Introduction

Venous thromboembolism (VTE), comprised of deep vein thrombosis (DVT) and pulmonary embolism (PE), is a clotting disorder affecting the venous system of the body [1, 2]. VTE can have serious disabling health consequences and remains a leading cause of preventable hospital death. Previous studies suggest that 5%-10% of in-hospital deaths are due to PE [3, 4]. Additionally, recent data have shown that post-discharge VTE may account for as much as 75% of VTE occurrences, representing a large proportion of preventable postoperative morbidity and mortality [5–8]. Furthermore, with an estimated annual cost of $13.5-$27.2 billion in the United States for 375,000 to 425,000 newly diagnosed, medically treated incident cases, VTE poses a significant socioeconomic burden on the United States healthcare system [9–11]. Therefore, appropriate management of VTE is of paramount importance, contributing to patient safety and well-being, and cost-effective care [12].

Extended VTE prophylaxis (eVTEp) has been recommended for patients undergoing major abdominal or pelvic surgery, with a combination of both mechanical and pharmacologic prophylaxis showing the greatest reduction in VTE occurrence [13, 14]. Specifically, in women undergoing significant gynecologic surgery without prophylaxis, the risk of DVT varies from 17% to 40%. For those undergoing surgery specifically for gynecologic cancer, the risk of DVT is notably higher in the absence of thromboprophylaxis [15]. Currently, the administration of eVTEp depends upon each patient’s perceived VTE risk, which is based on preoperative and perioperative factors such as age, presence of cancer, and history of VTE. There are, however, no generally agreed-upon criteria for selecting patients for eVTEp [16–21].

While the prescription of prophylaxis ought to be determined by VTE risk factors rather than the social identities of patients, current literature suggests racial disparities may exist in post-operative care with regard to incidence of VTE [22–24]. Moreover, there is limited evidence regarding racial disparities in the prescription of post-discharge or extended prophylaxis. To determine if racial disparities exist in the administration of eVTEp after hysterectomy, we investigated the rate of eVTEp prescription rates with respect to race. The findings of this project have the potential to inform evidence-based health policies and produce targeted strategies aimed at reducing racial inequities in terms of access and receipt of risk-appropriate eVTEp.

Methods

Statistical analyses

We described the baseline sociodemographic and clinical characteristics and explored bivariate associations between the covariates and eVTEp administration, followed by another baseline comparison of unadjusted rates of VTE risk factors by race using Pearson’s chi-squared tests and student’s t-tests (Tables 1 and 2). VTE risk scores as defined by Pannucci et al. were calculated with the mean and standard deviation reported, respectively, for Black and White patients (Table 3) [20]. To account for the fact that patients were clustered by hospitals, we fit a mixed-effects logistic regression model with fixed effects of risk factors including race, and random hospital effects (Table 4). Adjusted odds ratios (aOR) and 95% confidence intervals (CI) were reported. Statistical analyses were conducted in R 3.6.3 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a p-value less than 0.05.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline comparison of sociodemographic and clinical characteristics by extended venous thromboembolism (VTE) prophylaxis administration.

https://doi.org/10.1371/journal.pone.0318433.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Baseline comparison of venous thromboembolism (VTE) risk factors by race.

https://doi.org/10.1371/journal.pone.0318433.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Summary of Pannucci score by race.

https://doi.org/10.1371/journal.pone.0318433.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Multivariable mixed effect logistic regression model of extended venous thromboembolism (VTE) prophylaxis administration.

https://doi.org/10.1371/journal.pone.0318433.t004

Results

From January 2016 to February 2020, the MSQC collected data on 24,513 White and Black patients who underwent hysterectomy. Patients were divided based on the presence (1,107) or absence (23,406) of eVTEp administration. Of the group that received eVTEp, 954 (86.18%) were White and 153 (13.82%) were Black.

Baseline characteristics of all patients by eVTEp prescription are displayed in Table 1. There was a higher frequency of eVTEp prescription in patients older than 55 years of age, BMI > = 30, history of diabetes, CHF, CAD, bleeding disorder, cancer, ASA status >3, dependent or unknown functional status, history of hypertension, administration of postoperative VTE prophylaxis, personal history of VTE, family history of VTE, surgical time >360 minutes, LOS >1 day, minimally invasive surgical approach, and government/Medicaid or other health insurance.

Table 2 compares the frequency of VTE risk factors by race. Black patients are noted to have a greater frequency of BMI > = 40, history of diabetes, hypertension, longer surgical time, and length of stay, as compared to their White counterparts. Comparatively in this cohort, White patients were noted to have a greater frequency of age > = 60 and personal history of cancer.

Based on the cohort characteristics in Table 2, a score was calculated based on Pannucci et al. and presented in Table 3. White patients had an average risk score of 4.20 (SD = 2.17) as compared to Black patients with an average risk score of 3.88 (SD = 1.79), with a p-value of <0.001 indicating a statistically significant difference between scores.

The results of multivariable logistic regression analyses are summarized in Table 4. In the mixed-effects model adjusting for risk factors across hospitals, Black patients had approximately 25% lower odds (aOR = 0.776, 95% CI = 0.615–0.979) of receiving eVTEp than the White patients. Significant associations were also revealed between lower occurrence of eVTEp prescription and smoking history (aOR = 0.795, 95% CI = 0.649–0.975), perioperative VTE prophylaxis (aOR = 0.577, 95% CI = 0.349–0.954), and invasive surgical approach (aOR = 0.512, 95% CI = 0.396–0.662). In contrast, higher odds of eVTEp prescription were significantly associated with age 55–74 (aOR = 1.305, 95% CI = 1.081–1.575), CAD (aOR = 4.481, 95% CI = 3.239–6.200), bleeding disorder (aOR = 12.806, 95% CI = 8.426–19.462), unknown functional status (aOR = 7.775, 95% CI = 4.279–14.128), cancer (aOR = 5.687, 95% CI = 4.700–6.881), personal history of DVT/ PE (aOR = 9.832, 95% CI = 7.892–12.250), longer surgical time (coefficient = 1.175, 95% CI = 1.093–1.264), LOS of 2–9 days (aOR = 1.582, 95% CI = 1.218–2.056 for 2–4 days and aOR = 2.287, 95% CI = 1.571–3.330 for 5–9 days) and other insurance/uninsured/self-pay (aOR = 1.864, 95% CI = 1.167–2.976).

Discussion

For most patients undergoing benign gynecologic surgery, mechanical prophylaxis is appropriate for prevention of VTE [25]. However, pharmacological VTE prophylaxis may be beneficial for many patients such as those requiring prolonged immobilization or for gynecologic oncology patients the risk of VTE may persist well beyond 4 weeks [26, 27]. In this study, we sought to identify the relationship between race and eVTEp prescription after hysterectomy surgery. We found that Black patients were prescribed eVTEp at lower rates as compared to their White counterparts, even after risk adjustment with multivariable logistic regression.

There are no broadly accepted guidelines for prescribing eVTEp, and the etiology of this discrepancy is likely multifactorial. Prior studies suggest Black Americans are at a higher risk for VTE as compared to their White counterparts due to a higher prevalence of diabetes, obesity, elevated factor VIII, and genetic polymorphisms in genes coding for prothrombin [23, 28–30]. Indeed, this study’s findings correspond with prior literature as the Black cohort was noted to have a significantly greater proportion of obesity, diabetes, and prior VTE history as compared to their White counterparts. However, among the White cohort, this study noted significantly increased proportions of patients over the age of 75 and with cancer, variables classically associated with increased VTE incidence. Within this study, the incidence of post-discharge VTE was low (0.26% for PE, 0.12% for DVT). While not statistically significant, the proportion of patients with post-operative VTE was higher in the Black cohort (0.37% for PE, 0.17% for DVT) as compared to the White cohort (0.23% for PE, 0.11% for DVT).

Variations in the prescription of eVTEp may be attributed to many factors including, but not limited to, risk scores, physician clinical judgement, regional practices, and hospital protocol. In regard to VTE risk, this study identified a significant difference in the Pannucci VTE risk score between White and Black patients, with the former cohort having a higher average risk score (4.20 vs 3.88). Though the White cohort had a statistically significantly higher average risk score, this does not translate to clinical significance. The Pannucci VTE risk score is on a whole-point scale, and with the cohorts’ averages rounded to the nearest whole number, there would be no difference in risk score–both the Black and White cohorts would have a risk score of 4. Further, the Pannucci VTE risk score is on a 17-point scale, even if there were a whole-point difference in score, the risk score would categorize them in the same risk category (not high-risk). Taken together, the lack of significant clinical difference in risk scores between the White and Black cohorts and the higher prevalence of certain risk factors in the Black cohort necessitates a change in eVTEp prescription practices.

The analysis boasts a large cohort of 24,513 female patients undergoing hysterectomy with comprehensive risk adjustment. Consequently, significant disparities were identified regarding the administration of eVTEp with sufficient statistical power. Due to its retrospective design, the findings are susceptible to potential coding errors, missing data, and data entry mistakes. Additionally, there was a lack of information including specific indications for eVTEp, the specific eVTEp regimens prescribed, and patient compliance with prescriptions. While there was information regarding post-operative VTE, post-discharge VTE events and outcomes were not available. Lastly, provider-related causes such as surgeon sex, surgeon education, surgeon experience, and surgeon race, potentially having an impact on the disparities of the administration of eVTEp, were not available in the data.

Conclusions

This large retrospective cohort study showed a lower rate of prescription of eVTEp in Black patients compared to White patients after certain gynecological operations despite no clinically significant difference in risk scores between the cohorts. This study identifies a unique difference in care regarding risk-appropriate prophylaxis treatment among patients undergoing gynecologic surgery by racial group. While risk scores may be employed by physicians, this study suggests a need to explore eVTEp prescription practices in greater detail to ensure safe and equitable care is provided to all patients regardless of race.

Acknowledgments

The authors are grateful to Ms. Wenjing Weng at the Michigan Surgical Quality Collaborative for making the data accessible for this study.