Effect of text-messaging on treatment adherence practices among young people living with HIV in Niger State, Nigeria

Recent studies reveal that Youths living with HIV (YLHIV) enrolled in HIV treatment experience a higher rate of interruption in treatment, sub-optimal medication adherence, and greater HIV-related mortality compared with other age groups. Therefore, this study investigated the effect of text messaging on medication adherence and antiretroviral-therapy (ART) retention practices among YLHIV in Niger state, Nigeria. The study was a quasi-experimental design in two selected hospitals. One hospital was assigned to a once-weekly text message intervention (TMI) for six weeks, and the 2nd served as the control group (CG). Data were collected at baseline, immediate post-intervention, and at the sixth-week follow-up. Data were analyzed using descriptive and inferential statistics at a 0.05 level of significance. Findings showed that between baseline and immediate post-intervention, there was a significant increase (p < 0.05) in the mean score of knowledge, perception, medication-adherence, and ART retention practices (from 9.62±12.28 to 22.98±4.57; 32.21±30.01 to 79.04±4.13; 13.50±8.41 to 26.65±13.15; 6.62±3.53 to 15.15±2.42) but there was no significant (p > 0.05) difference in the CG (from 10.29±11.08 to 10.58±10.74, 34.81±25.15 to 35.81±24.80, 17.13±9.23 to 17.23±9.16, 9.00±5.59 to 9.00±5.59) respectively. Between baseline and 6th week follow-up, there was a significant difference (p < 0.05) with a large effect size for knowledge (1.6), perception (2.2), medication-adherence (2.2) and ART retention practices (3.1) but no significant difference (p > 0.05) in the CG for all measures with an effect size of 0. The text messages intervention effectively influenced knowledge and perception of YLHIV, thereby improving treatment adherence practices. In addition, findings from this study can be used to design projects or interventions for scaling up the use of mobile phone text messaging to promote the adoption of health innovations and influence mHealth policy dialogue and review .


INTRODUCTION
Human Immunodeficiency Virus (HIV) infection burden globally has accelerated in the past 40 years, negatively impacting all socioeconomic and ethnic groups, and significantly influencing health outcomes, welfare, employment, and criminal justice (Awofala and Ogundele, 2018).According to recent epidemiological statistics, HIV remains a public health issue that continues to drain the economy, having taken more than 36.3 million people's lives in the four decades since its discovery (UNAIDS, 2021).As of the end of 2021, 37.7 million were estimated to be living with HIV, with Sub-Saharan Africa accounting for 25.4 million PLHIV with 67 and 39% of the global HIV burden and newly infected people living with HIV (UNAIDS, 2021).Young people aged 15 to 24 comprise approximately a third of all new HIV infections in sub-Saharan Africa.They account for about 7% of persons living with HIV in Nigeria and often have poorer treatment outcomes than adults (Jahun et al., 2021).Nigeria is home to 9% of all HIV-positive persons (UNAIDS, 2014).The nation, which has already suffered from political instability and widespread political corruption because of almost 33 years of military dictatorship, seems to be on track to 'cure' the virus within a few (NACA, 2012).However, despite advances in institutional changes and political commitment to combat the illness, the nation has experienced a rise in the number of residents put on lifesaving antiretroviral treatment (AART) to improve their chances of survival (NACA, 2012).Medication adherence and antiretroviral therapy (ART) retentions are critical requirements for successful virologic suppression and improved treatment outcomes among youths.However, several factors influence youth regimen compliance, such as forgetfulness, low level of knowledge, distance to health facilities, fear of disclosure of HIV status, poor selfcare skills, and poor parental care and support, among others (Erah and Arute, 2008;Verma et al., 2020).Youth is a transition period from childhood dependence to adulthood independence and awareness of our interdependence as community members (Alawale et al., 2022).HIV treatment among the youths is characterized by loss of follow-up, poor adherence to ART, poor integration of Adolescent and youth-friendly services with intervention centers, adolescent perception of care, risk of treatment failure, and high mortality rates compared to young children and adults.A cross-sectional study conducted in four facilities as a baseline to an interventional program revealed that medication adherence among young people living with HIV is suboptimal (Asaolu and Agbede, 2022).
Routine Antiretroviral Therapy (ART) is tedious to some young people, while others find it inconvenient when people find out, fundamentally becoming an impediment to their everyday lives.For this population segment, timeliness, consistency in treatment, and maintenance of active medication often involve more than a simple time track for medication and hospital check-ups.It requires good motivation and additional effort to rise above the various factors that affect adherence (Luseno et al., 2019).Young people now have increased access to the internet and mobile phones, which have stimulated interest in exploring the role of health (Poushter, 2016, WHO, 2016).Access and ownership of mobile telecommunication devices have increased drastically in the past two and half decades, especially in low and middle-income countries (Linnemayr et al., 2007).Mobile communication technologies in medical care (mHealth) have been suggested as a method to improve the delivery of health services, especially to this youth population (Van der Kop et al., 2018).There are different variations of mHealth interventions, and they include short messaging services (SMS) reminders, gaming applications (apps), and interactive mobile apps (Cele and Archary, 2019).Text messaging is a unique mHealth variant because it allows researchers to personalize message content, promote bidirectional communication, and pair message timing to ART dosage schedules (Finitsis et al., 2014).Results from the previous study reported that using mobile phone text messaging reminders was feasible and acceptable in a group of eight HIV-infected young people, but that viral suppression declined for most individuals after the 12week intervention ended; moreover, no objective measure of adherence was used (Poushter, 2016,).Although there are few efficacy studies on the use of mobile phones to improve adherence, a rising body of research suggests that their use in chronic conditions is promising (Tanue et al., 2020).Therefore, this study investigated the effect of text messaging intervention on knowledge, perception, medication adherence practice, and ART retention among youths living with HIV in Niger State, Nigeria.

Study design, population
The study utilized a quasi-experimental involving patients receiving HIV care in two selected general hospitals in Niger State.Multistage sampling technique was employed in the selection of participating health facilities.One of the three senatorial districts (Niger East) in Niger State was purposively selected for this study.The two hub facilities providing free comprehensive HIV care and treatment were purposively selected-Rafi LGA (Control) and Suleja LGA (Intervention).Fifty-two (52) young people living with HIV from the two selected health facilities that met inclusion criteria and *Corresponding author.E-mail: gbengasaolu@gmail.com.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License consented to participate in the study were enrolled in the intervention arm.To be eligible, patients had to be at least 18 years old, enrolled on ART for at least 6 months, own or have access to a mobile phone, and be able to use simple text messaging.Exclusion criteria were previous assessment for ART eligibility, previous or current exposure to ART, and pregnancy.The study was conducted between December 2021 and March 2022.Ethical approval for the study was obtained from the Babcock University Health Research Ethical Committee (BUHREC) and the Niger State Ministry of Health Ethical Review Committee (NSMOH ERC).Signed consent forms were obtained from the respondents.Respondents on the intervention arm received one-way text messages containing key messages from the Intervention Module designed by the researcher.Prior notifications were given to the respondents for the text message group that the word "food" would be used instead of medication or drug when sending the messages to maintain confidentiality.So, the following messages were sent consecutively once weekly for six weeks.
1. Taking your food correctly, as advised, helps one to live a healthy life 2. Adherence is taking your food as advised at the correct time.3. Adherence is not taking breaks and taking your food as advised at the right time. 4. Adopting a healthy lifestyle and trying to avoid risky behaviors.Take your food. 5.You are critical.Do not play with your health.Take your food.6.You are so loved.Handle your health with great care.Take your food.
The control arm received the standard of care provided by the facilities but excluded text messaging during the twelve weeks of intervention and post-intervention.The study outcomes were knowledge, perception, medication adherence practice, and ART retention practice scores from each construct of the validated questionnaire administered at baseline, immediately postintervention, and six-week follow-up period.

Sample size determination
Previous studies documented that approximately 60% of participants would maintain >= 95% adherence without any intervention, while 85% of participants would maintain >=95% adherence with the intervention (Tanue et al., 2020).Hence, the mean difference between the treatment and control arm (D 2 ) was set at 0.3.Furthermore, the significance level (p-value) and Power of the study was set at 5 and 80%, respectively.Hence, the Standard normal deviation (z score) Zα and Z β were 1.960 and 0.84, respectively, at a Confidence Interval of 95%.The sample size was computed using the sample size for intervention trails described by Cohen (1992) and arrived at 47 participants for each study group.Ten percent was added to account for any attrition or loss to follow-up during the study.Fifty-two YPLHIV were assigned to each group, and there was a 100% completion rate.Therefore, a total of 104 participants were enrolled in the study and participated throughout the study.

Instrumentation
Three instruments were deployed for this study, and the first was a 67-item interviewer-administered questionnaire (Cronbach alpha was 0.96).The second was a visual analog scale that fed into the last item on the questionnaire.The third was a text message record log to document the timely dispatch and delivery of text messages to participants.The questionnaire contains four sections.Section A captured the socio-demographic characteristics of the participants, Asaolu et al. 43 partner HIV status, and social climate.The age of the participants was assessed in an open-ended question; sex, ethnicity, and other demographic variables were coded.Section B captured information on knowledge relating to medication adherence and ART retention practices and was assessed on a 25-point scale dichotomous question of Yes/No, I don't know.Section C captured information on the Perception of medication adherence and ART retention.The perception of the YLHIV on medication adherence and ART retention was assessed on an 82-point Likert scale which was categorized into four perception domains operationalized by variables of perceived seriousness of the consequences of poor adherence and retention (24-point scale); Perceived susceptibility to complications resulting from treatment failure (12-point scale); perception of benefit of taking recommended medication (30-point scale) and perception of self-efficacy to take recommended ART medications (16-point scale).The fourth section captured information on Medication adherence and ART retention practice.Medication Adherence was measured using the combination of 4response options Likert-type response categories, and three yes or no questions.At the same time, while ART retention was also measured using the combination of 4-response options Likert-type response categories, and two yes or no questions.Aggregating the seven items and six items in the sub-scale created a 27-point scale and 16-point scale of measurement for Medication adherence practice and ART retention practices, respectively.The average time to complete the questionnaire was estimated to be 25 min.The instrument was validated through a rigorous review by the Project Supervisor, a Professor of Public Health, and other faculty members at the School of Public and Allied Health.The contents of the questionnaire were strengthened with items extracted from peer-reviewed literature.The internal consistency of the questionnaire was assured through retesting with equivalent groups (n=10) away from the intervention area.Content and item analysis was conducted with a Cronbach alpha score of 0.97.

Data management and analysis
Data Analysis was accomplished by using the computer software STATA package version 15.0.Data collected from participants using the instruments were reviewed for completeness, edited, and coded using a coding guide designed before data collection and entered into the computer by research assistants.Computations involving frequency distributions, summaries of descriptive statistics and paired t-tests, were used to process the data collected and to test the validity of the leading hypothesis concerning knowledge, perception, medication adherence practice, and ART retention practice among clients who participated in the study.All statistical tests are set at p=0.05 level of significance cut-off.The decision rule applied was that if computed p ≤ 0.05, the null hypothesis will be rejected in favour of the alternative view; otherwise, do not reject.To standardize the magnitude of the impact accountable to the intervention conditions, since the p-value cannot estimate this change but only expresses that the change is present and is significant at a predetermined cut-off; hence the inclusion of Cohen"s d, also known as the effect size (ES) for the difference in means of two independent groups and the corresponding 95% confidence interval (95%CI).This became an effective tool to accurately estimate and compare the magnitude of the changes produced by the intervention across all variables of interest (Cohen, 1988;Nakagawa and Cuthill, 2007).

RESULTS
One hundred and four participants were engaged in the study, with fifty-two participants in each group.The two  3).Furthermore, between baseline and 6 weeks of follow-up, there was a significant increase (p<0.05) in the mean score of knowledge (from 9.62±12.28,38% to 23.85±3.07,95%), perception   2 and 3).

DISCUSSION
Overall, the intervention group had improved knowledge and medication adherence immediately after the intervention compared to the control group.This can be attributable to text messaging.

Table 1 .
Socio-demographic characteristics of the participants in the study for each arm of the intervention at baseline.

Table 2 .
Impact evaluation of variables at baseline and immediate post-intervention for text messages.
*Perception is a composite aggregate of the components; severity, susceptibility, benefits and self-efficacy.Source: Authors

Table 3 .
Impact evaluation of variables at baseline and immediate post-intervention for control group.Perception is a composite aggregate of the components; severity, susceptibility, benefits and self-efficacy. *