Computer programs reliable when assigning anticoagulants

Two computer programs were at least as adept as experienced medical staff at assigning anticoagulants and even more successful at assigning oral anticoagulant dosage (mainly warfarin) in most cases, according to the results of a recent trial published in the Journal of Thrombosis and Haemostasis.

A team of researchers at 32 sites in 13 countries in Europe, Australia and Israel, most based in the United Kingdom, conducted a 4.5-year randomized study at 32 sites in 13 countries to test the safety and effectiveness of the PARMA 5 and the DAWN AC computer-assisted dosing programs. The programs were compared with experienced staff at centers with an established interest in oral anticoagulation.

The centers recruited 13,052 patients, with 6,503 randomly assigned to the manual arm and 6,716 to the computer-assisted arm. One hundred and sixty-seven patients were excluded because of incomplete records.

Leon Poller, MD, said it was the largest study ever conducted in oral anticoagulation.

“We found that the computer dosage was as safe and effective as the dosage prescribed by doctors at experienced centers,” he told HemOnc Today. “We particularly picked experienced centers with a reputation of international expertise in anticoagulation.”

The report concluded that there was a greater incidence of clinical events in the manual dose arm (n=555) compared with the computer-assisted arm (n=513), but determined that difference was not statistically significant after adjusting for covariates. Likewise, the researchers found no significant differences in the results for patients with atrial fibrillation, mechanical heart valves or other indications.

However, the study team did find a significant difference in clinical events of bleeding or further thrombosis between the two arms for the 3,209 patients with deep vein thrombosis and/or pulmonary embolism. Patients in the computer-assisted arm had a lower adjusted incidence ratio (IRR=0.67%; 95% CI, 0.52-0.85) compared with patients in the manual arm (P <0.001).

The researchers also found a significant improvement in time in target International Normalized Ratio between the two study arms (1.2%) in the computer-assisted arm.

Poller cautioned that only two computer programs were evaluated in this study, and the results should not be taken as a blanket endorsement of all such systems. He went on to say that provided there is access to a good, validated computer program, technology should replace physicians when assigning oral anticoagulants. – by Jason Harris

From a medical perspective, these are experienced practitioners and experienced clinics comparing their performance to a software dosing algorithm. In the United States, the care of patients taking anticoagulants is much more fragmented than it is in Europe — we have a larger number of practitioners with less experience doing the dosing. The thing I would take away from this study is that these computer-assisted programs function pretty well when compared with clinicians that are as good at this as one gets to be.

This study did not take into account any pharmacogenomic information which is now beginning to work its way into the initial dosing of warfarin, so it is already outdated in some ways. If you combined these computer-assisted algorithms with a pharmacogenomic initial dosing strategy, you may have the best possible safety record.

–Nigel Key, MD

HemOnc Today Editorial Board member

For more information:
Poller L, Keown M, Ibrahim S, et al. An international multicenter randomized study of computer-assisted oral anticoagulant dosage vs. medical staff dosage. J Thromb Haemost. 2008;6:935-943.