Abstract
Background
Total hip arthroplasty (THA) and total knee arthroplasty (TKA) procedure volumes are increasing. Venous thromboembolism (VTE) remains a significant complication, with incidence rates between 0.45% and 5.30%. Enhanced rapid-recovery pathways and chemoprophylaxis evolution may correlate with decreased VTE events over time. This study analyzes recent trends of VTE after THA and TKA.
Methods
Adults undergoing THA or TKA between 2009 and 2022 were identified from the National Quality Surgical Improvement Program database using Current Procedural Terminology codes. VTE was defined as the occurrence of a deep vein thrombosis (DVT) and/or pulmonary embolism (PE) event in the same patient. The 30-day incidence data of VTE, DVT, and PE were trended over time. Multivariate regression analyses estimated the adjusted risk of events by year of surgery relative to 2009 and identified associated risk factors.
Results
A total of 382,515 THAs and 593,060 TKAs were included with 5713 DVTs and 3807 PEs observed. Trends of 30-day VTE decreased over the study period in THA (0.8%-0.5%, P < .001) and TKA (1.8%-0.9%, P < .001). Significant decreasing trends were observed for both DVT and PE following TKA (both P < .001) and for DVT (P < .001) following THA. Adjusted regression showed significantly lower odds of 30-day DVT (odds ratio 0.68, P < .01) and PE (odds ratio 0.59, P < .01) after TKA in 2022 compared to 2009.
Conclusions
The 30-day VTE incidence following THA and TKA has significantly decreased since 2009. Both DVT and PE have decreased in the TKA population, likely due to improved preoperative patient optimization and enhanced recovery pathways, despite shifts toward more selective antiplatelet chemoprophylaxis.
Keywords: Venous thromboembolism, Total hip arthroplasty, Total knee arthroplasty, Complications, Pulmonary embolism, Deep vein thrombosis
Introduction
Total hip arthroplasty (THA) and total knee arthroplasty (TKA) are among the most common and rapidly increasing surgical procedures in the United States [1,2]. As these procedures continue to rise in frequency, venous thromboembolism (VTE) remains one of the most serious potential complications, including both deep vein thrombosis (DVT) and pulmonary embolism (PE). With an overall 90-day mortality of 9.8% associated with VTE events [3], understanding current incidence rates and temporal trends is essential for improving patient care. The rates of symptomatic VTE after total joint arthroplasty (TJA) have been reported to be between 0.45% and 5.3% [[3], [4], [5], [6], [7]].
Although the relative number of patients who experience a symptomatic VTE is low, the morbidity and relative costs to treat these complications remain high. In addition to high readmission rates, patients are prone to longer lengths of hospitalization; especially when VTE occurs in the immediate postoperative period [8,9]. Moreover, the estimated financial burden to the healthcare system ranges from $16,644 to over $33,000 annually for a single postoperative VTE [[8], [9], [10]].
Prophylaxis against VTE following THA and TKA has evolved significantly over the past decade. While these procedures continue to increase in frequency, Clinical Practice Guidelines, early mobilization protocols, and chemoprophylaxis strategies have undergone substantial changes, potentially influencing VTE rates [5].
Prior analyses examined VTE rates and trends dating back to the original American College of Chest Physicians (ACCP) and American Academy of Orthopaedic Surgeons (AAOS) recommendations which included chemoprophylaxis against VTE after TJA [3,5,11,12]. However, most existing reports either focus on one of TKA and THA or evaluate data that precede full implementation of contemporary VTE prophylaxis protocols and the widespread shift toward outpatient arthroplasty. The purpose of this study was to examine the most recent annual trends in VTE rates following both THA and TKA in the United States using the National Quality Surgical Improvement Program (NSQIP) database, providing an updated assessment that reflects current clinical practice patterns and modern prophylaxis approaches. This updated assessment aims to determine whether observed decreases in VTE rates represent temporary improvements or sustained advances in patient safety, which prior studies would be unable to definitively establish.
Material and methods
Data source and sample population
This study is a retrospective analysis of data collected from the American College of Surgeons (ACS) NSQIP between 2009 and 2022. The NSQIP database consists of information on preoperative patient characteristics and comorbidities, perioperative variables, and 30-day postoperative complications for surgical procedures contributed by participating institutions of varying sizes across the United States. These data are collected by trained surgical reviewers according to specific criteria set by the ACS. Details regarding the design, data collection, and definitions of variables for the ACS-NSQIP are described extensively elsewhere [13].
We included patients aged 18 years or older undergoing a THA or a TKA in this study. Patients were identified using Current Procedural Terminology codes 27130 and 27447, respectively, between 2009 and 2022. We excluded cases with inadequate/missing information on variables of interest (n = 536), including participants with nonbinary sex designation (n = 266) and missing sex (n = 220), age (n = 1), anesthesia type (n = 33), American Society of Anesthesiologists (ASA) classification (n = 15), and hypertension status (n = 1) values. These exclusions were necessary as these variables were included in our multivariate models.
Study variables
The primary outcomes of interest were DVT, PE, and VTE. We defined DVT and PE according to NSQIP standardized definitions. For the purposes of this study, we defined VTE in this study as the occurrence of a DVT and/or PE event in a patient [14,15].
Age was categorized in groups as 18-49 years, 50-64 years, 65-79 years, and 80 years or older. Body mass index (BMI) was categorized as <25, 25-30, 30-35, 35-40, and >40. We recategorized the ASA physical status classification as 1, 2, 3, and 4 or higher. Race was also recategorized as non-Hispanic Whites, non-Hispanic Blacks, Hispanic, and other race. We also collected data on functional status, the presence of diabetes, smoking status, chronic obstructive pulmonary disease, the use of antihypertensive medication, and steroid use. We also assessed other perioperative variables including anesthesia types, total operation time, length of stay (LOS), and admission status.
Statistical analysis
All statistical analyses were performed using Stata MP version 18.0 (StataCorp LLC, College Station, TX). Patient characteristics were described using means or medians for continuous variables and proportions for categorical variables. Trends for VTE over time, for THA and TKA were described, and statistical tests of trend were performed using the ptrend function. This function performs a chi-square test for trend in proportions across ordered groups (years from 2009 to 2022) [16]. The test evaluates whether there is a statistically significant linear trend in the proportions of VTE events over the ordered time periods. The resulting P-values represent the statistical significance of these trends rather than comparisons between specific years. Multivariate regression analyses were performed to estimate the adjusted risk of PE and DVT events by year of surgery relative to 2009, while controlling for demographic factors, comorbidities, and perioperative variables. Predictors for DVT and PE were also evaluated using multivariate regression modeling, which included procedure type (THA vs TKA) as a covariate along with all demographic and clinical variables. In addition, we examined changes in hospital LOS over time during the study period, using multiple linear regression models to evaluate year-over-year changes in LOS for both THA and TKA. For all regression models, odds ratios (ORs) with 95% confidence intervals and P-values are reported.
Patient characteristics
A total of 975,575 patients—382,515 THAs and 593,060 TKAs—met inclusion criteria for the present study. The mean age of the cohorts was 65.6 years among patients undergoing THA, and 67.1 years among patients undergoing TKA. The majority of patients were in the 65-74 years age category (44.9% for THA and 52.3% for TKA). Overall, a large majority of patients were female and non-Hispanic White. The mean BMI was 30.0 kg/m2 in THA vs 32.8 kg/m2 in TKA (Table 1). Pertinent comorbidities are documented in Table 2. Hypertension was the most prevalent comorbidity (55.2% for THA and 64.3% for TKA), followed by diabetes mellitus (12.3% for THA and 18.2% for TKA). Majority of patients undergoing both THA and TKA were classified as ASA class II or III. General anesthesia (45.9% for THA and 41.3% for TKA) and spinal anesthesia (36.3% for THA and 39.8% for TKA) were the most common anesthetic modalities selected for THA and TKA.
Table 1.
General characteristics of the study population.
| Characteristic | THA, N = 382,515 | TKA, N = 593,060 |
|---|---|---|
| Age, mean (SD) | 65.6 (11.4) | 67.1 (9.4) |
| Age categories, n (%) | ||
| <50 y | 29,798 (7.8) | 18,927 (3.2) |
| 50-64 y | 139,991 (36.6) | 209,017 (35.2) |
| 65-79 y | 171,716 (44.9) | 310,353 (52.3) |
| ≥80 y | 41,010 (10.7) | 54,763 (9.2) |
| Gender, n (%) | ||
| Female | 210,384 (55.0) | 363,197 (61.2) |
| Male | 172,131 (45.0) | 229,863 (38.8) |
| Race, n (%) | ||
| Non-Hispanic White | 254,690 (66.6) | 382,352 (64.5) |
| Non-Hispanic Black | 28,845 (7.5) | 44,955 (7.6) |
| Hispanic | 12,840 (3.4) | 33,909 (5.7) |
| Other race | 86,140 (22.5) | 131,844 (22.2) |
| Body mass index, mean (SD) | 30.0 (6.9) | 32.8 (7.1) |
| Body mass index categories, n (%) | ||
| <25 | 80,351 (21.0) | 58,647 (9.9) |
| 25-29.9 | 126,027 (32.9) | 158,868 (26.8) |
| 30-34.9 | 97,912 (25.6) | 173,495 (29.3) |
| 35-39.9 | 51,287 (13.4) | 118,251 (19.9) |
| >40 | 26,938 (7.0) | 83,799 (14.1) |
THA, total hip arthroplasty; TKA, total knee arthroplasty.
All continuous variables are expressed as mean and standard deviation (SD). All categorical variables are expressed as frequencies (%).
Table 2.
Comorbidities and clinical characteristics.
| Clinical characteristic | THA, N = 382,515 | TKA, N = 593,060 |
|---|---|---|
| Medicated diabetes mellitus | ||
| Insulin | 10,962 (2.9) | 25,311 (4.3) |
| Non-insulin | 36,034 (9.4) | 82,694 (13.9) |
| Heart failurea | 3354 (0.9) | 4552 (0.8) |
| Current smokerb | 46,357 (12.1) | 46,211 (7.8) |
| Severe COPD | 15,000 (3.9) | 19,845 (3.3) |
| Immunosuppressive therapy | 14,776 (3.9) | 21,588 (3.6) |
| Medicated hypertension | 211,222 (55.2) | 381,280 (64.3) |
| Bleeding disorder | 8534 (2.2) | 11,961 (2.0) |
| Disseminated cancer | 1681 (0.4) | 751 (0.1) |
| Functional health status | ||
| Independent | 372,813 (97.5) | 582,992 (98.3) |
| Partially dependent | 7580 (2.0) | 6787 (1.1) |
| Totally dependent | 484 (0.1) | 226 (0.0) |
| Unknown | 1638 (0.4) | 3055 (0.5) |
| ASA classification | ||
| 1 | 13,015 (3.4) | 10,048 (1.7) |
| 2 | 195,968 (51.2) | 282,299 (47.6) |
| 3 | 164,648 (43.0) | 289,987 (48.9) |
| 4 | 8389 (2.2) | 9990 (1.7) |
| None assigned | 495 (0.1) | 736 (0.1) |
| Principal anesthesia technique | ||
| Epidural | 2162 (0.6) | 4522 (0.8) |
| General | 175,576 (45.9) | 245,128 (41.3) |
| Local | 87 (0.0) | 172 (0.0) |
| Monitored anesthesia care/ intravenous sedation | 58,026 (15.2) | 94,355 (15.9) |
| Monitored anesthesia care | 8 (0.0) | 50 (0.0) |
| None | 55 (0.0) | 126 (0.0) |
| Other | 182 (0.0) | 502 (0.1) |
| Regional | 7433 (1.9) | 12,307 (2.1) |
| Spinal | 138,944 (36.3) | 235,843 (39.8) |
| Unknown | 42 (0.0) | 55 (0.0) |
| Length of hospital stay (d)c | 2.2 (4.0) | 2.2 (3.3) |
| Duration of operation (min)c | 92.2 (39.0) | 91.6 (36.4) |
| DVT/thrombophlebitis requiring therapy | 1394 (0.4) | 4319 (0.7) |
| Pulmonary embolism | 1004 (0.3) | 2803 (0.5) |
| Inpatient/Outpatient | ||
| Inpatient | 316,214 (82.7) | 466,741 (78.7) |
| Outpatient | 66,301 (17.3) | 126,319 (21.3) |
ASA, American Society of Anesthesiologist; COPD, chronic obstructive pulmonary disease; DVT, deep vein thrombosis.
Heart failure documented at least 30 d before surgery.
Smoking documented within 1 y of surgery.
Mean (SD).
Results
VTE trends
A total of 5713 DVTs and 3807 PEs were found after THA and TKA between the years of 2009 and 2022 (Table 2). A significant decrease in incidence of DVT over the study period was observed in patients undergoing TKA (1.2%-0.6%; P for trend <.001), and a slight but significant downtrend for THA (0.4%-0.3%, P for trend <.001). A significant decrease in incidence of PE over the study period also occurred in patients undergoing TKA (0.8%-0.4%; P for trend <.001; Fig. 1). Overall, trends of 30-day VTE decreased over the study period in THA (0.8%-0.5%, P for trend <.001) and TKA (1.8%-0.9%, P for trend <.001; Fig. 2).
Figure 1.
Trends for DVT and PE classified by procedure from 2009 to 2022. DVT, deep venous thrombosis; PE, pulmonary embolism; THA, total hip arthroplasty; TKA, total knee arthroplasty. ∗P-value for trend <.001. †P-value for trend = .23.
Figure 2.
VTE trends classified by procedure type from 2009 to 2022. VTE, venous thromboembolism. ∗P-value for trend <.001.
Regression modeling
After adjusted regression, the odds of 30-day post-TKA DVT and PE show a trend of decreasing odds of events with increasing statistical significance over time (Table 3). The odds of post-TKA DVT (OR 0.68, P < .01) and PE (OR 0.59, P < .01) were significantly lower in 2022 than those in 2009 (Table 3). However, for THA, the odds of DVT and PE remained statistically unchanged across all study years when compared to 2009. Patients aged 80 years or older had the highest odds of both DVT and PE (OR 1.74 [1.48-2.04], P < .001 & OR 2.38 [1.93-2.92], P < .001 respectively; Table 4). Surgeries performed in the outpatient setting displayed lower odds of developing DVT or PE (OR 0.73 [0.67-0.80], P < .001 and OR 0.63 [0.56-0.71)] P < .001, respectively). Patients undergoing a TKA had significantly higher odds of developing both DVT (OR 2.04 [1.92-2.17]) and PE (OR 1.66 [1.65-1.79]) than those undergoing THA procedures (both P < .001). All variables utilized for regression modeling are described in Table 3, Table 4.
Table 3.
Adjusted regression models stratified by surgical intervention and VTE subtype.
| Year of operation (ref: 2009) | DVT TKA |
PE TKA |
DVT THA |
PE THA |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | |
| 2010 | 0.92 | 0.65-1.31 | .66 | 0.97 | 0.63-1.48 | .88 | 1.83 | 0.88-3.78 | .11 | 0.93 | 0.41-2.10 | .86 |
| 2011 | 0.82 | 0.60-1.13 | .23 | 0.72 | 0.49-1.08 | .11 | 1.12 | 0.56-2.25 | .76 | 0.74 | 0.35-1.53 | .41 |
| 2012 | 0.80 | 0.59-1.08 | .14 | 0.94 | 0.66-1.36 | .76 | 1.15 | 0.59-2.25 | .69 | 0.53 | 0.26-1.08 | .08 |
| 2013 | 0.88 | 0.66-1.18 | .40 | 0.84 | 0.59-1.21 | .36 | 1.22 | 0.63-2.35 | .55 | 0.76 | 0.39-1.50 | .44 |
| 2014 | 0.80 | 0.60-1.07 | .14 | 0.89 | 0.62-1.26 | .50 | 1.15 | 0.60-2.21 | .68 | 0.81 | 0.41-1.57 | .53 |
| 2015 | 0.73 | 0.55-0.97 | <.05 | 0.90 | 0.64-1.27 | .56 | 1.14 | 0.60-2.18 | .69 | 0.79 | 0.41-1.52 | .47 |
| 2016 | 0.69 | 0.52-0.91 | <.01 | 0.59 | 0.41-0.84 | <.01 | 1.14 | 0.60-2.17 | .69 | 0.74 | 0.38-1.42 | .36 |
| 2017 | 0.68 | 0.52-0.91 | <.01 | 0.64 | 0.45-0.90 | <.05 | 1.00 | 0.52-1.90 | .99 | 0.65 | 0.34-1.26 | .21 |
| 2018 | 0.62 | 0.46-0.82 | <.01 | 0.59 | 0.42-0.84 | <.01 | 0.97 | 0.51-1.85 | .93 | 0.69 | 0.36-1.32 | .26 |
| 2019 | 0.61 | 0.46-0.81 | <.01 | 0.54 | 0.38-0.76 | <.01 | 1.05 | 0.55-1.99 | .89 | 0.77 | 0.40-1.47 | .43 |
| 2020 | 0.66 | 0.50-0.89 | <.01 | 0.53 | 0.37-0.77 | <.01 | 1.03 | 0.54-1.97 | .94 | 0.71 | 0.37-1.38 | .31 |
| 2021 | 0.70 | 0.53-0.94 | <.05 | 0.61 | 0.42-0.87 | <.01 | 1.26 | 0.66-2.42 | .48 | 0.83 | 0.43-1.61 | .58 |
| 2022 | 0.68 | 0.51-0.91 | <.01 | 0.59 | 0.41-0.85 | <.01 | 1.15 | 0.60-2.20 | .68 | 0.86 | 0.44-1.65 | .64 |
COPD, chronic obstructive pulmonary disease; DVT, deep venous thrombosis; OR, odds ratio; PE, pulmonary embolism; THA, total hip arthroplasty; TKA, total knee arthroplasty; VTE, venous thromboembolism.
Adjusted for: sex, age, race/ethnicity, body mass index, year of operation, functional status, comorbidities—smoking, history of COPD, antihypertensive use, immunosuppressive therapy, American Society of Anesthesiologists physical status classification, admission status, hospital length of stay and duration of surgical procedure.
Table 4.
Adjusted regression model for risk factors for DVT and PE in the overall study population.
| Covariate | DVT |
PE |
||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | |
| Gender (ref: male) | ||||||
| Female | 0.90 | <0.001 | .85-.95 | 1.19 | <0.001 | 1.11-1.28 |
| Age (ref: <50 y) | ||||||
| 50-64 y | 1.06 | 0.44 | .92-1.22 | 1.16 | 0.12 | .96-1.40 |
| 65-79 y | 1.25 | <0.01 | 1.08-1.45 | 1.63 | <0.001 | 1.35-1.96 |
| 80 y and older | 1.74 | <0.001 | 1.48-2.04 | 2.38 | <0.001 | 1.93-2.92 |
| Race/Ethnicity (ref: non-Hispanic White) | ||||||
| Non-Hispanic Black | 1.35 | <0.001 | 1.23-1.48 | 1.67 | <0.001 | 1.50-1.86 |
| Hispanic | 1.24 | <0.001 | 1.11-1.39 | 1.35 | <0.001 | 1.17-1.55 |
| Other/unknown | 0.85 | <0.001 | .79-.91 | 1.26 | <0.001 | 1.16-1.36 |
| BMI (ref: < 25) | ||||||
| 25-29.9 | 1.13 | <0.05 | 1.03-1.23 | 1.31 | <0.001 | 1.16-1.48 |
| 30-34.9 | 1.15 | <0.01 | 1.05-1.26 | 1.61 | <0.001 | 1.43-1.82 |
| 35-39.9 | 1.10 | 0.08 | .99-1.21 | 1.77 | <0.001 | 1.56-2.02 |
| ≥40 | 0.93 | 0.2 | .82-1.04 | 1.70 | <0.001 | 1.47-1.96 |
| Year of operation (ref: 2009) | ||||||
| 2010 | 1.07 | 0.67 | .78-1.46 | 0.97 | 0.89 | .67-1.42 |
| 2011 | 0.86 | 0.29 | .64-1.14 | 0.71 | 0.05 | .50-1.00 |
| 2012 | 0.85 | 0.24 | .65-1.12 | 0.84 | 0.28 | .61-1.16 |
| 2013 | 0.93 | 0.59 | .71-1.21 | 0.82 | 0.22 | .60-1.12 |
| 2014 | 0.86 | 0.25 | .66-1.11 | 0.86 | 0.36 | .63-1.18 |
| 2015 | 0.80 | 0.08 | .61-1.03 | 0.87 | 0.38 | .64-1.19 |
| 2016 | 0.76 | <0.05 | .58-.98 | 0.62 | <0.01 | .45-.84 |
| 2017 | 0.73 | <0.05 | .56-.94 | 0.64 | <0.01 | .47-.86 |
| 2018 | 0.67 | <0.01 | .52-.87 | 0.61 | <0.01 | .45-.83 |
| 2019 | 0.68 | <0.01 | .53-.88 | 0.59 | <0.01 | .43-.80 |
| 2020 | 0.72 | <0.05 | .55-.94 | 0.58 | <0.01 | .42-.79 |
| 2021 | 0.79 | 0.08 | .61-1.03 | 0.66 | <0.05 | .48-.91 |
| 2022 | 0.75 | <0.05 | .58 - .98 | 0.66 | <0.01 | .48 - .90 |
| Comorbidities (ref: absence of) | ||||||
| Current smoker | 0.92 | 0.09 | .83-1.01 | 0.78 | <0.001 | .69-.89 |
| History of COPD | 1.20 | <0.01 | 1.06-1.36 | 1.53 | <0.001 | 1.33-1.77 |
| Use of antihypertensives | 0.95 | 0.07 | .89-1.00 | 0.92 | <0.05 | .86-.99 |
| Immunosuppressive therapy | 1.20 | <0.01 | 1.06-1.36 | 1.13 | 0.14 | .96-1.32 |
| Functional status (ref: independent) | ||||||
| Partially dependent | 1.27 | <0.01 | 1.06-1.53 | 1.07 | 0.55 | .85-1.35 |
| Totally dependent | 2.87 | <0.001 | 1.68-4.92 | 0.98 | 0.96 | .36-2.66 |
| ASA class (ref: class 1) | ||||||
| 2 | 0.91 | 0.35 | .75-1.11 | 1.11 | 0.43 | .85-1.45 |
| 3 | 1.08 | 0.45 | .89-1.31 | 1.20 | 0.19 | .92-1.56 |
| ≥4 | 1.20 | 0.17 | .93-1.55 | 1.43 | <0.05 | 1.03-1.99 |
| Surgery type (ref: THA) | ||||||
| TKA | 2.04 | <0.001 | 1.92-2.17 | 1.66 | <0.001 | 1.54-1.79 |
| Inpatient/Outpatient status (ref: inpatient) | ||||||
| Outpatient | 0.73 | <0.001 | .67-.80 | 0.63 | <0.001 | .56-.71 |
| Hospital length of staya | 1.03 | <0.001 | 1.02-1.03 | 1.03 | <0.001 | 1.03-1.04 |
| Duration of procedurea | 1.00 | <0.001 | 1.002-1.003 | 1.00 | <0.001 | 1.002-1.003 |
ASA, American Society of Anesthesiologists; BMI, body mass index; COPD, chronic obstructive pulmonary disease; DVT, deep vein thrombosis; PE, pulmonary embolism, THA, total hip arthroplasty; TKA, total knee arthroplasty.
Odds ratios (ORs) and 95% confidence intervals (CIs) from multivariate logistic regression models adjusted for all covariates listed in the table. Reference categories are indicated in parentheses.
Continuous variables.
Hospital LOS trends
Between 2009 and 2022, the mean hospital LOS decreased from 3.57 days to 1.25 days for TKA, and 3.67 days to 1.34 days for THA. Regression analysis showed a progressively decreasing LOS for both TKA and THA procedures. For TKA, we observed increasingly negative coefficients from 2011 onward (all P < .01), indicating consistent year-over-year reductions in hospital stays. Similarly, for THA procedures, we found progressively significant larger negative coefficients across most years (all P < .01), apart from 2011, which showed a nonsignificant reduction compared to 2009 (coefficient −0.13, P = .14).
Discussion
The present study reports on incidence and risk factors associated with VTE after primary TKA and THA in the United States between 2009 and 2022. Prior studies have analyzed VTE trends since the introduction of changes in guidelines and postoperative protocols; however, to the best of our knowledge no studies have included the most recently available data from the NSQIP database [6,11,12]. This study sought to provide an updated analysis of the most recent data and correlate the trends to associated risk factors using regression models. We found that the 30-day incidence of VTE significantly decreased over the study period. Furthermore, patients aged 80 years or older were found to have the highest odds of developing either DVT or PE. Procedures performed in the outpatient setting displayed lower odds of developing VTE. Our study highlights a significantly decreasing trend of 30-day incidence of VTE (both P for trend <.001) following both THA and TKA. Similarly, Agarwal et al. [11] found decreased rates of 90-day PE and DVT after THA from 2011 to 2019 using the PearlDiver database. Although the NSQIP uses standardized clinical data protocols and consistent definitions for variables, our study with a 30-day follow-up period through NSQIP complements the study by Agarwal et al. [11] with a 90-day postoperative period for inclusion through claims data and together provide a more comprehensive understanding of both short- and medium-term postoperative VTE trends.
Importantly, our present results build upon a prior study by Warren et al. [12], which analyzed the incidence of 30-day VTE and mortality in TJA using the NSQIP database from 2008 to 2016. They found no significant change in VTE incidence in THA patients after multivariate regression and a significant reduction in VTE incidence among TKA patients. Our study, which extends the trend timeline to 2022, confirms these findings and provides additional insights into the trends over the subsequent years with regression modeling. Overall, we observed a trend of decreasing VTE rates for THA patients, as well as a significant decrease in the incidence of DVT over the study period. For TKA patients, the trend of decreasing VTE rates persisted beyond 2016, with statistically significant reductions in both DVT and PE in our extended study period. In addition, after adjusted regression, the odds of 30-day DVT (OR 0.68, P < .01) and PE (OR 0.59, P < .01) after TKA were significantly lower in 2022 than in 2009. These significant reductions in overall VTE incidence as well as individual DVT and PE events since 2009 highlight the success of current treatment guidelines and the advancements in perioperative care and early mobilization protocols. Furthermore, the decreasing trends in VTE, especially DVT and PE rates in TKA, appear to show a sustained pattern of decrease coinciding with new guidelines championing the use of aspirin for VTE which also likely contributes to this trend.
Our analysis also identified several perioperative risk factors for VTE following THA and TKA including age >80, higher BMI, and surgical setting. These findings are in line with those reported by Kahn and Shivakumar [17] who found that the factors previously mentioned significantly influence VTE risk in orthopedic patients. The association of these characteristics with increased VTE incidence provides further support in favor of individualization and risk-stratification treatment strategies. Studies have shown that using risk-stratification protocols can achieve low overall incidences of symptomatic VTE. For example, Nam et al. [18] demonstrated that a risk-stratification protocol resulted in a VTE rate of 0.7% in routine-risk patients and 0.5% in high-risk patients, with a significant reduction in major bleeding events in the routine risk cohort. Risk stratification also allows for individualized anticoagulation regimens based on patient-specific risk factors. The Caprini score is often used to categorize patients into low- and high-risk groups where low-risk patients may receive less-aggressive prophylaxis such as aspirin, while high-risk patients may require more potent anticoagulants like Apixaban or Warfarin [19,20].
Our analysis of LOS at hospital trends revealed a substantial reduction over the study period, with the mean LOS decreasing from 3.67 to 1.34 days for THA and from 3.57 to 1.25 days for TKA between 2009 and 2022. This dramatic shortening of hospitalization aligns with the observed decrease in VTE rates and likely reflects the successful implementation of Enhanced Recovery After Surgery (ERAS) protocols and early mobilization strategies. The significant association between longer hospital stays and increased VTE risk (OR 1.03, P < .001 for both DVT and PE) suggests that factors contributing to a shorter LOS may play a protective role against thromboembolic events. The concurrent trends of decreasing LOS and VTE rates, despite shifts toward more selective chemoprophylaxis, highlight the multifactorial nature of successful VTE prevention in modern arthroplasty care.
VTE is universally recognized as a significant complication following TJA, but over the last several years, the implementation of multifaceted strategies including pharmacologic prophylaxis, mechanical prophylaxis, and early mobilization protocols has aimed to decrease the incidence of this complication. In addition, there have been pivotal changes in guidelines, particularly with the release of updated Clinical Practice Guidelines from the AAOS in 2011, suggesting pharmacologic or mechanical prophylaxis, and the ACCP in 2012, which endorsed aspirin as an acceptable antithrombotic for post-THA and TKA prophylaxis [11,21,22]. These guidelines have prompted increased use among providers as highlighted by a survey performed at the 2022 AAHKS annual meeting, which found 93% of members were using aspirin/mechanical devices as VTE prophylaxis following THA, and 94% following TKA [23]. Furthermore, mechanical prophylaxis, including the use of intermittent pneumatic compression devices and graduated compression stockings, has been shown to complement pharmacologic measures by promoting venous return and reducing stasis [24].
Early mobilization protocols, which encourage patients to begin ambulation shortly after surgery, including on the first postoperative day, have also played a critical role in VTE prevention. A study by Chandrasekaran et al. [25] demonstrated that the incidence of DVT and PE after TKA was lower in the group that mobilized early than in those who had strict bed rest on the first postoperative day (P = .03). The same study found that the greater the distance mobilized, the lower the odds of developing a thromboembolic complication (P < .01) [25]. Although each individual strategy is effective in reducing rates of VTE following TJA, it is crucial to analyze large-scale, population-level data to determine the cumulative effect of guidelines adjustments on current treatment protocols.
In addition to changes in chemoprophylaxis guidelines, the widespread adoption of ERAS protocols during our study period represents another significant shift in perioperative management. These comprehensive protocols formalize and standardize many elements beyond early mobilization, including preoperative patient optimization, multimodal pain-management strategies, and minimization of opioid use [26,27]. The adoption of aspirin for VTE prophylaxis, increasing from approximately 20% (TKA) and 15% (THA) to 95% for both among AAHKS members between 2012 and 2020 [28], coincides with the progressive decrease in VTE rates observed in our study. Concurrently, diagnostic approaches have evolved, with a significant shift away from routine postoperative ultrasound screening for asymptomatic DVT toward symptom-driven diagnostic testing [5,29]. Previous guidelines often recommended screening ultrasonography, particularly in high-risk patients, but more recent approaches have prioritized clinical surveillance and selective testing based on clinical suspicion [22,30]. This timeline of clinical practice change regarding the use of aspirin following the 2011 AAOS and 2012 ACCP guideline updates, combined with the institutional adoption of ERAS protocols and changes in screening practices, potentially influences the detection and reported rates of asymptomatic VTE and likely contributed significantly to our findings.
Limitations
While the present study provides a robust analysis of the annual trends of VTE rates following TJA, it is not without limitations. First, as with all large administrative database studies, there may be confounding variables present that may not have been adequately accounted for. However, regression models were utilized to limit their impact on the results and conclusions drawn. We used Current Procedural Terminology codes for identifying patients who received TKA and THA. This could have led to coding errors leading to misclassification bias, which could have affected our study’s findings.
In addition, the NSQIP database only tracks outcomes and events for 30 days postoperatively, which may not fully capture all instances of VTE following TJA. The ACCP considers patients at increased risk of VTE up to 90 days postoperatively [22]. As such, our analysis may underestimate the true incidence of DVT and PE following THA and TKA compared to other databases with longer follow-up periods.
The NSQIP does not collect any information on therapies for VTE prophylaxis; as such, the impact of increased use of aspirin or other therapies on VTE rates could not be directly evaluated. Variations in clinical practice and adherence to guidelines across different institutions that participate in NSQIP could also influence the observed trends and potentially limit the generalizability of our findings to all practice settings
Finally, since the inception of the NSQIP database, the number of participating hospitals and definitions of clinical variables collected have evolved over time affecting the consistency of data collected. Some studies have highlighted how these changes may affect collected data and their interpretation [31].
Conclusions
In conclusion, our study observed decreased incidence rates of VTE after TJA from 2009 to 2022 in the United States when utilizing the NSQIP database, highlighting the success of current VTE prophylaxis strategies. Continued efforts to refine and personalize prophylactic approaches are essential to further reduce the incidence of these potentially life-threatening complications in arthroplasty patients. Future studies should seek to characterize the relationship between various prophylactic strategies such as pharmacologic agent of choice, postoperative mobilization protocols, and surgical setting with VTE rates.
Conflicts of interest
JCS receives royalties from Corin USA; is in the speakers’ bureau/gave paid presentations for DePuy; is a paid consultant for DePuy; and is in the editorial board of Arthroplasty Today. AV is a consultant for DePuy, a Johnson & Johnson company. All other authors declare no potential conflicts of interest.
For full disclosure statements refer to https://doi.org/10.1016/j.artd.2025.101737.
CRediT authorship contribution statement
Hannah Mosher: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Data curation, Conceptualization. Hallie B. Remer: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Data curation, Conceptualization. Chukwuemeka U. Osondu: Writing – review & editing, Visualization, Validation, Project administration, Methodology, Formal analysis, Conceptualization. Kevin Smidt: Writing – review & editing, Project administration, Methodology, Investigation, Conceptualization. Alexander van der Ven: Writing – review & editing, Supervision, Methodology, Data curation, Conceptualization. Juan C. Suarez: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Conceptualization.
Disclaimer
The American College of Surgeons NSQIP and the hospitals participating in it are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.
Appendix.
Supplementary Table 1.
Mean hospital length of stay (d) and trend analysis using linear regression.
| Year of operation | TKA |
THA |
||||
|---|---|---|---|---|---|---|
| Mean LOS (d) | Regression coefficient | Regression P-value | Mean LOS (d) | Regression coefficient | Regression P-value | |
| 2009 | 3.57 | Ref. | - | 3.67 | Ref. | - |
| 2010 | 3.50 | −0.07 | .25 | 3.37 | −0.30 | <0.01 |
| 2011 | 3.38 | −0.19 | <.01 | 3.54 | −0.13 | .14 |
| 2012 | 3.37 | −0.21 | <.001 | 3.25 | −0.42 | <.001 |
| 2013 | 3.13 | −0.45 | <.001 | 3.11 | −0.56 | <.001 |
| 2014 | 3.00 | −0.57 | <.001 | 2.91 | −0.76 | <.001 |
| 2015 | 2.81 | −0.76 | <.001 | 2.74 | −0.94 | <.001 |
| 2016 | 2.54 | −1.03 | <.001 | 2.49 | −1.18 | <.001 |
| 2017 | 2.36 | −1.22 | <.001 | 2.31 | −1.36 | <.001 |
| 2018 | 1.97 | −1.60 | <.001 | 1.99 | −1.68 | <.001 |
| 2019 | 1.82 | −1.76 | <.001 | 1.85 | −1.82 | <.001 |
| 2020 | 1.53 | −2.04 | <.001 | 1.64 | −2.03 | <.001 |
| 2021 | 1.33 | −2.24 | <.001 | 1.43 | −2.25 | <.001 |
| 2022 | 1.25 | −2.32 | <.001 | 1.34 | −2.33 | <.001 |
Appendix A. Supplementary data
References
- 1.Kurtz S., Ong K., Lau E., Mowat F., Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg. 2007;89:780–785. doi: 10.2106/JBJS.F.00222. [DOI] [PubMed] [Google Scholar]
- 2.Kremers H.M., Larson D.R., Crowson C.S., Kremers W.K., Washington R.E., Steiner C.A., et al. Prevalence of total hip and knee replacement in the United States. J Bone Joint Surg Am. 2014;97:1386–1397. doi: 10.2106/JBJS.N.01141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zuke W.A., Chughtai M., Emara A.K., Zhou G., Koroukian S.M., Molloy R.M., et al. What are drivers of readmission for readmission-requiring venous thromboembolic events after primary total hip arthroplasty? An analysis of 544,443 cases. J Arthroplasty. 2022;37:958–965.e3. doi: 10.1016/j.arth.2022.01.024. [DOI] [PubMed] [Google Scholar]
- 4.Bateman D., Dow R., Brzezinski A., Bar-Eli H., Kayiaros S. Correlation of the caprini score and venous thromboembolism incidence following primary total joint arthroplasty-results of a single-institution protocol. J Arthroplasty. 2017;32:3735–3741. doi: 10.1016/j.arth.2017.06.042. [DOI] [PubMed] [Google Scholar]
- 5.Dubin J.A., Bains S.S., Hameed D., Remily E.A., Moore M.C., Mont M.A., et al. Trends in deep vein thrombosis prophylaxis after total knee arthroplasty: 2016 to 2021. J Arthroplasty. 2024;2023:1–5. doi: 10.1016/j.arth.2024.01.050. [DOI] [PubMed] [Google Scholar]
- 6.Dua A., Desai S.S., Lee C.J., Heller J.A. National trends in deep vein thrombosis following total knee and total hip replacement in the United States. Ann Vasc Surg. 2017;38:310–314. doi: 10.1016/j.avsg.2016.05.110. [DOI] [PubMed] [Google Scholar]
- 7.Pedersen A., Mehnert F., Sorensen H., Emmeluth C., Overgaard S., Johnsen S. The risk of venous thromboembolism, myocardial infarction, stroke, major bleeding and death in patients undergoing total hip and knee replacement: a 15-year retrospective cohort study of routine clinical practice. Bone Joint J. 2014;96-B:479–485. doi: 10.1302/0301-620X.96B4.33209. [DOI] [PubMed] [Google Scholar]
- 8.Shahi A., Chen A.F., Tan T.L., Maltenfort M.G., Kucukdurmaz F., Parvizi J. The incidence and economic burden of in-hospital venous thromboembolism in the United States. J Arthroplasty. 2017;32:1063–1066. doi: 10.1016/j.arth.2016.10.020. [DOI] [PubMed] [Google Scholar]
- 9.Spyropoulos A.C., Lin J. Direct medical costs of venous thromboembolism and subsequent hospital readmission rates: an administrative claims analysis from 30 managed care organizations. J Manag Care Pharm. 2007;13:475–486. doi: 10.18553/jmcp.2007.13.6.475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ruppert A., Steinle T., Lees M. Economic burden of venous thromboembolism: a systematic review. J Med Econ. 2011;14:65–74. doi: 10.3111/13696998.2010.546465. [DOI] [PubMed] [Google Scholar]
- 11.Agarwal A.R., Das A., Harris A., Campbell J.C., Golladay G.J., Thakkar S.C. Trends of venous thromboembolism after total hip arthroplasty in the United States: analysis from 2011 to 2019. J Am Acad Orthop Surg. 2023;31:E376–E384. doi: 10.5435/JAAOS-D-22-00708. [DOI] [PubMed] [Google Scholar]
- 12.Warren J.A., Sundaram K., Anis H.K., Kamath A.F., Higuera C.A., Piuzzi N.S. Have venous thromboembolism rates decreased in total hip and knee arthroplasty? J Arthroplasty. 2020;35:259–264. doi: 10.1016/j.arth.2019.08.049. [DOI] [PubMed] [Google Scholar]
- 13.ACS National Surgical Quality Improvement Program Acs. https://www.facs.org/quality-programs/data-and-registries/acs-nsqip/
- 14.Chiu A.K., Agarwal A.R., Hussain N., Gu A., Thakkar S.C., Golladay G.J. Trends in venous thromboembolism and chemoprophylaxis utilization in elective total knee arthroplasty from 2011 to 2020. J Arthroplasty. 2024;39:S212–S217.e1. doi: 10.1016/j.arth.2024.05.025. [DOI] [PubMed] [Google Scholar]
- 15.Salous A.K., Reyad A., Sweeney K., Mavanur A. A significant proportion of venous thromboembolism events in general surgical patients occurs after discharge: analysis of the ACS-NSQIP Essentials database. Perioper Med (Lond) 2019;8:18. doi: 10.1186/s13741-019-0131-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Royston P. PTREND: Stata module for trend analysis for proportions. Statistical Software Components. 2014. https://ideas.repec.org//c/boc/bocode/s426101.html
- 17.Kahn S.R., Shivakumar S. What’s new in VTE risk and prevention in orthopedic surgery. Res Pract Thromb Haemost. 2020;4:366–376. doi: 10.1002/rth2.12323. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Nam D., Nunley R.M., Johnson S.R., Keeney J.A., Clohisy J.C., Barrack R.L. The effectiveness of a risk stratification protocol for thromboembolism prophylaxis after hip and knee arthroplasty. J Arthroplasty. 2016;31:1299–1306. doi: 10.1016/j.arth.2015.12.007. [DOI] [PubMed] [Google Scholar]
- 19.Krauss E.S., Segal A., Dengler N., Cronin M., Pettigrew J., Simonson B.G. Utilization of the caprini score for risk stratification of the arthroplasty patient in the prevention of postoperative venous thrombosis. Semin Thromb Hemost. 2022;48:407–412. doi: 10.1055/s-0042-1742739. [DOI] [PubMed] [Google Scholar]
- 20.Krauss E.S., Cronin M., Dengler N., Simonson B.G., Enker P., Segal A. Lessons learned: using the caprini risk assessment model to provide safe and efficacious thromboprophylaxis following hip and knee arthroplasty. Clin Appl Thromb Hemost. 2020;26 doi: 10.1177/1076029620961450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mont M., Jacobs J., Boggio L. Preventing venous thromboembolic disease in patients undergoing elective hip and knee arthroplasty. J Am Acad Orthop Surg. 2011;19:768–776. doi: 10.5435/00124635-201112000-00007. [DOI] [PubMed] [Google Scholar]
- 22.Falck-Ytter Y., Francis C., Johanson N. Prevention of VTE in orthopedic surgery patients: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2 Suppl):e278S–e325S. doi: 10.1378/chest.11-2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Abdel M.P., Watts C.D., Houdek M.T., Lewallen D.G., Berry D.J. Epidemiology of periprosthetic fracture of the femur in 32 644 primary total hip arthroplasties. Bone Joint J. 2016;98-B:461–467. doi: 10.1302/0301-620X.98B4.37201. [DOI] [PubMed] [Google Scholar]
- 24.Geerts W.H., Bergqvist D., Pineo G.F., Heit J.A., Samama C.M., Lassen M.R., et al. Prevention of venous thromboembolism. Chest. 2008;133:381S–453S. doi: 10.1378/chest.08-0656. [DOI] [PubMed] [Google Scholar]
- 25.Chandrasekaran S., Ariaretnam S.K., Tsung J., Dickison D. Early mobilization after total knee replacement reduces the incidence of deep venous thrombosis. ANZ J Surg. 2009;79:526–529. doi: 10.1111/j.1445-2197.2009.04982.x. [DOI] [PubMed] [Google Scholar]
- 26.Soffin E., YaDeau J. Enhanced recovery after surgery for primary hip and knee arthroplasty: a review of the evidence. Br J Anaesth. 2016;117(suppl 3):iii62–iii72. doi: 10.1093/bja/aew362. [DOI] [PubMed] [Google Scholar]
- 27.Pellegrini V.J., Donaldson C., Farber D., Lehman E., Evarts C. The Mark Coventry Award: prevention of readmission for venous thromboembolism after total knee arthroplasty. Clin Orthop Relat Res. 2017;475:16–25. doi: 10.1097/01.blo.0000229357.19867.84. [DOI] [PubMed] [Google Scholar]
- 28.Abdel M.P., Meneghini R.M., Berry D.J. Current practice trends in primary hip and knee arthroplasties among members of the American association of hip and knee Surgeons: an update during the COVID-19 pandemic. J Arthroplasty. 2021;36:S40–S44.e3. doi: 10.1016/j.arth.2021.01.080. [DOI] [PubMed] [Google Scholar]
- 29.Abraham P., Ternisien C., Hubert L., Pidhorz L., Saumet J. Does venous microemboli detection add to the interpretation of D-dimer values following orthopedic surgery? Ultrasound Med Biol. 1999;25:637–640. doi: 10.1016/s0301-5629(99)00004-6. [DOI] [PubMed] [Google Scholar]
- 30.Lieberman J., Heckmann N. Venous thromboembolism prophylaxis in total hip arthroplasty and total knee arthroplasty patients: from guidelines to practice. J Am Acad Orthop Surg. 2017;25:789–798. doi: 10.5435/JAAOS-D-15-00760. [DOI] [PubMed] [Google Scholar]
- 31.Shultz B.N., Galivanche A.R., Ottesen T.D., Bovonratwet P., Grauer J.N. The use of older versus newer data in the national surgical quality improvement Program database influences the results of total hip arthroplasty outcomes studies. J Am Acad Orthop Surg Glob Res Rev. 2019;3 doi: 10.5435/JAAOSGlobal-D-19-00108. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.