1. Image: CDC, 2016, https://www.cdc.gov/malaria/
A randomized trial on the impact of text message reminders on adherence to antimalarial treatment
Julia Raifman, ScD
Heather Lanthorn, ScD
Slawa Rokicki, ScM
Günther Fink, PhD
SUMMARIZE: We look at the impact of sending a short or a long reminder message on self-reported ACT adherence
Image: CDC, 2016,

2. Study objectives Objectives
Problem Low adherence to antimalarial treatment
Objectives
Assess the impact of text message reminders on adherence to ACT regimens
Develop and test a scalable method of enrolling patients in a text message reminder system
We have some reason to think that text message reminders might improve adherence
Few studies on text message reminders among developing country populations
Image: CDC, 2016,

3. Study Setting & Population
Tamale, capital of Northern Region
Endemic malaria
85% mobile phone ownership1
Recruited from 69 public and private ACT vendors within 30 minutes of Tamale
ACTs: AL and AS+AQ
3X2 regimen!
1. International Telecommunications Union, 2012
Image: Ghana Web, 2012

4. Enrollment in text message system
Own phone
“Flashing”
Designed to represent feasible method of scale-up
Self-enrollment and own phone differentiates this study from past studies of text reminders
Outcome: Self-reported adherence validated by pill count Method: Logistic regression analysis

5. Randomization Enrollment 3 Days Day 5
All individuals purchasing ACTs from vendor
Receive flyer for text message system enrollment
Enrollment
Study enrollment, Enrollment message:
“Thank you for registering for mobile health information”
Randomization
Treatment
6 messages at 7:00 am and 7:00 pm for 60 hours
Control
No messages during ACT treatment
Long message
“Please take your MALARIA drugs. Even if you feel better, you must take all the tablets to kill all the malaria.”
Follow up Interview
Short message
“Please take your MALARIA drugs”
3 Days
Malaria prevention message:
“Sleep under a bed net to prevent malaria”
Day 5

6. Adherence among adults in the control group
Dose 1 Dose 2 Dose 3 Dose 4 Dose 5 Dose 6
100%
80%
60%
40%
20% 0%
Dose number
Percent of respondents completing dose
This is adherence among adults in the control group.
84% reported taking dose 4
74% reported taking dose 5
56% reported taking dose 6
Percent of respondents completing dose
95% Confidence Interval
Percent of respondents completing dose
95% Confidence Interval
Percent of respondents completing dose
95% Confidence Interval

7. Proportion completed treatment
Intervention Impact
N=1141
Proportion completed treatment
Adjusted OR
Control group
61.5%
331/538
Reference group
Either text message
66.4%
380/572
1.21
(0.94 – 1.56)
Short message
69.0%**
185/268
1.45**
(1.03 – 2.04)
Long message
64.1%
195/304
1.08
(0.80 – 1.47)
Covariates: Patient sex, patient age, male household head, household head educational attainment, household wealth quintile, and whether participants sought care at private hospitals or clinics or public hospitals or clinics.
* p < 0.1 ** p < 0.05 *** p <0.001

8. Discussion ACT adherence is low 58% of adults in control group
Impact of text message reminders depends on content
Short text message reminders increase the odds of ACT adherence by 38.5%
Pharmacists distributing flyers is a feasible distribution method

9. Acknowledgements Clinton Health Access Initiative
Innovations for Poverty Action-Ghana
Jessica Cohen, Harvard School of Public Health