The Wells score is only slightly better than a coin toss for predicting deep vein thrombosis (DVT) in hospitalized patients, researchers have found.
"The Wells score risk stratification is not sufficient to rule out DVT or influence management decisions in the inpatient setting," Dr. Patricia C. Silveira from Brigham and Women’s Hospital in Boston says.
Although the Wells score has been validated in outpatient and ED settings, it has not been studied in hospitalized patients in a large prospective trial, Dr. Silveira and her colleagues note in JAMA Internal Medicine, online May 18.
The team evaluated the utility of the tool for risk stratification of inpatients with suspected DVT in a prospective study that included more than 1,100 patients. About one in eight were found on lower-extremity venous duplex ultrasound studies (LEUS) to have proximal DVT and 9.2% to have distal DVT.
The incidence of proximal DVT in the low, moderate, and high Wells pretest probability groups was 5.9%, 9.5%, and 16.4%, respectively, a much narrower range than was previously reported for outpatients (3.0%, 16.6%, and 74.6%).
The AUC for the discriminatory accuracy of the Wells score for risk of proximal DVT identified on LEUS was only 0.6 (a coin toss would yield a predicted AUC of 0.5), the researchers found.
Results were even less informative for distal DVT, where low, moderate, and high pretest probability groups had DVT incidences of 7.4%, 9.1%, and 9.7%, respectively.
"Physician should use their clinical judgment to order lower extremity ultrasound studies for hospitalized patients with suspected DVT," Dr. Silveira concludes. "A new clinical decision rule might be useful to determine a patient's pre-test probability of DVT in the inpatient setting."
Dr. Erika Leemann Price from San Francisco Veterans Affairs Medical Center in California, who coauthored an invited commentary on the new report, told Reuters Health by email, "In the inpatient setting, the Wells score for DVT doesn't do a good job of telling us who has a DVT and who doesn't. Inpatients are different from outpatients in that they are at greater risk for DVT overall, but they also have multiple other comorbidities that can mimic the signs and symptoms of DVT.
"We don't currently have a validated clinical prediction model for DVT in the inpatient setting, although there is clearly a need for one that includes factors more predictive of VTE (venous thromboembolism) specifically in inpatients," Dr. Price says. "For now, if you are worried that your hospitalized patient may have a DVT, skip the Wells score and get an ultrasound." TH