Analysis of Ordering Practice of Thrombophilia Testing at a Single Large Teaching Hospital

Thrombophilia describes either an inherited or acquired predisposition to thrombosis. There are no well defined guidelines for thrombophilia testing. Thus, testing after a single episode of thrombosis in unselected patients has become a common yet often inappropriate practice. An ideal work-up includes: antithrombin (AT) activity, protein C (PC) activity, protein S (PS) activity and antigen (total and free), activated protein C resistance (APC-R), factor V Leiden (FVL) if the APC-R is abnormal, prothrombin gene mutation (PGM), factor VIII (FVIII) and antiphospholipid antibodies (aPL) (lupus anticoagulant, anti-cardiolipin, and anti-β₂ glycoprotein I, IgG, M and A).

The purpose of this retrospective observational study was to characterize the pattern of such thrombophilia testing at a single large teaching hospital.

All laboratory requests for thrombophilia testing in October and November of 2009 were included in this retrospective evaluation. The tests included were PC, PS, AT, FVIII, FVL, PGM, homocysteine and aPL. Clinical information was extracted from electronic medical records and analyzed for indication and timing of testing, completeness of the tests, whether the tests were performed on anticoagulation, and whether an abnormal test result was repeated to confirm the diagnosis.

The study consisted of 100 consecutive patients (69 females, 68 less than 50 years of age). 70 patients had a known event prior to thrombophilia testing, 34 had an inappropriate indication [provoked thrombosis, 1–2 pregnancy losses, pre-eclampsia] (Table 1). 88% of the patients who were tested had an incomplete workup for the indication. Of the 70 patients who had an identifiable event prior to testing for thrombophilia, 41 patients were tested at the time of the event (<1 week) when natural anticoagulants could be consumed and results will not alter the management. Of all 100 patients, 30 were tested after anticoagulant therapy was initiated, which is known to affect results of several laboratory tests listed above; 8 of these patients had abnormal results and only 2 had repeat testing to confirm. 53% of all patients in the cohort tested positive; 20 of which underwent repeat testing, with half showing normalization (Table 2). Of the 36 patients tested for the appropriate indications [unprovoked thrombosis, or ≥3 pregnancy losses], only 3 had complete testing; 20 were tested at the time of the acute event, and 13 were tested on anticoagulant.

Only 36% of patients had thrombophilia testing done for the appropriate thrombotic or obstetric indications. 10% had confirmed positive tests, while 33% had an abnormal result that remained unconfirmed. Thrombophilia testing at our center deviates markedly from what would be considered appropriate in terms of indication, completeness, timing from thrombotic event, testing off anticoagulation and necessity of repeat testing. Ideally testing should be performed on young patients with unprovoked events, off anticoagulation to allow factor levels to return to baseline, and repeated for all abnormal results. Building on these observations, this study aims to devise standards of practice to minimize cost and prevent misdiagnosis in thrombophilic patients.

No relevant conflicts of interest to declare.