Peripheral arterial disease (PAD) of the lower limbs is caused by atherosclerotic occlusive disease in which narrowing of arteries reduces blood flow to the lower limbs. PAD is common; it is estimated to affect 236 million individuals worldwide. Advanced age, smoking, hypertension, diabetes and concomitant cardiovascular disease are common factors associated with increased risk of PAD. Complications of PAD can include claudication pain, rest pain, wounds, gangrene, amputation and increased cardiovascular morbidity and mortality. It is therefore clinically important to use diagnostic tests that accurately identify PAD. Accurate and timely detection of PAD allows clinicians to implement appropriate risk management strategies to prevent complications, slow progression or intervene when indicated. Toe-brachial index (TBI) and toe systolic blood pressure (TSBP) are amongst a suite of non-invasive bedside tests used to detect PAD. Both TBI and TSBP are commonly utilised by a variety of clinicians in different settings, therefore a systematic review and meta-analysis of their diagnostic accuracy is warranted and highly relevant to inform clinical practice. To (1) estimate the accuracy of TSBP and TBI for the diagnosis of PAD in the lower extremities at different cut-off values for test positivity in populations at risk of PAD, and (2) compare the accuracy of TBI and TSBP for the diagnosis of PAD in the lower extremities. Secondary objectives were to investigate several possible sources of heterogeneity in test accuracy, including the following: patient group tested (people with type 1 or type 2 diabetes, people with renal disease and general population), type of equipment used, positivity threshold and type of reference standard. The Cochrane Vascular Information Specialist searched the MEDLINE, Embase, CINAHL, Web of Science, LILACS, Zetoc and DARE databases and the World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov trials registers to 27 February 2024. We included diagnostic case-control, cross-sectional, prospective and retrospective studies in which all participants had either a TSBP or TBI measurement plus a validated method of vascular diagnostic imaging for PAD. We needed to be able to cross-tabulate (2 x 2 table) results of the index test and the reference standard to include a study. To be included, study populations had to be adults aged 18 years and over. We included studies of symptomatic and asymptomatic participants. Studies had to use TSBP and TBI (also called toe-brachial pressure index (TBPI)), either individually, or in addition to other non-invasive tests as index tests to diagnose PAD in individuals with suspected disease. We included data collected by photoplethysmography, laser Doppler, continuous wave Doppler, sphygmomanometers (both manual and aneroid) and manual or automated digital equipment. Two review authors independently completed data extraction using a standardised form. We extracted data to populate 2 x 2 contingency tables when available (true positives, true negatives, false positives, false negatives). Where data were not available to enable statistical analysis, we contacted study authors directly. Two review authors working independently undertook quality assessment using QUADAS-2, with disagreements resolved by a third review author. We incorporated two additional questions into the quality appraisal to aid our understanding of the conduct of studies and make appropriate judgements about risk of bias and applicability. Eighteen studies met the inclusion criteria; 13 evaluated TBI only, one evaluated TSBP only and four evaluated both TBI and TSBP. Thirteen of the studies used colour duplex ultrasound (CDU) as a reference standard, two used computed tomography angiography (CTA), one used multi-detector row tomography (MDCT), one used angiography and one used a combination of CDU, CTA and angiography. TBI was investigated in 1927 participants and 2550 limbs. TSBP was investigated in 701 participants, of which 701 limbs had TSBP measured. Studies were generally of low methodological quality, with poor reporting of participant recruitment in regard to consecutive or random sampling, and poor reporting of blinding between index test and reference standard, as well as timing between index test and reference standard. The certainty of evidence according to GRADE for most studies was very low. Whilst a small number of diagnostic test accuracy studies have been completed for TBI and TSBP to identify PAD, the overall methodological quality was low, with most studies providing a very low certainty of evidence. The evidence base to support the use of TBI and TSBP to identify PAD is therefore limited. Whilst both TBI and TSBP are used extensively clinically, the overall diagnostic performance of these tests remains uncertain. Future research using robust methods and clear reporting is warranted to comprehensively determine the diagnostic test accuracy of the TBI and TSBP for identification of PAD with greater certainty. However, conducting such research where some of the reference tests are invasive and only clinically indicated in populations with known PAD is challenging.
Read full abstract