Efforts are under way to greatly expand medical male circumcision (MC) for HIV prevention in sub-Saharan Africa.1–6 MC reduces the risk of female-to-male HIV transmission by about 60%1–3 and may reduce the risk of male-to-female transmission by 46%.7 Resumption of sexual activity during the healing period may increase the risk for HIV acquisition in HIV-negative men.8 Similarly, in HIV-positive men, early resumption of sexual activity may increase risk for transmission to sex partners.9,10 After MC, men are advised to abstain from sexual activity for at least 6 weeks to allow for complete wound healing.11,12 Earlier resumption of sexual activity may also increase the risk of postoperative surgical complications.13 Despite these risks, the proportion of circumcised men reporting resumption of sexual activity before the end of the recommended 6-week abstinence period has been high. For example, approximately 31% of circumcised men reported resuming sex early in Kenya and 24% in Zambia.10,14
The low cost of mobile phone communication in resource-limited settings presents a unique avenue to potentially change behavior, improve health outcomes, and magnify the impact of public health interventions.15 During the past decade, access to mobile phones has increased substantially in developing countries. The International Telecommunications Union estimates that approximately 79% of the population in developing countries has access to a mobile phone. The use of mobile phone technology to complement public health interventions (mHealth)16,17 is gaining traction across Africa, especially as part of larger health communication strategies. Although no empirical data were available to support it, the Joint United Nations Programme on HIV/AIDS recommends the use of short messaging service (SMS) as part of strategic communication toward accelerating the scale-up of MC and providing high-quality services.18,19 Mobile phones are also acceptable for providing HIV-related information.20 Text messages have been effective in varied settings for a variety of interventions, including increasing adherence to antiretroviral treatment,21,22 reducing HIV risk behaviors,23 improving uptake of vaccination,24,25 promoting smoking cessation,26 and shortening notification times for infant HIV test results.27 We previously reported that text messaging improves attendance at postoperative clinic appointments after male circumcision for HIV prevention.28 However, SMS interventions per se have not been universally successful. For example, the Cameroon Mobile Phone SMS (CAMPS) trial found no effect of text messaging on antiretroviral adherence.29
Our aim was to determine whether a series of 1-way educational and reminder SMS messages could significantly reduce the frequency of self-reported resumption of sexual activity before 42 days after MC compared with a control condition of usual care.
Trial Design and Participants
We conducted a 2-arm, parallel, randomized controlled trial conducted at 12 sites in Nyanza province, Kenya. There were 2 main endpoints: (1) attendance at the scheduled 7-day postoperative clinic visit, and (2) self-reported resumption of sexual activity before 42 days after MC. Results related to the first endpoint have been reported elsewhere.28
Men who had undergone MC were approached during the 30-minute postcircumcision recovery period and assessed for eligibility. Those who were at least 18 years old, owned mobile phones, had the phone in their possession at enrollment, and were willing to respond to a phone interview after 6 weeks were randomized in a 1:1 ratio to receive either text messages or usual care.
Randomization and Masking
An independent biostatistician generated a block randomization scheme with variable block sizes of 4–16 participants, stratified by clinic. Consecutive, sequentially numbered, sealed, opaque envelopes were used to assign participants to study arms. Because the intervention arms included receiving SMS versus not receiving study SMS messages, participants were not masked to study group assignment. However, clinic staff members were unaware of the participants' group assignments.
At enrollment, baseline characteristics and participant mobile phone numbers were collected. After randomization, participants sent a registration text message containing the study site and participant identification number. For those in the SMS arm, the registration message also included the preferred time of day and language (English, Kiswahili, or Dholuo) for receiving intervention messages. All participants received US $0.25 worth of airtime as reimbursement to cover the cost of the registration SMS. Usual care consisted of a “minimum package” that included HIV testing and counseling, screening and treatment for sexually transmitted infections, condom promotion and provision, risk reduction and safe sex counseling, the MC procedure, and postoperative review 7 days after surgery.30
To automate the SMS intervention, we used RapidSMS, an open source framework for SMS applications.31 Text messages were sent at the time and in the language preferred by the participant. For the first 7 days, messages were sent once each day. Thereafter, messages were sent on days 8, 14, 21, 28, 35, 41, and 42 postprocedure. Table 1 summarizes the text message protocol. The message timing schedule was adapted from a pilot study conducted to assess the acceptability of the messages among circumcised men in Nyanza province, Kenya. After 42 days, study staff called participants in both randomization arms to conduct a brief structured interview. Participants were asked whether they had resumed sexual activity, which was defined to include sexual intercourse or masturbation per the Kenya MC guidelines.32 For those who reported sexual activity, additional questions were asked to determine the number of sexual partners, total number of sex acts, and number of sex acts with a condom. If participants were unreachable by phone on our first attempt, we tried calling again once each day for 1 week. Those who could still not be reached were considered as lost to follow-up.
All participants provided written informed consent. This trial was approved by the Human Subjects Division at the University of Washington, the Ethical Review Committee of the Kenya Medical Research Institute, and Institutional Review Board #3 at the University of Illinois at Chicago.
The primary outcome was the proportion of men reporting early resumption of sex (ie, before 42 days post-MC). We assumed that 22.5% of men in the control group would report early resumption.1 To have at least 80% power to detect a reduction to 15% or less reporting early resumption of sex in the SMS group, and allowing for 11% loss to follow-up at 42 days, we enrolled 600 men in each arm.
We assessed the effect of study arm on the primary outcome using relative risk (RR) regression (Poisson model, robust error variance).33 We performed (1) a complete case analysis that included only men with follow-up at the end of the 42-day period; and (2) a multiple imputation by chained equations analysis that imputed missing outcome data under missing-at-random assumptions, a method considered superior to complete case analysis in the presence of missing outcome data.34 The multiple imputation analysis used baseline covariate information (age, ethnicity, employment status, education level, marital status, age at first sex, number of sex partners in the month before MC, source of referral to MC clinic, whether HIV testing was done before MC, whether condoms were issued after MC, and cadre of the health worker who performed the MC procedure) and allowed us to use data from all 1200 men. Last, we performed a sensitivity analysis that examined the composite outcome of early resumption or missing outcome.
We conducted secondary analyses using the complete case dataset. We used adjusted RR regression to estimate associations between other covariates and early self-reported resumption of sex. Variables included in this model were selected based on known or suspected risk factors for early resumption of sex and included the following: study arm, age, marital status, number of sex partners in the month before MC, level of education, and employment status. Secondary analyses were repeated using the imputed dataset.
We compared time to self-reported resumption of sex (in weeks from time of circumcision) by study arm using Kaplan–Meier analysis and a log-rank test. We also compared the number of sex partners, total sex acts, and unprotected sex acts by study arm and compared using Wilcoxon rank sum tests. Among men who resumed early, the proportions of men reporting any unprotected acts versus using protection for all sex acts in each arm was compared using RR regression.
All tests were 2-sided at the 5% significance level. Stata IC v. 10 was used for analysis. Multiple imputation used the Stata “ice” package with 5 imputed datasets.35,36
Role of the Funding Source
The sponsors had no role in study design, data collection, analysis, interpretation, writing of the report, or decision to submit the manuscript for publication. The corresponding author had full access to all data and had final responsibility for the decision to submit for publication.
We assessed 3572 men for eligibility at 12 sites in Nyanza province, Kenya, between September 2010 and April 2011. Of these, we randomly assigned 1200 to receive either the intervention or usual care (Fig. 1). Data on self-reported resumption of sex before 42 days were available for 491 (81.8%) and 493 (82.2%) men in the SMS and usual care groups, respectively.
Participant characteristics were similar at baseline in SMS and usual care groups28 and remained similar when restricted to only participants with nonmissing outcome data (Table 2). However, several baseline characteristics of participants with endpoint data differed from the baseline characteristics of participants lost to follow-up (see Table, Supplemental Digital Content, http://links.lww.com/QAI/A439). Specifically, participants who were lost were significantly younger, less likely to be employed, and spent more time traveling to clinic. Greater loss to follow-up was also significantly associated with having fewer sex partners in the month before MC, fewer years of education, unmarried status, and more likely to be referred to MC by a friend.
In the complete case analysis, self-reported resumption of sex before 42 days postcircumcision was reported by 139 of 491 (28.3%) men in the SMS group and 124 of 493 (25.2%) men in the control group (RR for early resumption = 1.13, 95% CI: 0.91 to 1.38; P = 0.3). Results were essentially unchanged when multiple imputation methods were used to account for losses to follow-up at day 42 (RR: = 1.13, 95% CI: 0.92 to 1.38, P = 0.2). Similarly, when we examined a composite outcome of resumed or missing outcome, the proportion of men who resumed sex in the SMS group remained slightly higher (248/600; 41.3%) than in the control group (231/600; 38.5%), but again the difference was not statistically significant (RR = 1.07, 95% CI: 0.93 to 1.23, P = 0.3).
In the complete case analysis, men were more likely to report early resumption of sex (Table 3) if they were married or had a live-in sexual partner [adjusted relative risk (aRR): 1.57, 95% CI: 1.18 to 2.08, P < 0.01]; had 1 (aRR: 1.50, 95% CI: 1.07 to 2.12, P = 0.02) or more than 1 (aRR: 1.81, 95% CI: 1.24 to 2.66, P < 0.01) sexual partner(s) in the month before MC; had primary school or lower education level (aRR: 1.62, 95% CI: 1.33 to 1.97, P < 0.001); or were employed (aRR: 1.35, 95% CI: 1.05 to 1.72, P = 0.02). Compared with men younger than 21 years, there was higher likelihood of early resumption of sex in men aged 21–30 years (aRR: 1.58, 95% CI: 1.01 to 2.47, P = 0.05); 31–40 years (aRR: 1.91, 95% CI: 1.18 to 3.09, P < 0.01); and older than 40 years (aRR: 1.76, 95% CI: 1.04 to 2.97, P = 0.03).
A Kaplan–Meier plot showing the cumulative incidence of self-reported resumption of sex is shown in Figure 2. There was no significant difference between the groups in time to resumption of sex (log-rank test, P = 0.3). Of the 263 men who reported resuming sex early, 236 (90%) reported their first sexual activity after 28 days or later.
The median number of sexual partners within 42 days after MC did not differ between the SMS and control arms [median = 1, interquartile range (IQR): 1–1 versus median = 1, IQR: 1–1, P = 0.7]. Similarly, the median number of sex acts was the same in the intervention and control arms (median = 0, IQR: 0–1 versus median = 0, IQR: 0–0, P = 0.3).
Although the median number of unprotected sex acts was zero in both intervention and control groups, the distributions in the 2 groups differed significantly (Wilcoxon rank sum text, P = 0.007). We explored this finding further by comparing risk of any unprotected sex in analyses restricted to the subgroup of men who reported early resumption of sex. In this subgroup, 87 of 135 (64.4%) in the SMS arm and 58 of 122 (47.5%) in the control arm reported unprotected sex. Data on number of sex acts, number of unprotected sex acts, and number of sex partners for 6 men (4 SMS arm, 2 control arm) who reported resuming sex were missing and were excluded. After adjusting for age, employment, marital status, age at first sex, educational level, source of referral to the clinic, number of sex partners in the month before MC, and whether condoms were issued at the clinic after MC, men in the SMS arm remained significantly more likely to report unprotected sex within the first 42 days after MC (aRR: 1.30, 95% CI: 1.05 to 1.61, P = 0.02). Men were also more likely to report unprotected sex during this period if they were married (aRR: 1.61, 95% CI: 1.05 to 2.48, P = 0.03), or were referred to the clinic from a voluntary counseling and testing center (aRR: 1.96, 95% CI: 1.25 to 3.10, P < 0.01).
In this large randomized trial in Kenya, SMS messages with educational and reminder content sent to men after MC did not reduce self-reported resumption of sexual activity before the end of the 42-day prescribed period of abstinence. To our knowledge, this is the first trial to assess the effect of text messages on resumption of sex after MC for HIV prevention. Although the intervention was not successful, additional analyses of data from this trial provide new insights about factors associated with early resumption of sexual activity after MC.
There is a growing body of evidence that text message reminders can increase attendance at clinic appointments.37–39 Similarly, we have shown improvement in clinic attendance after male circumcision.28 Our SMS intervention did not lead to delayed self-reported resumption of sexual intercourse as intended and might have been more successful if it had been supported by formative qualitative research grounded in a behavioral change framework.40–42 Other SMS interventions have also provided conflicting results. For example, the CAMPS trial found no effect of text messaging on antiretroviral adherence.29 This finding is somewhat surprising, as 2 randomized trials in Kenya had demonstrated a significant benefit.21,22 The success of SMS interventions may be related to specifics of the intervention and to contextual factors. Designing future interventions with a firm basis in behavioral theory may make it easier to predict where, when, and how SMS interventions are likely to be successful.
Our study did not formally assess acceptability of the intervention or whether participants would have preferred more interactive 2-way text messaging. In other studies such as the WelTel Kenya1 trial,43 2-way messaging was associated with significant improvements in antiretroviral adherence and virologic suppression. Further support for the acceptability of 2-way messaging comes from the CAMPS trial where half of participant responses were messages of gratitude.44
The high overall proportion who reported resuming sex within 42 days (26.7%) was similar to the proportion in a randomized trial population in South Africa1 and in observational studies in Kenya (31%; 37.7%)14,45 and Zambia (24%).10 In contrast, fewer men reported resumption of sex earlier than 42 days in the randomized trials of MC for HIV prevention in Kisumu, Kenya (3.9%), and Rakai, Uganda (5.4%).8 This striking difference may be explained by the intense individually tailored risk-reduction counseling given to participants in Kisumu and Rakai at each study visit. In contrast, the South African Trial did not include any counselling during the 42-day abstinence period, and MC programs on-going today typically include counselling only at the time of circumcision.
In our study, the great majority (90%) of participants who resumed sex early did so more than 28 days after MC. This may be related to men's own perception of completion of wound healing. In a separate study by our group, more than 70% of men had completely healed wounds by the end of the fourth week postcircumcision.45 Of note, the SMS message we sent on day 28 advised men not to have sex until they were fully healed. If men in the intervention arm perceived their wounds to be fully healed by day 28, then this message may have inadvertently supported a decision to resume sex before 42 days for this group. Also, none of the messages explicitly told men to delay sex until 6 weeks, except the day 8 message. The 41 day message may have tacitly alluded to sex or may have seemed to encourage resumption by use of the word “if” instead of “when”. Future interventions aimed at delaying resumption of sex should focus on risk reduction throughout the risk period after MC, and specifically encourage abstinence until after day 42 regardless of the appearance of the wound. In this regard, messages between days 28 and 42 should address the issue of not assuming that the wound is healed based on self-assessment. Although many men have achieved complete wound healing by 4 weeks, nearly one-third are not completely healed at this point.13,46 Moreover, men’s own assessment of wound healing agrees with that of a clinician only 65% of the time.47 As such, what men perceive as healed may not be “fully healed”. In the absence of follow-up visits for certification of wound healing by a clinician, the conservative period of a minimum of 6 weeks allows for the majority of wounds to heal.
This study included a number of limitations. First, men younger than 18 years and those who did not possess phones were excluded, limiting generalizability of our results. However, access to mobile phones is increasing substantially in Kenya. Between December 2011 and March 2012, mobile phone subscribers increased from 28.08 million to 29.2 million.48 Second, we recognize that resumption of sex before 42 days was a self-reported outcome. In this context, our results may be affected by social desirability bias, with underreporting of early resumption of sex. If men were worried about sexual performance after MC, they may also have overreported resumption of sex. Third, we did not ask men in the intervention arm if they received the text messages, or how many messages were received. It is possible that some may not have received the SMS due to changing phone numbers, network failures, or other reasons. Fourth, we did not collect data on phone sharing by study participants. Some men may share phones and could have missed the SMS messages as a result. Fifth, we did not specifically test for literacy in participants. However, only 30 (2.5%) study participants reported having no education. Nonetheless, we recognize that failure to receive messages, sharing of phones, and challenges with literacy may have contributed to the lack of efficacy observed in this trial. Finally, the 18% loss to follow-up at day 42 was higher than anticipated. Being unreachable by phone may have been associated with failure to receive intervention messages for men in the SMS arm.
Despite these limitations, our study identified important risk factors for early resumption of sexual activity after MC, providing guidance to inform the development of interventions to improve outcomes after adult MC for HIV prevention. Specifically, and consistent with previous studies by Herman-Rolof et al14 and Odoyo-June et al,47 we found that early resumption of sex was associated with being married or living with a sexual partner, reporting 1 or more sexual partners in the month before MC, older age, employment, and education level. Men living with sexual partners likely have greater opportunity to resume sex, may simply feel more comfortable resuming sex earlier than advised, or find it harder to abstain. Involvement of female partners in scale-up of MC might address this problem.49,50
Men with primary or lower level of education were significantly more likely to report early resumption of sex. It is possible that such men were less likely to understand the postoperative care instructions. Additional education and counselling for this group should be considered as MC programs continue to expand. Lower health literacy is associated with poor HIV treatment adherence.51,52 In-depth qualitative studies would be helpful for understanding the reasons for early resumption of sex and could inform the development of interventions aimed at promoting sexual abstinence until 42 days after MC.
Although we did not find a reduction in early self-reported resumption of sex after MC with this SMS intervention, these data highlight the need for further research using low-cost technologies to reduce early resumption of sexual activity and maximize the benefits of MC for HIV prevention. Where SMS interventions have been successful in promoting antiretroviral adherence and smoking cessation, the lingering question has been why they worked.53,54 Future interventions with mobile phone technologies could be strengthened by developing a foundation based on behavioral theory to explain the mechanism by which they might reduce early resumption of sex after MC.
As MC programs continue to expand across sub-Saharan Africa, resumption of sex before 42 days after the procedure will continue to be an issue to address and a possible threat to significant reduction in HIV incidence. Our findings point to several key risk factors for early self-reported resumption of sex that MC programs should take into account as they develop interventions to address this issue.
The authors thank the Director, KEMRI, for permission to publish this work. The authors also thank the Impact Research, Development Organization, and the Nyanza Reproductive Health Society for their help with setting up study sites. They also thank Dimagi, Inc. for the software set-up and technical support. They are grateful to the study staff, and the young men of Nyanza province, who are eager to fight the scourge of HIV by undergoing voluntary medical male circumcision.
1. Auvert B, Taljaard D, Lagarde E, et al.. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PLoS Med. 2005;2:e298.
2. Bailey RC, Moses S, Parker CB, et al.. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet. 2007;369:643–656.
3. Gray RH, Kigozi G, Serwadda D, et al.. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet. 2007;369:657–666.
4. Mills E, Cooper C, Anema A, et al.. Male circumcision for the prevention of heterosexually acquired HIV infection: a meta-analysis of randomized trials involving 11,050 men. HIV Med. 2008;9:332–335.
5. Gray R, Kigozi G, Kong X, et al.. The effectiveness of male circumcision for HIV prevention and effects on risk behaviors in a posttrial follow-up study. AIDS. 2012;26:609–615.
6. Bailey RC, Moses S, Parker CB, et al.. The protective effect of adult male circumcision against HIV acquisition is sustained for at least 54 months: results from the Kisumu, Kenya Trial. Vienna, Austria: International AIDS Society; 2010.
7. Hallett TB, Alsallaq RA, Baeten JM, et al.. Will circumcision provide even more protection from HIV to women and men? New estimates of the population impact of circumcision interventions. Sex Transm Infect. 2011;87:88–93.
8. Mehta SD, Gray RH, Auvert B, et al.. Does sex in the early period after circumcision increase HIV-seroconversion risk? Pooled analysis of adult male circumcision clinical trials. AIDS. 2009;23:1557–1564.
9. Wawer MJ, Makumbi F, Kigozi G, et al.. Circumcision in HIV-infected men and its effect on HIV transmission to female partners in Rakai, Uganda: a randomised controlled trial. Lancet. 2009;374:229–237.
10. Hewett PC, Hallett TB, Mensch BS, et al.. Sex with stitches: assessing the resumption of sexual activity during the postcircumcision wound-healing period. AIDS. 2012;26:749–756.
13. Kigozi G, Gray RH, Wawer MJ, et al.. The safety of adult male circumcision in HIV-infected and uninfected men in Rakai, Uganda. PLoS Med. 2008;5:e116.
14. Herman-Roloff A, Bailey RC, Agot K. Factors associated with the early resumption of sexual activity following medical male circumcision in Nyanza province, Kenya. AIDS Behav. 2012;16:1173–1181.
15. Thirumurthy H, Lester RT. M-health for health behaviour change in resource-limited settings: applications to HIV care and beyond. Bull World Health Organ. 2012;90:390–392.
16. WHO Department of Knowledge Management and Sharing. Knowledge management and health. Available at: http://www.who.int/kms/en/
. Accessed December 27, 2012.
17. U.S. Department of Health and Human Services Health Information Technology and Quality Improvement. What is mHealth? Available at: http://www.hrsa.gov/healthit/mhealth.html
. Accessed December 27, 2012.
18. UNAIDS. Joint Strategic Action Framework to Accelerate the Scale-Up of Voluntary Medical Male Circumcision for HIV Prevention in Eastern and Southern Africa 2012–2016. Geneva, Switzerland: UNAIDS; 2011.
20. Lester RT, Gelmon L, Plummer FA. Cell phones: tightening the communication gap in resource-limited antiretroviral programmes? AIDS. 2006;20:2242–2244.
21. Lester RT, Ritvo P, Mills EJ, et al.. Effects of a mobile phone short message service on antiretroviral treatment adherence in Kenya (WelTel Kenya1): a randomised trial. Lancet. 2010;376:1838–1845.
22. Pop-Eleches C, Thirumurthy H, Habyarimana JP, et al.. Mobile phone technologies improve adherence to antiretroviral treatment in a resource-limited setting: a randomized controlled trial of text message reminders. AIDS. 2011;25:825–834.
23. Reback CJ, Grant DL, Fletcher JB, et al.. Text messaging reduces HIV risk behaviors among methamphetamine-using men who have sex with men. AIDS Behav. 2012;16:1993–2002.
24. Stockwell MS, Kharbanda EO, Martinez RA, et al.. Effect of a text messaging intervention on influenza vaccination in an urban, low-income pediatric and adolescent population: a randomized controlled trial. JAMA. 2012;307:1702–1708.
25. Kharbanda EO, Stockwell MS, Fox HW, et al.. Text message reminders to promote human papillomavirus vaccination. Vaccine. 2011;29:2537–2541.
26. Free C, Knight R, Robertson S, et al.. Smoking cessation support delivered via mobile phone text messaging (txt2stop): a single-blind, randomised trial. Lancet. 2011;378:49–55.
27. Seidenberg P, Nicholson S, Schaefer M, et al.. Early infant diagnosis of HIV infection in Zambia through mobile phone texting of blood test results. Bull World Health Organ. 2012;90:348–356.
28. Odeny TA, Bailey RC, Bukusi EA, et al.. Text messaging to improve attendance at post-operative clinic visits after adult male circumcision for HIV prevention: a randomized controlled trial. PLoS One. 2012;7:e43832.
29. Mbuagbaw L, Thabane L, Ongolo-Zogo P, et al.. The Cameroon Mobile Phone SMS (CAMPS) trial: a randomized trial of text messaging versus usual care for adherence to antiretroviral therapy. PLoS One. 2012;7:e46909.
30. National AIDS and STI Control Programme MoH, Kenya. Kenya National Strategy for voluntary medical male circumcision October 2009. Nairobi, Kenya: National AIDS and STI Control Programme, Ministry of Health; 2009.
31. RapidSMS—an open source framework for SMS and Web integrated applications. 2012; Available at: http://www.dimagi.com/rapidsms/
. Accessed July 30, 2012.
33. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159:702–706.
34. Little RJ, D'Agostino R, Cohen ML, et al.. The prevention and treatment of missing data in clinical trials. N Engl J Med. 2012;367:1355–1360.
35. Royston P. Multiple imputation of missing values: update of ice. Stata J. 2005;5:527–536.
36. Royston P. Multiple imputation of missing values: further update of ice, with an emphasis on categorical variables. Stata J. 2009;9:466–477.
37. Free C, Phillips G, Watson L, et al.. The effectiveness of mobile-health technologies to improve health care service delivery processes: a systematic review and meta-analysis. PLoS Med. 2013;10:e1001363.
38. Car J, Gurol-Urganci I, de Jongh T, et al.. Mobile phone messaging reminders for attendance at healthcare appointments. Cochrane Database Syst Rev. 2012;7:CD007458.
39. Guy R, Hocking J, Wand H, et al.. How effective are short message service reminders at increasing clinic attendance? A meta-analysis and Systematic review. Health Serv Res. 2012;47:614–632.
40. Free C, Phillips G, Galli L, et al.. The effectiveness of mobile-health technology-based health behaviour change or disease management interventions for health care consumers: a systematic review. PLoS Med. 2013;10:e1001362.
41. A reality checkpoint for mobile health: three challenges to overcome. PLoS Med. 2013;10:e1001395.
42. Tomlinson M, Rotheram-Borus MJ, Swartz L, et al.. Scaling up mHealth: where is the evidence? PLoS Med. 2013;10:e1001382.
43. van der Kop ML, Karanja S, Thabane L, et al.. In-depth analysis of patient-clinician cell phone communication during the WelTel Kenya1 antiretroviral adherence trial. PLoS One. 2012;7:e46033.
44. Mbuagbaw L, Thabane L, Ongolo-Zogo P. Opening communication channels with people living with HIV using mobile phone text messaging: Insights from the CAMPS trial. BMC Res Notes. 2013;6:131.
45. Odoyo-June E, Rogers JH, Jaoko W, et al.. Wound healing and resumption of sex following medical circumcision of adult males in Kisumu, Kenya. Paper presented at: 16th International Conference on AIDS and STIs in Africa; December 4-8, 2011; Addis Ababa, Ethiopia. Abstract #4312.
46. Rogers JH, Odoyo-June E, Jaoko W, et al.. Time to complete wound healing in HIV-positive and HIV-negative men following medical male circumcision in Kisumu, Kenya: a prospective cohort study. PLoS One. 2013;8:e61725.
47. Odoyo-June E, Rogers JH, Jaoko W, et al.. Factors associated with resumption of sex before complete wound healing in circumcised HIV-positive and HIV-negative men in Kisumu, Kenya. J Acquir Immune Defic Syndr. 2013;62:465–470.
49. Agot K, Ohaga S, Ayieko B, et al.. We too are shareholders: why women must be meaningfully involved in the rollout of medical male circumcision in Africa. Paper presented at: XIX International AIDS Conference; July 22-27, 2012; Washington DC. Abstract #MOPDE0107.
50. Odoyo-June E, Rogers JH, Jaoko W, et al.. Sex before wound healing and condom use among newly circumcised HIV-positive and HIV-negative men in Kisumu, Kenya. Paper presented at: XIX International AIDS Conference; July 22-27, 2012; Washington DC. Abstract #MOPE170.
51. Kalichman SC, Pope H, White D, et al.. Association between health literacy and HIV treatment adherence: further evidence from objectively measured medication adherence. J Int Assoc Physicians AIDS Care (Chic). 2008;7:317–323.
52. Kalichman SC, Catz S, Ramachandran B. Barriers to HIV/AIDS treatment and treatment adherence among African-American adults with disadvantaged education. J Natl Med Assoc. 1999;91:439–446.
53. Chi BH, Stringer JS. Mobile phones to improve HIV treatment adherence. Lancet. 2010;376:1807–1808.
54. Bennett DA, Emberson JR. Text messaging in smoking cessation: the txt2stop trial. Lancet. 2011;378:6–7.