DVT and PE prophylaxes not needed in all patients after knee arthroscopy

Study finds few post-arthroscopic events, but notes certain patients are at risk.

Issue: August 2009

In our current medicolegal environment, we want not only what is best for our patients but also to be able to provide them with care that is based on facts. There are isolated cases and experiences that can alter any of our chosen and recommended treatments, but we must have large numbers to really assess risk factors.
Gregory B. Maletis, MD, has studied a potential problem following arthroscopic surgery that, at times, has touched most of us performing large numbers of this type of surgery. I invited him to share his insights and investigations with all of us this month.

Douglas W. Jackson, MD
Chief Medical Editor

Douglas W. Jackson, MD: You recently reported on the venous thromboembolism (VTE) incidence following 20,950 knee arthroscopy cases. What were the inclusion criteria for your study and your methodology of establishing DVT’s and the occurrence of pulmonary embolus (PE)?

Gregory B. Maletis, MD: Using the Kaiser Permanente integrated health care electronic medical record system, patients who underwent arthroscopic knee surgery within a 27-month period were identified. We searched the administrative database for procedure codes for arthroscopy with meniscal repair or debridement, synovectomy, or anterior cruciate ligament reconstruction and identified 21,794 patients. Exclusions were a previous history of DVT or PE, because we couldn’t be sure if a DVT or PE identified after surgery was a new finding or a residual change from a previous event. Pharmacy records were evaluated and patients were excluded if they received prophylactic treatment with anticoagulation medication given on the day of surgery. We also excluded patients who received any anticoagulation medication for any other medical condition within the 2 weeks prior to surgery. This left a cohort of 20,950 patients.

In order to identify patients with a symptomatic VTE, the diagnosis codes for thromboembolism were used and the electronic medical record was screened for thromboembolic events. Patients were screened for a 90-day postoperative period for inpatient stays or procedures, emergency room encounters, urgent care encounters or outpatient visits. Also, the electronic records were screened for any lower-extremity ultrasound test that had been performed or for any order of heparin, low molecular-weight heparin (LMWH), or warfarin sodium within the 90-day postoperative period. This was done to ensure that charts on all potential patients with a typical test used to diagnose a DVT or those who were being treated with an anticoagulation medication typically used to treat a DVT or PE would be captured. A duplex ultrasound consistent with a thrombus within the vein was required to be considered a positive DVT. A chest CT scan or a ventilation perfusion scan with moderate or high probability was required for the confirmation of a PE.

Jackson: What was the incidence of DVT and PE in your study group?

Maletis: We found that 52 patients (0.25%) suffered a DVT within 90 days of surgery; 41 were considered to have a proximal DVT (popliteal vein and proximal), and in 11 patients the thrombus was in a calf vein. A pulmonary embolus was noted in 35 patients (0.17%). Two patients were diagnosed with both a PE and DVT. The venous thromboembolism rate (DVT+PE) was 0.41%. This rate was found to be significantly higher in patients 50 years of age and older (0.51%) compared with those younger than 50 years (P = .04). We did not find significant differences based on sex or procedure code.

Eighteen patients died during the 90-day follow-up period. The mean age of this group was significantly older, 66 years vs. 44 years. Of those who died, 11 had arthroscopic surgery for irrigation and debridement of septic arthritis and each had multiple medical comorbidities. There was one autopsy-proven PE that resulted in the death of a patient who had an initial arthroscopy followed by an open osteochondral transplant. The cause of death in the remaining patients was not related to a PE.

Jackson: This is a large experience. What were some of the shortcomings of your study?

Maletis: The study is retrospective and relies on the accuracy of the administrative database. We have attempted to minimize any errors by taking advantage of the multiple administrative databases and validating thromboembolic events using chart review methodology. For example, the radiology database was screened for any patient who may have had a lower-leg ultrasound, and the pharmacy database was screened for any patient who may have been placed on anticoagulation medication postoperatively suggesting thromboembolism treatment. These charts were screened to assure that no positive events would be missed.

Gregory B. Maletis, MD
Gregory B. Maletis

The information within the database is limited and therefore does not let us comment on factors that may potentially affect the risk of thromboembolism such as length of surgery, tourniquet use, smoking history, body mass index or over-the-counter medication use such as aspirin or NSAIDs.

This study evaluated only symptomatic patients, which most surgeons would agree is the most clinically relevant group. Any patient having an asymptomatic DVT or PE would not have been identified. Patients may have been missed if they sought care outside of the Kaiser Health Plan system but because it is a prepaid system, they would most likely have had follow-up care and treatment within the plan and therefore were captured within the database search. Overall, 451 patients (2.1%) left the health plan within 90 days of their procedure and their data were not available.

Jackson: What guidelines would you recommend based on this data?

Maletis: The current literature has been inconclusive regarding the overall benefit of thromboprophylaxis after knee arthroscopy. Some studies have shown a decrease in the rate of DVT in patients receiving prophylaxis although many of the DVTs that are prevented may be clinically insignificant. In addition, there is a risk of bleeding and the added expense of the medication.

The Cochrane review found no strong evidence to conclude that thromboprophylaxis is effective in preventing thromboembolic events in people undergoing knee arthroscopy. The American College of Chest Physicians (ACCP) Evidence-Based Clinical Practice Guidelines (8th edition) states that unless patients have thromboembolic risk factors, no additional thromboprophylaxis other than early mobilization is necessary. In our large cohort of patients we found the rate of symptomatic DVT and PE to be very low after knee arthroscopy. We did find a higher risk in patients over the age of 50 years.

Our findings are in agreement with the ACCP guidelines and we do not routinely use, nor recommend, chemoprophylaxis in patients undergoing knee arthroscopy unless they have a prior history of DVT or PE or are at otherwise high risk. We do use sequential compression devices on the contralateral limb during surgery and in the recovery room and we attempt to mobilize patients rapidly. In patients with a prior history of DVT or PE we prefer to use adjunctive chemoprophylaxis.

For more information:

Gregory B. Maletis, MD, can be reached at 1011 Baldwin Park Blvd., Baldwin Park, CA 91706; 626-851-5904; e-mail: Gregory.B.Maletis@kp.org. He has no direct financial interest in any companies or products mentioned in this article.

Reference:

Maletis GB, Reynolds S, Inacio MCS. Incidence of thromboembolism after knee arthroscopy. Paper #340. Presented at the American Academy of Orthopaedic Surgeons 76th Annual Meeting. Feb. 25-28, 2009. Las Vegas.