The benefits of transcutaneous electrical nerve stimulation (TENS) for pain relief have not been reliably established, as most systematic reviews find poor methodological quality in many studies. The paradox within the evidence base for TENS is that despite identified sources of bias that may lead to an overestimation of treatment effects, no benefits for TENS can be clearly demonstrated. Conventional assessments of quality assume a single direction of bias, and little work has been undertaken examining other directions of bias. Our hypothesis was that low fidelity in studies (bias leading to an underestimation of treatment effects) may account for inconclusive findings. We included 38 studies from 3 recently published Cochrane systematic reviews that examined TENS for acute, chronic, and cancer pain. We extracted data relating to treatment allocation, application of TENS and to the assessment of outcomes. We quantified these data and judged this against standardised assessment criteria using a "traffic light" approach based on the number of studies reaching the standard. We identified significant sources of potential bias in both directions in relation to study design and implementation fidelity that have not been quantified previously. Suboptimal dosing of TENS and inappropriate outcome assessment were particularly prevalent weaknesses indicating low fidelity. We propose criteria for judging directions of bias in future studies of TENS that may be adapted to assess other trials in which implementation fidelity is important, such as other nonpharmacological interventions for pain. Poor implementation fidelity was identified as a significant source of bias in systematic reviews of TENS studies and might explain lack of consistent treatment effects of TENS in pain. Here, criteria for assessing methodology are proposed for use in designing future clinical trials of TENS.
Copyright © 2010 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved.