ABSTRACT
Objective
Venous thromboembolism (VTE) remains an important cause of maternal morbidity and mortality, yet the diagnosis of deep vein thrombosis (DVT) during pregnancy is challenging because clinical symptoms overlap with physiological gestational changes and D‐dimer specificity is reduced. This study evaluated the association between complete blood count‐derived systemic inflammatory indices and pregnancy‐associated DVT.
Methods
This retrospective case‐control study included 36 pregnant women with Doppler ultrasonography‐confirmed DVT and 36 healthy pregnant controls matched for maternal age and gestational age. Neutrophil‐to‐lymphocyte ratio (NLR), platelet‐to‐lymphocyte ratio (PLR), monocyte‐to‐lymphocyte ratio (MLR), systemic immune‐inflammation index (SII), systemic inflammation response index (SIRI), and aggregate index of systemic inflammation (AISI) were calculated from routine complete blood counts. Univariable and multivariable Firth penalized logistic regression analyses were performed, adjusting for body mass index, history of cesarean delivery, and varicose veins. Elastic net penalized regression was used for variable selection. Gestational age‐stratified sensitivity analyses, false discovery rate correction, receiver operating characteristic (ROC) analyses, and a supplementary trimester‐specific D‐dimer threshold analysis were also performed.
Results
Women with DVT had higher NLR, SII, SIRI, AISI, and D‐dimer levels than controls, whereas PLR and MLR did not differ significantly. After false discovery rate correction, NLR, SII, SIRI, and AISI remained statistically significant. In multivariable Firth regression, NLR, SII, and D‐dimer were independently associated with DVT. Elastic net analysis identified NLR and SII as the most informative inflammatory indices. ROC analyses showed moderate discriminatory performance for NLR and SII, either alone or combined with D‐dimer. The gestational age‐stratified analyses showed a consistent direction of association, while trimester‐specific D‐dimer thresholds increased specificity but substantially reduced sensitivity in this cohort.
Conclusion
NLR and SII were associated with pregnancy‐associated DVT and may reflect the thrombo‐inflammatory component of this condition. However, the findings are exploratory and hypothesis‐generating. These indices should be considered only as adjunctive markers within the existing diagnostic workflow, not as standalone diagnostic tools. Prospective studies using standardized sampling and pregnancy‐adapted D‐dimer strategies are needed before clinical implementation.
Keywords: deep vein thrombosis, inflammation biomarkers, neutrophil‐to‐lymphocyte ratio, pregnancy, systemic immune‐inflammation index, Venous thromboembolism
1. Introduction
Venous thromboembolism (VTE) represents an important cause of morbidity and mortality in pregnant and postpartum women. It has been estimated that VTE in pregnancy and postpartum period has a risk incidence ranging from 1 per 1000 to 1 per 2000 cases, representing a 4‐ to 5‐fold increased risk in comparison with non‐pregnant women [1, 2, 3]. Interestingly, the increased risk in these pregnant women can be attributed to the prothrombotic state of pregnancy, with increased concentrations of coagulation factors, impaired fibrinolytic function, and stasis of veins, fully meeting all requirements of Virchow's triad [4]. Other associated factors, like advanced age of the mother, obesity, immobilization, multiple pregnancy, caesarean section, antecedent VTE events, and thrombophilias, make significant contributions to VTE risk in pregnant cases [5, 6, 7]. Thus, pregnancy and postpartum period are considered to attain the state of high VTE risk. In turn, occurrence of deep vein thrombosis (DVT) in pregnant women can lead to potentially lethal events in mothers due to the development of fatal pulmonary embolism and represents the source of post‐thrombotic sequelae in the long run [8].
Clinical manifestations of DVT usually include swelling, pain, tenderness, and warmth in the affected limb, whereas pulmonary embolism commonly presents with dyspnea, tachypnea, tachycardia, and chest pain. Many of these symptoms may also occur physiologically during pregnancy, which complicates the clinical diagnosis of DVT in pregnant women. Diagnostic evaluation is further challenged by concerns regarding ionizing radiation with some imaging modalities in pregnancy [9]. Although compression ultrasonography is generally considered safe and is the first‐line imaging modality for suspected DVT, its diagnostic performance may be limited in cases of pelvic or iliac vein thrombosis. In addition, D‐dimer levels increase physiologically during pregnancy because of the hypercoagulable state, which reduces the specificity of conventional nonpregnant cutoffs [10, 11, 12]. Therefore, D‐dimer alone has limited value for ruling out VTE during pregnancy.
Current evidence indicates that VTE is not solely a disorder of the coagulation system; inflammation and immune activation also play important roles in thrombus formation [13, 14]. Endothelial activation, leukocyte‐platelet interactions, and the release of inflammatory cytokines are key mechanisms linking systemic inflammation to venous thrombosis. Accordingly, there is growing interest in practical and readily available biomarkers that may reflect thrombo‐inflammatory activity. Complete blood count‐derived indices, including the neutrophil‐to‐lymphocyte ratio (NLR), platelet‐to‐lymphocyte ratio (PLR), monocyte‐to‐lymphocyte ratio (MLR), systemic immune‐inflammation index (SII), systemic inflammation response index (SIRI), and aggregate index of systemic inflammation (AISI), have been evaluated as low‐cost markers of systemic inflammation in obstetric and gynecological conditions [15, 16, 17, 18].
Recent evidence from non‐obstetric populations further supports the clinical relevance of composite inflammation‐related indices. SIRI has been associated with stroke risk in elderly patients with hypertension [19], while a large‐scale cross‐sectional analysis reported associations between inflammatory indices and metabolic dysfunction‐associated fatty liver disease among hypertensive patients [20]. In critically ill patients with sepsis, SIRI has also been shown to improve prognostic prediction [21]. In addition, other inflammation‐related composite biomarkers, such as the lactate dehydrogenase‐to‐albumin ratio, have been associated with outcomes in acute vascular disease [22]. Although these findings cannot be directly extrapolated to pregnancy‐associated DVT, they support the broader concept that composite inflammatory markers may capture clinically relevant systemic inflammatory activity across different disease settings.
Pregnancy is a unique physiological state characterized by hypercoagulability, venous stasis, and gestational changes in inflammatory and immune responses. Therefore, evaluating whether complete blood count‐derived inflammatory indices are associated with pregnancy‐associated DVT may provide useful exploratory information. The aim of this study was to investigate the association between systemic inflammation‐related hematological indices and DVT during pregnancy and to explore their potential adjunctive discriminatory performance within the existing diagnostic pathway.
2. Materials and Methods
This retrospective case‐control study was conducted at Ankara Etlik Zübeyde Hanım Women's Health Training and Research Hospital, a tertiary referral center, between January 2016 and August 2022. Institutional ethics committee approval was obtained (Protocol No: 14/09 Date: 24.10.2022), and the study was conducted in accordance with ethical principles and the Declaration of Helsinki. Due to the study's retrospective nature, informed consent was waived.
2.1. Participants and Study Groups
During the study period, all pregnant women who were diagnosed with DVT were eligible for inclusion. The diagnosis of DVT was established based on compatible clinical symptoms, such as lower extremity pain, swelling, tenderness, or unilateral limb asymmetry, together with confirmatory findings on lower extremity Doppler ultrasonography. Imaging findings considered consistent with acute thrombosis included venous non‐compressibility, visualization of intraluminal thrombus, or abnormal venous flow patterns. All cases were identified through retrospective review of hospital medical records and corresponding radiology reports. As the control group, healthy singleton pregnancies matched for maternal age and gestational age who presented to our hospital for routine antenatal monitoring within the specified study period were included. Exclusion criteria included a prior history of VTE, multiple pregnancy, chronic liver or renal disease, autoimmune or chronic inflammatory disorders, cardiovascular disease (including hypertension, coronary artery disease, or hyperlipidemia), active or recent infection (such as urinary tract infection, upper respiratory tract infection, pelvic inflammatory disease, current or previous COVID‐19 infection, or suspected or confirmed chorioamnionitis), smoking, use of corticosteroids or anti‐inflammatory medications, and incomplete or inaccessible clinical or laboratory data.
2.2. Study Parameters and Data Collection
Data were obtained from patient files and the hospital electronic record system. Demographic and obstetric variables (maternal age, body mass index (BMI), gravidity, parity, abortion, and previous cesarean section history), obstetric parameters, complete blood count parameters, pregnancy outcome, and pregnancy complications were recorded. For women with DVT, laboratory parameters recorded at the time of diagnostic evaluation were used. For controls, laboratory parameters obtained during routine antenatal assessment at the matched gestational age were recorded. All laboratory analyses were performed in the same institutional laboratory according to routine standardized procedures. Because of the retrospective design, the exact clock time of blood sampling, fasting status, and some preanalytical conditions could not be uniformly verified for all participants.
Systemic inflammation indices were calculated as follows: NLR, the neutrophil count divided by the lymphocyte count; PLR, the platelet count divided by the lymphocyte count; MLR, the monocyte count divided by the lymphocyte count. SII was calculated as (platelet count × neutrophil count) divided by the lymphocyte count; SIRI, the product of the neutrophil and monocyte counts divided by the lymphocyte count; and AISI were calculated by dividing the product of the neutrophil, monocyte, and platelet counts by the lymphocyte count.
2.3. Size Calculation
In the study by Tort et al., titled “Evaluation of SII in acute DVT: A propensity‐matched study”, the standardized mean differences (Cohen's d) between patients with acute DVT and controls were reported as 1.11 for NLR and 1.22 for SII [23]. Taking the smaller effect size for NLR (Cohen's d = 1.11) as a conservative estimate, an a priori power analysis was performed using G*Power version 3.1.9.7 for a two‐sided independent samples t‐test with a significance level of α = 0.05 and a desired power of 95% (1−β = 0.95). This analysis indicated that a minimum of 23 participants per group (46 in total) would be required to detect a difference of this magnitude between groups.
2.4. Statistical Analysis
All data analyses were performed using Stata version 19.5 (StataCorp, College Station, TX, USA). Normality of the continuous variables was tested using the Shapiro‐Wilk test. For normally distributed variables, comparison between groups was made using the independent sample t‐test, and data represented as mean ± standard deviation; whereas in non‐normally distributed variables, comparison was made using the Mann‐Whitney U test, and the result was demonstrated as median (25th‐75th percentiles). To determine factors related to DVT, logistic regressions were performed. In consideration of small sample size and separation, Firth penalized logistic regressions were used. Inflammatory indices were first explored in Firth logistic regressions using an observation‐wide approach, and multivariable Firth logistic regressions were performed following that. In the multivariable analysis, covariates included in the model were pre‐specified clinical factors: BMI, cesarean section history, and presence of varicose veins. Data are presented in terms of odds ratio (OR) and adjusted odds ratio (aOR) with corresponding 95% confidence intervals (CI). In order to test multicollinearity between inflammatory indices or to figure out which parameters contain most useful information, elastic net penalized logistic regressions were employed as variable selection procedure. For elastic net analysis, 10‐fold cross‐validation was applied with mixing parameter value of alpha = 0.5. All the included continuous variables in elastic net analysis were scaled (z‐scores) before analysis. For evaluation of diagnostic performance, receiver operating characteristic (ROC) curve analyses were performed for individual inflammatory biomarkers, D‐dimer, and clinically meaningful combined models identified through elastic net–based variable selection, using predicted probabilities derived from standard binary logistic regression models. Performance was presented in terms of area under the curves (AUC) with corresponding 95% confidential intervals. Optimal cutoffs were identified using the Youden index (sensitivity + specificity − 1). For D‐dimer, the primary ROC analysis used the conventional fixed threshold of >500 ng/mL to maintain comparability with routine clinical use and previous literature. However, because D‐dimer levels increase physiologically across gestation, this threshold was considered only a clinically familiar reference benchmark and not a pregnancy‐specific diagnostic cutoff. To address gestational age‐related variation in D‐dimer interpretation, a supplementary sensitivity analysis was performed using trimester‐specific upper reference limits reported in the literature: >721 ng/mL for the first trimester, >1653 ng/mL for the second trimester, and >2256 ng/mL for the third trimester [10]. These thresholds were used only as a reference‐interval‐based sensitivity approach and should not be interpreted as validated diagnostic cutoffs for pregnancy‐associated DVT. Corresponding values of sensitivities, specificities, positive likelihood ratio (LR+), and negative likelihood ratio (LR−) are presented.
To assess whether the observed associations were influenced by gestational age at blood sampling, a gestational age‐stratified sensitivity analysis was performed. Given the limited number of first‐ and second‐trimester observations, the primary stratified analysis used two gestational age categories: <28 weeks and ≥28 weeks. Within each stratum, inflammatory indices were compared between women with DVT and controls using the Mann‐Whitney U test. Trimester‐specific descriptive comparisons were also provided as supplementary exploratory analyses.
Because multiple complete blood count‐derived inflammatory indices were evaluated, the Benjamini‐Hochberg false discovery rate procedure was applied as an additional correction for multiple comparisons among NLR, PLR, MLR, SII, SIRI, and AISI. Unadjusted p‐values were reported in the main tables for consistency with the primary analyses, while FDR‐adjusted q‐values were provided in the supplementary material. A two‐sided p‐value <0.05 was considered statistically significant for primary analyses, and an FDR‐adjusted q‐value <0.05 was considered statistically significant in corrected comparisons. All tests were two‐sided in nature; whereas significance level was set at p < 0.05.
3. Results
A total of 72 pregnant women were included in the analysis, comprising 36 patients diagnosed with DVT and 36 healthy pregnant controls. Baseline demographic, obstetric, and neonatal characteristics of the study population are presented in Table 1. Maternal age, body mass index, gravidity, parity, number of abortuses, and gestational age at blood sampling were similar between the groups (p > 0.05 for all). A history of previous cesarean delivery was more frequent among women with DVT (p = 0.050). Varicose veins were present only in the DVT group. Gestational age at delivery and birth weight were lower in the DVT group (p = 0.004 and p = 0.023, respectively). Neonatal intensive care unit admission occurred more frequently among infants born to mothers in the DVT group (p = 0.028).
TABLE 1.
Comparison of demographic, obstetric, and neonatal characteristics between DVT and control groups.
| Variables | DVT (n = 36) | Control (n = 36) | p‐value |
|---|---|---|---|
| Maternal Age (years) | 28 (26–32) | 28 (25–33) | 0.595 a |
| BMI (kg/m2) | 28.19 ± 4.24 | 29.81 ± 6.01 | 0.193 b |
| Gravidity | 3 (2–4) | 3 (2–3) | 0.665 a |
| Parity | 2 (1–2) | 1 (1–2) | 0.185 a |
| Number of abortions | 0 (0‐0) | 0 (0–1) | 0.131 a |
| History of Previous C/S | 17 (47.2%) | 9 (25.0%) | 0.050 c |
| Gestational age at DVT diagnosis and blood sampling (weeks) | 32 (26–35) | 32 (26–35) | 0.991 a |
| Presence of varicose veins | 12 (33.3%) | — | <0.001 c |
| Gestational week at birth (weeks) | 38 (36—39) | 39 (38—40) | 0.004 a |
| Birth weight (g) | 3010 (2520–3363) | 3340 (2890–3495) | 0.023 a |
| Delivery Mode | 0.238 c | ||
| Vaginal | 15 (41.7%) | 21 (58.3%) | |
| C/S | 21 (58.3%) | 15 (41.7%) | |
| Apgar score at first minute | 9 (9‐9) | 9 (9‐9) | 0.018 a |
| Apgar score at fifth minute | 10 (9–10) | 10 (10‐10) | 0.009 a |
| NICU admission | 8 (22.2%) | 1 (2.8%) | 0.028 c |
The Mann–Whitney U test was used for comparisons between groups. Data are presented as median (interquartile range).
Independent samples t‐test was used for comparisons between groups. Data are presented as mean ± standard deviation.
Categorical variables were compared using the chi‐square or Fisher's exact test, as appropriate. Results are shown as n (%).
Abbreviations: BMI, Body mass index; C/S, Cesarean section; DVT, Deep Vein Thrombosis; DVT, Deep vein thrombosis; NICU, Neonatal intensive care unit.
Gestational age‐stratified sensitivity analyses are summarized in Table S1. In the <28‐week stratum, NLR and SII were higher in women with DVT than in controls. Similar findings were observed in the ≥28‐week stratum. The direction of association for the main inflammatory indices was therefore generally consistent across gestational age strata, suggesting that the overall findings were not solely driven by differences in gestational age at blood sampling. However, because the number of participants in the earlier gestational age strata was limited, these results should be interpreted as exploratory.
Comparisons of hematological and inflammatory parameters measured at gestational age‐matched blood sampling are shown in Table 2. Compared with controls, women with DVT had higher leukocyte and neutrophil counts and lower hematocrit levels (p = 0.009, p = 0.001, and p = 0.013, respectively). NLR (p < 0.001), SII (p < 0.001), SIRI (p = 0.003), AISI (p = 0.008), and D‐dimer levels (p = 0.004) were higher in the DVT group. Platelet count, monocyte count, eosinophil count, PLR, and MLR did not differ significantly between groups (p > 0.05 for all).
TABLE 2.
Comparison of hematological and inflammatory parameters between DVT and control groups at gestational age‐matched blood sampling.
| Variables | DVT (n = 36) | Control (n = 36) | p‐value |
|---|---|---|---|
| Hemoglobin (g/dL) | 10.9 ± 1.7 | 11.5 ± 1.5 | 0.124 a |
| Leukocyte count (103/µL) | 10.8 ± 4.0 | 8.8 ± 2.1 | 0.009 a |
| Hematocrit (%) | 33.1 ± 4.7 | 35.8 ± 4.3 | 0.013 a |
| Neutrophil count (103/µL) | 8.56 ± 3.51 | 6.17 ± 2.20 | 0.001 a |
| Lymphocyte count (103/µL) | 1.52 (1.17–1.81) | 1.66 (1.44–2.28) | 0.021 b |
| Monocyte count (103/µL) | 0.48 (0.42–0.54) | 0.49 (0.40–0.65) | 0.761 b |
| Eosinophil count (103/µL) | 0.05 (0.03–0.15) | 0.07 (0.04–0.12) | 0.748 b |
| Platelet count (103/µL) | 234 ± 75 | 241 ± 70 | 0.689 a |
| NLR | 5.24 (3.98–7.09) | 3.88 (2.36–4.38) | <0.001 b |
| PLR | 143.83 (123.94–176.31) | 124.42 (108.61–167.45) | 0.062 b |
| MLR | 0.33 (0.25–0.42) | 0.28 (0.22–0.38) | 0.182 b |
| SII | 1174.48 (888.57–1688.99) | 769.47 (610.43–972.13) | <0.001 b |
| SIRI | 2.60 (1.89–3.68) | 1.61 (1.05–2.67) | 0.003 b |
| AISI | 584.27 (434.04–891.29) | 352.95 (239.49–537.98) | 0.008 b |
| D‐dimer (ng/mL) | 590 ± 148 | 475 ± 176 | 0.004 a |
Independent samples t‐test was used for comparisons between groups. Data are presented as mean ± standard deviation.
The Mann–Whitney U test was used for comparisons between groups. Data are presented as median (interquartile range).
Abbreviations: AISI, Aggregate Index of Systemic Inflammation; DVT, Deep Vein Thrombosis; MLR, Monocyte‐to‐Lymphocyte Ratio; NLR, Neutrophil‐to‐Lymphocyte Ratio; PLR, Platelet‐to‐Lymphocyte Ratio; SII, Systemic Immune‐Inflammation Index; SIRI, Systemic Inflammation Response Index.
After applying false discovery rate correction for multiple comparisons among complete blood count‐derived inflammatory indices, NLR, SII, SIRI, and AISI remained statistically significant, whereas PLR and MLR did not. These corrected findings support the robustness of the main inflammatory‐index results while also indicating that not all evaluated indices provided independent statistical evidence after adjustment for multiple testing (Table S2).
Univariable and multivariable Firth penalized logistic regression analyses examining associations between inflammatory indices and DVT status are shown in Table 3. In the univariable model NLR (OR = 1.569, 95% CI = 1.181–2.083, p = 0.002), SII (OR = 1.002, 95% CI = 1.000–1.003, p = 0.006), and D‐dimer levels (OR = 1.004, 95% CI = 1.001–1.007, p = 0.007) were associated with DVT. In the multivariable model adjusted for body mass index, history of cesarean delivery, and presence of varicose veins, these associations remained statistically significant for NLR (aOR = 1.629, 95% CI = 1.205–2.201, p = 0.001), SII (aOR = 1.002, 95% CI = 1.001–1.003, p = 0.004), and D‐dimer (aOR = 1.003, 95% CI = 1.000–1.007, p = 0.049).
TABLE 3.
Univariable and multivariable Firth penalized logistic regression analyses of inflammatory indices in pregnant women with DVT.
| Variables | Univariable | Multivariable | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p‐value | aOR | 95% CI | p‐value | |
| NLR | 1.569 | 1.181–2.083 | 0.002 | 1.629 | 1.205–2.201 | 0.001 |
| PLR | 1.007 | 0.998–1.016 | 0.130 | 1.006 | 0.678–6.058 | 0.199 |
| MLR | 1.305 | 0.117–14.595 | 0.829 | — | — | — |
| SII | 1.002 | 1.000–1.003 | 0.006 | 1.002 | 1.001–1.003 | 0.004 |
| SIRI | 1.207 | 0.944–1.543 | 0.134 | 1.212 | 0.924–1.591 | 0.165 |
| AISI | 1.001 | 0.999–1.002 | 0.158 | 1.001 | 0.999–1.002 | 0.184 |
| D‐dimer (ng/mL) | 1.004 | 1.001–1.007 | 0.007 | 1.003 | 1.000–1.007 | 0.049 |
Note: Multivariable analyses were adjusted for body mass index, history of cesarean delivery, and presence of varicose veins.
Abbreviations: AISI, Aggregate Index of Systemic Inflammation; aOR, Adjusted Odds Ratio; CI, Confidence Interval; NLR, Neutrophil‐to‐Lymphocyte Ratio; OR, Odds Ratio; NLR, Neutrophil‐to‐Lymphocyte Ratio; SII, Systemic Immune‐Inflammation Index; SIRI, Systemic Inflammation Response Index.
Results of the elastic net penalized logistic regression analysis are presented in Table 4. Among the inflammatory indices entered into the model, NLR and SII were selected, whereas PLR, SIRI, and AISI were not retained.
TABLE 4.
Elastic net penalized logistic regression results for inflammatory indices.
| Variable | Standardized coefficient | Selected (Yes/No) |
|---|---|---|
| NLR (z‐score) | 0.633 | Yes |
| PLR (z‐score) | 0 | No |
| SII (z‐score) | 0.268 | Yes |
| SIRI (z‐score) | 0 | No |
| AISI (z‐score) | 0 | No |
Note: Elastic net penalized logistic regression was performed using 10‐fold cross‐validation (α = 0.5). Continuous variables were standardized prior to analysis. Coefficients represent standardized elastic net weights and should not be interpreted as odds ratios.
Abbreviations: AISI, Aggregate Index of Systemic Inflammation; NLR, Neutrophil‐to‐Lymphocyte Ratio; SII, Systemic Immune‐Inflammation Index; SIRI, Systemic Inflammation Response Index.
The ROC curve analyses for individual inflammatory indices, D‐dimer, and combined models are presented in Table 5 and Figure 1. The area under the curve (AUC) for NLR was 0.764 (95% CI: 0.649–0.856), with a sensitivity of 61.1% and specificity of 80.6% at a cutoff value >4.586. The SII yielded an AUC of 0.751 (95% CI: 0.635–0.845), with a sensitivity of 77.8% and specificity of 75.0% at a cutoff value >848.65. For D‐dimer, using a cutoff value of >500, the AUC was 0.703 (95% CI: 0.584–0.805), with a sensitivity of 72.2% and specificity of 63.9%.
TABLE 5.
Primary ROC analysis of selected inflammatory indices, D‐dimer, and combined models for pregnancy‐associated DVT.
| Variables | Cut‐off | AUC (95% CI) | Sensitivity (%) | Specificity (%) | LR+ | LR− | Youden index | p‐value |
|---|---|---|---|---|---|---|---|---|
| NLR | >4.586 | 0.764 (0.649–0.856) | 61.1 | 80.6 | 3.14 | 0.48 | 0.417 | <0.001 |
| SII | >848.65 | 0.751 (0.635–0.845) | 77.8 | 75.0 | 3.11 | 0.30 | 0.528 | <0.001 |
| D‐dimer (ng/mL) | >500 | 0.703 (0.584–0.805) | 72.2 | 63.9 | 2.00 | 0.43 | — | 0.001 |
| D‐dimer + NLR | — | 0.813 (0.703–0.895) | 86.1 | 69.4 | 2.82 | 0.20 | 0.556 | <0.001 |
| D‐dimer + SII | — | 0.809 (0.699–0.892) | 97.2 | 58.3 | 2.33 | 0.05 | 0.556 | <0.001 |
| D‐dimer + NLR + SII | — | 0.815 (0.706–0.897) | 88.9 | 69.4 | 2.91 | 0.16 | 0.583 | <0.001 |
Abbreviations: AUC, area under the receiver operating characteristic curve; CI, confidence interval; LR−, negative likelihood ratio; LR+, positive likelihood ratio; NLR, Neutrophil‐to‐Lymphocyte Ratio; ROC, Receiver operating characteristic; SII, Systemic Immune‐Inflammation Index.
FIGURE 1.
The ROC curves of NLR, SII, D‐dimer, and combined models for the diagnosis of DVT in pregnant women.
Combined models incorporating D‐dimer with selected inflammatory indices demonstrated higher AUC values. The model including D‐dimer and NLR had an AUC of 0.813 (95% CI: 0.703–0.895), with a sensitivity of 86.1% and specificity of 69.4%. The combination of D‐dimer and SII resulted in an AUC of 0.809 (95% CI: 0.699–0.892), with a sensitivity of 97.2% and specificity of 58.3%. The model combining D‐dimer, NLR, and SII had an AUC of 0.815 (95% CI: 0.706–0.897), with a sensitivity of 88.9% and specificity of 69.4%. Corresponding positive and negative likelihood ratios and Youden index values for each model are detailed in Table 5.
A supplementary gestational age‐specific D‐dimer threshold analysis was performed using trimester‐specific pregnancy reference limits (Table S3). When the first‐trimester threshold of >721 ng/mL was applied, D‐dimer positivity was observed in 2 of 5 women with DVT and 1 of 5 controls, corresponding to a sensitivity of 40.0% and specificity of 80.0%. Using the second‐trimester threshold of >1653 ng/mL, 1 of 7 DVT cases and none of the 7 controls were classified as D‐dimer positive, yielding a sensitivity of 14.3% and specificity of 100.0%. In the third trimester, application of the >2256 ng/mL threshold identified 1 of 24 DVT cases and none of the 24 controls as D‐dimer positive, corresponding to a sensitivity of 4.2% and specificity of 100.0%. Overall, when trimester‐specific thresholds were applied according to gestational age at blood sampling, D‐dimer positivity was observed in 4 of 36 DVT cases and 1 of 36 controls, resulting in an overall sensitivity of 11.1% and specificity of 97.2%. These findings suggest that applying trimester‐specific pregnancy reference limits markedly reduces D‐dimer positivity and increases specificity compared with the conventional fixed threshold, but at the cost of substantially lower sensitivity in this cohort; therefore, this analysis should be interpreted as exploratory.
4. Discussion
In this study, we evaluated the association between complete blood count‐derived inflammatory indices and pregnancy‐associated DVT, as well as their exploratory discriminatory performance. Our findings indicate that NLR, SII, and D‐dimer levels are associated with the presence of DVT. All three parameters retained their statistical significance in the multivariate analyses even after adjusting for clinically meaningful variables such as body mass index, history of cesarean section, and presence of varicose veins. Furthermore, elastic net analysis revealed that NLR and SII were the main parameters containing discriminative information among the inflammatory indices. Furthermore, ROC analyses evaluating NLR and SII, both alone and in combination with D‐dimer, showed moderate discriminatory performance for the presence of DVT.
Current evidence suggest that inflammation and coagulation processes are closely related in the development of VTE [24, 25, 26, 27]. It has been reported that major cytokines and acute phase reactants, such as interleukin‐6, interleukin‐8, and C‐reactive protein (CRP), are significantly elevated at the time of diagnosis, while these parameters show a decreasing trend during antithrombotic treatment and clinical recovery [28]. This pattern suggests that the observed increase in biomarkers reflects the systemic inflammatory response accompanying thrombosis. Similarly, in biomarker panels including adhesion molecules reflecting endothelial dysfunction and cellular interactions, and metalloproteinases involved in matrix remodeling, P‐selectin, VCAM‐1, and various metalloproteinase levels have been shown to be higher in DVT patients compared to healthy controls [29, 30]. Furthermore, the parallel changes over time in D‐dimer and high‐sensitivity CRP levels in the acute phase following DVT support the notion that coagulation and inflammation processes are closely interlinked [28].
Against this pathophysiological background, interest in composite hematological indices, which can be easily obtained from routine complete blood counts and are accepted as reflecting systemic inflammatory status, has increased in recent years. Parameters such as the neutrophil/lymphocyte ratio (NLR), platelet/lymphocyte ratio (PLR), monocyte/lymphocyte ratio (MLR), SII, SIRI, and AISI have been increasing, suggesting that these parameters may be associated with the presence and clinical characteristics of various thromboembolic events, particularly DVT and pulmonary embolism. These indices are attracting attention as readily available and low‐cost biomarkers in the thrombotic process, where inflammation and coagulation are intertwined. Numerous studies in non‐pregnant populations have reported that inflammatory indices derived from complete blood counts may be associated with DVT [31, 32, 33, 34]. For example, Tort et al. reported that NLR, PLR, and SII levels were significantly higher in patients diagnosed with DVT compared to the control group, and ROC analyses yielded AUC values of 0.797 for NLR, 0.788 for PLR, and 0.861 for SII [23]. Similarly, in their study, Tural et al. found higher NLR (AUC: 0.814) and PLR (AUC: 0.621) values in patients with DVT and reported these parameters as independent variables in logistic regression analyses [35]. On the other hand, a comprehensive review published by Murariu‐Gligor et al. evaluated in detail the potential roles of inflammatory indices obtained from complete blood counts in the diagnosis and prognosis of VTE [36]. This review particularly emphasized that composite indices such as NLR and SII were higher in acute DVT and PE cases compared to healthy controls and could reflect the inflammatory burden associated with thrombosis. Furthermore, it was reported that these indices demonstrated moderate to good discriminatory performance in ROC analyses in some studies and that adding them to existing clinical risk scores could increase predictive power. However, it was underscored that the majority of the studies included in the review were cross‐sectional or retrospective in nature and that inflammatory indices reflect the systemic inflammatory response accompanying the acute thrombotic process rather than playing a causal role in thrombosis development. Therefore, it was stated that it would be more appropriate to consider these indices as complementary parameters supporting existing assessments rather than as independent diagnostic tools in the clinical decision‐making process.
The diagnosis and differential diagnosis of DVT and PE in the obstetric population is more complex than in non‐pregnant individuals, which has led to a relatively limited number of studies in this field in the literature. In Topkara and Çelen's study evaluating DVT in pregnant women, inflammatory indices such as NLR, PLR, SII, and SIRI did not show significant differences between cases diagnosed with DVT and the control group [37]. However, the definition of the control group in this study appears noteworthy in terms of interpreting the results. Although the Methods section states that 37 pregnant women diagnosed with DVT were compared with 37 healthy pregnant women, examination of the tables reveals that the analyses were performed between the 37 cases diagnosed with DVT and a group of 88 individuals who underwent Doppler ultrasonography due to suspected DVT but were excluded from the DVT diagnosis. It is thought that this group, clinically suspected of having DVT, may not represent a completely healthy population in terms of inflammatory response, and that this situation may have made it difficult to reveal possible differences between the groups in terms of inflammatory indices. This finding demonstrates that the definition of the control group in obstetric studies is critical for interpreting results, as inflammatory indices are parameters that reflect systemic inflammation.
On the other hand, while data on the evaluation of inflammatory indices in studies directly addressing lower extremity DVT during pregnancy are extremely limited, indirect evidence is available through other venous thrombotic conditions associated with pregnancy. In this context, in a study examining patients diagnosed with cerebral venous sinus thrombosis (CVST) during pregnancy and the postpartum period, Başaran et al. found that the SII and SIRI were significantly higher compared to the healthy control group and that these indices showed moderate discriminatory performance in ROC analyses [38]. Although CVST and lower extremity DVT are different pathologies in terms of anatomical location, clinical findings, and diagnostic approaches, it is known that both conditions develop on a common pathophysiological basis in pregnancy, such as increased hypercoagulability, venous stasis, and thrombo‐inflammatory mechanisms. Therefore, the high levels of inflammatory indices found in CVST cases support the notion that the systemic inflammatory response may be a common component in different clinical phenotypes of venous thrombosis during pregnancy. However, the fact that other inflammatory indices such as NLR and PLR were not evaluated in this study did not allow for a comparison of the relative contributions of these parameters.
Given this literature framework, the findings of our study are generally consistent with the existing evidence. The comparison of patients diagnosed with DVT during pregnancy with healthy pregnant women provided a meaningful assessment in terms of revealing the relationship between hematological indices reflecting the inflammatory response and the presence of thrombosis. Our findings suggest that some indices reflecting the inflammatory load change in association with the presence of DVT and that this relationship may also be valid in the obstetric population. In particular, the fact that NLR and SII remained associated with DVT even after adjusting for clinically meaningful factors suggests that these indices may reflect the interaction between systemic inflammation and thrombosis. Conversely, the loss of significance of different inflammatory indices in regression models suggests that these parameters may be largely interrelated and carry similar signals in terms of clinical information. This finding is consistent with previous studies highlighting the problem of multicollinearity among inflammatory indices. The elastic net‐based variable selection approach highlighting NLR and SII supports that these two indices contain relatively more discriminative information. When ROC analyses were evaluated, it was observed that models considering NLR and SII both alone and in combination with D‐dimer showed moderate performance in distinguishing the presence of DVT. This moderate discriminatory performance is biologically plausible in pregnancy, because physiological leukocytosis, immune adaptation, and gestational changes in coagulation and inflammatory markers can create substantial overlap between women with and without DVT. However, the lack of statistically significant superiority among different models suggests that inflammatory indices may be more appropriately considered as complementary criteria to clinical evaluation and existing laboratory findings rather than as independent diagnostic tools for DVT diagnosis in pregnancy. In this context, considering the limitations of D‐dimer, which is physiologically elevated and less specific during pregnancy, it appears more rational to evaluate inflammatory indices not as alternatives to D‐dimer, but as parameters that support and complement the clinical decision‐making process in thrombosis assessment.
The supplementary analyses provided further context for interpreting these findings. The gestational age‐stratified analysis showed a generally consistent direction of association for the main inflammatory indices, particularly NLR and SII, across the <28‐week and ≥28‐week strata, suggesting that the overall results were not driven solely by gestational age at sampling. In addition, the persistence of NLR, SII, SIRI, and AISI after false discovery rate correction supports the robustness of the inflammatory signal, although elastic net selection indicated that NLR and SII carried the most informative discriminatory contribution. In contrast, the trimester‐specific D‐dimer threshold analysis showed that applying pregnancy reference limits increased specificity but substantially reduced sensitivity in this cohort. This finding reinforces that gestational age‐adapted D‐dimer interpretation may reduce false‐positive classification, but should not be used in isolation to exclude pregnancy‐associated DVT.
From a clinical workflow perspective, the inflammatory indices evaluated in this study should not be interpreted as tools to rule in or rule out DVT in pregnancy. Their potential role, if validated, would be limited to adjunctive risk assessment in pregnant women who already have clinical suspicion of DVT or recognized thrombotic risk factors. In such patients, elevated NLR or SII may support the overall clinical impression of a thrombo‐inflammatory state, but these indices cannot replace structured clinical assessment, appropriate interpretation of D‐dimer, or confirmatory compression ultrasonography. Therefore, the role of these markers should be framed as exploratory and complementary within the existing diagnostic pathway rather than as independent diagnostic tests.
This study has several limitations. First, its retrospective and single‐center design limits causal inference and reduces the generalizability of the findings to different clinical settings. Although Firth penalized logistic regression was used to improve model stability in the setting of a limited sample size, statistical power remained restricted, particularly for subgroup and gestational age‐stratified analyses. Therefore, the sensitivity analyses should be interpreted as exploratory. Second, inflammatory indices were measured at a single time point and do not reflect longitudinal changes during pregnancy or dynamic changes after treatment. Third, although gestational age at sampling was matched between cases and controls, the exact clock time of blood sampling, fasting status, and some preanalytical conditions could not be uniformly verified because of the retrospective design. This lack of complete preanalytical standardization may have introduced measurement variability or detection bias. Fourth, the conventional D‐dimer threshold of >500 ng/mL has limited specificity during pregnancy, and the trimester‐specific D‐dimer thresholds used in the sensitivity analysis were based on published reference intervals rather than validated diagnostic cutoffs for pregnancy‐associated DVT. Fifth, the inflammatory indices evaluated in this study are composite parameters derived from routine complete blood counts, and more specific inflammatory or endothelial biomarkers, such as cytokine profiles, selectins, or markers of endothelial dysfunction, were not analyzed. Finally, although correction for multiple comparisons was applied, the findings remain hypothesis‐generating and require validation in larger prospective multicenter cohorts with standardized sampling protocols.
In conclusion, this study found that NLR and SII were associated with the presence of DVT during pregnancy and may reflect the thrombo‐inflammatory component of pregnancy‐associated venous thrombosis. Nevertheless, the retrospective design, limited sample size, moderate discriminatory performance, and potential variability in laboratory sampling restrict the clinical applicability of these findings. Therefore, these indices should be interpreted as exploratory and adjunctive markers rather than independent diagnostic tools. Their possible role should be limited to supporting clinical assessment in pregnant women with suspected or high‐risk DVT, and they should not replace established diagnostic approaches such as appropriate D‐dimer interpretation and compression ultrasonography. Prospective multicenter studies with standardized laboratory protocols, gestational age‐stratified analyses, and pregnancy‐specific D‐dimer interpretation are needed before these markers can be incorporated into routine clinical decision‐making.
Funding Statement
This research received no specific grant from any funding agency.
Conflicts of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have influenced to influence the work reported in this paper.
Supporting information
Supporting Information: Table S1: Gestational age‐stratified comparison of inflammatory indices between DVT and control groups. Table S2: False discovery rate‐corrected comparisons of inflammatory indices between DVT and control groups. Table S3: Gestational age‐specific D‐dimer threshold analysis using trimester‐specific pregnancy reference limits.
Acknowledgments
The authors has nothing to report.
Data Availability Statement
Due to hospital policies, patient data and study materials cannot be shared publicly. However, the data are available from the corresponding author upon reasonable request.
References
- 1. Bourjeily G., Paidas M., Khalil H., Rosene‐Montella K., and Rodger M., “Pulmonary Embolism in Pregnancy,” Lancet 375, no. 9713 (2010): 500–512, 10.1016/S0140-6736(09)60996-X. [DOI] [PubMed] [Google Scholar]
- 2. Morris J. M., Algert C. S., and Roberts C. L., “Incidence and Risk Factors for Pulmonary Embolism in the Postpartum Period,” Journal of Thrombosis and Haemostasis 8, no. 5 (2010): 998–1003, 10.1111/j.1538-7836.2010.03794.x. [DOI] [PubMed] [Google Scholar]
- 3. Bremme K. A., “Haemostatic Changes in Pregnancy,” Best Practice & Research Clinical Haematology 16, no. 2 (2003): 153–168, 10.1016/s1521-6926(03)00021-5. [DOI] [PubMed] [Google Scholar]
- 4. Greer I. A., “Thrombosis in Pregnancy: Updates in Diagnosis and Management,” Hematology 2012 (2012): 203–207, 10.1182/asheducation-2012.1.203. [DOI] [PubMed] [Google Scholar]
- 5. Krenitsky N., Friedman A. M., Yu K., et al., “Trends in Venous Thromboembolism and Associated Risk Factors during Delivery Hospitalizations From 2000 to 2018,” Obstetrics & Gynecology 139, no. 2 (2022): 223–234, 10.1097/AOG.0000000000004648. [DOI] [PubMed] [Google Scholar]
- 6. Sultan A. A., Tata L. J., West J., et al., “Risk Factors for First Venous Thromboembolism around Pregnancy: a Population‐based Cohort Study from the United Kingdom,” Blood 121, no. 19 (2013): 3953–3961, 10.1182/blood-2012-11-469551. [DOI] [PubMed] [Google Scholar]
- 7. Hwang H. G., Lee J. H., and Bang S. M., “Incidence of Pregnancy‐Associated Venous Thromboembolism: Second Nationwide Study,” Thrombosis and Haemostasis 123, no. 9 (2023): 904–910, 10.1055/a-2019-0607. [DOI] [PubMed] [Google Scholar]
- 8. Ernst D. M., Oporto J. I., Zuñiga P. A., Pereira J. I., Vera C. M., and Carvajal J. A., “Maternal and Perinatal Outcomes of a Venous Thromboembolism High‐Risk Cohort Using a Multidisciplinary Treatment Approach,” International Journal of Gynecology & Obstetrics 154, no. 3 (2021): 500–507, 10.1002/ijgo.13628. [DOI] [PubMed] [Google Scholar]
- 9. ACOG , “ACOG Practice Bulletin No. 196 Summary: Thromboembolism in Pregnancy,” Obstetrics & Gynecology 2018;132(1):243–248, 10.1097/AOG.0000000000002707. [DOI] [PubMed] [Google Scholar]
- 10. Siennicka A., Kłysz M., Chełstowski K., et al., “Reference Values of D‐Dimers and Fibrinogen in the Course of Physiological Pregnancy: the Potential Impact of Selected Risk Factors—A Pilot Study,” BioMed Research International 2020 (2020): 3192350, 10.1155/2020/3192350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kawaguchi S., Yamada T., Takeda M., et al., “Changes in D‐dimer Levels in Pregnant Women According to Gestational Week,” Pregnancy Hypertension: An International Journal of Women's Cardiovascular Health 3, no. 3 (2013): 172–177, 10.1016/j.preghy.2013.03.003. [DOI] [PubMed] [Google Scholar]
- 12. Makowska A., Treumann T., Venturini S., and Christ M., “Pulmonary Embolism in Pregnancy: a Review for Clinical Practitioners,” Journal of Clinical Medicine 13, no. 10 (2024): 2863, 10.3390/jcm13102863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Rabinovich A., Cohen J. M., Cushman M., and Kahn S. R., BioSOX Investigators , “Association between Inflammation Biomarkers, Anatomic Extent of Deep Venous Thrombosis, and Venous Symptoms after Deep Venous Thrombosis,” Journal of Vascular Surgery: Venous and Lymphatic Disorders 3, no. 4 (2015): 347–353.e1, 10.1016/j.jvsv.2015.04.005. [DOI] [PubMed] [Google Scholar]
- 14. Matos M. F., Lourenço D. M., Orikaza C. M., Bajerl J. A. H., Noguti M. A. E., and Morelli V. M., “The Role of IL‐6, IL‐8 and MCP‐1 and Their Promoter Polymorphisms IL‐6 ‐174GC, IL‐8 ‐251AT and MCP‐1 ‐2518AG in the Risk of Venous Thromboembolism: a Case‐Control Study,” Thrombosis Research 128, no. 3 (2011): 216–220, 10.1016/j.thromres.2011.04.016. [DOI] [PubMed] [Google Scholar]
- 15. Ozkan S., Firatligil F. B., Sucu S., et al., “Can Inflammatory Biomarkers Based on First Trimester Complete Blood Count Parameters Predict Placental Abruption ?: a Case‐control Study,” Journal of Reproductive Immunology 164 (2024): 104279, 10.1016/j.jri.2024.104279. [DOI] [PubMed] [Google Scholar]
- 16. Firatligil F. B., Sucu S. T., Tuncdemir S., et al., “Evaluation of Systemic Immune‐Inflammation Index for Predicting Late‐Onset Fetal Growth Restriction,” Archives of Gynecology and Obstetrics 310, no. 1 (2024): 433–439, 10.1007/s00404-024-07453-x. [DOI] [PubMed] [Google Scholar]
- 17. Dereli M. L., Savran Üçok B., Özkan S., et al., “The Importance of Blood‐Count‐Derived Inflammatory Markers in Predicting Methotrexate Success in Patients With Tubal Ectopic Pregnancy,” International Journal of Gynecology & Obstetrics 167, no. 2 (2024): 789–796, 10.1002/ijgo.15696. [DOI] [PubMed] [Google Scholar]
- 18. Dereli M. L., Sucu S., Kurt A., Kumcu E., Mat E., and Turhan Çağlar A., “The Role of Blood Count‐Derived Inflammatory Indices in the Short‐Term Prediction of Delivery in Women with Threatened Preterm Labor,” American Journal of Reproductive Immunology 94, no. 1 (2025): e70126, 10.1111/aji.70126. [DOI] [PubMed] [Google Scholar]
- 19. Cai X., Song S., Hu J., et al., Journal of Inflammation Research 16 (2023): 4821–4832, 10.2147/JIR.S433190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Shen D. I., Cai X., Hu J., et al., Journal of Inflammation Research 18 (2025): 1623–1638, 10.2147/JIR.S503648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Xu T., Song S., Zhu K. E., et al., “Systemic Inflammatory Response Index Improves Prognostic Predictive Value in Intensive Care Unit Patients with Sepsis,” Scientific Reports 15, no. 1 (2025): 1908, 10.1038/s41598-024-81860-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mutailifu S., Zhu Q., Wang M., Zhang D., Song S., and Li N., Journal of Inflammation Research 18 (2025): 6281–6292, 10.2147/JIR.S515010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Tort M., Sevil F. C., Sevil H., and Becit N., “Evaluation of Systemic Immune‐Inflammation Index in Acute Deep Vein Thrombosis: a Propensity‐Matched,” Journal of Vascular Surgery: Venous and Lymphatic Disorders 11, no. 5 (2023): 972–977.e1, 10.1016/j.jvsv.2023.02.008. [DOI] [PubMed] [Google Scholar]
- 24. Borgel D., Bianchini E., Lasne D., Pascreau T., and Saller F., “Inflammation in Deep Vein Thrombosis: a Therapeutic Target?,” Hematology 24, no. 1 (2019): 742–750, 10.1080/16078454.2019.1687144. [DOI] [PubMed] [Google Scholar]
- 25. Cocoi A. F., Pop D., Cocoi M., Serban A. M., and Vida‐Simiti L. A., “Involvement of Inflammatory Markers in Pathogenesis of Venous Thromboembolism,” Romanian Journal of Laboratory Medicine 25, no. 3 (2017): 227–236, 10.1515/rrlm-2017-0019. [DOI] [Google Scholar]
- 26. von Brühl M. L., Stark K., Steinhart A., et al., “Monocytes, Neutrophils, and Platelets Cooperate to Initiate and Propagate Venous Thrombosis in Mice in Vivo,” Journal of Experimental Medicine 209, no. 4 (2012): 819–835, 10.1084/jem.20112322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Branchford B. R. and Carpenter S. L., “The Role of Inflammation in Venous Thromboembolism,” Frontiers in Pediatrics 6 (2018): 142, 10.3389/fped.2018.00142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Roumen‐Klappe E. M., den Heijer M., van Uum S. H. M., van der Ven‐Jongekrijg J., van der Graaf F., and Wollersheim H., “Inflammatory Response in the Acute Phase of Deep Vein Thrombosis,” Journal of Vascular Surgery 35, no. 4 (2002): 701–706, 10.1067/mva.2002.121746. [DOI] [PubMed] [Google Scholar]
- 29. Gremmel T., Ay C., Seidinger D., Pabinger I., Panzer S., and Koppensteiner R., “Soluble p‐Selectin, D‐Dimer, and High‐Sensitivity C‐Reactive Protein after Acute Deep Vein Thrombosis of the Lower Limb,” Journal of Vascular Surgery 54, no. 6 Suppl (2011): 48S–55S, 10.1016/j.jvs.2011.05.097. [DOI] [PubMed] [Google Scholar]
- 30. Mosevoll K. A., Lindås R., Tvedt T. H. A., Bruserud Ø., and Reikvam H., “Altered Plasma Levels of Cytokines, Soluble Adhesion Molecules and Matrix Metalloproteases in Venous Thrombosis,” Thrombosis Research 136, no. 1 (2015): 30–39, 10.1016/j.thromres.2015.04.002. [DOI] [PubMed] [Google Scholar]
- 31. Hu C., Zhao B., Ye Q., Zou J., Li X., and Wu H., “The Diagnostic Value of the Neutrophil‐to‐Lymphocyte Ratio and Platelet‐to‐Lymphocyte Ratio for Deep Venous Thrombosis: a Systematic Review and Meta‐Analysis,” Clinical and Applied Thrombosis/Hemostasis 29 (2023): 10760296231187392, 10.1177/10760296231187392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Yılmaz M. and Gunesli A., “Can Hematological Parameters Predict the Diagnosis of Deep Venous Thrombosis?,” Ankara Medical Journal 20, no. 3 (2020): 653–662, 10.5505/amj.2020.65668. [DOI] [Google Scholar]
- 33. Akgul E., Bircan N., Yumun G., and Vural A. H., “Clinical Value of Neutrophil to Lymphocyte Ratio in the Region of Lower Extremity Deep Venous Thrombosis,” European Research Journal 0, no. 0 (2016): 00, 10.18621/eurj.2017.5000199818. [DOI] [Google Scholar]
- 34. Kahraman E., Cetin S., Cetin M., and Ulgen A., “Two Sides of the Coin: Coagulation and Inflammation in Deep Vein Thrombosis—a Prospective Study on D‐dimer and SIRI,” Frontiers in Medicine 12 (2025): 1604286, 10.3389/fmed.2025.1604286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Tural K., Kara F., Avcı S., and Erdoğdu H. İ., “Can Complete Blood Cell Count Parameters Predict Deep Vein Thrombosis?,” Acta Clinica Croatica 59, no. 4 (2020): 661–666, 10.20471/acc.2020.59.04.12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Murariu‐Gligor E. E., Mureșan S., and Cotoi O. S., “From Cell Interactions to Bedside Practice: Complete Blood Count‐Derived Biomarkers with Diagnostic and Prognostic Potential in Venous Thromboembolism,” Journal of Clinical Medicine 14, no. 1 (2025): 205, 10.3390/jcm14010205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Topkara S. and Çelen Ş., “Prediction of Deep Vein Thrombosis in Pregnant Women by Platelet Indices: a Retrospective Case‐Control Study,” Journal of Obstetrics and Gynaecology Research 51, no. 1 (2025): e16212, 10.1111/jog.16212. [DOI] [PubMed] [Google Scholar]
- 38. Başaran E., Tanacan A., Farisoğulları N., et al., “The Utility of the Systemic Immune‐Inflammation Index (SII) and the Systemic Immune‐Response Index (SIRI) in Predicting Pregnancy‐related Cerebral Venous Sinus Thrombosis: an Experience from a Tertiary Center,” Jinekoloji‐Obstetrik ve Neonatoloji Tıp Dergisi 22, no. 3 (2025): 255–261, 10.38136/jgon.1578867. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information: Table S1: Gestational age‐stratified comparison of inflammatory indices between DVT and control groups. Table S2: False discovery rate‐corrected comparisons of inflammatory indices between DVT and control groups. Table S3: Gestational age‐specific D‐dimer threshold analysis using trimester‐specific pregnancy reference limits.
Data Availability Statement
Due to hospital policies, patient data and study materials cannot be shared publicly. However, the data are available from the corresponding author upon reasonable request.