Effectiveness of Standardized Venous Thromboembolism Prophylaxis Order Sets in Patients Undergoing Abdominal or Pelvic Surgery for Cancer

In our tertiary academic medical center, standardized pre-printed physician order sets were implemented in 2008 for inpatient venous thromboembolism (VTE) risk stratification and prophylaxis on the day of admission. Electronic medical records were not utilized during the study period. Recommendations for pharmacologic prophylaxis were derived from the 2008 American College of Chest Physicians (ACCP) guidelines. Patients undergoing abdominal or pelvic surgery for cancer were specifically identified in the highest risk category with enoxaparin, fondaparinux, or warfarin (goal INR 2–3) provided as options for pharmacologic VTE prophylaxis. Sequential compression devices (SCDs) were also recommended. This retrospective analysis evaluates adherence to risk stratification and recommendations for pharmacologic VTE prophylaxis in patients undergoing abdominal or pelvic surgery for cancer.

Using the hospital administrative database, all patients ≥ 18 years hospitalized between January 1, 2009 and December 31, 2010 were identified by ICD-9 codes for cancer and surgery performed. Patients undergoing an abdominal or pelvic surgery for cancer lasting >45 minutes were included. Exclusion criteria were age >90 years, patients requiring therapeutic anticoagulation, length of hospitalization <24 hours or >30 days, transplant recipients, patients requiring multiple visits to the operating room during hospitalization, laparoscopic or transurethral procedures, and esophagectomy. Primary outcomes included adherence to order set risk stratification and pharmacologic VTE prophylaxis recommendations, and timing, type, and dose of pharmacologic VTE prophylaxis prescribed. Chart review of physician orders for mechanical and pharmacologic prophylaxis for each day of hospitalization was completed on all patients. Operative reports, medication administration records, and progress notes were reviewed if clarification of physician orders was required. Administration of the prescribed medication during hospitalization was not verified.

A total of 773 hospital admissions representing 767 patients met inclusion criteria. The mean length of hospitalization was 7 +/− 5 days during the study period. The mean age was 62 +/− 12 years and 51% were female. Pharmacologic prophylaxis was prescribed 5645 of 6147 patient days (91.8%). Enoxaparin was prescribed 630 (9.8%) patient days, and fondaparinux in 15 patient days (0.002%). Although not indicated as an option for this patient population in our order set, low dose unfractionated heparin (LDUH) was prescribed 4991 (81.2%) of the patient days and was to be dosed every 8 hours (80%) or every 12 hours (19.8%). Warfarin was not prescribed. Mechanical thromboprophylaxis with SCDs was ordered 93.7% of the patient encounters on the day of surgery.

The standardized pre-printed physician order set was utilized for prescribing prophylaxis at the time of admission for 630 (82.1%) patient encounters. Clinicians risk stratified only 350 (45.3%) of these patient encounters per the order set guidelines. Of those, 131 (37.4%) were appropriately deemed high risk based upon “abdominal/pelvic cancer undergoing operative procedure”. In the high risk group, LDUH was prescribed in 90 (68.7%) patient encounters, and enoxaparin as recommended in 30 (22.9%). Eleven (8.4%) were not prescribed pharmacologic prophylaxis on the day of admission.

Completion of standardized pre-printed physician order sets with appropriate risk stratification in patients undergoing major, open abdominal or pelvic surgery for cancer is underutilized in our institution. Although pharmacologic prophylaxis is prescribed for the majority of inpatient days, LDUH is primarily ordered despite preferred recommendations in favor of enoxaparin or fondaparinux. Thus, an opportunity exists to implement an educational intervention to improve adherence to evidence based order set risk stratification and pharmacologic VTE prophylaxis recommendations.

No relevant conflicts of interest to declare.