Abstract
Background
In October 2008, CMS instituted a new “No payment for preventable complications“ programme and has released a list of conditions for which it intends to expand the programme in 2009. Although not reimbursing for preventable complications is justifiable, some of the proposed target conditions are lacking in both adequate diagnostic testing accuracy and preventability.
Aims
This study examines the effects of imperfect sensitivity and specificity of diagnostic testing, the prevalence of condition, and the rate of surveillance on the ratio of numbers of DVT/PE diagnosed and those that actually occur.
Methods
Given that proximal DVT following orthopaedic surgery are not preventable (incidence 4-10% despite prophylaxis), and that the accepted screening test (duplex ultrasonongraphy) has a sensitivity of 96.5% and specificity of 94%, up to 2.5 times as many DVT will be diagnosed as actually occur. Since the new CMS policy would withhold payment when a complication is identified, hospitals would be encouraged to decrease screening for DVT, which would save them lost reimbursement but would result in an increase in incidence of unrecognised – and therefore untreated – DVT after major orthopaedic surgical procedures. The study examines the likely CMS savings in the United States for implementation of this proposed policy (estimated $300M annually), as well as the potential increase in undiagnosed DVT likely to result from implementation of the policy. The overall goal of this study is to provide information to better enable the reader to draw their own conclusions on whether non-payment for DVT after an orthopaedic procedure is a fiscal policy that makes sense and improves the health and healthcare of Americans.
