Reliability of treadmill testing in peripheral arterial disease: a meta-regression analysis

A meta-regression analysis was conducted to identify the most reliable treadmill protocol for the assessment of patients with peripheral arterial disease (PAD).

Treadmill testing is the main assessment method to evaluate walking ability in patients with PAD in clinical studies. Reported treadmill protocols are continuous (C) and graded (G) protocols and outcome measurements are initial claudication distance (ICD) and absolute claudication distance (ACD). Variety in protocols might hamper the ability to compare results of different studies. Ideally, future studies should use a protocol with highest reliability.

We searched PubMed and EMBASE (until February 2008) and we hand searched the reference lists. Trials assessing reliability of treadmill testing were identified. Inclusion criteria were the use of a C- or G-protocol, repetition of this protocol, and a retrievable intraclass correlation coefficient (ICC). We identified eight studies in which 658 patients were included.

For ICD, the estimated reliabilities of the C- and G-protocol (as assessed by the ICC) were 0.85 (95% confidence interval [CI]: 0.82-0.88) and 0.83 (95% CI: 0.80-0.85), respectively, without dependency of the reliability on velocity or grade. For ACD, the reliability was significantly better for the G-protocol (0.95, 95% CI: 0.94-0.96) than for the C-protocol. Moreover, the reliability of the C-protocol was dependent on grade of the treadmill (0%, 10%, and 12%) with a mean ICC of 0.76 (95% CI: 0.54-0.88), 0.89 (95% CI: 0.86-0.91), and 0.91 (95% CI 0.88-0.92), respectively.

Treadmill assessment has the highest reliability when using a G-protocol together with the ACD as the primary outcome measure.