Initiative Improves Adherence to Venous Thromboembolism Prophylaxis Recommendations, But Will It Actually Lower Risk of VTE?

Researchers present surprising new data at Hospital Medicine 2013 about venous thromboembolism prophylaxis implementation, along with novel approaches to integrating primary care and hospital management of very ill patients.

Surprising new data about venous thromboembolism prophylaxis implementation, along with novel approaches to integrating primary care and hospital management of very ill patients, were among the research highlights presented during a plenary session on Day 2 of the Society for Hospital Medicine’s 2013 annual conference, held May 17-19 at the National Harbor’s Gaylord Nelson Convention Center in Ft. Washington, MD. The plenary session featured three presentations highlighting innovative research, and featured participation by Stanford University’s Kelley Skeff, MD, PhD, a pioneer in improving medical teaching effectiveness.

The comprehensive care physician: A new physician model to care for the most ill?

As the hospitalist role has evolved, primary care physicians (PCPs) have become increasingly responsible for ambulatory care, resulting in role specialization that may confer some benefits to patients and providers, but which may also carry some disadvantages, especially for more medically fragile patients, for whom the connection with a PCP who knows them can have significant positive impact on inpatient care. Optimizing care for these individuals also targets the area where the most health care dollars are spent -- 85% of health care expenditures are incurred by 25% of the population of the US.

The University of Chicago’s David Meltzer, MD, PhD, spoke of his organization’s work funded by the Center for Medicare and Medicaid Innovation. This grant is trialing a care model that provides a tailored approach to general medical care, with low-use patients receiving care from a more traditional hospitalist/PCP model, and high-use patients receiving care from a comprehensive care physician (CCP), or primary care hospitalist. In this model, all hospital patients still receive care from physicians with significant hospital experience and presence, and physicians are still free to specialize. This approach is compatible with the patient-centered medical home model, and should help eliminate readmission penalties.

Using a model where CCPs assume care for patients who are expected to spend 10 or more days per year in the hospital, these CCPs spend mornings on the hospital wards, have some buffer time, and then move to the clinic in the afternoon. Patient panel per physician is capped at about 200, allowing for intensive treatment of these at-risk patients. Team-based coordinated care, post-discharge follow-up and IT support, clinically relevant continuing education and training, and psychosocial support for providers are all part of the care model as well. Currently, 135 patients are enrolled, and preliminary qualitative and quantitative results are “positive” with no data yet published.

Real-time dashboards, pay-for-performance improve hospitalist venous thromboembolism prophylaxis rates

Henry Michtalik, MD, MPH, MHS, of Johns Hopkins University reported on a two-pronged QI effort aimed at improving physician compliance with venous thromboembolism (VTE) prophylaxis recommendations in practice. Approximately two million Americans experience VTE each year, with over half occurring during hospitalization or within 30 days post-discharge. Each VTE event costs between $10,000 and $20,000, and reduction of VTE is an important quality improvement target. With the movement toward value based purchasing emphasizing transparency in performance metrics, physician adherence to VTE prophylaxis guidelines becomes especially important.

The study design sequentially examined the effects of a physician dashboard and a pay for performance (P4P) program to improve appropriate VTE prophylaxis rates among hospitalized patients at a tertiary care medical center. The hypothesis that both interventions would increase appropriate implementation of VTE prophylaxis by hospitalists was supported. At baseline, physicians had use only of decision support at the point of CPOE to guide implementation of VTE prophylaxis for inpatients. Interventions included a dashboard that showed both individual and group performance, followed by a pay for performance program. Analysis involved determining the percent of physicians who demonstrated compliance with risk-appropriate VTE prophylaxis at baseline and as the interventions were implemented, and included methods to account for potential confounders.

The rate of physician improvement in compliance tripled with the introduction of the dashboard, and improved again when P4P was introduced. After both interventions were initiated, all physicians studied met minimum compliance criteria of 80%, and approximately half achieved the maximum measured compliance rate (>95%). For the 19 physicians studied, the total P4P payout, approximately $12,000, amounted to less than the locale-specific cost of a single VTE event. The program was felt to be cost-effective.

Venous thromboembolism incidence not associated with hospitals’ pharmacologic prophylaxis rates

Scott Flanders, MD, of the University of Michigan then presented results of the Michigan Hospital Safety Consortium’s work to determine whether hospital compliance with pharmacologic venous thromboembolism (VTE) prophylaxis is associated with a decreased rate of VTE. The consortium, comprised of 35 diverse hospitals in the state, collected data regarding risk factors for VTE and pharmacologic prophylaxis for 800 patients per site per year.

Through a combination of medical record review and phone interviews, researchers determined VTE outcomes, following subjects to 90 days after discharge. High-risk patients (Caprini score ≥ 2) were included. Hospitals were placed into performance tertiles according to pharmacologic prophylaxis rates: high ≥ 80%, moderate <80%, low <65%. In discussing results which Skeff termed “dropping a bomb,” Flanders revealed that although there was significant stratification of performance rates, there was no association between performance rate and risk of VTE, either during or subsequent to hospitalization.

In further elaboration of these surprising results, Flanders noted that AHRQ recommendations for VTE prophylaxis for inpatients, for example, may result in recommending near-universal prophylaxis. Further research should focus on continued exploration of unintended as well as intended consequences of practice recommendations, and should also continue to focus on clinical outcomes.