Cost-effectiveness of the WelTel mHealth program to improve
adherence to antiretroviral therapy in Kenya
Anik R. Patel1, Richard T. Lester1, R. Scott Braithwaite2, Mia L. van der Kop1,3, Zafar Zafari1, Carlo A. Marra1
1. University of British Columbia, Canada 2. New York University, USA 3. Karolinska Institutet, Sweden
Objective
To evaluate the cost-effectiveness of the WelTel
text-messaging program to improve combination
antiretroviral therapy (cART) adherence.
Introduction
Mobile phone use in Africa has surged and dramatic
growth in this sector has drawn the attention of
global health efforts.1
This mobile technology boom has fueled an
emerging market of mobile health (mHealth)
solutions to address a number of health challenges
across Africa.
Figure 1: Mobile phone penetration and market size across Africa.1
Methods
Methods (Continued)
The WelTel Program
Model Inputs
WelTel is an mHealth engagement program that
connects health providers and patients. A weekly text
message is sent to HIV patients that reads ‘Mambo?’,
Kiswahili for ‘How are you?’ Patients can respond
‘Sawa’ meaning ‘Fine’ or ‘Shida’ meaning ‘Problem’.
Table 2: Utility and characteristic inputs9
If they respond ‘Shida’, or do not respond within
48hrs, clinic staff follow up with a phone call and
respond to the problem with the appropriate action.
WelTel Kenya Trial
A randomized controlled trial comparing WelTel to
standard HIV care was conducted at three health
clinics in and around Nairobi, Kenya.
† Kenyan cohort characteristic data provided by Academic Model Providing Access to
Healthcare initiative (AMPATH)9
Table 3: Kenyan cost inputs9
The study involved HIV patients newly initiating
cART. Participants were from a wide range of socioeconomic, age, and rural or urban groups.
The primary outcomes were proportion adherent by
self-report and proportion of viral suppression after
one year of follow-up. (Table 1) Adherence was
defined as having taken greater than 95% of cART
doses. A level of 95% adherence is a critical value to
lower risk treatment failure.8
Table 1: WelTel Kenya trial results3
† Kenyan cost data provided by AMPATH9
Results
Table 4: ICER of WelTel compared to
standard care at different program costs and
levels of adherence of non-adherent patients.
† 273 patients in the SMS group and 265 in the control group
* Relative risk and 95% Confidence Interval
HIV Economic Model
mHealth Evidence
Based on high-quality evidence, a recent Cochrane
review concluded that text-messaging improves
adherence to antiretroviral therapy (cART) among
individuals infected with HIV.2
In the review, the WelTel Kenya trial was described
as a strong source of evidence for mhealth
effectiveness with low risk of bias.3
mHealth Assessment
The Sixtieth World Health Assembly of 2007
expressed concerns of inappropriate investments in
health technology resulting in resource waste. They
urged member states to formulate national strategies
for assessment of health technologies.5
Economic evaluations are essential for local payers
to make funding decisions. Economic evaluation of
mhealth is lacking.2,4
Adherence outcomes were transformed into lifetime
costs and benefits through the use of a decision
analytic model that is described in detail elsewhere.8
Briefly, the model calculates the costs of lifetime care
of simulated HIV patients using relationships between
clinical markers and HIV progression.
The simulation also calculates quality adjusted life
years (QALY) based on the specified inputs. The
Incremental Cost Effectiveness Ratio (ICER)
summarizes the incremental lifetime costs relative to
the QALYs gained.
The simulation derives outputs using inputs derived
from published literature and the trial. (Table 2 and 3)
An increase in the proportion of adherent
patients (by Weltel compared to standard care)
results in lower lifetime HIV morbidity and greater
survival -- two major factors driving the ICER.
Model Description
mHealth Sustainability
Table 5: WHO cost-effective thresholds in Kenya10
Highly Cost-Effective
$ 2,154 USD/QALY
Cost-Effective
$ 6,461 USD/QALY
Average level of adherence among non-adherent
patients is a key driver of the ICER and the program
is more cost-effective when levels are low.
Adherence levels below 50% were not modeled
because our results suggest the program is
highly cost-effective already.
Discussion
“At present, the vast majority of mHealth
implementers in low and middle-income countries
are dependent on short-term, grant-based funding
continue to operate.”
“All too often, mHealth implementations – even
those showing positive changes in behaviours or
health outcomes – do not survive because of their
dependence on this form of financing.”
- mhealth Alliance sustainability report6
* Per person lifetime WelTel program costs discounted at 3%
† Average lifetime level of adherence among the non-adherent patients. A
conservative assumption was made that WelTel had no impact on level of
adherence in the average patient. Proportion of adherent vs non-adherent
based on trial results.
§ 2008 US Dollars per QALY based on difference in proportion adherent
ß Confidence interval - based on the treatment effect on proportion adherent
WelTel is cost-effective or very cost-effective at
improving cART adherence. The ICER is dependent
upon the baseline level of adherence in a population.
Many regions in Kenya would benefit from addition
of such a program to their HIV strategy. Findings
may be generalizable to other African countries.
Figure 2: The model simulates individual HIV infected
patients through the health states and events shown
above. Transition probabilities through clinical stages of
HIV and probabilities of HIV/AIDS events came from
published literature. Lifetime costs, utility decrement and
survival time are the key outputs of the simulation. The
model was validated and calibrated using data from
several Kenyan and Ugandan HIV cohorts.7,9
Loss to follow-up and transmission benefits of the
program were not modeled in this analysis.
References
1. mHealth Africa Website: Allo Africa: Available at: http://www.mhealthafrica.com/infographic-1-allo-africa: Accessed Nov 6, 2013.
2. Horvath, Tara, et al. (2012). "Mobile phone text messaging for promoting adherence to antiretroviral therapy in patients with HIV infection."Cochrane Database Systematic Review 3.
3. Lester, R. T., Ritvo, P., et al. (2010). Effects of a mobile phone short message service on antiretroviral treatment adherence in Kenya (WelTel Kenya1): a randomised trial. The
Lancet 376(9755), 1838-1845.
4. Déglise, C., et.al. (2012). Short message service (SMS) applications for disease prevention in developing countries. Journal of Medical Internet Research, 14(1), e3.
5. Resolution WHA60.29. Health technologies. In: Sixtieth World Health Assembly, Geneva, 14–23 May 2007. Resolutions and decisions. Geneva, World Health Organization, 2007.
http://www.who.int/medical_devices/resolution_wha60_29-en1.pdf, accessed November 1, 2013.
6. mhealth Alliance website: Sustainable financing report: http://www.mhealthalliance.org/images/content/sustainable_financing_for_mhealth_report.pdf. Accessed November 1, 2013.
7. Braithwaite, R. S., Roberts, M. S., et al. (2008). Influence of alternative thresholds for initiating HIV treatment on quality-adjusted life expectancy: a decision model. Annals of internal
medicine, 148(3), 178-185.
8. Wood E, Hogg RS, et al. (2003) Effect of medication adherence on survival of HIV-infected adults who start highly active antiretroviral therapy when the CD4+ cell count is 0·200 to
0·350 × 10(9) cells/L. Ann Intern Med; 139: 810–16.
9. Braithwaite, R. S., Nucifora, K. A., et. al. (2011). Alternative antiretroviral monitoring strategies for HIV-infected patients in east Africa: opportunities to save more lives?. Journal of
the International AIDS Society, 14(1), 38.
10. WHO Choice Website: Cost-effectiveness thresholds http://www.who.int/choice/en/. Accessed November 1, 2013