Diagnostic devices in the undergraduate curriculum.

Andreas Demetriades (May 2004 JRSM1) comments on the difficulty educators face in deciding which clinical techniques deserve preservation in an age of imaging technologies. Even Laennec faced opposition from those who saw the stethoscope as a threat to traditional practice. Today a case in point is the diagnosis of varicose veins. Management decisions depend on assessment of valvular incompetence, for which medical students are taught the tap test, the cough test and the Trendelenburg tourniquet test. Yet in practice vascular surgeons use hand-held doppler probes and perform duplex scans as necessary. We reviewed the published work regarding accuracy of these clinical tests. In 1984 McIrvine et al.2 compared clinical tests for saphenofemoral incompetence with findings at surgery in 105 patients. The simplest and most accurate method was the tourniquet test plus hand-held doppler (sensitivity 0.90, specificity 0.45). More recently, De Palma et al.3 found the sensitivities of clinical tests and hand-held doppler to be about the same (0.48) but the addition of doppler to clinical tests raised the sensitivity and increased positive predictive value to 83%. Singh et al.4 asked surgeons ‘hypothetically’ to commit to one of a selection of management protocols having seen findings at preoperative assessment. Inappropriate surgery would have been performed in 20% of limbs on the basis of clinical examination with a tourniquet and in 13% on the basis of tourniquet examination plus hand-held doppler. Kim et al.,5 using duplex scanning as their gold standard, found that the sensitivities of clinical tests ranged from 0.18 to 0.97 and specificities from 0.20 to 0.92. Hand-held doppler was far superior, with sensitivity 0.97 and specificity 0.73 at the saphenofemoral junction. On the above evidence, clinical tests alone are of low value in assessment of varicose veins and medical education should embrace hand-held doppler.