The factors influencing adherence among HIV positive adolescents were investigated from the Social Ecological Model, (SEM) theoretical perspective of Individual, Interpersonal, Organizational/Institutional, and Community Levels. The SEM theoretical perspective stresses that individual behavior is influenced by factors at different levels as classified by Sallis and Owen [25] . The study was conducted between August and December 2018.

The study defined Non-adherence to cART as being unable to answer correctly all the responses on the 4 item Morisky drug compliance scale within the past one month from the date of data collection.

In this study, when the respondent correctly answered 12 or more questions on knowledge it was regarded as having high level of knowledge. The respondent was said to have low level of knowledge if they answered 1 to 11 questions on knowledge correctly.

Social support was defined as adequate when the respondent answered correctly 5 questions and above on the social support section. The respondents who had inadequate social support were those who only managed to correctly answer between 1 and 4 questions correctly.

Stigmatization was regarded as high when the respondent had 1 or 2 scores and as low with the score of 3 or 4 on the stigmatization section. Responses from the questionnaires were assigned numerical codes for entry and analysis using the SPSS software computer package version 20. Pearson’s Chi square and fisher’s exact statistical tests were used to determine the association between dependent (adherence to cART) and independent variables (knowledge of HIV and disease progression, social support, stigmatization and cultural beliefs). The fisher’s exact test was used where the criteria for using the Pearson’s chi square was not met. The binary logistic regression test was used to come up with the true predictors and to adjust confounders of adherence to cART among HIV positive adolescents. The cut off point for statistical significance was set at 5%.

The study used a descriptive quantitative cross sectional research design. The design was selected because data was collected at one point and no interventions or manipulations were carried out on both the environment and the respondents.

The study population consisted of HIV positive adolescents on cART treatment receiving care at Adult Infectious Diseases Centre, Lusaka. The study targeted males and females adolescents aged between 16 and 19 years on cART for at least three months enrolled for care at Adult Infectious Diseases Centre. These were adolescents who had transitioned from children to adult hospital soon after attaining their 16^th birthday. Parents and guardians after this hospital transition seem to reduce on accompanying these HIV positive adolescents for the clinical follow ups.

A probability sampling method was adopted using systematic sampling method. The k^th subject on a scheduled booking was asked to participate in the study using N/n. Therefore, every 5^th subject of HIV positive adolescents on cART who was coming for scheduled visits was requested to participate in the study [26] . This method provided all the subjects in the population with equal chance to be included in the study. A sample size is a small fraction of the population to represent the target population under study [26] . The sample size was calculated according to Krejcie and Morgan formula for the limited or finite population [27] . The final sample size was 173. The questionnaire with the adopted Morisky 4 item scale was used to determine level of adherence to cART among the eligible study participants. The validity and reliability of the instrument was insured by pretesting the instrument and which was also reviewed by team of experts.

Almost half 49.1% (85) of the study respondents were males. Slightly above half 53.2% (92) of the respondents were within the age group range of 16 - 17 years.

Slightly above half 52.6% (91) of the respondents had attained secondary school education. The majority 99.4% (172) of the respondents were living with their family i.e. biological parents, grandparents and uncle or aunty. Above one third 38.1% (66) respondents had both biological parents still alive. The majority of the respondents had either their father 24.9% (43) or mother 17.9% (31) dead and 19.1% (33) had reported that both of their parents had died.

Generally in this study 44% (76) of the respondents were found to be non adherent to cART (Figure 1).

Most respondents 59.5% (103) were found to have low level knowledge about HIV and disease progression (Figure 2).

The majority of the respondents 94.8% (164) reported to have adequate social support and only 5.2% (9) had inadequate social support (Figure 3).

The social demographic characteristics showed that 50.9% (88) of the respondents interviewed were female. The slightly high proportion of female respondents may be due to older male partners infecting them in contexts of poverty, manipulation and exploitation without condom use [30] . To support this view the current study found that 1.7% (3) of the HIV positive adolescents respondents got the infection from being raped by males in their home set ups. Men in such relationships often have multiple sexual partners and acquire sexually transmitted infections, which they pass on to adolescent females [23] . The high female rates may be related to unprotected heterosexual relations with multiple partners [31] . An additional factor contributing to the sex imbalance in adolescent HIV prevalence may be the higher risk of AIDS-related mortality in male adolescents compared to females, which may be related to the lower proportion of males who receive cART [10] .

Similar to the current study, higher rates of HIV among females than males have been reported in sub-Saharan Africa [32] . According to UNICEF reports, females aged 15 to 17 years in the sub-Saharan region have up to four times higher prevalence rates of HIV reported than among their male counterparts [33] . The similar high prevalence among females may be due to similar inter-cultural belief systems and practices which promote HIV transmission to the girl child. The practices that include high cases of early marriages to female adolescents and adult HIV infected men practicing sexual intercourse rituals with minor girl to try and cure the infection [3] .

Contrary to findings in the current study, the highest HIV prevalence rate in the United States and parts of Europe was reported among males [34] . In the developed settings the prevalence among male adolescents could be higher because of injectable illicit drugs abuse using same needle by several group members. The use of the same needle among several adolescents within the same gang, could contribute to high HIV rates among males [35] .

The study also revealed that more than half (53.2% (92)) of the respondents were within the age group 16 - 17 years. This statistics suggest that most of the perinatally infected adolescents are now living longer attributed to improved HIV/AIDS care services. Most (52.6% (91)) of the respondents had attained secondary school education. This is contrary to the beliefs that HIV diagnosis is a death sentence and there was no need to pursue career or educational goals [36] . Literature has shown that with the advances in effectiveness and availability of cART, HIV infected children are now surviving to adolescence and emerging as a new group in the global HIV/AIDS pandemic [24] . It has been also cited that despite the fact that HIV positive adolescents find it a challenge living with the HIV, they have remained positive and focused with their education [24] .

Respondents in the current study (44% (76)) were found to be non adherent to their prescribed cART (Figure 1). It must be emphasized that adherence to cART is critical for successful viral suppression because non adherence could lead to an increase in HIV viremia, risk of treatment failure, and accumulating resistance mutations [37] . The 44% non adherence rate is very alarming and pauses as a threat to the maintenance of future cART regimen among HIV positive adolescents due to possible occurrences of related drug resistance and virologic failure.

Similar to the current study, in Thailand slightly higher (48.4%) suboptimal adherence to cART among adolescents was reported [24] . Despite difference in geographical locations it has been noted that there could be some similar characteristics which exist and affecting the pattern of cART adherence among adolescents [38] . It could be said that non adherence to cART is becoming a global phenomenon and collective strategic efforts need to be developed to prevent the risk of development of treatment failure. Several studies in sub-Saharan Africa have also reported poor adherence among adolescents which is of significant concern given the limited cART options available and the risk of drug resistance [38] [39] . Non-adherence to cART has been perceived as an emerging public health concern [39] .

Among the respondents that were found to be non adherent to cART, slightly above half 51% (39) were males. These findings in the current study confirms the growing body of evidence indicating that male adolescents face challenges in accessing HIV services more often than their female counterparts. The males also experience worse treatment outcomes, including high mortality [40] . With these challenges among male adolescents it is likely that they may not adhere to their cART. Elsewhere evidence has also shown sex-specific differences outcomes that male adolescents have worse adherence compared to their female counterparts [41] . Globally more emphasis has been placed on protection of the girl child and as a result the male child seems to be disadvantaged much more in health related matters [30] . However, it was found that gender was independent of adherence to cART (p-value of 0.760). This could be attributed to the fact that the pathogenesis of HIV is similar despite the person’s gender.

The current study found that majority of the respondents 57% (43) who where non adherent were aged between 16 and 17 years old. This coincides with the age when adolescents are transitioned from pediatrics to adults HIV care. The higher non adherence to cART occurring during this period could be attributed to some challenges associated with implementation of transitioning process of care. A study in West Africa observed that transitioning of care from pediatrics to adult should be a gradual process according to the maturity of the adolescents with evidence of psychological stability and readiness of independent care [30] .

On a contrary to the current study, another Zambian study found that there was suboptimal adherence to cART mainly due to increase in age [18] . This could have been that as adolescents increase in age, they are expected to take up more gender roles and responsibilities in society. It has been documented that the successful fulfillment of new gender roles and responsibilities come with challenges that may disrupt the routine daily activities including medication timings [42] .

Adherence to cART was found to be independent (p-value 0.428) of the age of the adolescent in the current study. The adolescents’ age did not significantly influence adherence and there was a minimal difference in the pattern of adherence to cART between older and younger adolescents. It can be argued that despite the age of the adolescents they experience similar influences of adherence to cART. Although some studies have shown high rates of cART adherence among HIV-infected adolescents, a lower rate of adherence among HIV-infected adolescents has been reported more often, and expectations have been that adherence will diminish over time [24] [42] . More cohort studies need to be conducted to identify the non enhancers of adherence to cART.

In the current study, majority (59.5% (10)) of the respondents were found to have low level knowledge about HIV and disease progression (Figure 2). Earlier CDC had highlighted that low level of knowledge among adolescents could be due to limited access to health information and little sources of information about HIV and disease progression [35] . The health information could easily be shared in adolescents-friendly health corners, which according to Zambian consolidated treatment and prevention guidelines are very limited in most health care centers [12] . Limited adolescent-friendly corners may therefore contribute to low level of knowledge about HIV and disease progression. Efforts to improve knowledge and maintain adherence to cART, have been noted where adolescents training curriculum have been developed by the Zambian Ministry of Health in collaboration with International Center for AIDS Care and Treatment Programs [12] . This is aimed at empowering adolescents and caregivers with youth-friendly HIV care information. Also caregiver support was mentioned [29] as a major facilitator of adherence among adolescents with a chronic disease, but this needs a lot of education on disease etiology among all those in this age group. This initiative could also mean that most interventions to promote adherence to cART among adolescent patients can easily be universally applied to enhance of knowledge on HIV and disease progression.

The findings in the current study are consistent with CSO report were it was also reported that the comprehensive HIV knowledge among adolescents was also very low that is, 39% and 45% for females and males respectively [3] . It could be argued that adolescents depend on healthcare providers for information about their health and cannot independently seek to acquire more knowledge about their health. Therefore health care providers must be proactively providing health related massages for the adolescents to acquire more knowledge about HIV and the disease progression.

Contrary to the current study, conducted in Zambia found 56.7% of the adolescents had knowledge about their HIV status [43] . Despite the reported knowledge it was found that adolescents had challenges with maintaining the desired 95% adherence threshold. The 95% adherence to cART has been strongly linked to viral suppression in most studies.

Similar to the current study UNICEF reported that the overall level of knowledge of HIV and AIDS remained very low among older adolescents aged 15 - 19 [23] . Interestingly to note that similar findings were reported in Ghana that 47% of adolescents on cART had no knowledge about the HIV disease and its progression [44] . These similarities could mean that adolescents have common interests and aspiration despite geographical variation. It could also mean that there is very little or no impact on the current trends of practice to enhance knowledge on HIV and disease progression. This is because low level of knowledge has kept on being reported hence the need for studies on impact assessment on the HIV care models implementation.

However, when adherence to cART was associated with knowledge of HIV and disease progression in the current study, the relationship was found to be independent (p-value of 0.585) (Table 1). Having Knowledge about HIV and disease progression alone cannot prompt improved adherence to cART. This is evident in some more recent studies were despite high knowledge among HIV positive adolescents report [24] [44] [45] , it was found that they were non adherence to cART. It could be argued that acquisition of high level knowledge alone is not enough to promote adherence to cART among HIV positive adolescents. A holistic approach in devising interventions must be adapted as the most accurate measure to promoting adherence to cART among HIV positive adolescents. A combination of personal, interpersonal, community and societal interventions as prescribed by social ecological model levels could be the most effective approach to improving adherence to cART among adolescents.

The current study reported that understanding the reason why the respondents were taking cART was found to be strongly dependent (p-value 0.006) with adherence to cART. Having understanding could mean having the ability to interpret and explain information about HIV and being able to estimate and anticipate undesired outcomes when there is poor adherence to cART. It is expected that those respondents who had understanding of the reason they were taking cART, also knew the consequences of non adherence. Therefore, they were compelled to be adherent to their cART to prevent disease progression and drug resistance.

Similar to the current study some studies elsewhere have also shown that having understanding of the benefits for taking cART could have a positive influence on cART adherence among HIV positive adolescents [46] . These finding suggest that promotion of adherence to cART, requires deeper level of knowledge about HIV and disease progression.

The current study established that 58% (56) of the respondents (Table 2) who were adherent to cART had low level knowledge. This could mean that the adolescents who acquire more knowledge feel that it is enough to prevent the disease progression. This is evident by higher pattern of adherence to cART by the respondents who had low knowledge 62% (47). The respondents who had low knowledge of HIV and disease progression had more than half times chance more likely to be adherent to cART than those who had high level of knowledge about HIV and disease progression (Table 2). This could be due to the caregiver’s strictness in ensuring that the adolescents are adherent to cART who may

have low knowledge of HIV and disease progression. The adolescents who had low level of knowledge could also be depending on their buddies and healthcare providers for the instructions and care hence the higher rate of adherence to cART among those reported to have low level knowledge.