Effect of Enhanced Recovery After Surgery compliance on postoperative venous thromboembolism

Abstract

Background

Implementing Enhanced Recovery After Surgery (ERAS) guidelines has been demonstrated to reduce complications; however, it is unknown if ERAS may influence incidence of postoperative venous thromboembolism, a particularly challenging complication. The objective of this study was to examine the association between ERAS compliance and venous thromboembolism across multiple surgery types.

Methods

This retrospective cohort study included adult patients undergoing one of seven ERAS-guided surgeries between 2017 and 2021 at nine hospitals in Alberta, Canada, that implemented ERAS guidelines. The exposure was overall ERAS compliance (categorized as low, moderate, high) and compliance with each ERAS element. The primary outcome was venous thromboembolism within 30 days of surgery. Secondary outcomes included 30-day hospital readmission, emergency department visits and healthcare costs.

Results

Of the 8118 included patients, most had colorectal (52.8%) and gynaecologic (26.1%) surgery. There were 118 (1.5%) patients who experienced a postoperative venous thromboembolism. ERAS compliance was associated with developing a venous thromboembolism; each unit increase in the ERAS compliance score was associated with a 23% decrease in the occurrence of venous thromboembolism. More patients with venous thromboembolism had low (11.0%) or moderate (44.1%) overall ERAS compliance compared with those with no venous thromboembolism (5.6% and 34.5% respectively, P = 0.001). Using logistic regression analysis, the overall ERAS compliance score and American Society of Anesthesiologists class remained significant risk factors for developing a venous thromboembolism.

Conclusions

ERAS compliance was associated with decreased odds of postoperative venous thromboembolism across multiple surgical disciplines, highlighting the importance of improving ERAS compliance to decrease postoperative venous thromboembolism.

Introduction

Venous thromboembolism (VTE) is a known perioperative complication that is associated with significant rates of patient morbidity and mortality^1–3. Developing a VTE contributes to substantial healthcare costs due to an increase in emergency department (ED) visits, hospital readmission and costs associated with VTE management^1,3. It is estimated that VTE events cost the USA healthcare system approximately 7–10 billion euros annually. Patients undergoing surgery are at an increased risk of VTE due to patient characteristics, disease processes and surgical risk factors^4–6. Approximately one-third of VTE-related deaths in the USA are due to perioperative VTE⁷. Studies have demonstrated that postoperative VTE increases the risk of perioperative fatality in surgical patients^8,9, and accounts for up to 50% of perioperative deaths¹⁰. While VTE prophylaxis has reduced the risk of postoperative VTE⁷, VTE still affects an estimated 2% of patients undergoing surgery¹⁰ and appropriate use of VTE prophylaxis is variable^11–14. Given that VTE is a relatively common surgical complication, with significant morbidity rates and costs, it is important to find interventions that decrease the risk of postoperative VTE.

Enhanced Recovery After Surgery (ERAS) is a multimodal, evidence-based approach to perioperative care¹⁵ that includes preoperative, intraoperative and postoperative elements that improve surgical recovery^15,16. ERAS has been shown to decrease duration of hospital stay and healthcare costs across multiple surgical disciplines^16–20. ERAS implementation is also associated with lower risk of perioperative complications^17,18,21; however, it is unclear if ERAS compliance influences the development of postoperative VTE. The objective of this study was to examine the association between ERAS compliance and VTE across multiple surgery types.

Methods

Setting

This study was conducted in the province of Alberta, Canada. Alberta is the third most populated province in the west of Canada. Healthcare is delivered in a publicly funded, universal system and health services are delivered by a single health services provider, Alberta Health Services (AHS). At the time of writing, most surgeries occur within acute care hospitals. A unique personal healthcare number is assigned to each citizen at birth or upon immigration to the province and follows patients throughout their lifetime.

Participants (cohort)

This multicentre (n = 9) retrospective cohort study included adult (18 years and over) patients who had one of seven ERAS-guided surgeries in Alberta, Canada, and a record in the ERAS Interactive Audit System (EIAS; Encare, Kista, Sweden) between January 2017 and September 2021. The surgical procedures included colorectal, breast reconstruction, liver, pancreas, gynaecology, urology, and major head and neck surgeries. This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary, Alberta, Canada (REB-21-1021).

Data sources and measurement

The cohort was identified in the EIAS, a provincially maintained repository of information related to ERAS elements. EIAS captures information about patient demographic characteristics (including age, sex, body mass index (BMI), American Society of Anesthesiologists (ASA) class, Charlson Co-morbidity Index), diagnosis, surgical and anaesthetic details (type of surgery, location, surgical complexity). The cohort identified in EIAS was deterministically linked to additional population-based data sources including: the Discharge Abstract Database (DAD)²², which contains demographic, administrative, diagnostic and procedural data on all patients discharged from hospital; Physician Claims, which contains administrative data on physician billings for publicly insured healthcare services; and the National Ambulatory Care Reporting System (NACRS)²³, which includes data on ED visits. These provincial databases ensure that patients who presented for care anywhere in the province were captured by the study.

Variables

The primary outcome was development of VTE within 30 days of surgery (dichotomous). The incidence of VTE was classified using a combination of data from the EIAS database (presence of deep vein thrombosis (DVT) and/or pulmonary embolism (PE)) and International Classification of Diseases, 10th revision, Canadian edition (ICD-10-CA) codes for DVT and PE in the DAD using a previously validated methodology¹¹ (Table S1, Supplementary material). The presence of bleeding while in hospital was also identified, using previously identified ICD-10-CA codes¹¹, as bleeding is a potential consequence of VTE prophylaxis and treatment. Secondary outcomes included 30-day hospital readmission (dichotomous, abstracted from DAD), 30-day ED visits (dichotomous, abstracted from NACRS) and cumulative cost of physician visits (continuous).

The independent variable was ERAS compliance, which, for each element, was recorded in EIAS (dichotomous—compliant or not). The overall ERAS compliance score was a count variable and converted to a percentage (elements compliant/total possible elements × 100). The overall ERAS score was categorized for analysis as low (≤ 52% compliant), moderate (53–71% compliant) or high (≥ 72% compliant), based on prior studies assessing compliance and outcomes in ERAS^16,21,24. Variables that could be confounders of the association between VTE and ERAS compliance included age (continuous), sex (dichotomous), body mass index (BMI, continuous), ASA class (categorical; low = I–II and high = III–IV), surgery type (categorical), year surgery was completed, surgical complexity (dichotomous; low or high) and cancer diagnosis (dichotomous).

Statistical methods

Demographic characteristics of the cohort were reported using descriptive statistics; mean with standard deviation (mean(s.d.)) for normally distributed variables, and median with interquartile range (i.q.r.) for non-parametric variables. Differences between patients with and without VTE were compared using the chi-square test (or Fisher’s exact test if the assumptions of the chi-square test were violated) for categorical variables, and two-sided Student’s t tests for continuous variables. The association between VTE and ERAS compliance was explored using stepwise backwards logistic regression, adjusting for potential confounders including cancer diagnosis, surgery year, ASA group, surgical complexity, bleeding, age, sex, BMI and surgery type. A P value of < 0.050 was considered significant, a priori, for all analyses. To account for multiple comparisons, a Bonferroni correction was applied separately to the univariable analyses of demographic characteristics and for the univariable analyses of outcomes. All analyses were conducted in STATA-17²⁵.

Results

Cohort characteristics

There were 8118 patients included in the study (Fig. S1, Supplementary material). Baseline demographics are outlined in Table 1. Briefly, the median age of the cohort was 61 (i.q.r. 50–70) years old, most were female (n = 4951, 61.0%), and the median BMI was 27.7 kg/m² (i.q.r. 24–32.2). The most common surgical types were colorectal (n = 4288, 52.8%) and gynaecologic surgery (n = 2117, 26.1%), followed by pancreas (n = 559, 6.9%) and liver surgery (n = 4485.5%). Nearly half of the cohort had a cancer diagnosis (n = 3642, 44.9%) (Table 1).

Primary outcome (VTE)

There were 118 (1.5%) patients who experienced a postoperative VTE. Of those, 26 (22.0%) patients had a PE, 63 had a DVT (53.4%) and the remaining 29 (24.6%) patients had an unspecified VTE. The highest incidence of VTE was among patients undergoing pancreas surgery (2.3%) (Table S2, Supplementary material). More patients with a VTE had a high ASA score (44.7%) than those without a VTE (30.8%) (Table 1) but there were no other differences in patient characteristics between those who did and did not develop a VTE. There were also no differences in the proportion of patients that experienced bleeding in-hospital between those with a VTE and those without a VTE.

ERAS compliance

The overall ERAS compliance was high with 59.7% (n = 4847) of the cohort having high ERAS compliant care (≥72%) and 35.6% (n = 2809) having moderate ERAS compliant care (53–71%) (Table 2). The elements with the highest compliance included preoperative smoking recommendations (n = 6554, 99.4%), antibiotic prophylaxis guidelines (n = 7875, 98.4%), nasogastric tube guidelines (n = 7235, 96.3%) and forced heating guidelines (n = 7443, 96.3%). Elements with the lowest overall compliance included termination of intravenous (i.v.) drip (n = 3608, 44.9%), early mobilization guidelines (n = 4224, 53.1%) and preoperative oral carbohydrate guidelines (n = 5202, 66.7%).

Association between outcome and exposure

ERAS compliance differed between patients who did and did not develop a VTE (Table 2). More patients who developed a VTE had low (11.0%) or moderate (44.1%) ERAS compliance compared with those who did not develop a VTE (5.6% and 34.5% respectively, P = 0.001). When the association between compliance for each ERAS element and developing a VTE was examined, it was found that there were only differences between those who developed a VTE and those who did not with regards to compliance with termination of urinary drainage recommendations (57.9% versus 77.3%, P < 0.001). There was no difference in thrombosis prophylaxis compliance between the two groups.

Using stepwise backwards logistic regression, controlling for surgical complexity and ASA group, there was a 23% decrease in the occurrence of VTE for each one unit increase in the total compliance score (OR 0.79, 95% c.i. 0.71 to 0.88, P < 0.001) (Fig. 1). Overall, ERAS compliance and ASA category were associated with developing a VTE, an association that remained significant even after statistically controlling for cancer diagnosis, surgery year, bleeding, age, surgery type, BMI and sex (Table 3). The odds of developing a VTE were higher among patients with a high ASA class compared with patients with a low ASA class (OR 1.77, 95% c.i. 1.20 to 2.60, P = 0.004) and the odds of developing a VTE were lower among patients receiving higher ERAS-compliant care compared with those with lower ERAS compliant care (OR 0.79, 95% c.i. 0.71 to 0.88, P < 0.001).

Secondary outcomes

Patients who developed a VTE had higher rates of 30-day hospital readmission (26.3% versus 9.3%, P < 0.001) (Table 2). There was no difference in 30-day ED visits between patients with and without a VTE (4.2% versus 3.2%, P = 0.515). Experiencing a VTE was associated with significantly increased healthcare costs, with a median cost of 17 423 euros compared with 13 178 euros (P < 0.001).

Discussion

This multicentre study across several surgical types found that 1.5% of surgical patients developed a postoperative VTE, which resulted in higher 30-day hospital readmission and healthcare costs compared with patients who did not develop a VTE. An increase in ERAS compliance was associated with lower odds of developing VTE, with a 23% reduction in VTE for each one unit increase in the total ERAS compliance score. ERAS compliance remained an independent predictor of VTE even after adjusting for other variables such as cancer diagnosis, surgery year, bleeding, age, sex, BMI and surgery type. Compliance with the termination of urinary drainage guideline was the only ERAS element associated with decreased VTE.

Evidence suggests that ERAS compliance is associated with improved perioperative outcomes, including decreased complications and duration of hospital stay^16,18,21. A recent meta-analysis found a mean reduction in duration of hospital stay by 1.88 days for ERAS-guided surgeries across multiple disciplines compared with controls²⁶. The study also found a 30% reduction in risk of complications with ERAS-guided surgery²⁶. To our knowledge, only one other study has evaluated ERAS compliance and VTE; a study of patients undergoing ERAS-guided gynaecologic surgery identified a significant association between compliance with VTE prophylaxis, net fluid balance and development of postoperative VTE²⁷. In the present study, developing a VTE was not associated with compliance with VTE prophylaxis, suggesting that other factors are influencing the development of VTE. The present study found compliance with the early removal of urinary catheters, which is supported across multiple surgical disciplines including colorectal surgery^28,29, gynaecology³⁰, and head and neck surgery³¹, was associated with a reduction in developing a VTE. The early removal of urinary catheters is associated with decreased rates of urinary tract infections^29,32, shorter duration of hospital stay³³ and increased early mobilization^34,35, which may explain the observed decrease in VTE.

In this study, 1.5% of patients experienced a postoperative VTE, which is similar to other published estimates that include multiple surgical disciplines^5,10. VTE is the cause of up to 50% of perioperative deaths¹⁰, and contributes to the patient morbidity rate including hospital readmission, increased major bleeding, recurrent VTE and chronic complications such as postthrombotic syndrome^8,36,37. VTE is also known to contribute significantly to healthcare costs; it is estimated that hospital readmissions due to postoperative VTE cost a median of 7850 euros in excess compared with those who did not have a VTE³⁶. The present study demonstrated a median increase in healthcare cost of 4245 euros for patients who experienced a VTE. This increase in healthcare cost is likely attributable to testing, treatment and readmission for patients who experience a VTE. While the findings of the present study reflect those of the previous study, the cost was substantially lower in this study, likely due to the differences between publicly funded and privately funded costs.

Given the significant patient morbidity rate, mortality rate and healthcare costs associated with VTE, it is essential to find strategies to further decrease the risk of postoperative VTE. The findings from the present study suggest that ERAS compliance may be a modifiable factor that can be used to reduce the odds of developing a VTE^38,39. Despite ample evidence supporting the benefits of ERAS in reducing postoperative complications, duration of hospital stay and healthcare costs, gaps in ERAS implementation persist. Perceived barriers to ERAS implementation and compliance include resource limitations and resistance to change^40,41. Successful ERAS implementation requires a committed multidisciplinary team and ERAS facilitator, which requires upfront investment and resources⁴². However, ERAS implementation is associated with overall healthcare savings, with a potential return on investment as high as 7.3 for every dollar spent⁴³. Future research should consider how to overcome barriers to ERAS implementation to improve ERAS compliance and consequently developing VTE after surgery⁴⁴.

To the authors' knowledge, this is the first multicentre study to demonstrate an association between ERAS compliance and postoperative VTE across multiple surgical disciplines. The strengths of this study include the use of a large, multicentre cohort spanning multiple surgical disciplines. The cohort included multiple surgical types and a diverse patient population, allowing for broad applicability of these results. The use of two databases to identify VTE events improved the accuracy of estimating VTE incidence in the cohort. Furthermore, the use of provincial databases ensured that patients who may have presented to another hospital within the province were captured in this study.

As with most retrospective cohort studies that use existing data sets, this study has some limitations, which are largely related to the availability of variables of interest. For example, the out of hospital mortality rate was not available, therefore, if a patient died at home (for example, from a VTE), the death of these patients would not have been identified in the study. Also, details of the VTEs experienced were not available (for example the proportion that were symptomatic and diagnosed in the community versus hospital). Similarly, subclinical VTE may not have been diagnosed and included in the study, potentially underestimating the incidence of VTE. There are other potential factors, such as postoperative infection, that may be associated with the development of VTE that were not evaluated in the present study and are potential areas for future research. Guidelines including the American College of Chest Physicians guidelines and Society of Hematology guidelines recommend extended pharmacologic VTE prophylaxis for patients at high risk of thrombosis^4,7; however, data on use of extended VTE prophylaxis was unavailable for the cohort. It is possible that compliance with posthospital extended VTE prophylaxis may have influenced the development of VTE and may not have been captured in this study. VTE prophylaxis prescribing patterns, and compliance with VTE prophylaxis regimens, is an important area of consideration for future research.

In this multicentre retrospective cohort study, ERAS compliance was significantly associated with decreased odds of VTE. Future efforts should focus on improving ERAS adoption, which will translate to a reduction in complications such as VTE.

Funding

This work is supported by a Canadian Institutes of Health Research (CIHR) Project Grant. The funding source had no role in the design and conduct of the study.

Acknowledgements

K.M.S. and G.N. contributed equally to this study.

Disclosure

G.N. reports speaker and advisory board fees from Pfizer, AstraZeneca, GSK, and Smith & Nephew, outside the submitted work. G.N. is Physician Lead for ERAS Alberta and Co-Chair of Enhanced Recovery Canada. The authors declare no other conflicts of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

Data will be made available upon reasonable request to the corresponding author.

Author contributions

Kristin Black (Conceptualization, Methodology, Visualization, Writing—original draft, Writing—review & editing), Abby Thomas (Formal analysis, Investigation, Methodology, Visualization, Writing—review & editing), Khara Sauro (Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Writing—original draft, Writing—review & editing) and Gregg Nelson (Conceptualization, Funding acquisition, Methodology, Supervision, Writing—original draft, Writing—review & editing).

References

Author notes