ICU patients have a 5-10% risk of deep venous thrombosis, even with appropriate prophylaxis. Ultrasound screening can detect many of these asymptomatic and clinically unsuspected DVTs. However, since many DVTs disappear without incident, and complications can result from additional testing and treatment with anticoagulation, and all of this costs money, the best approach to prevention and screening for deep venous thrombosis/pulmonary embolism in the ICU is unclear.
Using a Markov decision analytic model, Sud et al compared weekly leg ultrasound screening against a hypothetical program to increase adherence to DVT prophylaxis. Their probabilities and assumptions about the incidence of DVT and PE, operating characteristics of ultrasound, rates of complications, etc. came from systematic literature reviews (their own and others’).
The result of their modeling was that:
- Ultrasound screening resulted in 3 fewer PEs per 1,000 patients, but two additional hemorrhages, and cost $223,800 per quality-adjusted life year gained ($50,000 considered cost-effective).
- Increase in adherence with thromboprophylaxis by 10% resulted in 16 fewer DVTs, 1 fewer PE, 1 additional HIT and bleeding episode, and cost $27,953 per QALY.
There are a huge number of assumptions here, so I would have to ask Markov what he thought of their methodology. It does give me pause before deciding to roam around the ICU with the portable ultrasound, screening asymptomatic patients.
Sud S et al (E-PROTECT and Canadian Critical Care Trials Group). Screening and Prevention of Venous Thromboembolism in Critically Ill Patients: A Decision Analysis and Economic Evaluation. Am J Resp Crit Care Med 2011; 184:1289-1298.