Background
Venous thromboembolism events, including deep venous thrombosis and pulmonary embolism are the most common cause of preventable deaths in hospitalized patients in the United States. Although the risk of venous thromboembolism events in benign gynecologic surgery is generally low, the potential for venous thromboembolism events in urogynecologic population is significant because most patients undergoing the pelvic organ prolapse surgery have increased surgical risk factors.
Objective
This study aimed to investigate the incidence and risk factors for venous thromboembolism events within 30 days after different routes of the pelvic organ prolapse surgery in a large cohort population using the American College of Surgeons-National Surgical Quality Improvement Program.
Study Design
This retrospective cohort study used Current Procedural Terminology codes to identify pelvic organ prolapse repairs with and without concurrent hysterectomy performed during 2011–2017 in the American College of Surgeons-National Surgical Quality Improvement Program database. Demographics, preoperative length of hospital stay, operative time, preoperative comorbidities, smoking status, American Society of Anesthesiologists classification system scores, along with other variables were collected. Postoperative 30-day complications, including readmission, reoperation, and mortality, were collected. The incidence rates of venous thromboembolism, as defined by American College of Surgeons-National Surgical Quality Improvement Program, were compared among different surgical routes. Descriptive statistics were used, and logistic regression was performed to identify associations.
Results
Among 91,480 pelvic organ prolapse surgeries identified, 63,108 were analyzed: 43,279 (68.6%) were performed vaginally, 16,518 (26.2%) laparoscopically, and 3311 (5.2%) abdominally. A total of 34,698 (55.0%) underwent a concurrent hysterectomy. Of 63,108 subjects, 133 developed venous thromboembolism within 30 days after surgery (0.21%; 95% confidence interval, 0.18–0.25; P <.0001). More than half (60%) of venous thromboembolism events occurred within 10 days after surgery. For all surgical routes, older age ( P <.041), higher body mass index ( P =.002), race or ethnicity ( P =.04), longer operating time ( P <.0001), inpatient status ( P <.0001), American Society of Anesthesiologists 3 or 4 ( P <.0001 ) , having preoperative renal failure ( P =.001), and chronic steroid use ( P =.02) were significantly associated with venous thromboembolism. In addition, in the vaginal pelvic organ prolapse repair group, concurrent hysterectomy ( P =.03) and preoperative dyspnea ( P =.01) were associated with development of venous thromboembolism. In the abdominal pelvic organ prolapse repair, concurrent hysterectomy ( P =.005) and hypertension requiring medication ( P =.04) were also independently associated with venous thromboembolism development ( Table 1 ). The incidence of venous thromboembolism was highest in abdominal repairs (0.72%), followed by laparoscopic repairs (0.25%) and vaginal repairs (0.16%). After adjusting for confounders, abdominal compared with vaginal approach (adjusted odds ratio, 3.27; 95% confidence interval, 1.93–5.41; P <.0001), longer operative time (adjusted odds ratio, 1.005; 95% confidence interval, 1.003–1.006; P <.0001), older age (adjusted odds ratio, 1.020; 95% confidence interval, 1.00–1.037; P =.015), greater body mass index (adjusted odds ratio, 1.04; 95% confidence interval, 1.01–1.07; P =.0006), American Society of Anesthesiologists 3 or 4 (adjusted odds ratio, 1.55; 95% confidence interval, 1.03–2.31; P =.03), and preoperative renal failure (adjusted odds ratio, 8.87; 95% confidence interval, 1.16–44.15; P =.04) remained significantly associated with developing venous thromboembolism. Neither laparoscopic repair (compared with vaginal repair) nor concurrent procedures (hysterectomy, antiincontinence procedure, vaginal mesh insertion) were found to be significantly associated with the development of venous thromboembolism. The abdominal pelvic organ prolapse repairs were associated with an increased hazard of venous thromboembolism (hazard ratio, 3.27; 95% confidence interval, 1.96–5.45; P <.0001). Venous thromboembolism development was associated with 30-day mortality, readmission, and reoperation (all P <.0001).
Conclusion
The overall incidence of venous thromboembolism after pelvic organ prolapse repairs based on a recent, large cohort database was very low, confirming the finding in previous smaller cohort studies. The highest venous thromboembolism risk was associated with abdominal route, and more than 60% of venous thromboembolism events occurred within 10 days after surgery. Thus, focus should be placed on risk-reducing strategies in the immediate postoperative period, with greater emphasis on patients undergoing abdominal surgery.
Venous thromboembolism (VTE) events, including deep venous thrombosis (DVT) and pulmonary embolism (PE) are the most common cause of preventable deaths in hospitalized patients in the United States. The risk of postoperative VTE events in patients undergoing major gynecologic surgery ranges from 15% to 40% in the absence of thromboprophylaxis.
Why was the study conducted?
This study aimed to investigate the incidence and risk factors for venous thromboembolism (VTE) events within 30 days after different routes of prolapse surgeries, using the recent American College of Surgeons-National Surgical Quality Improvement Program database.
Key findings
Of 63,108 prolapse surgeries, the incidence of postoperative VTE events was very low (0.21%) and differed by surgical approach. In addition, VTE was associated with abdominal compared with vaginal route, longer operative time, older age, and other medical comorbidities. VTE development within 30 days after prolapse repairs was a significant predictor of higher 30-day mortality, readmission, and reoperation. Moreover, more than 60% of VTE events occurred within 10 days after surgery.
What does this add to what is known?
Abdominal route of prolapse repairs was associated with the highest VTE risk and >60% of VTE occurred within 10 days after surgery. Concurrent procedures (hysterectomy, antiincontinence procedure, vaginal mesh insertion) were not found to be significantly associated with the development of VTE. Our larger sample size with more recent outcome data provides more accurate estimate for the incidence of postoperative VTE.
| Variable | Vaginal (n=43,279) | Laparoscopic (n=16,518) | Abdominal (n=3311) | P value |
|---|---|---|---|---|
| Demographics | ||||
| Age, mean (SD), y | 61.33 (13.15) | 56.11 (12.25) | 57.88 (11.65) | <.0001 |
| BMI, mean (SD), kg/m 2 | 28.45 (6.10) | 28.50 (6.13) | 28.36 (5.98) | .39 |
| BMI≥30, n (%), kg/m 2 | 14,612 (33.8) | 5554 (33.6) | 1094 (33.0) | .68 |
| Race or ethnicity, n (%) | <.0001 | |||
| White non-Hispanic | 26,980 (62.3) | 12,308 (74.5) | 2164 (65.4) | |
| Hispanic | 4043 (9.3) | 1336 (8.1) | 362 (10.9) | |
| Black | 1672 (3.9) | 956 (5.8) | 166 (5.0) | |
| Asian | 1316 (3.0) | 449 (2.7) | 65 (2.0) | |
| Unknown or other | 9268 (21.4) | 1469 (8.9) | 554 (16.7) | |
| Clinical variables | ||||
| Smoker, n (%) | 3972 (9.2) | 1780 (10.8) | 340 (10.3) | <.0001 |
| Concurrent hysterectomy, n (%) | 19,244 (44.5) | 13,311 (80.6) | 2143 (64.7) | <.0001 |
| Concurrent antiincontinence procedure, n (%) | 15,899 (36.7) | 5443 (33.0) | 1062 (32.1) | <.0001 |
| Concurrent vaginal mesh insertion (%) | 3797 (8.8) | 722 (4.4) | 203 (6.1) | <.0001 |
| Total operating time, mean (SD), min | 103.73 (60.78) | 177.05 (80.54) | 169.26 (77.57) | <.0001 |
| Preoperative LOS ≥1 d, n (%) | 338 (0.8) | 68 (0.4) | 26 (0.8) | <.0001 |
| Inpatient, n (%) | 17,887 (41.3) | 5244 (31.7) | 2938 (88.7) | <.0001 |
| ASA, n (%) a | <.0001 | |||
| Class 1 or 2 | 32,122 (74.2) | 13,109 (79.4) | 2583 (78.0) | |
| Class 3 or 4 | 11,109 (25.7) | 3394 (20.5) | 725 (21.9) | |
| Unknown | 48 (0.1) | 15 (0.1) | 3 (0.1) | |
| Comorbidities | ||||
| Diabetes, n (%) | <.0001 | |||
| Noninsulin dependent | 4110 (9.5) | 1137 (6.9) | 292 (8.8) | |
| Insulin dependent | 904 (2.1) | 250 (1.5) | 55 (1.7) | |
| Respiratory morbidities | ||||
| Dyspnea, n (%) b | 1447 (3.3) | 450 (2.7) | 103 (3.1) | .0006 |
| COPD, n (%) | 987 (2.3) | 198 (1.2) | 80 (2.4) | <.0001 |
| Preoperative ventilator dependent, n (%) | 2 (0.0) | 0 (0.0) | 0 (0.0) | .63 |
| Cardiac morbidities | ||||
| HTN, n (%) | 18,174 (42.0) | 5472 (33.1) | 1260 (38.1) | <.0001 |
| CHF, n (%) | 42 (0.1) | 6 (0.0) | 4 (0.1) | .051 |
| Bleeding disorders, n (%) | 400 (0.9) | 109 (0.7) | 18 (0.5) | .001 |
| Renal failure, n (%) c | 29 (0.1) | 10 (0.1) | 2 (0.1) | .96 |
| Preoperative steroid use, n (%) | 862 (2.0) | 266 (1.6) | 49 (1.5) | .002 |
| Preoperative weight loss, n (%) | 40 (0.1) | 6 (0.0) | 3 (0.1) | .09 |
| Preoperative transfusion, n (%) | 32 (0.1) | 10 (0.1) | 3 (0.1) | .79 |
a Sixty-six charts missing information regarding the ASA class
b Dyspnea at rest or moderate exertion
Pelvic organ prolapse (POP) affects millions of women; approximately 200,000 inpatient surgical procedures for prolapse are performed annually in the United States. The lifetime risk of a woman undergoing surgery for POP is estimated to be 13 %. The observed incidence of VTE among patients undergoing urogynecologic surgery has been shown to be consistent with that of patients undergoing benign gynecologic surgery (0.1%–2.2%).
Previous studies , , looking at a smaller cohort have found that risk factors associated with increased VTE events in POP repairs include obesity, increased length of stay, American Society of Anesthesiologists (ASA) classification score of 3 or higher, history of cancer, history of hormone replacement therapy (period of 3 months before surgery), and history of previous VTE.
The risk of VTE in benign gynecologic surgery is generally low ; however, the potential for VTE in those undergoing urogynecologic surgery is significant because most patients undergoing POP surgery are elderly, and advanced age is associated with increase in surgical complications. ,
To the best of our knowledge, only a few studies have investigated incidence or risk factors for developing VTE after POP repair. , , Among those previous studies, 2 studies described risk factors for developing VTE after POP repairs, with respect to the route of surgery, patient comorbidities, concomitant surgeries, and other perioperative complications, using relatively older and smaller cohorts from the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database. ,
The primary aim of this retrospective cohort study was to investigate the incidence of VTE events defined by the presence of either PE or DVT requiring therapy in the first 30 days after different routes of POP repairs, using the ACS-NSQIP database with a large national cohort, including newer data on this complication in more recent years. The secondary aim of this study was to investigate the risk factors for developing VTE after different routes of POP repairs.
Materials and Methods
Data collection
This study used data from the ACS-NSQIP, a peer-controlled database comprising data collected prospectively by trained medical chart reviewers at participating hospitals nationally. NSQIP clinical reviewers sample a random selection of cases from their hospital caseload.
All major surgeries are sampled as determined by the Current Procedural Terminology (CPT) code. Excluded cases include minor cases, patients younger than 18 years, ASA class 6, trauma cases, transplant cases, and those involving hyperthermic intraperitoneal chemotherapy. Cases are sampled through a systematic 8-day cycle scheduling scheme, which ensures that every eligible case performed at the hospital has an equal chance of being selected for review. For each case, data are collected for 30 days postoperatively using a variety of methods, including medical chart reviews and letters or phone calls to patients if necessary. The ACS-NSQIP uses several programs to assure quality of their data, including web-based training modules and annual recertification programs for the designated chart reviewers and interrater reliability auditing, which involves the review of both randomly selected medical charts and those at increased risk of reporting errors, and which to date has revealed an overall disagreement rate of 2.3%. As of 2017, there are 708 participating institutions; this number has increased yearly from 315 in 2011. ,
The database was queried to identify cases of patients who underwent POP procedures from 2011 to 2017 using the specific CPT codes ( Appendix A ). Data from this period were specifically requested because (1) the authors wanted to compare the newly available database with the findings of a previous study by Mueller et al, which used the ACS-NSQIP database from 2006 to 2010 to investigate VTE in reconstructive pelvic surgery; and (2) the authors desired to analyze the most recent data available to represent the most current US population. At the time of this study, the NSQIP database was only available up to the year of 2017.
The surgical approaches of the POP procedures were grouped as vaginal, laparoscopic, or abdominal. In addition, the database was queried for hysterectomies using the specific CPT codes ( Appendix A ). Hysterectomies in which a concurrent POP repair was identified in the “other” and “concurrent” procedure fields were also included in the study.
Patients of male sex or those who received a surgical procedure specific to males; patients who had a postoperative diagnosis or a concurrent procedure related to malignancy; patients who had a preoperative diagnosis of sepsis, shock, wound infection, or contaminated or dirty wound; and patients who underwent antiincontinence procedure only without POP repair were excluded from the study ( Appendix B ). The detailed inclusion and exclusion criteria are presented in Figure 1 .
The cases were grouped by different routes of surgery (vaginal vs abdominal vs laparoscopic or robotic). It must be noted that robotic-assisted laparoscopic hysterectomy cannot be distinguished from laparoscopic hysterectomy, using CPT codes. Given that prolapse repairs may include multiple routes during a single procedure, the investigators had predetermined the definitions of each route for POP repairs. The reason why the route of hysterectomy was prioritized when assigning the route of POP repair was because the investigators believed that hysterectomy is the most critical portion of POP repair when it is performed at the time of POP repair surgery. The current database showed 77.8% agreement between the route of POP repair and the route of hysterectomy. The investigators had predetermined the definitions of each route for POP repairs as follows: (1) if the case included hysterectomy, the route of surgery was defined as the route of hysterectomy; (2) if the case did not involve a hysterectomy but had an apical suspension procedure, then the route of surgery was defined as the route of apical suspension; and (3) in all other cases, the route of surgery was defined as the route of the POP repair that was identified for that case. Patient demographic information, preoperative morbidities, preoperative length of hospital stay, smoking status, ASA classification score, intraoperative factors such as operative time, and postoperative inpatients vs outpatient status were recorded for each case. Furthermore, 30-day postoperative readmission, reoperation, and mortality rate were collected for secondary outcome measures.
The institutional review board (IRB) at our institution reviewed this study and deemed it exempt from formal review (IRB-18-2529) because all data previously existed and were deidentified .
Statistical analysis
Descriptive statistics (mean, standard deviation [SD]; SD for continuous variables; n [%] for categorical variables) were calculated to characterize the study groups according to the demographic and preclinical variables. Univariate analyses (chi-square test for categorical variables and Student’s t -test for continuous variables) were then performed to examine the association between demographics, preclinical variables, and route of surgery with the occurrence of primary and secondary outcomes.
A multivariate logistic regression was fit to assess the independent association between demographic and clinical covariates and VTE occurrence. Predictors with P<.30 in univariate analyses were entered a multivariate logistic regression model. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were calculated to quantify the relative change in odds of VTE occurrence associated with each covariate independent of other risk factors.
In addition, time-to-VTE events were characterized using descriptive statistics. Kaplan-Meier plots and a log-rank test were performed to compare time-to-VTE between different surgical routes. A Cox proportional hazards model was then developed to describe the independent association between demographic and clinical covariate and incidence of VTE. As with the multivariate model, predictors with P<.30 in univariate analyses were entered in the Cox proportional hazards model. Hazard ratios (HRs) and 95% CIs were calculated from the model to quantify the relative change in incidence of VTE events associated with each covariate. All statistical analyses were 2-tailed; P<.05 was considered to be statistically significant. All analyses were performed using SAS version 9.4 (Cary, NC) and R version 3.6.1 (Vienna, Austria).
Results
A total of 91,480 POP repair surgeries were identified through the NSQIP database from 2011 to 2017. Of these cases, 248 involved a male subject or a male-specific procedure, 4463 involved a procedure or postoperative diagnosis related to malignancy, 837 had a preoperative diagnosis of sepsis or shock or infected or contaminated wound, and 22,915 had an antiincontinence procedure without prolapse repair. Some of these cases had met more than 1 exclusion criteria: the crossovers include (1) 9 cases had both a male-related code and a malignancy, (2) 3 cases had both a male-related code and preoperative infection; and (3) 79 cases had both a malignancy and a preoperative infection. Of note, none of the cases met all 3 exclusion criteria. After applying the exclusion criteria, 63,108 cases were included in this study for analysis. Of these procedures, 43,279 (68.6%) were performed vaginally, 16,518 (26.2%) laparoscopically, and 3311 (5.2%) abdominally. A total of 34,698 cases (55.0%) underwent a concurrent hysterectomy ( Figure 1 ).
Cohorts had similar demographics in terms of body mass index (BMI), preoperative ventilator use and transfusion, and having preoperative congestive heart failure (CHF), renal failure, and chronic weight loss. Compared with other cohorts, patients who underwent vaginal POP repair were older; had more concomitant antiincontinence procedures; and comorbidities such as diabetes, ASA 3 or 4, hypertension (all P<. 0001), and dyspnea ( P =.0006), bleeding disorder ( P =.001), and chronic steroid use ( P =.002). Patients who underwent abdominal POP repair were more likely Hispanic; had more inpatient status and chronic obstructive pulmonary disease (COPD) (all P <.0001). Patient who underwent laparoscopic POP repair were more likely white and smoker, had concurrent hysterectomy, and longer operating time (all P <.0001) ( Table 1 ) .
Regardless of the route of surgery, a total of 133 cases (0.21%; 95% CI, 0.18–0.25; P <.0001) developed VTE in the first 30 days after POP repairs. Univariate analyses for association between demographic or clinical covariates and VTE, grouped by different routes of POP surgery showed that for all surgical routes, older age (62.10±14.04 vs 59.78±13.05 years; P =.04), higher BMI (30.08±7.84 vs 28.46±6.10 kg/m 2 ; P =.002), race or ethnicity (white, 0.24%; black, 0.36%; Hispanic, 0.10%; Asian, 0.16%; other or unknown, 0.14%; P =.04), longer operating time (172.48±102.61 vs 126.27±75.18 minutes, P <.0001), inpatient status (0.28% vs 0.16%; P <.0001), higher ASA classification (class 3 or 4: 0.33% vs 0.17%; P <.0001), having preoperative renal failure (2.44% vs 0.63%; P =.001), and chronic steroid use (0.51% vs 0.21%; P =.02) were significantly associated with the development of VTE. In addition, in the vaginal POP repair group, concurrent hysterectomy ( P =.03) and preoperative dyspnea ( P =.01) were associated with the development of VTE; in the abdominal POP repair group, concurrent hysterectomy ( P =.005) and hypertension requiring medication ( P =.04) were also independently associated with the development of VTE ( Table 2 ) .
| Route Variable | Vaginal (n=43,279) | Laparoscopic (n=16,518) | Abdominal (n=3311) | Total (N=63,108) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No VTE (n=43,211) | VTE (n=68) | P value | No VTE (n=16,477) | VTE (n=41) | P value | No VTE (n=3287) | VTE (n=24) | P value | No VTE (n=62,975) | VTE (n=133) | P value | |
| Age, mean (SD), y | 61.33 (13.15) | 65.31 (13.72) | .01 | 56.11 (12.24) | 56.34 (14.10) | .90 | 57.84 (11.65) | 62.83 (12.10) | .04 | 59.78 (13.05) | 62.10 (14.04) | .04 |
| BMI, mean (SD), kg/m 2 | 28.45 (6.10) | 29.63 (7.26) | .11 | 28.50 (6.12) | 32.05 (9.73) | .0002 | 28.36 (5.98) | 27.96 (4.74) | .75 | 28.46 (6.10) | 30.08 (7.84) | .002 |
| BMI≥30, n (%), kg/m 2 | 14,587 (33.8) | 25 (36.8) | .60 | 5531 (33.6) | 23 (56.1) | .002 | 1087 (33.1) | 7 (29.2) | .69 | 21,205 (33.7) | 55 (41.4) | .06 |
| Race or ethnicity, n (%) | .10 | .75 | .36 | .04 | ||||||||
| White non-Hispanic | 26,933 (62.3) | 47 (69.1) | 12,275 (74.5) | 33 (80.5) | 2146 (65.3) | 18 (75.0) | 41,354 (65.7) | 98 (73.7) | ||||
| Hispanic | 4039 (9.3) | 4 (5.9) | 1334 (8.1) | 2 (4.9) | 362 (11.0) | 0 (0.0) | 5735 (9.1) | 6 (4.5) | ||||
| Black | 1666 (3.9) | 6 (8.8) | 954 (5.8) | 2 (4.9) | 164 (5.0) | 2 (8.3) | 2784 (4.4) | 10 (7.5) | ||||
| Asian | 1314 (3.0) | 2 (2.9) | 449 (2.7) | 0 (0.0) | 64 (1.9) | 1 (4.2) | 1827 (2.9) | 3 (2.3) | ||||
| Unknown or other | 9259 (21.4) | 9 (13.2) | 1465 (8.9) | 4 (9.8) | 551 (16.8) | 3 (12.5) | 11,275 (17.9) | 16 (12.0) | ||||
| Clinical variables | ||||||||||||
| Smoker, n (%) | 3969 (9.2) | 3 (4.4) | .17 | 1773 (10.8) | 7 (17.1) | .19 | 335 (10.2) | 5 (20.8) | .09 | 6077 (9.6) | 15 (11.3) | .53 |
| Concurrent hysterectomy, n (%) | 19,205 (44.4) | 39 (57.4) | .03 | 13,277 (80.6) | 34 (82.9) | .70 | 2134 (64.9) | 9 (37.5) | .005 | 34,616 (55.0) | 82 (61.7) | .12 |
| Concurrent antiincontinence procedure, n (%) | 15,870 (36.7) | 29 (42.6) | .31 | 5429 (32.9) | 14 (34.1) | .87 | 1058 (32.2) | 4 (16.7) | .10 | 22,357 (35.5) | 47 (35.3) | .97 |
| Concurrent vaginal mesh insertion, n (%) | 3794 (8.8) | 3 (4.4) | .2033 | 721 (4.4) | 1 (2.4) | .5446 | 201 (6.1) | 2 (8.3) | .6517 | 4716 (7.5) | 6 (4.5) | .19 |
| Total operating time, mean (SD), min | 103.67 (60.71) | 141.21 (89.24) | <.0001 | 176.97 (80.42) | 209.88 (114.60) | .009 | 169.06 (77.44) | 197.21 (91.48) | .08 | 126.27 (75.18) | 172.48 (102.61) | <.0001 |
| Preoperative LOS ≥1 d, n (%) | 337 (0.8) | 1 (1.5) | .52 | 68 (0.4) | 0 (0.0) | .68 | 25 (0.8) | 1 (4.2) | .06 | 430 (0.7) | 2 (1.5) | .25 |
| Inpatient, n (%) | 17,856 (41.3) | 31 (45.6) | .48 | 5224 (31.7) | 20 (48.8) | .02 | 2917 (88.7) | 21 (87.5) | .85 | 25,997 (41.3) | 72 (54.1) | .003 |
| ASA, n (%) a | .01 | <.0001 | .39 | .0001 | ||||||||
| Class 1 or 2 | 32,082 (74.2) | 40 (58.8) | 13,083 (79.4) | 26 (63.4) | 2567 (78.1) | 16 (66.7) | 47,732 (75.8) | 82 (61.7) | ||||
| Class 3 or 4 | 11,081 (25.6) | 28 (41.2) | 3380 (20.5) | 14 (34.1) | 717 (21.8) | 8 (33.3) | 15,178 (24.1) | 50 (37.6) | ||||
| Unknown | 48 (0.1) | 0 (0.0) | 14 (0.1) | 1 (2.4) | 3 (0.1) | 0 (0.0) | 65 (0.1) | 1 (0.8) | ||||
| Comorbidities | ||||||||||||
| Diabetes, n (%) | .54 | .08 | .81 | .65 | ||||||||
| Noninsulin dependent | 4106 (9.5) | 4 (5.9) | 1132 (6.9) | 5 (12.2) | 290 (8.8) | 2 (8.3) | 5528 (8.8) | 11 (8.3) | ||||
| Insulin dependent | 902 (2.1) | 2 (2.9) | 248 (1.5) | 2 (4.9) | 55 (1.7) | 0 (0.0) | 1205 (1.9) | 4 (3.0) | ||||
| Respiratory morbidities | ||||||||||||
| Dyspnea, n (%) b | 1441 (3.3) | 6 (8.8) | .01 | 448 (2.7) | 2 (4.9) | .40 | 102 (3.1) | 1 (4.2) | .76 | 1991 (3.2) | 9 (6.8) | .02 |
| COPD, n (%) | 985 (2.3) | 2 (2.9) | .72 | 197 (1.2) | 1 (2.4) | .47 | 79 (2.4) | 1 (4.2) | .58 | 1261 (2.0) | 4 (3.0) | .41 |
| Preoperative ventilator dependent, n (%) | 2 (0.0) | 0 (0.0) | .96 | 0 (0.0) | 0 (0.0) | N/A | 0 (0.0) | 0 (0.0) | N/A | 2 (0.0) | 0 (0.0) | .95 |
| Cardiac morbidities | ||||||||||||
| HTN, n (%) | 18,145 (42.0) | 29 (42.6) | .91 | 5460 (33.1) | 12 (29.3) | .60 | 1246 (37.9) | 14 (58.3) | .04 | 24,851 (39.5) | 55 (41.4) | .66 |
| CHF, n (%) | 42 (0.1) | 0 (0.0) | .80 | 6 (0.0) | 0 (0.0) | .90 | 4 (0.1) | 0 (0.0) | .86 | 52 (0.1) | 0 (0.0) | .74 |
| Bleeding disorder, n (%) | 400 (0.9) | 0 (0.0) | .43 | 108 (0.7) | 1 (2.4) | .16 | 18 (0.5) | 0 (0.0) | .72 | 526 (0.8) | 1 (0.8) | .92 |
| Renal failure, n (%) c | 28 (0.1) | 1 (1.5) | <.0001 | 10 (0.1) | 0 (0.0) | .87 | 2 (0.1) | 0 (0.0) | .90 | 40 (0.1) | 1 (0.8) | .002 |
| Preoperative steroid use, n (%) | 860 (2.0) | 2 (2.9) | .57 | 262 (1.6) | 4 (9.8) | <.0001 | 49 (1.5) | 0 (0.0) | .55 | 1171 (1.9) | 6 (4.5) | .02 |
| Preoperative weight loss, n (%) | 40 (0.1) | 0 (0.0) | .80 | 6 (0.0) | 0 (0.0) | .90 | 3 (0.1) | 0 (0.0) | .88 | 49 (0.1) | 0 (0.0) | .75 |
| Preoperative blood transfusion, n (%) | 32 (0.1) | 0 (0.0) | .82 | 10 (0.1) | 0 (0.0) | .87 | 3 (0.1) | 0 (0.0) | .88 | 45 (0.1) | 0 (0.0) | .76 |
a Sixty-six charts missing information regarding the ASA status
b Dyspnea at rest or moderate exertion
Rates of VTE were highest in abdominal surgical route (0.72%; 24/3311), followed by laparoscopic (0.25%; 41/16,518), and lowest in the vaginal repairs (0.16%; 68/43,279) ( Table 3 ; Figure 2 ). Among 133 VTE cases, 12.8% (17/133) had developed both DVT and PE. There was statistically significant difference in VTE rates between different routes of repair: abdominal vs vaginal ( P <.0001), laparoscopic vs vaginal ( P =.03), and abdominal vs laparoscopic ( P <.0001).
| Surgical route | DVT or PE | DVT and PE | ||||
|---|---|---|---|---|---|---|
| N | Row (%) | Col (%) | N | Row (%) | Col (%) | |
| VAG | 68 | 0.16 | 51.13 | 9 | 0.02 | 52.94 |
| LSC | 41 | 0.25 | 30.83 | 5 | 0.03 | 29.41 |
| ABD | 24 | 0.72 | 18.05 | 3 | 0.09 | 17.65 |
| P values | ||||||
| Overall | <.0001 | .06 | ||||
| VAG vs LSC | .03 | .71 | ||||
| VAG vs ABD | <.0001 | .06 | ||||
| LSC vs ABD | <.0001 | .27 | ||||
