This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
research methodsrandomized controlled trialsalcoholInternetbrief intervention
Two challenging issues in Internet intervention research, as well as in other behavioral intervention trials, are ensuring that participants receive the intervention (adherence) and that their outcomes are captured at follow-up (retention) [1]. The interesting analysis presented by Murray et al [2] demonstrated that, at least in their study sample, the participant adherence and retention were positively related.
One issue to consider is whether this finding can be replicated in other study samples. It is possible that research involving, for example, different recruitment methods or with higher (or lower) retention rates, might not display this same positive relationship. To that purpose, results were examined from an Internet intervention trial that employed a proactive telephone recruitment method and obtained complete follow-up data for 86% of participants [3-5]. As with the Murray et al study [2], adherence (measured by the number of intervention participants logging onto a brief alcohol intervention, where N=92; 57 participants logged onto the intervention and 35 participants did not log on) and retention were strongly positively related (retention at 3-months: logged onto intervention=100%, did not log on=80%, P<.001; retention at 6-months: logged onto intervention=100%; did not log on=80%, P<.001; retention at 12-months: logged onto intervention=96%; did not log on=74.3%, P=.002; Fisher’s Exact Tests).
Given that the positive relationship between adherence and retention can be replicated, what are the implications of this finding? From one perspective, the fact that these two key issues are related could underline the increased importance of obtaining a good retention rate. This is because the positive relationship of adherence to retention implies that a confound in the interpretation of the results is more likely as loss to follow-up (or reduced adherence to the intervention) increases. Alternatively, it could be argued that this positive relationship might reduce the importance of obtaining a good retention rate. This is because traditional intent-to-treat analysis assumes that participants who are lost to follow-up do not make any change in their behavior from baseline to follow-up (and are included as imputed values in the analysis based on this assumption). If it is then assumed that only those participants who accessed the intervention will actually make a change in their behavior, then the fact that participants who adhere to the intervention are more likely to follow-up can only increase the likelihood that participants who are lost to follow-up are less likely to have made a change in their behavior (thus validating the intent-to-treat analysis assumption). Determining which of these implications is correct is important, particularly in a field where low retention rates are an unfortunate reality in many research trials [1].
None declared.
EysenbachGThe law of attrition200571e1110.2196/jmir.7.1.e1115829473v7e11PMC1550631MurrayEWhiteIRVaragunamMGodfreyCKhadjesariZMcCambridgeJAttrition revisited: adherence and retention in a web-based alcohol trial2013158e16210.2196/jmir.233623996958v15i8e162PMC3815435CunninghamJAWildTCCordingleyJvan MierloTHumphreysKA randomized controlled trial of an internet-based intervention for alcohol abusers2009121041220233210.1111/j.1360-0443.2009.02726.x19922569ADD2726PMC2779998CunninghamJAWildTCCordingleyJVan MierloTHumphreysKTwelve-month follow-up results from a randomized controlled trial of a brief personalized feedback intervention for problem drinkers2010064532586210.1093/alcalc/agq00920150170agq009PMC2857148CunninghamJAWildTCHumphreysKWho uses online interventions for problem drinkers?201110413261410.1016/j.jsat.2011.03.00321632197S0740-5472(11)00057-2PMC3166539