Assessing Level of Agreement Between 3 Risk Stratification Models to Prevent Venous Thromboembolism in Hospitalized Medical Patients

1Henry Ford Hospital, Detroit, MI

Meeting: Hospital Medicine 2011, May 10-13, Dallas, Texas.

Abstract number: 113

Background:

Venous thromboembolism (VTE) is a common preventable cause of inpatient mortality. Pharmacologic methods are the preferred mode of prophylaxis according to guidelines. Various risk‐assessment tools are available to assist physicians in identifying patients who should receive VTE prophylaxis when hospitalized. The objective of this study was to determine the level of agreement between 3 published VTE risk stratification models.

TABLE 1 Distribution of Patients Based on Risk Assessment Method

Methods:

Adult patients newly admitted to the general medical service were assessed and were categorized as low, moderate, or high risk for VTE based on the Caprini, Kucher, and Maynard risk stratification models. The level of agreement between the 3 models was determined using the weighted kappa statistic.

Results:

One hundred and eighty‐four patients were assessed, and data were analyzed on 128 patients. Those who were admitted with deep vein thrombosis or pulmonary embolism (11 patients) were excluded. An additional 45 patients (24%) had to be excluded because of missing clinical data that prevented calculation of 1 or more of the scores or assignment to a risk category. There was poor agreement between the Caprini and Kucher (к = 0.205), Caprini and Maynard (к = 0.165), and Kucher and Maynard risk assignments (к = 0.051). The Maynard scale identified most patients (91%) as moderate risk, whereas the Kucher scale scored 76% of patients as low risk. The Caprini scale never classified a patient at a lower risk when compared with the Kucher scale. The rates of pharmacologic prophylaxis and of any prophylaxis did not correlate with the level of VTE risk as determined by the 3 scales.

Conclusions:

We found poor agreement between 3 published risk assessment models for VTE. The rates of prophylaxis in all risk categories were generally high in our institution, where we follow an “opt‐out” policy, and did not correlate with the level of risk as determined by the 3 models. Future studies are needed to validate the most useful tool in clinical practice.

TABLE 2 Proportion of Patients in Each Risk Category Receiving Prophylaxis

Disclosures:

S. Shah ‐ none; S. Patel ‐ none; J. Rana ‐ none; H. Singh ‐ none; E. Sabia ‐ none; D. Paje ‐ none; S. Kaatz ‐ none

To cite this abstract:

Shah S, Patel S, Rana J, Singh H, Sabia E, Paje D, Kaatz S. Assessing Level of Agreement Between 3 Risk Stratification Models to Prevent Venous Thromboembolism in Hospitalized Medical Patients. Abstract published at Hospital Medicine 2011, May 10-13, Dallas, Texas. Abstract 113. Journal of Hospital Medicine. 2011; 6 (suppl 2). https://www.shmabstracts.com/abstract/assessing-level-of-agreement-between-3-risk-stratification-models-to-prevent-venous-thromboembolism-in-hospitalized-medical-patients/. Accessed July 21, 2019.

« Back to Hospital Medicine 2011, May 10-13, Dallas, Texas.