Proximal and isolated distal deep vein thrombosis and Wells score accuracy in hospitalized patients.

ABSTRACT

Deep vein thrombosis (DVT) is an important cause of morbidity and mortality in hospitalized patients. The Wells score for DVT pretest probability (PTP) was validated in outpatients, but its utility for inpatients is unclear. The aim of this study was to establish the prevalence of inpatient proximal and distal DVT and the Wells score performance in inpatients. A single-center cross-sectional study was conducted in a university hospital. During 183 days, all inpatients with suspected lower-extremity DVT were evaluated with the Wells score and whole-leg ultrasound. Among 634 inpatients (age 77.5 ± 13.8 years, males 39.3%), 507 (80.0%) were from medical wards and 127 (20.0%) from surgical wards. During the study period, there were 11,662 hospital admissions in the surgical/medical services. Whole-leg ultrasound detected 128 DVTs (20.2%); 51 (39.8%) were proximal and 77 (60.1%) were isolated distal DVTs. Estimated DVT prevalence in hospital setting was 1.09% (95% CI 0.93-1.31), and isolated distal DVT prevalence was 0.66% (95% CI 0.53-0.82). DVT frequency in low-, moderate-, and high-PTP groups was 9.8%, 24.3%, and 41.5%, respectively (p = 0.001). The area under the receiver operating characteristic curve for the Wells score was 0.67 ± 0.03 for all DVTs and 0.75 ± 0.04 for only proximal DVTs. A high PTP had a sensitivity of 24% (95% CI 14-37%) and a specificity of 93% (95% CI 91-95%) for proximal DVT diagnosis. In hospitalized patients, isolated distal DVT has a higher incidence than expected, and the Wells score accuracy for proximal DVT is similar to that found in outpatients.