Oral Presentation Australian Diabetes Society and the Australian Diabetes Educators Association Annual Scientific Meeting 2017

Determining the diagnostic accuracy of the ankle brachial index for detecting peripheral arterial disease in people with diabetes. A systematic review and meta-analysis. (#157)

Vivienne H Chuter 1 , Alex L Barwick 2 , Peta E Tehan 1 , Angela T Searle 1 , Lucy Leigh 3 , Nathan A Johnson 4 5 , Stephen M Twigg 4 6
  1. Discipline of Podiatry, University of Newcastle, Ourimbah, NSW, Australia
  2. School of Health and Human Sciences, Southern Cross University, Gold Coast, QLD, Australia
  3. Clinical Research Design, IT and Statistical Support, Hunter Medical Research Institue, Newcastle, NSW, Australia
  4. The Charles Perkins Centre, University of Sydney, Sydney, NSW, Australia
  5. Faculty of Health Sciences, University of Sydney, Lidcombe, NSW, Australia
  6. Sydney Medical School, University of Sydney, Sydney, NSW, Australia

Background: In diabetes cohorts, diagnostic accuracy of the ankle-brachial index (ABI) for peripheral arterial disease (PAD) has been reported to be variable. This review systematically evaluated the literature investigating the diagnostic accuracy of the ABI for PAD in people with diabetes including utilising meta-analysis where possible.

Methods:  A database search of EBSCO Megafile Premier, EMBASE, and The Cochrane Library was conducted to 30th December, 2016.  Prospective and retrospective investigations of diagnostic accuracy of the ABI for PAD in people with diabetes using an imaging reference standard were eligible. Study quality was assessed with the revised Quality Assessment of Diagnostic Accuracy Studies tool.

Results: In total 6249 citations were retrieved, and after serial systematic exclusions, 22 studies included. Ten studies compared ABI to angiography, and 12 studies to colour Doppler ultrasonography (CDU). Six studies were retrospective. Primary sources of bias were: inadequate reporting of blinding and risk of spectrum bias. A meta-analysis (mean sensitivity, specificity and receiver operating characteristic [ROC] curve and area under the curve [AUC]) was performed for studies using CDU. For studies using angiography, a standard summary ROC curve was calculated and the AUC derived from this.  For CDU studies, sensitivity was 0.56 (95%CI:0.43 to 0.67), specificity was 0.88 (95%CI:0.77 to 0.94) and the AUROC was 0.66 (95%CI:0.52 to 0.74). For angiography studies, sensitivity ranged from 0.50 to 0.91 and specificity 0.40 to 0.89. The AUROC was 0.79 (95%CI:0.68 - 0.90).

Conclusion: Diagnostic accuracy of the ABI for PAD is better compared to angiography than CDU. This is likely to be due to high incidence of disease in study populations undergoing angiography. These findings suggest the ABI has limited use as a screening tool in community-based diabetes populations especially due to low sensitivity, but, has clinical utility where there is a strong suspicion of disease.