Sage Journals: Discover world-class research

Abstract

Objective: This study investigates risk factors for lower extremity deep vein thrombosis (LEDVT) after gynecologic laparoscopy and develops a predictive model. Methods: The Delphi method was used in expert opinion to identify risk factors for LEDVT. Patients undergoing gynecologic laparoscopic surgery from January 2018 to June 2023 were divided into LEDVT and non-LEDVT groups and matched as a ratio of 1:3. A nomogram model for LEDVT was established and compared to Caprini and G-Caprini scores. Results: A total of 191 patients with LEDVT and 573 matched patients without LEDVT were included. Multivariate logistic regression analysis showed that older age (OR = 1.146, P < 0.001), history of varicose veins (OR = 6.721, P = 0.011), gynecological malignancies (OR = 8.053, P < 0.001), longer operation time (OR = 1.010, P < 0.001), longer postoperative bed rest time (OR = 99.406, P < 0.001), without postoperative anticoagulation (OR = 0.282, P = 0.042), higher D-dimer (OR = 1.145, P = 0.015), lower direct bilirubin (OR = 0.715, P = 0.006) and lower hemoglobin (OR = 0.982, P = 0.016) are independent risk factors for LEDVT in patients undergoing gynecologic laparoscopic surgery. A nomogram model was established with an area under the receiver operating characteristic curve of 0.839, sensitivity of 78.0%, and specificity of 76.1%. Validation set showed that this predictive model had an ROC of 0.954, sensitivity of 92.6% and specificity of 96.7%. Conclusion: A nomogram model that includes age, history of varicose veins, gynecologic malignancies, operation time, postoperative bed rest time, postoperative anticoagulation, D-dimer, direct bilirubin, and hemoglobin could effectively predict the risk of LEDVT in patients undergoing gynecologic laparoscopic surgery. The current nomogram model potentially helps to reduce the incidence of LEDVT and improve the quality of life.

Keywords

gynecology laparoscopic surgery lower extremity deep vein thrombosis risk factors nomogram model

Introduction

Deep vein thrombosis (DVT) is the pathological formation of blood clots within deep veins, obstructing the lumen and impairing venous return. It represents one of the most severe complications following gynecologic surgery.¹ Postoperative physiological alterations, including shifts in hormone levels, increased blood viscosity, and dyslipidemia, predispose patients to DVT.² Furthermore, some patients develop pulmonary embolism (PE), a critical concern. One significant risk factor is compression of pelvic veins by large masses, compounded by vascular endothelial injury during surgical lymphadenectomy. These factors promote DVT formation, which may subsequently lead to PE and mortality.³ Studies indicate that PE secondary to lower extremity DVT (LEDVT) accounts for 40% of postoperative deaths in this population.⁴ While gynecologic laparoscopic surgery offers advantages over traditional laparotomy, such as reduced tissue trauma and faster recovery,⁵ LEDVT incidence remains substantial (9.2%-15.6%). Alarmingly, up to 46% of these patients develop PE.^1,6 LEDVT often progresses rapidly with minimal early symptoms, leading to underdiagnosis. Failure to detect and treat LEDVT promptly compromises postoperative recovery and poses life-threatening risks.³

Currently, two scoring models are recommended for preventing DVT and PE following gynecologic surgery. The first is the Caprini risk assessment model, which stratifies patients into low, intermediate, high, and very high-risk groups based on over 40 thrombosis risk factors. It then guides corresponding preventive interventions for each stratification. Subsequent clinical validation studies indicate that the Caprini model demonstrates reliable predictive value for high-risk patients with LEDVT in orthopedic, thoracic, and plastic surgery.⁷ However, the model's complexity and time-consuming nature due to its extensive risk factors limit widespread clinical adoption. The second model is the Gynecological Caprini (G-Caprini) score. Developed using data from Chinese gynecological surgeries, this model stratifies postoperative risk levels.⁸ Compared to the Caprini model, the G-Caprini is simpler and easier for healthcare staff to implement rapidly. Nevertheless, confirmatory studies validating its clinical efficacy remain lacking.⁹ The G-Caprini identifies six independent risk factors for LEDVT after gynecologic surgery: age ≥ 50 years, hypertension, varicose veins, operation time ≥ 3 h, postoperative bed rest ≥ 48 h, and laparotomy.⁴ A limitation arises when applying the G-Caprini model to laparoscopic surgery patients: classifying laparotomy as a risk factor inherently expands the scope of considered risk sources. This correspondingly reduces the predicted probability of LEDVT for laparoscopic patients, potentially leading to missed optimal prevention opportunities. Consequently, neither of these widely used LEDVT scoring models appears fully suitable for assessing and predicting LEDVT risk specifically in patients undergoing laparoscopic surgery.

In venous thromboembolism (VTE) risk assessment, most tools employ scoring models that coarsely stratify patients’ VTE probability. In contrast, the nomogram is a visual model that graphically represents, quantifies, and scores specific predictors based on multivariate analysis results. This model demonstrates higher clinical utility than traditional approaches.^10–14 By translating complex regression equations into intuitive graphs with point-based scoring systems, nomograms enable clinicians to efficiently calculate disease probability and predict patient prognosis.¹⁵ Consequently, nomogram models are gaining significant attention as superior alternatives to conventional VTE risk scoring tools.¹⁶

Therefore, this study integrates a systematic literature review with the Delphi expert consensus method to identify independent risk factors for LEDVT following gynecologic laparoscopic surgery. We subsequently construct a nomogram prediction model to develop a clinically adaptable LEDVT risk assessment tool with high sensitivity and specificity. This model will establish an evidence-based identification system for postoperative LEDVT incidence, enabling early preventive interventions to reduce thrombotic events and improve patient outcomes.

Methods

Expert Correspondence

Through systematic review of international clinical guidelines and expert consensus literature, 124 potential risk factors for LEDVT following gynecologic laparoscopic surgery were initially identified. These comprised: 25 demographic and clinical characteristics, 21 laparoscopic surgery-specific parameters, and 78 preoperative laboratory parameters (measured within five days prior to surgery). Following expert group discussions, two Delphi consensus rounds were conducted using structured questionnaires. Each questionnaire contained three sections: 1. Introduction: Study background, objectives, and completion guidelines. 2. Assessment Matrix: Proposed risk factors with expert rating fields. Importance was evaluated using 5-point Likert scales (1 = least important, 5 = critically important).¹⁷ 3. Open-Ended Feedback: Sections for proposing factor additions/deletions with justifications. Additionally, experts completed a demographic survey capturing age, professional experience, academic qualifications, clinical position, self-rated judgment basis, and familiarity with LEDVT risk indicators.

Using deliberative sampling, the research team selected nationally representative Delphi panelists based on three inclusion criteria: (1) minimum bachelor's degree with intermediate or higher professional title; (2) ≥ 10 years of specialized clinical experience in vascular medicine or gynecology; and (3) demonstrated research interest with voluntary participation. Experts completing the first Delphi round were invited to continue participation in the second round.

During February 2023, questionnaires were electronically distributed and collected via WeChat. Following analysis of first-round responses, the research team revised the questionnaire by adding, deleting, and modifying indicators according to expert feedback. The updated instrument was then distributed for the second Delphi round. The consultation process concluded after two rounds when expert opinions achieved response stability. Item deletion criteria required meeting either of the following: (1) mean Likert importance score <4.0, or (2) coefficient of variation (CV) ≥ 0.25.

Design of the Study

Based on Delphi consensus results, potential risk factors for LEDVT following gynecologic surgery were identified. This case-control study included patients undergoing gynecologic laparoscopic surgery at our institution from January 2018 to June 2023. Participants were stratified into:

LEDVT group: Patients developing postoperative LEDVT; Control group: Randomly selected non-LEDVT patients (1:3 ratio matched).

The nomogram developed in this study underwent external validation using a cohort of patients who underwent gynecologic laparoscopic surgery at our institution between October and December 2023. This validation cohort enabled comparative assessment of three LEDVT risk prediction models: the newly developed nomogram, the Caprini score, and the G-Caprini model.

Inclusion and Exclusion Criteria

Inclusion criteria: (1) Undergoing gynecologic laparoscopic surgery (including diagnostic/therapeutic procedures);(2) Availability of complete perioperative clinical documentation;(3) Absence of preoperative LEDVT;(4) No contraindications to anticoagulation therapy

Exclusion criteria:(1) Significant preexisting comorbidities contraindicating general anesthesia;(2) Preoperative use of hormonal medications or coagulation-altering agents;(3) Preoperative diagnosis of LEDVT or PE;(4) Pregnancy or active lactation.

Diagnostic Criteria

Lower extremity compression ultrasonography (CUS) serves as the primary noninvasive diagnostic modality for DVT. This technique comprehensively evaluates proximal veins (common femoral, femoral deep/superficial, and popliteal veins) and distal veins (anterior/posterior tibial, peroneal, soleal, and gastrocnemius veins). LEDVT diagnosis requires meeting any of the following criteria: (1) venous lumen dilation with loss of compressibility, (2) absence of flow signals, (3) intraluminal filling defects, or (4) absent flow augmentation during distal limb compression.^18,19 Preoperatively, quantitative D-dimer levels were measured. Patients with levels ≥0.5 mg/L underwent bilateral lower extremity Doppler ultrasonography to exclude DVT. Only those without ultrasonographic evidence of LEDVT were enrolled.

Statistical Analysis

Statistical analyses of expert consultation data were performed using SPSS 26.0. Continuous variables are presented as mean ± standard deviation (SD) with CV. Expert engagement was quantified through questionnaire response rate. The authority coefficient (Cr) was calculated as Cr = (Ca + Cs)/2, where Ca represents judgment basis and Cs denotes familiarity level. Consensus consistency was evaluated using Kendall's coefficient of concordance (W).

Univariate analysis employed binary logistic regression. Variables with P-values <0.05 were entered into multivariate logistic regression to identify independent risk factors for LEDVT. Using R software, we constructed a nomogram prediction model and evaluated its discriminative ability via the concordance index (C-index). Internal validation utilized 1000 bootstrap resamples to generate calibration curves and assess model fit. Predictive performance was comprehensively evaluated through: (1) receiver operating characteristic (ROC) curve analysis (calculating AUC), (2) decision curve analysis (DCA), and (3) computation of maximum Youden index, sensitivity, specificity, positive/negative predictive values, and Cohen's kappa. Statistical significance was defined as two-sided P < 0.05.

Results

Basic Information of Experts

The Delphi panelists (n = 18) demonstrated the following characteristics: Age: Range 30–62 years (mean 46.11 ± 7.53); Clinical experience: Range 13–39 years (mean 22.33 ± 9.80); Highest degree: Doctorate (n = 7, 38.9%), Master's (n = 8, 44.4%), Bachelor's (n = 3, 16.7%); Professional title: Chief physician (n = 9, 50.0%), Deputy chief physician (n = 8, 44.4%), Attending physician (n = 1, 5.6%); Administrative position: Department director (n = 8, 44.4%), Deputy director (n = 4, 22.2%).

Degree of Activity and Authority of Experts

In the first Delphi round, all 18 distributed questionnaires were returned (100% response rate). The second round similarly achieved a 100% response rate, with all 18 questionnaires returned by the same expert cohort. Expert motivation was high, reflected in response rates of 1.00 for both rounds. The expert authority coefficient (Cr), calculated as Cr = (Ca + Cs)/2 (where Ca represents judgment basis and Cs represents familiarity level), was 0.93 for both rounds (Ca = 0.98, Cs = 0.88). This high authority coefficient indicates strong consensus among expert opinions, enhancing the reliability of the study's conclusions.

The Degree of Coordination of Expert Opinions

Kendall's concordance coefficients for the two rounds were 0.780 (P < 0.001) and 0.574 (P < 0.001), respectively. This high authority coefficient and significant concordance indicate strong consensus among expert opinions, enhancing the reliability of the study's conclusions.

Final Results of Expert Correspondence

After two rounds of expert correspondence, the results of the mean value and coefficient of variation of the quantitative determination of the degree of concentration of expert opinions are shown in Table 1.

Table 1.

Means and Coefficients of Variation to Quantify the Degree of Concentration of Expert Opinions After Two Rounds of Expert Correspondence.

Characteristics	Significance (x ± s)	Coefficient of Variation
Age	4.56 ± 0.511	0.112
BMI	4.89 ± 0.323	0.066
History of hypertension	4.11 ± 0.676	0.165
History of diabetes	4.06 ± 0.725	0.179
History of pelvic and lower extremity surgery	4.50 ± 0.515	0.114
History of lower extremity varicose veins	5 ± 0	0
Operation time	4.28 ± 0.461	0.108
Intraoperative hemorrhage	4.17 ± 0.515	0.124
Intraoperative blood transfusion	4.44 ± 0.511	0.115
Gynecological malignancies	5.00 ± 0	0
Postoperative pain score >4	3.56 ± 0.616	0.173
Postoperative bed rest time	4.72 ± 0.575	0.122
Postoperative pneumatic pressure treatment	4.83 ± 0.55	0.106
Postoperative anticoagulation	5 ± 0	0
Platelet count	4.11 ± 0.676	0.165
International normalized ratio	4.78 ± 0.428	0.090
D-dimer	5 ± 0	0
Direct bilirubin	3.50 ± 0.707	0.202
Hemoglobin	4.11 ± 0.676	0.165

Abbreviations: BMI, body mass index.

Risk Factors for LEDVT by Binary Logistic Regression Models

Baseline characteristics of the two groups are compared in Table 2. Univariate analysis revealed statistically significant differences in: age, body mass index (BMI), hypertension history, lower extremity varicose veins history, operation time, intraoperative hemorrhage, gynecological malignancies, postoperative pain score >4, postoperative bed rest duration, postoperative anticoagulation, D-dimer levels, direct bilirubin levels, and hemoglobin levels. Multivariate analysis (Table 3) identified the following independent risk factors for postoperative LEDVT: Older age (adjusted OR = 1.146, 95% CI 1.097–1.198; P < 0.001); History of lower extremity varicose veins (adjusted OR = 6.721, 95% CI 1.553–29.079; P < 0.05); Gynecologic malignancies (adjusted OR = 8.053, 95% CI 2.899–22.374; P < 0.05); Longer operation time (adjusted OR = 1.010, 95% CI 1.006–1.014; P < 0.05); Prolonged postoperative bed rest (adjusted OR = 99.406, 95% CI 19.976–494.664; P < 0.05); Elevated D-dimer levels (adjusted OR = 1.145, 95% CI 1.027–1.276; P = 0.015) Conversely, protective factors included:Postoperative low-molecular-weight heparin (LMWH) anticoagulation (adjusted OR = 0.282, 95% CI 0.083–0.958; P = 0.042); Higher direct bilirubin levels (adjusted OR = 0.715, 95% CI 0.563–0.909; P = 0.006); Higher hemoglobin levels (adjusted OR = 0.982, 95% CI 0.967–0.997; P = 0.016).

Table 2.

Comparison of Included Patients Between the Two Groups.

Characteristics	Non-LEDVT Group	LEDVT Group	P
Number (%)	573(75)	191 (25)
Age, mean (SD), years	43.62(9.707)	51.82(9.874)	<0.001
BMI, mean (SD), kg/m²	23.25(3.523)	24.12(3.053)	0.001
History of hypertension, n (%)			0.001
Yes	71(62.3)	43(37.7)
No	502(77.2)	148(22.8)
History of diabetes, n (%)			0.128
Yes	33(66)	17(34)
No	540(75.6)	174(24.4)
History of pelvic and lower extremity surgery, n (%)			0.403
Yes	93(72.1)	36(27.9)
No	480(75.6)	155(24.4)
History of lower extremity varicose veins, n (%)			<0.001
Yes	11(25)	33(75)
No	562(78.1)	158(21.9)
Operation time, Median (25%, 75%), min	135(99,189.50)	180(130,280)	<0.001
Intraoperative hemorrhage, Median (25%, 75%), ml	30(20,50)	50(20,100)	<0.001
Intraoperative blood transfusion, Median, n (%)			0.002
Yes	18(52.9)	16(47.1)
No	555(76.0)	175(24.0)
Gynecological malignancies, n (%)			<0.001
Yes	97(53)	86(47)
No	476(81.9)	105(18.1)
Postoperative pain score>4, n (%)			0.040
Yes	3(37.5)	5(62.5)
No	570(75.4)	186(24.6)
Postoperative bed rest time, n (%)			<0.001
< 24h	352(86.1)	57(13.9)
24h-48h	195(71.2)	79(28.8)
> 48h	26(32.1)	55(67.9)
Postoperative pneumatic pressure treatment, n (%)			0.700
Yes	475(75.3)	156(24.7)
No	98(73.7)	35(26.3)
Postoperative anticoagulation, n (%)			0.001
Yes	549(76.2)	171(23.8)
No	24(54.5)	20(45.5)
Platelet count, Median (25%,75%), × 10⁹/L	270(227.5,320)	275(230,322)	0.598
International normalized ratio, Median (25%, 75%), %	0.95(0.91,0.99)	0.94(0.90,0.98)	0.035
D-dimer, Median (25%, 75%), mg/L FEU	0.23(0.15,0.38)	0.45(0.23,1.12)	<0.001
Direct bilirubin, Median (25%, 75%), umol/L	2.1(1.5,3.3)	1.9(1.4,2.8)	0.006
Hemoglobint, Median (25%, 75%), g/L	122(109,131)	118(95,129)	0.005

Abbreviations: BMI, body mass index.

Table 3.

Univariate and Multivariate Analysis of Risk Factors for LEDVT by Binary Logistic Regression Models.

Risk Factors	Univariate Analysis		Multivariate Analysis
Risk Factors	OR (95% CI)	P	adjusted OR (95% CI)	P
Age (year)	1.149(1.116–1.116)	<0.001	1.146(1.097–1.198)	<0.001
BMI (kg/m2)	1.081(1.03–1.136)	0.002	-	-
History of hypertension	2.141(1.381–3.321)	0.001	-	-
history of diabetes	1.652(0.877–3.114)	0.12	-	-
History of previous pelvic and lower extremity surgery	1.211(0.782–1.877)	0.391	-	-
History of lower extremity varicose veins	11.684(5.383–25.362)	<0.001	6.721(1.553–29.079)	0.011
Operation time (min)	1.011(1.008–1.013)	<0.001	1.01(1.006–1.014)	<0.001
Intraoperative hemorrhage	1.001(1–1.002)	0.001	-	-
Gynecologic malignancies	14.951(7.39–30.248)	<0.001	8.053(2.899–22.374)	<0.001
Postoperative pain score > 4	5(1.195–20.922)	0.028	-	-
Postoperative bed rest time			-	-
< 24h	1		1	-
24h-48h	3.315(2.103–5.223)	<0.001	3.347(1.742–6.430)	<0.001
> 48h	32.657(14.487–73.615)	<0.001	99.406(19.976–494.664)	<0.001
Postoperative pneumatic pressure treatment	0.912(0.583–1.426)	0.686	-	-
Postoperative anticoagulation	0.352(0.184–0.673)	0.002	0.282(0.083–0.958)	0.042
Platelet count	1.001(1–1.002)	0.206	-	-
International normalized ratio	0.423(0.044–4.069)	0.456	-	-
D-dimer	1.177(1.031–1.343)	0.016	1.145(1.027–1.276)	0.015
Direct bilirubin	0.844(0.73–0.976)	0.022	0.715(0.563–0.909)	0.006
Hemoglobin	0.983(0.974–0.992)	<0.001	0.982(0.967–0.997)	0.016

Abbreviations: BMI, body mass index; All listed covariates in the model were not found to have multicollinearity.

Establishment of a Nomogram Model

We developed a nomogram to predict postoperative LEDVT based on identified independent risk factors (Figure 1). Using R software, we created a dynamic web-based nomogram calculator (https://whgoug.shinyapps.io/DynNomapp/). The predictive model demonstrated an AUC of 0.839 (95% CI: 0.805–0.872), with a maximum Youden index of 0.541 (Figure 2). Performance metrics included 78.0% sensitivity, 76.1% specificity, a PPV of 0.521, NPV of 0.912, and kappa statistic of 0.464. Calibration analysis revealed a C-index of 83.9% (Figure 3). DCA indicated clinical utility across threshold probabilities of 0.06–1.0 (Figure 4).

Figure 1.

Nomogram Model for Predicting the Risk of LEDVT.

Figure 2.

ROC Curve of the Predictive Model.

Figure 3.

Calibration Curve of the Nomogram Model.

Figure 4.

Decision Curve of the Nomogram Model.

Evaluation of Nomogram Model, Caprini Score and G-Caprini Model

The nomogram underwent external validation and was compared against both the Caprini score and G-Caprini model to assess discrimination, calibration, clinical utility, sensitivity, specificity, PPV, and NPV (Table 4). Superior predictive performance was visually confirmed through comparative ROC analysis (Figure 5) and DCA (Figure 6).

Figure 5.

The Comparison of Nomogram Model, Caprini Score and G-Caprini Score for ROC Curve.

Figure 6.

The Comparison of Nomogram Model, Caprini Score and G-Caprini Score for Decision Curve.

Table 4.

Prediction Performance of the Nomogram Model, Caprini and G-Caprini on the Test Set.

Model	AUC	95% CI	Sensitivity	Specificity	PPV	NPV	Accuracy Rate
Nomogram	0.954	0.904–1.000	0.926	0.967	0.714	0.993	0.963
G-Caprini	0.825	0.727–0.922	0.741	0.854	0.312	0.973	0.845
Caprini	0.793	0.695–0.891	0.519	0.947	0.467	0.956	0.912

Abbreviations: PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve; CI, confidence interval.

Discussion

The current study identified age, history of varicose veins, gynecologic malignancies, operation time, postoperative bed rest duration, postoperative anticoagulation, D-dimer level, direct bilirubin level, and hemoglobin level as independent risk factors for LEDVT following gynecologic laparoscopy. A nomogram model demonstrated an AUC of 0.839, with 78.0% sensitivity and 76.1% specificity. External validation yielded an AUC of 0.954, sensitivity of 92.6%, and specificity of 96.7%.

DVT is a common complication following gynecologic surgery. In the acute stage, thrombus detachment may cause life-threatening pulmonary embolism.²⁰ DVT comprises 40% of peripheral vascular disease cases, with mortality rates reaching 10%. Prolonged postoperative bed rest often induces blood hypercoagulability, leading to DVT. Key clinical manifestations include lower extremity pain, swelling, superficial varicose veins, and altered skin temperature.²¹ While laparoscopic surgery requires pneumoperitoneum establishment, its DVT risk—influenced by pelvic surgery and patient positioning—is comparable to laparotomy. Patients undergoing gynecologic laparoscopy face DVT risks due to postoperative blood stasis from prolonged bed rest, vascular intimal injury from extended intravenous medication, surgical trauma, and coagulation system activation from blood loss. Therefore, identifying DVT risk factors, guiding early clinical prevention, and reducing DVT incidence are critical for improving patient prognosis.²

This study identified age, history of varicose veins, pathologically confirmed gynecologic malignancies, operation time, postoperative bed rest duration, postoperative anticoagulation, D-dimer levels, direct bilirubin levels, and hemoglobin levels as independent risk factors for DVT following gynecologic laparoscopy.

Our data indicate that patients aged ≥50 years have twice the risk of postoperative DVT compared to those under 50 years, with the risk doubling for every 10-year increase in age.⁶

Varicose veins in the lower extremities involve tortuosity and dilation of superficial veins caused by venous stasis and vascular wall thinning. This condition is frequently associated with prolonged physical labor or standing occupations²²; its most severe postoperative complication is DVT. Recent studies indicate that lower extremity varicose veins arise from multiple factors, including smoking, reduced vascular elasticity, elevated venous pressure, venous hypertension, vascular wall hypoxia, and systemic inflammatory responses.^23,24 Chinese research confirms that varicose veins increase postoperative DVT risk (adjusted OR = 4.6), with incidence rates of 29.2% in patients with varicose veins versus 8.5% in those without.⁶

Malignant tumors exhibit vascular walls with reduced elasticity and increased fragility, predisposing to venous stasis. Concurrently, tumor cells secrete procoagulant particles and elevate circulating tissue factor levels. Growth factors and cancer procoagulants activate the coagulation cascade, inducing a hypercoagulable state that elevates DVT risk. Furthermore, tumor infiltration compresses surrounding veins and tissues, impeding blood flow. Malignant tumors also release substantial coagulation-promoting substances that enhance platelet aggregation, platelet adhesion, and coagulation factor activity, thereby accelerating DVT formation.

D-dimer is a key biomarker for detecting coagulation abnormalities.²⁵ As a specific fibrinolysis derivative of cross-linked fibrin, elevated plasma D-dimer levels indicate enhanced secondary fibrinolytic activity, serving as a molecular marker for hypercoagulability and hyperfibrinolysis. Studies demonstrate that postoperative D-dimer elevation is an independent risk factor for DVT, with increased levels after abdominal surgery showing predictive value for DVT progression.²⁶

This study identified direct bilirubin as an independent risk factor for VTE following gynecologic laparoscopy. While the pathogenesis underlying elevated serum bilirubin in VTE patients remains incompletely understood, in vitro experiments demonstrate that direct bilirubin enhances proliferation, migration, and angiogenesis in human umbilical vein endothelial cells, potentially through ERK1/2 and Akt pathway activation.⁶ Contrastingly, prior research indicates direct bilirubin acts as a protective factor for macrovascular disease in type 2 diabetes mellitus (T2DM) patients.²⁷ Proposed mechanisms include: 1. Antioxidant activity: Bilirubin protects against lipid peroxidation damage.²⁸ It reduces vascular risk by inhibiting oxidized LDL (ox-LDL) formation.²⁹ 2. Anti-inflammatory effects: By binding cell membranes, bilirubin disrupts surface receptors, suppressing tumor necrosis factor-α and interleukin-6 secretion, thereby lowering cardiovascular risk. 3. Heme oxygenase-1 (HO-1) modulation: Bilirubin upregulates HO-1 expression, enhancing its anti-inflammatory, antioxidant, and anti-apoptotic functions to confer vascular protection.³⁰ 4. Endoplasmic reticulum (ER) stress regulation: As an immunomodulator, bilirubin mitigates ER stress-induced apoptosis and promotes cellular regeneration.³¹ Given the absence of large-scale clinical studies linking direct bilirubin to VTE, further validation of these findings is warranted.

Previous research identified elevated D-dimer, reduced hemoglobin levels (g/L), prolonged bed rest duration, hemodynamic and coagulation parameter abnormalities, and altered hemorheology as primary risk factors for DVT development.³² Specifically, hemoglobin deficiency significantly increases lower extremity DVT risk following fractures.³³ Meta-analyses demonstrate that LMWH reduces DVT incidence by 51%–70% and PE by 64%–70% versus placebo, though with increased bleeding risk.^34,35 Additionally, LMWH achieves a 30% greater reduction in VTE compared to unfractionated heparin.

Surgical trauma and resultant perfusion alterations are significant contributors to postoperative VTE. Identified risk factors for DVT include advanced age, elevated BMI, comorbidities, disease pathology, prolonged postoperative bed rest, intraoperative pneumoperitoneum pressure, abnormal preoperative lower extremity ultrasound findings, and surgical positioning.³ Notably, frail elderly patients exhibit reduced mobility compounded by declining organ function, which alters circulating blood components and hemodynamics. This induces a hypercoagulable state with diminished local blood flow, substantially increasing DVT susceptibility.³⁶ Furthermore, prolonged bed rest due to pain or restricted mobility promotes venous stasis in lower extremities, accelerating DVT formation.³

Scholars have also employed nomogram models to predict peripherally inserted central catheter (PICC)-related thrombosis risk in cancer patients, demonstrating high predictive performance that supports clinical risk assessment.^9,37 Our analysis of thrombosis risk predictions across diverse cohorts indicates that nomograms consistently achieve robust predictive accuracy. This provides a foundation for identifying high-risk populations and formulating individualized clinical management strategies.

This study has several limitations. First, its retrospective design inherently introduces selection bias. Second, although the nomogram demonstrated superior predictive value compared to guideline-recommended models, its PPV was suboptimal—likely reflecting sample size constraints. Third, while hereditary thrombophilia contributes to DVT risk, our Delphi-validated model focused on dynamically monitorable parameters for early postoperative intervention. Future studies may integrate genetic screening in high-risk subgroups. This suggest that it needs to continuously update and improve based on the actual situation, and develop an assessment tool suitable for gynecological laparoscopy-related DVT.

Conclusions

In summary, this study identified advanced age, history of varicose veins, gynecologic malignancies, prolonged operation time, extended postoperative bed rest duration, omission of postoperative anticoagulation, elevated D-dimer levels, reduced direct bilirubin, and decreased hemoglobin levels as independent risk factors for DVT following gynecologic laparoscopic surgery. The nomogram integrating these factors demonstrates robust predictive accuracy for DVT, enabling early risk stratification and targeted prophylactic interventions to mitigate thrombosis and related surgical complications.

Footnotes

Acknowledgments

We are grateful to the patients who made this study possible.

ORCID iD

Yuping Zhao

Ethics Statement

The requirement of informed consent was waived due to the retrospective nature of the study. The study was conducted in accordance with the Declaration of Helsinki. Patient identity could not be identified in the publication.

Ethical Approval and Consent to Participate

This retrospective study was approved by the Institutional Review Board of Fujian Maternity and Child Health Hospital (2022YJ028–02).

Consent for Publication

All the authors listed have approved the manuscript that is enclosed.

Author Contributions

Chenyin Liu performed the experiments, prepared figures and/or tables, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.

Rong Zhang performed the experiments, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Yanbin Lin performed the experiments, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Xiumei Fang performed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Ying Fu performed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Shuiling Zu conceived and designed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Yuping Zhao conceived and designed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Nursing Research Special Fund of Fujian Maternal and Child Health Hospital (YCXH 22-10 and YCXH 22-18). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

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 data that support the findings of this study are available from the corresponding author, Yuping Zhao, upon reasonable request.

References

Insin

Vitoopinyoparb

Thadanipon

, et al. Prevention of venous thromboembolism in gynecological cancer patients undergoing major abdominopelvic surgery: A systematic review and network meta-analysis. Gynecol Oncol. 2021;161(1):304–313. doi:10.1016/j.ygyno.2021.01.027

Tian

. Risk factors of deep vein thrombosis of lower extremity in patients undergone gynecological laparoscopic surgery: What should we care. BMC Womens Health. 2021;21(1):130. doi:10.1186/s12905-021-01276-7

Geng

Dong

Fang

Tian

. The best evidence for the prevention and management of lower extremity deep venous thrombosis after gynecological malignant tumor surgery: A systematic review and network meta-analysis. Front Surg. 2022 Mar 22;9:841275. doi:10.3389/fsurg.2022.841275

Lang

Wang

. Chinese Expert consensus of prevention of deep venous thrombosis and pulmonary embolism after gynecologic surgery. Zhonghua Fu Chan Ke Za Zhi. 2017 Oct 25;52(10):649–653. doi:10.3760/cma.j.issn.0529-567X.2017.10.001

Vedantham

. Using it wisely: Catheter-directed thrombolysis for deep vein thrombosis. Lancet Haematol. 2020;7(1):e6–e7. doi:10.1016/S2352-3026(19)30205-4

Zhai

, et al. Predicting of venous thromboembolism for patients undergoing gynecological surgery. Medicine (Baltimore). 2015;94(39):e1653. doi:10.1097/MD.0000000000001653

Yago

Yamaki

Sasaki

, et al. Application of the Caprini risk assessment model for evaluating postoperative deep vein thrombosis in patients undergoing plastic and reconstructive surgery. Ann Vasc Surg. 2020 May;65:82–89. doi:10.1016/j.avsg.2019.10.082

Chen

Dai

Song

, et al. Establishment of a risk assessment tool for pregnancy-associated venous thromboembolism and its clinical application: Protocol for a prospective observational study in Beijing. BMC Pregnancy Childbirth. 2019;19(1):294. doi:10.1186/s12884-019-2448-7

Jiang

Zhang

Wang

. A nomogram model can predict the risk of venous thromboembolism in postoperative patients with gynecological malignancies. Int J Gynaecol Obstet. 2022;158(3):689–699. doi:10.1002/ijgo.14061

10.

Quan

Hao

, et al. A simple nomogram for assessing the risk of IgA vasculitis nephritis in IgA vasculitis Asian pediatric patients. Sci Rep. 2022;12(1):16809. doi:10.1038/s41598-022-20369-3

11.

Wang

, et al. Estrogen/progesterone receptor expression and cancer antigen 125 level as preoperative predictors to estimate lymph node metastasis in endometrioid endometrial cancer. Int J Gynecol Pathol. 2024;43(4):316–325. doi:10.1097/PGP.0000000000000984

12.

Wen

Wang

Bian

, et al. Nomogram to predict 6-month mortality in acute ischemic stroke patients treated with endovascular treatment. Front Neurol. 2024 Jan 5;14:1330959. doi:10.3389/fneur.2023.1330959

13.

Xiao

Cao

Zhen

Chen

Huang

. Nomogram analysis based on clinical and sonographic characteristics for the assessment of postpartum stress urinary incontinence. J Ultrasound Med. 2023;42(11):2591–2601. doi:10.1002/jum.16295

14.

Zhou

Wei

Zhang

Shen

Yang

. Establishment and validation of a nomogram model for prediction of diabetic nephropathy in type 2 diabetic patients with proteinuria. Diabetes Metab Syndr Obes. 2022 Apr 8;15:1101–1110. doi:10.2147/DMSO.S357357

15.

Chen

Zheng

Yang

Dai

Yuan

. A nomogram model to predict deep vein thrombosis risk after surgery in patients with hip fractures. Indian J Orthop. 2024;58(2):151–161. doi:10.1007/s43465-023-01074-3

16.

Ling

Huang

. Nomogram based on high-density lipoprotein cholesterol for the occurrence of preoperative deep vein thrombosis in patients with intertrochanteric femur fracture: A retrospective study. J Orthop Surg Res. 2024;19(1):22. doi:10.1186/s13018-023-04497-8

17.

Zhu

Yang

Zhou

. Construction of nursing care quality evaluation indicators for post-anaesthesia care unit in China. J Clin Nurs. 2023;32(1-2):137–146. doi:10.1111/jocn.16207

18.

Jin

Zheng

, et al. Nomogram for predicting pulmonary embolism in gynecologic inpatients with isolated distal deep venous thrombosis. Int J Gynaecol Obstet. 2024;164(1):324–333. doi:10.1002/ijgo.15050

19.

Sugimachi

Tajiri

Kinjo

, et al. Incidence and predictors of deep venous thrombosis after abdominal oncologic surgery: Prospective Doppler ultrasound screening. J Surg Res. 2012;178(2):657–661. doi:10.1016/j.jss.2012.06.002

20.

Chen

Hou

Wang

. Machine learning-based model for prediction of deep vein thrombosis after gynecological laparoscopy: A retrospective cohort study. Medicine (Baltimore). 2024;103(1):e36717. doi:10.1097/MD.0000000000036717

21.

Lorchaivej

Suprasert

Srisuwan

Rujiwetpongstorn

. Prevalence and risk factor of post-operative lower extremities deep vein thrombosis in patients undergoing gynecologic surgery: A single-institute cross-sectional study. Thromb J. 2022;20(1):14. doi:10.1186/s12959-022-00376-0

22.

Dua

Neiva

Sutherland

. Does previous varicose vein surgery alter deep vein thrombosis risk after lower limb arthroplasty? Orthop Surg. 2012;4(4):222–226. doi:10.1111/os.12003

23.

Rusinovich

. Cardiac Doppler in patients with primary varicose veins of lower extremities. Phlebology. 2020;35(1):62–66. doi:10.1177/0268355519848895

24.

Tamura

Maruyama

Sakurai

. Effectiveness of endovenous radiofrequency ablation for elderly patients with varicose veins of lower extremities. Ann Vasc Dis. 2019;12(2):200–204. doi:10.3400/avd.oa.19-00002

25.

Jensen

Chahin

Amin

, et al. Qualitative slow blood flow in lower extremity deep veins on doppler sonography: Quantitative assessment and preliminary evaluation of correlation with subsequent deep venous thrombosis development in a tertiary care oncology center. J Ultrasound Med. 2017;36(9):1867–1874. doi:10.1002/jum.14220

26.

Zhang

Liu

Cheng

Yang

Zhang

. Risk factors and treatment of venous thromboembolism in perioperative patients with ovarian cancer in China. Medicine (Baltimore). 2018;97(31):e11754. doi:10.1097/MD.0000000000011754

27.

Vitek

. The role of bilirubin in diabetes, metabolic syndrome, and cardiovascular diseases. Front Pharmacol. 2012 Apr 3;3:55. doi:10.3389/fphar.2012.00055

28.

Soto Conti

. Bilirubin: The toxic mechanisms of an antioxidant molecule. Arch Argent Pediatr. 2021 Feb;119(1):e18–e25. doi:10.5546/aap.2021.eng.e18

29.

Kwon

Lee

. Direct bilirubin is associated with low-density lipoprotein subfractions and particle size in overweight and centrally obese women. Nutr Metab Cardiovasc Dis. 2018;28(10):1021–1028. doi:10.1016/j.numecd.2018.05.013

30.

Kishimoto

Niki

Saita

, et al. Blood levels of heme oxygenase-1 versus bilirubin in patients with coronary artery disease. Clin Chim Acta. 2020 May;504:30–35. doi:10.1016/j.cca.2020.01.030

31.

Bai

Cheng

Yang

, et al. C1q/TNF-related protein 9 protects diabetic rat heart against ischemia reperfusion injury: Role of endoplasmic reticulum stress. Oxid Med Cell Longev. 2016;2016(1):1902025. doi:10.1155/2016/1902025

32.

Xue

Peng

. Analysis of risk factors for deep venous thrombosis in patients with gynecological malignant tumor: A clinical study. Pak J Med Sci. 2019 Jan-Feb;35(1):195–199. doi:10.12669/pjms.35.1.365

33.

Xiong

Cheng

. Association between red blood cell indices and preoperative deep vein thrombosis in patients undergoing total joint arthroplasty: A retrospective study. Clin Appl Thromb Hemost. 2022 Jan-Dec;28:10760296221149029. doi:10.1177/10760296221149029

34.

Gould

Garcia

Wren

, et al. Prevention of VTE in nonorthopedic surgical patients: Antithrombotic therapy and prevention of thrombosis, 9th ed: American college of chest physicians evidence-based clinical practice guidelines. Chest. 2012;141(2):e227S–e277S. doi:10.1378/chest.11-2297

35.

Koch

Bouges

Ziegler

Dinkel

Daures

Victor

. Low molecular weight heparin and unfractionated heparin in thrombosis prophylaxis after major surgical intervention: Update of previous meta-analyses. Br J Surg. 1997 Jun;84(6):750–759.

36.

Zhao

Wang

, et al. A nomogram model for predicting lower extremity deep vein thrombosis after gynecologic laparoscopic surgery: A retrospective cohort study. PeerJ. 2023 Sep21;11:e16089. doi:10.7717/peerj.16089

37.

Wang

Cui

Lyu

Lang

. A nomogram model to predict the portal vein thrombosis risk after surgery in patients with pancreatic cancer. Front Surg. 2023 Dec19;10:1293004. doi:10.3389/fsurg.2023.1293004

Development and Empirical Study of a Lower Extremity Deep Vein Thrombosis Prevention Protocol for Patients Undergoing Gynecologic Laparoscopic Surgery

Abstract

Keywords

Introduction

Methods

Expert Correspondence

Design of the Study

Inclusion and Exclusion Criteria

Diagnostic Criteria

Statistical Analysis

Results

Basic Information of Experts

Degree of Activity and Authority of Experts

The Degree of Coordination of Expert Opinions

Final Results of Expert Correspondence

Risk Factors for LEDVT by Binary Logistic Regression Models

Establishment of a Nomogram Model

Evaluation of Nomogram Model, Caprini Score and G-Caprini Model

Discussion

Conclusions

Footnotes

Acknowledgments

ORCID iD

Ethics Statement

Ethical Approval and Consent to Participate

Consent for Publication

Author Contributions

Funding

Declaration of Conflicting Interests

Data Availability Statement

References