Effectiveness of bedside investigations to diagnose peripheral artery disease among people with diabetes mellitus: A systematic review.

The accurate identification of peripheral artery disease (PAD) in patients with diabetes and foot ulceration is important, in order to inform timely management and to plan intervention including revascularisation. A variety of non-invasive tests are available to diagnose PAD at the bedside, but there is no consensus as to the most useful test, or the accuracy of these bedside investigations when compared to reference imaging tests such as magnetic resonance angiography, computed tomography angiography, digital subtraction angiography or colour duplex ultrasound. Members of the International Working Group of the Diabetic Foot updated our previous systematic review, to include all eligible studies published between 1980 and 2018. Some 15 380 titles were screened, resulting in 15 eligible studies (comprising 1563 patients, of which >80% in each study had diabetes) that evaluated an index bedside test for PAD against a reference imaging test. The primary endpoints were positive likelihood ratio (PLR) and negative likelihood ratio (NLR). We found that the most commonly evaluated test parameter was ankle brachial index (ABI) <0.9, which may be useful to suggest the presence of PAD (PLR 6.5) but an ABI value between 0.9 and 1.3 does not rule out PAD (NLR 0.31). A toe brachial index >0.75 makes the diagnosis of PAD less likely (NLR 0.14-0.24), whereas pulse oximetry may be used to suggest the presence of PAD (if toe saturation < 2% lower than finger saturation; PLR 17.23-30) or render PAD less likely (NLR 0.2-0.27). We found that the presence of triphasic tibial waveforms has the best performance value for excluding a diagnosis of PAD (NLR 0.09-0.28), but was evaluated in only two studies. In addition, we found that beside clinical examination (including palpation of foot pulses) cannot reliably exclude PAD (NLR 0.75), as evaluated in one study. Overall, the quality of data is generally poor and there is insufficient evidence to recommend one bedside test over another. While there have been six additional publications in the last 4 years that met our inclusion criteria, more robust evidence is required to achieve consensus on the most useful non-invasive bedside test to diagnose PAD.