Comparison of hand-held acoustic Doppler with point-of-care portable color Doppler ultrasound in the assessment of venous reflux disease.

OBJECTIVE

Diagnostic-quality portable color Doppler ultrasound (PCD) offers convenient point-of-care venous reflux disease (VRD) diagnosis. Philips Lumify (Philips N.V., Best, The Netherlands), a high-fidelity broadband linear array transducer (4-12 MHz frequency), connects through a web-enabled smartphone or tablet to cloud software and offers B-mode and color Doppler imaging without pulsed wave Doppler capability. The aims of the study were to compare hand-held acoustic Doppler (HHD) vs PCD diagnostic performance using conventional duplex ultrasound (DUP) as the "gold standard" for VRD assessment, to assess effects of body mass index (BMI) and disease severity on diagnostic performance of HHD and PCD, and to determine whether PCD offers any diagnostic improvement over HHD in VRD assessment.

METHODS

There were 241 patients (65 male, 176 female; mean age, 55.5 ± 15.5 years; mean BMI, 32.2 ± 7.9 kg/m). DUP (447 legs), PCD (262 legs), and HHD (217 legs) studied the great saphenous vein at above-knee (AK) and below-knee (BK) levels. A phlebologist performed HHD, whereas PCD and DUP were performed sequentially (PCD first) by an experienced technologist and interpreted independently. PCD was done blinded to DUP results. DUP findings were analyzed blinded to HHD and PCD results. Venous reflux was dichotomously assessed as <2 seconds and >2 seconds.

RESULTS

HHD improves from moderate to good sensitivity from AK level (68%) to BK level (94%) but suffers poor specificity that declines significantly from AK level (50%) to BK level (12%; P < .05). HHD positive predictive value exceeds its negative predictive value (NPV) and remains unchanged from AK level (71%) to BK level (72%). HHD NPV remains consistently poor at AK (48%) and BK (42%) levels. PCD has similar sensitivity from AK level (69%) to BK level (74%), better AK level (79%) vs BK level (58%) specificity (P < .05), similar positive predictive value for AK (76%) and BK levels (78%), and better NPV for AK level (72%) vs BK level (53%; P < .05). BMI range (<30 kg/m vs ≥ 30 kg/m) did not influence diagnostic performance of HHD and PCD significantly. HHD and PCD specificity was higher for Clinical, Etiology, Anatomy, and Pathophysiology (CEAP) class <4 compared with CEAP class ≥4 (P < .05).

CONCLUSIONS

The relative diagnostic performance of HHD and PCD is highly dependent on insonation level. PCD advantages compared with HHD are marginally greater specificity at AK and BK levels and better NPV at AK level. Compared with HHD, PCD's disadvantage is lower sensitivity at BK level. Both HHD and PCD have higher specificity at AK level than at BK level. Overall, PCD offers only moderate sensitivity and specificity, making it inadequate for exclusion of significant venous reflux. Neither obesity nor CEAP class significantly influenced the general diagnostic performance of PCD or HHD.