Introduction: The presence of peripheral artery disease (PAD) confers a significantly increased risk of failure to heal and major lower limb amputation for people with diabetes‐related foot ulcer (DFU). Determining performance of non‐invasive bedside tests for predicting likely DFU outcomes is therefore key to effective risk stratification of patients with DFU and PAD to guide management decisions. The aim of this systematic review was to determine the performance of non‐invasive bedside tests for PAD to predict DFU healing, healing post‐minor amputation, or need for minor or major amputation in people with diabetes and DFU or gangrene.Methods: A database search of Medline and Embase was conducted from 1980 to 30 November 2022. Prospective studies that evaluated non‐invasive bedside tests in patients with diabetes, with and without PAD and foot ulceration or gangrene to predict the outcomes of DFU healing, minor amputation, and major amputation with or without revascularisation, were eligible. Included studies were required to have a minimum 6‐month follow‐up period and report adequate data to calculate the positive likelihood ratio (PLR) and negative likelihood ratio for the outcomes of DFU healing, and minor and major amputation. Methodological quality was assessed using the Quality in Prognosis Studies tool.Results: From 14,820 abstracts screened 28 prognostic studies met the inclusion criteria. The prognostic tests evaluated by the studies included: ankle‐brachial index (ABI) in 9 studies; ankle pressures in 10 studies, toe‐brachial index in 4 studies, toe pressure in 9 studies, transcutaneous oxygen pressure (TcPO₂) in 7 studies, skin perfusion pressure in 5 studies, continuous wave Doppler (pedal waveforms) in 2 studies, pedal pulses in 3 studies, and ankle peak systolic velocity in 1 study. Study quality was variable. Common reasons for studies having a moderate or high risk of bias were poorly described study participation, attrition rates, and inadequate adjustment for confounders. In people with DFU, toe pressure ≥30 mmHg, TcPO₂ ≥25 mmHg, and skin perfusion pressure of ≥40 mmHg were associated with a moderate to large increase in pretest probability of healing in people with DFU. Toe pressure ≥30 mmHg was associated with a moderate increase in healing post‐minor amputation. An ABI using a threshold of ≥0.9 did not increase the pretest probability of DFU healing, whereas an ABI <0.5 was associated with a moderate increase in pretest probability of non‐healing. Few studies investigated amputation outcomes. An ABI <0.4 demonstrated the largest increase in pretest probability of a major amputation (PLR ≥10).Conclusions: Prognostic capacity of bedside testing for DFU healing and amputation is variable. A toe pressure ≥30 mmHg, TcPO₂ ≥25 mmHg, and skin perfusion pressure of ≥40 mmHg are associated with a moderate to large increase in pretest probability of healing in people with DFU. There are little data available evaluating the prognostic capacity of bedside testing for healing after minor amputation or for major amputation in people with DFU. Current evidence suggests that an ABI <0.4 may be associated with a large increase in risk of major amputation. The findings of this systematic review need to be interpreted in the context of limitations of available evidence, including varying rates of revascularisation, lack of post‐revascularisation bedside testing, and heterogenous subpopulations.

As a progressive disease process, early diagnosis and ongoing monitoring and treatment of lower limb peripheral artery disease (PAD) is critical to reduce the risk of diabetes‐related foot ulcer (DFU) development, non‐healing of wounds, infection and amputation, in addition to cardiovascular complications. There are a variety of non‐invasive tests available to diagnose PAD at the bedside, but there is no consensus as to the most diagnostically accurate of these bedside investigations or their reliability for use as a method of ongoing monitoring. Therefore, the aim of this systematic review was to first determine the diagnostic accuracy of non‐invasive bedside tests for identifying PAD compared to an imaging reference test and second to determine the intra‐ and inter‐rater reliability of non‐invasive bedside tests in adults with diabetes. A database search of Medline and Embase was conducted from 1980 to 30 November 2022. Prospective and retrospective investigations of the diagnostic accuracy of bedside testing in people with diabetes using an imaging reference standard and reliability studies of bedside testing techniques conducted in people with diabetes were eligible. Included studies of diagnostic accuracy were required to report adequate data to calculate the positive likelihood ratio (PLR) and negative likelihood ratio (NLR) which were the primary endpoints. The quality appraisal was conducted using the Quality Assessment of Diagnostic Accuracy Studies and Quality Appraisal of Reliability quality appraisal tools. From a total of 8517 abstracts retrieved, 40 studies met the inclusion criteria for the diagnostic accuracy component of the review and seven studies met the inclusion criteria for the reliability component of the review. Most studies investigated the diagnostic accuracy of ankle ‐brachial index (ABI) (N = 38). In people with and without DFU, PLRs ranged from 1.69 to 19.9 and NLRs from 0.29 to 0.84 indicating an ABI <0.9 increases the likelihood of disease (but the extent of the increase ranges from a small to large amount) and an ABI within the normal range (≥0.90 and <1.3) does not exclude PAD. For toe‐brachial index (TBI), a threshold of <0.70 has a moderate ability to rule PAD in and out; however, this is based on limited evidence. Similarly, a small number of studies indicate that one or more monophasic Doppler waveforms in the pedal arteries is associated with the presence of PAD, whereas tri‐ or biphasic waveform suggests that PAD is less likely. Several forms of bedside testing may also be useful as adjunct tests and 7 studies were identified that investigated the reliability of bedside tests including ABI, toe pressure, TBI, transcutaneous oxygen pressure (TcPO₂) and pulse palpation. Inter‐rater reliability was poor for pulse palpation and moderate for TcPO_2. The ABI, toe pressure and TBI may have good inter‐ and intra‐rater reliability, but margins of error are wide, requiring a large change in the measurement for it to be considered a true change rather than error. There is currently no single bedside test or a combination of bedside tests that has been shown to have superior diagnostic accuracy for PAD in people with diabetes with or without DFU. However, an ABI 1.3, TBI of <0.70, and absent or monophasic pedal Doppler waveforms are useful to identify the presence of disease. The ability of the tests to exclude disease is variable and although reliability may be acceptable, evidence of error in the measurements means test results that are within normal limits should be considered with caution and in the context of other vascular assessment findings (e.g., pedal pulse palpation and clinical signs) and progress of DFU healing.