equal weight to heterogeneous risk factors, which may not reflect their true relative contributions to SSI risk. The operational definition of malnutrition using albumin < 4.1 g/dL may also have broadened the prevalence of this component and influenced the overall distribution of RiskN. Accordingly, the construct validity of RiskN as a biological or statistical stratification measure is limited. Therefore, RiskN should be interpreted as an exploratory, workflow-oriented stratification tool rather than a quantitative predictive score. Finally, while SSI was defined using CDC-based timing windows, outcome capture in routine clinical datasets can be incomplete if post-discharge events are managed elsewhere; thus, SSI incidence may be underestimated. In addition, the cohort was clinically heterogeneous, comprising cervical, thoracic, thoracolumbar, and lumbar procedures with differing surgical approaches and levels of complexity. Because these procedure types likely differ in soft-tissue exposure, operative burden, and baseline SSI risk, the pooled estimates should be interpreted cautiously and primarily as descriptive observations within a real-world implementation cohort.

Future studies should include standardized post-discharge surveillance and ideally compare ciNPWT versus standard dressings using prospective or well-adjusted comparative designs (e.g., propensity-weighted analyses), focusing on clinically meaningful endpoints such as deep SSI, reoperation, readmission, and cost-effectiveness [24]. Despite these limitations, our findings contribute practical, real-world evidence regarding selective ciNPWT implementation in high-risk spine surgery patients under the Japanese clinical and reimbursement environment. This study also provides one of the few real-world descriptions of reimbursement-linked selective ciNPWT use in spine surgery, which may be relevant for healthcare systems where resource allocation is constrained. The overall SSI rate of 3.8% and the numerical differences across RiskN strata may provide descriptive insights to inform future hypothesis-driven studies and potential refinement of patient selection strategies. In particular, patients with higher RiskN may represent a subgroup of interest for future hypothesis-driven studies.

Conclusions

In this single-center retrospective cohort of 104 reimbursable high-risk spine surgery patients managed with prophylactic ciNPWT using PICO 7, the overall SSI rate was 3.8%. Within this selectively treated “high-risk–only” implementation strategy, SSI was uncommon overall but remained observed, with a numerically higher breakthrough SSI rate in patients with greater cumulative risk burden (RiskN). PICO 7 use was feasible across a range of spine procedures and incision sizes, with straightforward dressing selection in routine practice.

This study does not establish the effectiveness of ciNPWT; rather, it provides a descriptive benchmark of residual SSI occurrence in a selectively treated high-risk population. Its primary contribution is to describe real-world ciNPWT implementation under a reimbursement-linked selective-use framework. Further comparative studies are required to determine incremental benefit, cost-effectiveness, and optimal indications.

Authors’ contributions

AH was responsible for the organization and coordination of this study. AH, DS and NH were responsible for data analysis. AH, DS, NH, and MW developed the study design. All authors contributed to the writing of the final manuscript. All authors contributed to the management or administration of this study. All authors approved the manuscript for publication and agree to be accountable for all aspects of the work, ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

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

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Ethics Committee, Institutional Review Board, and the Profit Reciprocity Committee of Tokai University School of Medicine. Due to the retrospective nature of the study, the Research Ethics Committee of Tokai University School of Medicine waived the need to obtain informed consent　(Approval No. 25R-314). All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

Not applicable. This study used anonymized retrospective clinical data, and no individual patient information or identifiable images are included in this manuscript.

Competing interests

The authors declare no competing interests.