[ARCHIVED] Adherence & Retention

I wonder if your finding on reminder frequency is due to 'reminder fatigue'? Weingart has demonstrated an 'alert fatigue' phenomenon when it comes to the use of computerized drug alerts.

http://archinte.ama-assn.org/cgi/content/full/169/3/305

http://archinte.ama-assn.org/cgi/content/full/163/21/2625

Hi Alex,

Yes, I imagine reminder fatigue or habituation likely played a role. Here's an excerpt from the Discussion relating to this point:

"It is also interesting to note that weekly reminders
improved adherence, whereas daily reminders did not.
Habituation, or the diminishing of a response to a
frequently repeated stimulus, may explain this finding.
Daily messages might also have been considered intrusive.
Further research is needed to distinguish the mechanisms
as to why the weekly messages were most efficacious."

It would be interesting to see how reminders tied to specific missed doses (enabled through real-time monitoring) perform.

Best,
Jessica

Jessica,

Thanks for posting this very important and interesting study. The study design and use of MEMS caps provides important new insight into how adherence behaviour change may be elicited through SMS to patients. It is very interesting that the greater effect observed with short, weekly messages versus longer, more frequent messages is consistent with the patients’ preferences we found in focus group discussion leading up to our WelTel trial. The discussion Alex raises around habituation versus a feeling of intrusion is similarly interesting – the adherence levels among those receiving daily messages seemed to drop very quickly to the level of the controls. Our sense was that daily frequency was considered too intrusive to patients. Do the authors of this study have a sense of what the participants’ preferences were?

Also, why was the trial analysis trumped at January 31st, 2008 providing data on only 431 patients out of the 720 enrolled? Was that a planned stopping point or do we expect additional reporting on these subjects to follow?

This study provides additional good news that low cost cell phone interventions may help ART adherence in a setting where every positive contribution will have a massive effect if brought to ...

Jessica,

Thanks for posting this very important and interesting study. The study design and use of MEMS caps provides important new insight into how adherence behaviour change may be elicited through SMS to patients. It is very interesting that the greater effect observed with short, weekly messages versus longer, more frequent messages is consistent with the patients’ preferences we found in focus group discussion leading up to our WelTel trial. The discussion Alex raises around habituation versus a feeling of intrusion is similarly interesting – the adherence levels among those receiving daily messages seemed to drop very quickly to the level of the controls. Our sense was that daily frequency was considered too intrusive to patients. Do the authors of this study have a sense of what the participants’ preferences were?

Also, why was the trial analysis trumped at January 31st, 2008 providing data on only 431 patients out of the 720 enrolled? Was that a planned stopping point or do we expect additional reporting on these subjects to follow?

This study provides additional good news that low cost cell phone interventions may help ART adherence in a setting where every positive contribution will have a massive effect if brought to scale. However, the overall high levels of non-adherence and treatment interruptions observed are concerning and should be a call to action for all of us.

Kind regards,
Richard

Alex,
Thanks for your link to the articles on alert fatigue. I think fatigue or habituation may well be the explanation for the low efficacy of the daily reminders. Our study did was not designed to test the optimal frequency of reminders, but it certainly appears that somewhat less frequent reminders (such as weekly) are better. I think Jessica's proposal to compare these results to those that would be obtained by reminders tied to specific missed doses (enabled through real-time monitoring) is interesting - but one of my concerns would be that such reminders might be much more costly to implement than 1-way messages sent at a certain frequency.


Richard, thanks for your comments. Unfortunately the study did not include extensive qualitative interviews, which might have further illuminated why the daily reminders were not effective. Our analysis of the data has indicated that there was a very early spike in the adherence of the group that received daily reminders, which would be consistent with efficacy during the first few weeks of reminders but tuning out of reminders subsequently. We would like to conduct additional focus group discussions with some of the study participants to verify that this is the case ...

Alex,
Thanks for your link to the articles on alert fatigue. I think fatigue or habituation may well be the explanation for the low efficacy of the daily reminders. Our study did was not designed to test the optimal frequency of reminders, but it certainly appears that somewhat less frequent reminders (such as weekly) are better. I think Jessica's proposal to compare these results to those that would be obtained by reminders tied to specific missed doses (enabled through real-time monitoring) is interesting - but one of my concerns would be that such reminders might be much more costly to implement than 1-way messages sent at a certain frequency.


Richard, thanks for your comments. Unfortunately the study did not include extensive qualitative interviews, which might have further illuminated why the daily reminders were not effective. Our analysis of the data has indicated that there was a very early spike in the adherence of the group that received daily reminders, which would be consistent with efficacy during the first few weeks of reminders but tuning out of reminders subsequently. We would like to conduct additional focus group discussions with some of the study participants to verify that this is the case.

Regarding the tial analysis, we restricted to participants enrolled before Jan 31, 2008 in order to examine a uniform cohort that had at least 48 weeks of potential follow-up. The initially planned stopping point was 18 months of intervention for all participants but budgetary constraints kept us from achieving this. Future analysis will examine all 721 patients, with the limitation being that some of these patients had less than 48 weeks of follow-up (while others, of course, had much more than 48 weeks of follow-up).

I also think you bring up a good point about the overall high levels of non-adherence. The most obvious explanation for this might stem from our reliance on MEMS as opposed to self-reports, but it would be equally important for us to consider the many other barriers to adherence - only some of which can be addressed by the cell phone intervention. Nonetheless, I think further implementation and evaluation of 2-way reminders such as the one in your recent article as well as 1-way reminders, as well as the optimal content and frequency of such reminders, might show that we can achieve higher levels of adherence than did the short, weekly 1-way reminders in our study. The identify of the message sender could also prove to be important - something that has not be carefully explored yet.

Best regards
Harsha