Back to 2015 Annual Meeting

---

Prevention of Postoperative Venous Thromboembolism after Hospital Discharge: A Risk Assessment and Adherence Protocol for Thoracic Surgical Patients
Krista J. Hachey, Daniel Choi, Philip D. Hewes, Emma Pinjic, Hiran C. Fernando, Virginia R. Litle
Boston Medical Center, Boston, MA

Objective: Thoracic surgical patients are among the highest risk for postoperative venous thromboembolism (VTE) and the mortality rate with VTE can reach up to 20%, particularly in lung resection patients. A retrospective study in our Thoracic Surgery Division demonstrated a VTE rate of 5.17% for lung resection patients, with one third of events occuring post-discharge. The Caprini VTE Risk Assessment Model (RAM) has been validated in multiple surgical specialties to aid in patient selection for extended course prophylaxis to prevent post-discharge VTEs. Per RAM protocol, patients at high risk for VTEs (Caprini scores 9+) are candidates for 30 days of total post-operative anticoagulation, which has been shown to reduce VTE rates by 60%. Those deemed intermediate risk (scores 5-8) are candidates for 10 days of total post-operative prophylaxis. We have previously demonstrated the predictive ability of the Caprini RAM in thoracic surgical patients and hypothesize that implementation of the Caprini protocol is feasibile for providers and patients. Furthermore, we predict that adherence to the RAM recommendations could reduce the rate of post-discharge VTEs without increased bleeding complications. The primary aim of this prospective study is to assess provider and patient adherence to Caprini RAM recommendations. The second aim is to evaluate VTE outcomes and adverse bleeding events associated with extended course prophylaxis.

Design: IRB-approved single institution prospective cohort study of all surgical patients within the Division of Thoracic Surgery. Funded study initiated in July 2014 with ongoing accrual.

Setting: Tertiary care academic medical center, largest safety net hospital in New England.

Patients: All thoracic surgical patients were included with the exception of bronchoscopic/endoscopic-only cases.

Intervention(s): All staff on the thoracic surgical service received training on risk stratification using a Caprini RAM integrated into the electronic medical record. Per protocol, risk screening was to begin at the time of hospital admission after a surgical event, throughout hospitalization, and at the time of discharge when decisions regarding extended course prophylaxis are made. Study interventions included provider adherence monitoring via chart review documenting completion of risk assessment and appropriate prophylaxis. Patient adherence was evaluated at the first post-operative clinic visit via interviews and patient knowledge questionnaires. Outcomes monitoring was conducted using a combination of chart review and patient assessment at the first post-operative visit.

Main Outcome Measure(s): Main outcomes measured include provider adherence to the RAM protocol and patient adherence to prophylaxis recommendations as well as knowledge regarding VTE prevention. Secondary outcomes include clinically significant VTE events and bleeding complications.

Results: Out of a total of 85 patients thus far, 21.4% scored high risk and 46.4% were intermediate risk by the RAM (fig). Adherence to risk assessment prior to discharge by the inpatient care team was 99% (84/85). Appropriate prophylaxis was prescribed by the care team in 97.6% (83/85) of cases, with one non-adherent case due to an increased bleeding risk identified by the care team. Of patients discharged on extended course prophylaxis, 83.8% (46/49) were adherent to daily lovenox injections. Only 65.3% (32/49) of patients demonstrated a clear understanding of extended course prophylaxis had been prescribed to manage VTE risk. There have been no post-discharge VTEs, including within the subset of lung resection patients (n=21). However, the overall VTE rate is 2.3% (2/85). There has been one clinically significant inpatient PE with DVT identified in an esophagectomy patient with a Caprini score of 10 (high risk) prior to symptom onset with PEA arrest. A second inpatient PE was identified near a surgical staple line in a lobectomy patient (Caprini score 6, intermediate risk) during a work up for hypoxia with imaging findings consistent with aspiration pneumonia. Although not thought to be the cause of hypoxia, the thrombus was treated with therapeutic anticoagulation; no DVT was identified on duplex screening ultrasound. There have been no adverse bleeding events on extended course prophylaxis.

Conclusions: This prospective quality improvement study demonstrates that routine VTE risk assessment using an electronic Caprini scoring system is achievable by a multidisciplinary thoracic surgical service with high rates of adherence. Furthermore, patient adherence to recommendations is also acceptable. We propose that incorporation of oral anticoagulants such as rivaroxaban (Xarelto) into RAM recommendations may improve patient adherence to extended course prophylaxis. Deficits in patient knowledge are apparent and provide an area for improvement. There have been no outpatient VTE events since the institution of the RAM protocol, however, inpatient VTEs occurred despite routine anticoagulant prophylaxis. Therefore, Caprini risk assessment throughout the perioperative period may also be useful for targeted screening ultrasound in highest risk patients to help prevent symptomatic inpatient VTEs.

---

Back to 2015 Annual Meeting