Predictors of loss to follow up from antiretroviral therapy among adolescents with HIV/AIDS in Tanzania.

Access to Antiretroviral Therapy (ART) is threatened by the increased rate of loss to follow-up (LTFU) among adolescents on ART care. We investigated the rate of LTFU from HIV care and associated predictors among adolescents living with HIV/AIDS in Tanzania. A retrospective cohort analysis of adolescents on ART from January 2014 to December 2016 was performed. Kaplan-Meier method was used to determine failure probabilities and the Cox proportion hazard regression model was used to determine predictors of loss to follow up. A total of 25,484 adolescents were on ART between 2014 and 2016, of whom 78.4% were female and 42% of adolescents were lost to follow-up. Predictors associated with LTFU included; adolescents aged 15-19 years (adjusted hazard ratio (aHR): 1.57; 95% Confidence Interval (CI); 1.47-1.69), having HIV/TB co-infection (aHR: 1.58; 95% CI, 1.32-1.89), attending care at dispensaries (aHR: 1.12; 95% CI, 1.07-1.18) or health center (aHR: 1.10; 95% CI, 1.04-1.15), and being malnourished (aHR: 2.27; 95% CI,1.56-3.23). Moreover, residing in the Lake Zone and having advanced HIV disease were associated with LTFU. These findings highlight the high rate of LTFU and the need for intervention targeting older adolescents with advanced diseases and strengthening primary public facilities to achieve the 2030 goal of ending HIV as a public health threat.

Introduction

In 2018, an estimate of 1.6 million adolescents aged 10 to 19 were living with HIV worldwide [1]. World Health Organization (WHO) reported HIV counseling and testing, linkage to care and treatment to be a challenge [2]. Of all patients enrolled in HIV care, only 23% were retained [3]. A review done in Sub-Saharan African countries showed the rate of loss to follow up (LTFU) among adolescents ranged between 15% and 54% [4]. In 2015, Tanzania National AIDS Control Program (NACP) reported the rate of LTFU among 75, 596 patients aged 15–24 years to be 23.5% [5]. Similarly, a study in Nigeria reported the probabilities of LTFU among adolescents living with HIV (ADLHIV) were 3.6%,6.9,% and 35.9% at 6,12, and 25 months respectively [6]. Predictors associated with increased risk of LTFU among ADLHIV reported in various studies included; ADLHIV aged 15–19 years, female adolescents, those diagnosed with HIV/TB co-infection, those with malnutrition, adolescents who attended clinics at primary facilities and having advanced WHO clinical stage [7-12]. Also increased risk of LTFU was observed among adolescents who had prior exposure to ART [13] and those who attended clinics at public health facilities [14].

In Tanzania, various studies and routine assessments of HIV services have been done among adults but limited data exist among adolescents group [15]. Routine data indicates a high rate of loss to follow-up among adolescents globally and in Tanzania. To achieve the 2030 goal of ending the HIV epidemic as a public health threat, identification of predictors of LTFU on ART is urgently needed to inform effective strategies of retention in care among adolescents living with HIV/AIDS.

Methods

Study design and population

We performed a retrospective cohort study of routinely collected data obtained from the Care and Treatment Clinics (CTC3) macro database at the National AIDS Control Program (NACP). This study included adolescents who were enrolled and initiated ART at age 10–19 with either vertical or horizontal HIV/AIDS infection from January 2014 to December 2016. Adolescents whose records had missing details of enrollment date and last appointment date in the CTC3 macro database were excluded from the study. Adolescents who were transferred out and died were censored at the time of their last visit.

Study setting

The data involved all ADLHIV enrolled in ART care from 2014 to 2016 in all 26 regions of Tanzania mainland available in CTC3 database. Tanzania mainland is the largest country in East Africa covering 947,300 square kilometers with an estimate of 55 million population. In 2020, population of adolescents aged 10–19 years in Tanzania mainland was projected to be 13,206,921 based on the 2012 national census [16] with the HIV prevalence of 5.8% [17]. In Tanzania, adolescents can access ART care in health facilities which are categorized into three levels. These levels include; primary health facilities also known as dispensaries, secondary health facilities (health centers), and tertiary health facilities (hospitals). The Tanzania HIV guideline requires adolescents to attend HIV clinics at least once a month for the first three months of the ART treatment and thereafter once every three months depending on their adherence status [18]. To reduce the burden of TB in HIV patients all ART clinics are integrated with TB/HIV services [18]. TB/HIV Co-infection refers to HIV patients who were also diagnosed with TB infection during routine TB screening. In Tanzania, all people living with HIV (PLHIV) are screened for TB on every clinic visit, and those not infected are provided with Isonized Preventive Therapy (IPT) to prevent them from developing active TB.

Data source

Data were extracted from CTC3 database at NACP combining all patients’ records attending CTC.

Standard patient management form known as CTC2 cards are used to capture patients’ records in the facilities. The CTC2 cards have information on; patient’s socio-demographic characteristics, WHO clinical stage, weight, CD4 counts, ART regimen, ART initiation date, test results (TB screening), nutrition status, height, ART adherence status, mortality records, and haemoglobin level. In each CTC clinic, patients’ records are directly exported to the CTC3 database through a computer. All patients are recorded by their unique identification number at the time of enrolment.

Study variables

The main outcomes in this study were loss to follow-up from HIV care (ART) and retention on ART care. The loss to follow-up was defined as any of the adolescents who failed to attend at least one clinic visit within 90 days of their scheduled appointment. Retention to care was defined as the state of an adolescent being alive, actively attending scheduled appointments at the clinics, and receiving ART care at the end of follow-up period. Independent variables of the study included; sex which was classified as male and female, and age was grouped as 10–14 years and 15–19 years. Marital status was grouped as never married/single and cohabiting/married, types of health facilities were categorized as dispensary, health centers, and hospitals. Also, facility ownership was classified as public and private, WHO clinical stages were grouped into stage I, II, III, and IV. Geographical zones were grouped as coastal, central, lake, northern, southern highland and western zones.

Nutrition status was categorized into two groups according to BMI scales. BMI was obtained from adolescent’s weight in kilograms divided by the square of height in meters which were assessed at every clinic visits. Participants with BMI of <18.5kg/m2 (underweight) or 25.0–29.9kg/m2 (obesity/overweight) were classified as “malnutrition”, while those with BMI between 18.5–24.9kg/m2 (normal weight) were classified as “no malnutrition”.

ART regimens were categorized as first-line and second-line based on Tanzania HIV guideline [18]. First-line regimens included; Tenofovir (TDF) 300 mg / Lamivudine (3TC) 300 mg / Efavirenz (EFV) 600mg. Alternative first line regimen used were; Tenofovir (TDF) + Emtricitabine (FTC) + Dolutegravir (DTG), Abacavir (ABC) + Lamivudine (3TC) + Efavirenz (EFV) and Zidovudine (AZT)+Lamivudine(3TC) +Nevirapine (NVP). Also, the second-line regimens included; Zidovudine (AZT), Tenofovir (TDF), Abacavir (ABC), Lamivudine (3TC), Emtricitabine (FTC), Atazanavir boosted by Ritonavir (ATV/r), Lopinavir boosted by Ritonavir (LPV/r) and Dolutegravir (DTG). Other independent variables included; prior exposure to ART during enrolment (yes or no) and HIV/TB Co-infection (yes or no).

Statistical analysis

Stata 15 IC (StataCorp, College Station, TX, USA) was used for analysis. Patients’ characteristics were summarized using proportion for categorical variables, mean and standard deviation for continuous variables. Overall and covariate specific rate of loss to follow up per 1000 person-months (pm) was determined. Kaplan-Meier curves were used to determine failure probabilities among adolescents who were enrolled on ART. The log-rank test was used to test for statistical significance of the difference in the failure probability curves. Predictors of loss to follow up on ART were determined using Cox proportion hazard regression model. Variables with a p-value ≤ 0.2 in the bivariate analysis and other potential confounders were included in multivariate analysis. Crude and adjusted hazard ratios with their respective 95% confidence interval were reported.

Ethical consideration

The study and analyses of the CTC3 patients’ records were covered by ethical approval from Muhimbili University of Health and Allied Sciences (MUHAS) with Ref. No. DA.287/298/01A. Since this study used secondary data obtained from the CTC3 database available at the Tanzania National AIDS Control programs, the institutional review board (IRB) waived the need for consent from parents and guardians of the adolescents. The informed consent was not required because patients’ records were anonymized before access and adolescent’s records were identified by their unique registration numbers. However, permission to access data was obtained from the National AIDS Control Program.

Results

A total of 25,880 records were obtained from the CTC-3 database. About 396(1.5%) records with no ART enrollment and last appointment date were excluded from this study. Of these, 25,484 (98.5%) records were eligible for this study based on inclusion criteria (Fig 1). About 177(0.7%) records were deceased and 74(1.7%) records were transferred out on their last appointment date. A total of 14,498(56.8%) records were retained duringfollow-uow up time of the study.

Fig 1

Flow chart for the selection procedure of the study subjects.

Baseline characteristics of the participants

Three-quarter (78.36%) of the adolescents on ART were female. Above seventy percent (72.09%) of the study cohort were aged between 15–19 years with a median (Interquartile range, IQR) age of 17 (14–18) years. At enrollment over ninety percent of adolescent had no HIV/TB Co-infection and more than one-third (45.96%) were in WHO stage I. Adolescents who were on the first line regimen accounted for 98.84% of the study population. Above 90% of adolescents on ART were termed to have normal nutrition status (Table 1).

Table 1

Characteristics of adolescents on ART (N = 25,484).

Characteristics	Number	Percentage
Sex
Male	5,515	21.64
Female	19,969	78.36
Age groups (years)
10–14	7,112	27.91
15–19	18,372	72.09
Median (IQR) age	17(14–18)
Marital status
Never married/Single	12,963	50.87
Cohabiting/Married	7,701	30.22
Missing	4,820	18.91
Types of Health Facility
Dispensary	9,114	35.76
Health Centre	8,573	33.64
Hospital	7,228	28.36
Missing	569	2.23
Facility Ownership
Public	20,075	78.77
Private	4,840	18.99
Missing	569	2.23
WHO Clinical Stage
I	11,712	45.96
II	5,594	21.95
III	6,409	25.15
IV	1,176	4.61
Missing data	593	2.33
Prior Exposure on ART
Yes	9,968	39.11
No	15,516	60.89
HIV/TB Co-Infection
Yes	366	1.44
No	23,296	91.41
Missing	1,822	7.15
Geographical zone
Coastal	5,620	22.05
Central	2,249	8.83
Lake	7,506	29.45
Northern	2,031	7.97
Southern Highland	6,632	26.02
Western	877	3.44
Missing	569	2.23
ART Regimens
Firstline Regimen	25,188	98.84
Second line Regimen	296	1.16
Nutritional Status
Malnutrition	94	0.37
No Malnutrition	23,114	90.70
Missing	2,276	8.93

Rate of loss to follow up among adolescents on ART

The follow-up time of 8,100 person-months at risk (pmr) was accumulated from 25,484 adolescents who were on ART during the study period. A total of 10,735 (42.20%) were LTFU with an overall rate of 1.33 per 1000 (95% CI; 38–40) pmr (Table 2). Rate of LTFU among adolescent on ART at 3,6,12 and 24 were 0.16, 5.60, 1.91 and 0.58 per 1000 pmr respectively. The highest rates were among adolescent aged 15–19 (1.63, 95% CI; 1.59–1.66); female adolescents (1.50, 95% CI; 1.47–1.53); with WHO stage I (1.72, 95% CI; 1.68–1.77) and residing in Lake zone (1.59, 95% CI; 1.54–1.64). Likewise, a high rate of LTFU was observed among malnourished adolescents (2.04, 95% CI; 1.52–2.74) and those in first-line regimen (1.35, 95% CI; 1.32–1.38) (Table 2). Also, the median time of LTFU from ART care was 5.59 months. Likewise, the overall failure probabilities of LTFU among adolescents on ART were 11.2%, 17% and 24% at 3, 6 and 12 months respectively. Failure probability was high among adolescents aged 15–19 years and those with HIV/TB co-infection (Fig 2).

Table 2

Rate of loss to follow up among adolescents on ART.

Characteristics	Person-Month (pm)	No of loss to follow up among Adolescents on ART	Rate of Loss to follow up Per 1000pm	95% CI
				Lower	Upper
Loss to follow up rate /Tanzania Mainland	8,100	10,735	1.33	1.31	1.36
3 Month	115.65	19	0.16	0.10	0.26
6 Months	966.70	5418	5.60	5.46	5.76
12 Months	1700	3194	1.91	1.84	1.98
24 Months	3000	1744	0.58	0.56	0.61
Sex
Male	1900	1558	0.80	0.76	0.84
Female	6100	9177	1.50	1.47	1.53
Age Group (Years)
10–14	2600	1860	0.72	0.69	0.75
15–19	5500	8875	1.63	1.59	1.66
Types of Health Facility
Dispensary	2700	4165	1.56	1.51	1.61
Health Center	2700	3696	1.36	1.32	1.41
Hospital	2500	2763	1.12	1.07	1.16
Missing	191.5	111	0.58	0.48	0.69
Facility Ownership
Public	6200	8725	1.39	1.37	1.42
Private	1600	1899	1.17	1.12	1.23
Missing	191.50	111	0.58	0.48	0.69
Marital Status
Never Married/Single	515	1,214	2.36	2.22	2.49
Cohabiting/Married	841	1,395	1.66	1.57	1.75
Missing	317	585	1.85	1.70	2.00
WHO Clinical Stage
I	3400	5858	1.72	1.68	1.77
II	1800	2026	1.10	1.05	1.15
III	2300	2177	0.96	0.92	1.00
IV	406	453	1.12	1.02	1.22
Missing	138.41	221	1.59	1.39	1.82
Nutrition Status
Malnutrition	21.59	44	2.04	1.52	2.74
No Malnutrition	7400	9915	1.35	1.32	1.37
Missing	675.82	776	1.15	1.07	1.23
ART Regime
First line Regime	7900	10684	1.35	1.32	1.38
Second line Regime	137.52	51	0.37	0.28	0.49
HIV/TB Co-Infection
Yes	91.77	154	1.36	1.33	1.38
No	7400	10084	1.68	1.43	1.96
Missing	539.35	497	0.92	0.84	1.00
Prior Exposure on ART
Yes	3200	4021	1.27	1.23	1.31
No	4900	6714	1.37	1.34	1.41
Geographical Zones
Central Zone	701.47	975	1.39	1.31	1.48
Coastal Zone	1800	2342	1.29	1.25	1.35
Lake Zone	2200	3517	1.59	1.54	1.64
Northern Zone	719.79	819	1.13	1.06	1.21
Southern Highland	2200	2614	1.20	1.16	1.25
Western Zone	269.07	357	1.33	1.19	1.47
Missing	191.45	111	0.58	0.48	0.69

Fig 2

Failure probability among adolescents in HIV/TB Co-infection status and age group.

Predictors associated with LTFU among adolescents

In bivariate analysis predictors associated with increased risk of LTFU were; age 15–19 years (crude hazard ratio (cHR): 2.24; 95% CI, 2.13–2.35), attending clinics at dispensaries (cHR: 1.39; 95% CI, 1.33–1.46) and health centers (cHR: 1.22; 95% CI, 1.16–1.28) and attending clinics at public facilities (cHR: 2.24; 95% CI, 2.13–2.35). Also, residing in central (cHR: 1.22; 95% CI, 1.11–1.34), lake zone (cHR: 1.39; 95% CI, 1.29–1.51) and HIV/TB co-Infection (cHR: 1.29; 95% CI, 1.10–1.51).

In multivariate Cox proportion hazard analysis, predictors associated with increased risk of LTFU were age 15–19 years (aHR: 1.57; 95% CI, 1.47–1.69), HIV/TB co-infection (aHR: 1.58; 95% CI, 1.32–1.89), attending clinic at dispensary (aHR: 1.12; 95% CI, 1.07–1.18) and health centers (aHR: 1.10; 95% CI, 1.04–1.15). Likewise, attending clinics at public facilities (aHR: 1.08; 95% CI, 1.02–1.14), residing in Lake zone (aHR: 1.09; 95% CI, 1.02–1.18) and being malnourished (aHR: 2.27; 95% CI, 1.56–3.23) were associated with increased risk of LTFU. Also, increased risk of LTFU was observed among female adolescents (aHR: 1.16; 95% CI, (1.09–1.25), having WHO stage III (aHR: 1.22; 95% CI, 1.11–1.37) and stage IV (aHR: 1.22; 95% CI, 1.11–1.35). Lower risk of LTFU was noticed among patients taking second-line regimen with 40% reduced risk of LTFU compared to those on first-line regimen (aHR: 0.40; 95% CI, (0.30–0.53)) (Table 3).

Table 3

Bivariate and multivariate analysis of loss to follow up among adolescent on ART.

Variable	Bivariate analysis		Multivariate analysis
	Crude Hazard Ratio (95%CI)	p-value	Adjusted Hazard Ratio (95% CI)	p-value
Sex
Male	Reference		Reference
Female	1.89(1.79–1.96)	<0.001	1.16(1.09–1.25)	<0.001
Age Group (Years)
10–14	Reference		Reference
15–19	2.24(2.13–2.35)	<0.001	1.57(1.47–1.69)	<0.001
Types of Health Facility
Hospital	Reference		Reference
Dispensary	1.39(1.33–1.46)	<0.001	1.12(1.07–1.18)	<0.001
Health Center	1.22(1.16–1.28)	<0.001	1.10(1.04–1.15)	<0.001
Facility Ownership
Private	Reference		Reference
Public	1.18(1.13–1.25)	<0.001	1.08(1.02–1.14)	0.004
Marital Status
Never married/Single	Reference		Reference
Cohabiting/Married	0.57(0.55–0.60)	<0.001	0.89(0.69–1.14)	0.347
WHO Clinical Stage
I	Reference		Reference
II	0.67(0.61–0.74)	<0.001	1.03(0.93–1.14)	0.642
III	1.04(0.94–1.15)	0.327	1.22(1.11–1.37)	<0.001
IV	1.18(1.06–1.32)	0.013	1.22(1.11–1.35)	<0.001
Nutrition Status
No Malnutrition	Reference		Reference
Malnutrition	1.56(1.16–2.08)	<0.001	2.27(1.56–3.23)	<0.001
ART Regimen
Firstline Regimen	Reference		Reference
Secondline Regimen	0.28(0.22–0.37)	<0.001	0.40(0.30–0.53)	<0.001
Geographical Zone
Northern zone	Reference		Reference
Central Zone	1.22(1.11–1.34)	<0.001	1.07(0.97–1.18)	0.149
Coastal Zone	1.14(1.06–1.24)	0.001	1.04(0.96–1.13)	0.343
Lake Zone	1.39(1.29–1.51)	<0.001	1.09(1.02–1.18)	0.030
Southern Highland Zone	1.06(0.98–1.15)	0.134	0.90(0.83–1.0.97)	0.010
Western Zone	1.17(1.04–1.33)	0.012	0.89(0.79–1.02)	0.100
HIV/TB Co-Infection
No	Reference		Reference
Yes	1.29(1.10–1.51)	0.002	1.58(1.32–1.89)	<0.001
Prior Exposure on ART
No	Reference		Reference
Yes	0.92(0.88–0.96)	<0.001	0.90(0.87–0.94)	<0.001

Discussion

This study bid to determine the rate and predictors of LTFU among adolescents on ART in Tanzania. We identified a high rate of LTFU among ADLHIV on ART in our study. Moreover, predictors of LTFU on ART were; HIV/TB Co-Infection, attending clinics at dispensaries, health centers, being malnourished, WHO stage II, III, and IV, Second-line ART regimen and Lake Zone.

Our study observed a high rate of loss to follow up among adolescents compared to WHO 90-90-90 targets., Nevertheless, this rate of LTFU among adolescents was observed to be low compared to the study conducted in Myanmar (69%) [19]. However compared to our results, a low proportion of loss to follow up on ART was reported in a study conducted in Ethiopia (13.6%) [20]. Also, the rate of LTFU was high during the first months of ART initiation as compared to the study conducted in Kenya, Ethiopia, and Zimbabwe [20-22]. The disparity of median time in the previous studies might be due to different times of follow up, a study conducted in Ethiopia and Kenya had a follow-up time of eight and four years respectively [20, 21]. Viral load suppression, being a mobile group (shifting of schools), fear of discrimination, and fear of disclosure is anticipated to cause dropout in the first six months of ART care. Strengthening the follow-up system and establishment of youth HIV support groups are highly encouraged.

Increased risk of LTFU among adolescents with WHO stage III and IV observed in our findings was similar to the previous studies conducted in West Africa, Tanzania, Uganda, and Zambia [23, 24]. Also, studies conducted in Ethiopia and Uganda showed an increased risk of LTFU among adolescents with WHO stage III and IV [25, 26]. No association between LTFU and the WHO stage was observed in the study conducted in Kenya [21]. In our study, an increased risk of LTFU among adolescents who were in WHO stage III and IV might be due to the death of the disease unbeknownst to the health system. Patients with advanced WHO stage might have severe malnutrition or undiagnosed infections such as tuberculosis resulting in an increased mortality rate [27]. Nevertheless, patients with a worse prognosis at baseline are more likely to be loss to follow up [27]. These findings call for early identification of HIV-infected individuals and early initiation of ART.

Furthermore, a prominent risk of LTFU in ART care among adolescents with TB infection observed in our study corroborated with the study conducted in Ethiopia [25]. A study in sub-Saharan Africa revealed no association between LTFU and HIV/TB co-infection [28]. Increased LTFU among HIV/TB adolescents in our study might be due to medication-related issues such as adverse effects, pill burden, or complexity of drug regimen. To ensure close follow-up among adolescents in ART care it is essential to strengthen TB screening for early diagnosis and the use of home-based care workers.

Adolescents attending clinics at dispensaries and health centers were more likely to be LTFU. Similar findings were reported in studies conducted in Uganda, rural Zimbabwe, and South Africa [14, 22, 29]. Other preceding studies elaborated adolescents who attended clinics in hospitals were more likely to dropout [25, 30]. However, a study conducted in sub-Saharan countries explained no association between type of health facility and LTFU [28]. In our study, the high risk of LTFU among adolescents at primary health facilities might be attributed to inadequate human resources resulting in long waiting time. Also, non-adherence to ART care might be caused by poor quality of adolescents’ HIV services at primary health facilities. This calls for greater investment in healthcare workforce and establishment of integrated adolescents’ ART clinics at the primary health facilities.

We also identified high risk of LTFU among adolescents who attended clinics at public health facilities. These findings were similar to a study conducted in Myanmar which showed an increased risk of LTFU among patients treated in public sectors [31]. However, a study conducted in Uganda explained no significant association between facility ownership and LTFU [32]. The high risk of LTFU in our results might be due to patients’ workload and long waiting time at the public facilities compared to private facilities. There is a need for establishing youth-friendly ART clinics, scheduling clinic visits on school holidays, and facilitating transport to clinics. Also, follow-up through telephone calls and shortening waiting time might improve retention on ART.

Adolescents on second-line ART regimen had reduced risk of LTFU, our findings concur with a study conducted in Ethiopia [33]. However, our results contrast with the studies conducted in Myanmar [34] and Nigeria [35] which reported an increased risk of LTFU among patients on a second-line regimen. A study conducted in Nigeria stated an increased risk of LTFU might be caused by adverse effects obtained from second-line drugs [35]. The reduced risk among adolescents in the second-line found in our study could be due to the close follow-up given to this group after treatment failure.

We identified that adolescents aged 15–19 years had a higher risk of LTFU from ART care. The results corresponded to previous studies conducted in Ethiopia, South Africa, and sub-Saharan Africa [20, 30, 36]. Growing independent, fear of stigma, peer pressure, discrimination, and being a mobile group with the shifting of schools might have contributed to LTFU among older adolescents [8, 37]. Nevertheless, there is a need for conducting age-specific interventions to reduce LTFU among adolescents. The increased risk of LTFU among female adolescents observed in our study was a consistency to the previous findings in Uganda and a study conducted on MTCT-Plus programs in 9 different countries [38, 39]. However, studies in Tanzania, Ethiopia, and Malawi reported a high risk of LTFU on ART among male adolescents [40-42]. Nevertheless, a study conducted in 15 ART programs in Africa, Asia, and South America reported no association between gender and LTFU [43]. A study conducted in Kenya reported that the reasons for high LTFU among females included; family commitments, high transport costs, and work commitments [44]. In our study, the reason for gender gap is that more females accessed HIV testing services than males, especially during Antenatal Clinic (ANC) services. Therefore, the high LTFU among female adolescents in our study might be attributed to challenges of securing childcare to attend follow-up clinic visits or undocumented transfer to other HIV care clinics. This calls for strengthening the linkage to HIV care and counseling among female adolescents and sustained outreach after delivery.

It was observed that adolescents who resided in the Lake zone were more likely to be LTFU in care. The variation of LTFU risk was also reported in various studies across different geographical zones [11, 25, 45]. The differences in retention in care could be due to variations of social, cultural, and religious practices [25, 35]. We speculated socio-cultural and economic activities such as early marriages, mining, and fishing activities might have led to a high risk of LTFU among adolescents. A study conducted in Lake zone reported that local beliefs that HIV-like illnesses were attributed to witchcraft, thus majority of HIV patients preferred traditional healers [46]. These beliefs might have resulted in poor attendance of adolescents in ART clinics. There is a need for region-specific interventions that will prioritize areas with a high risk of LTFU and ensure close follow up of adolescents in ART.

This study had several limitations due to the use of routine data from the existing database. We did not ascertain treatment failure as an outcome of LTFU from ART care using CD4 counts and viral load counts. However, a strong link between LTFU and treatment failure published in various studies in Tanzania and elsewhere provide justification for addressing LTFU among adolescents [25, 47]. Also, we were not able to assess if adolescents presented for care at other clinics during the two-year follow-up time. However, the strength of this study was the inclusion of adolescents in HIV care from the whole country making our study results generalizable to all HIV-infected adolescents on ART care in Tanzania. Additionally, the longitudinal nature of this analysis provided an opportunity to assess rate and time of LTFU and associated predictors.

Conclusion

Due to high rate of LTFU obtained in our findings than in other studies, retention in care among adolescents in ART is still of major concern in Tanzania. There is a need for targeted interventions for adolescents; aged 15–19 years, HIV/TB co-infected, with WHO stage IV, and those residing in the Lake zone. Also, attention among adolescents attending clinics at public facilities, dispensaries, and health centers is highly needed. There is a need for designing integrated clinics for adolescents, especially in primary health facilities to increase retention in care. Generally, interventions that will develop strategies for reducing LTFU among adolescents especially in the first six months in ART are highly warranted to be able to achieve the 2030 goal of ending the HIV epidemic as a public health threat.

23 Dec 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Brian C. Zanoni, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. You indicated that you had ethical approval for your study. In your Methods section, please ensure you have also stated whether you obtained consent from parents or guardians of the minors included in the study or whether the research ethics committee or IRB specifically waived the need for their consent.

3. Thank you for stating the following in the Acknowledgments Section of your manuscript:

Our special appreciation goes to Tanzania field epidemiology and Laboratory training program for its funding support.

Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

We acknowledge the financial support from the Tanzania Field Epidemiology and Laboratory Training Program(TFELTP) during development and preparation of this study. TFELTP had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

4. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

Additional Editor Comments:

This is an interesting manuscript addressing an important topic of LTFU among adolescents living with HIV. However, there are several critical weaknesses identified by the reviewers that would need to be addressed before this manuscript could be considered for publication in PLoS One. The methodology needs clarification as indicated by reviewer #3. In addition, it is unclear how deaths were separated from LTFU - especially given some of the finding. Can deaths be miss-classified as LTFU? Also the definitions of variables and outcomes needs to be further clarified. If resubmitting please address these and the issues by the reviewers in a detailed response letter with the revised manuscript.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Will need some grammatical edits

Strengths: large data set; through and sound analysis,

Areas for improvement:

Major:

Explain in methods where the various settings adolescent can access care (dispensary, health center, hospital) and perhaps a line about what that means to someone not familiar with infrastructure of care delivery in Tanzania)

Discussion includes reviews of other references which support or contradict the findings. But what I want to see is thoughtful ideas or references about why each of the factors such as living in the lake zone or getting care at a dispensary vs a hospital is causing diverse outcomes with loss to follow up. In short I want more in the discussion.

Reviewer #2: Thank you for the opportunity to review this paper on loss to follow up in adolescents living with HIV.

I think it is an important issue that the authors address but I think the paper would benefit from more information being included especially in the methods section.

Methods

1. What is the HIV prevalence in the adolescent age group in Tanzania?

2. How was malnutrition defined in this population?

3. What are first and second line regimens in Tanzania?

4. Please describe in more detail the different levels of care "dispensaries" vs "clinics"

5. Please also describe how many times adolescents are seen when starting treatment.

If follow up was over two years - surely adolescents were not TB/HIV connected for two years? Was only the baseline variable taken into account or were the variables assessed at every visit? (for malnutrition too?)

6. Was viral load available.. if yes why is it not presented?

7. Were the authors able to assess if adolescents presented for care at other clinics during the two year time period?

Results

1. How many died?

2. How many transferred out?

Reviewer #3: This is an interesting article covering an important topic. Given the high rate of LTFU seen among adolescents with HIV in this study, this is clearly an important focus to help understand associations with LTFU. A major strength is the use if a national dataset from a country with a large number of adolescents living with HIV. I thought more could have been done to discuss the findings in depth and to consider directions for addressing this challenge.

I have a few comments to strengthen the manuscript further:

1. In the abstract, it would be helpful to specify and delineate bivariate and multivariate associations with LTFU. Consider removing the description of less significant associations (lines 31-32), and removing some redundant descriptions (lines 36-38). Instead, could add more to the implications of the findings in the abstract. For example, that LTFU was high, and that in multivariate analysis, LTFU was most associated with XXXX factors. These findings point to: needed resources? Funding? Interventions? Integration of care? Adolescent-friendly services? See comments below on strengthening the discussion. These can inform how the abstract could be revised to emphasize key conclusions or interpretations. It is stated that ‘novel’ interventions are needed, but perhaps strategies might include those that are not necessarily so novel. See thoughts below on the discussion. Findings may prompt consideration of strategies including: greater investments in healthcare workforce, provision of quality adolescent-friendly services, and targeting support for those with most advanced illness or TB/HIV.

2. In the introduction, the list of predictors is confusing because some include e.g. ‘sex’ or ‘age’ and ‘WHO Stage,’ where it’s not actually clear what associations have been seen previously (e.g. older adolescence? Female sex? Advanced WHO stage?). Recommend revising to be more specific about which associations have been noted previously.

3. My most important point to revise for clarity relates to the inclusion criteria. It is stated that the study included adolescents “10-19 years initiated and enrolled on ART from 2014 to 2016”. Would add clarity to understand: does this mean this cohort only included those initiating ART at age 10 onwards? Or would it include adolescents who initiated ART before age 10? It’s an important distinction because if the study focuses on those who start ART age 10 or later, the sample may be primarily made up of adolescents with horizontal infection, or those with advanced perinatal infection, and this would exclude likely most adolescents with perinatal HIV who would start ART before age 10. This would also have implications for interpreting findings and contrasting with other studies in the discussion section. I may be mistaken in how I am reading this, but having more clarity here would avoid others potentially misinterpreting the criteria as well.

4. Would add specifics regarding “TB history” and “TB/HIV co-infection” definitions. Is this referring to having a diagnosis of TB disease? Does this include others requiring TB preventive therapy? Would add the details of how this was defined, and how it would have been documented in the routine data. Further, would similarly provide a definition for how nutritional status was determined.

5. For the variable “prior exposure on ART”, is this at the time of enrollment in care? This comes back to my earlier question, if these are patients who initiated care in 2014-2016 at age 10-19.

6. For the section on bivariate analyses (starting line 136) would give the crude hazard risks, not just the p-values.

7. It’s unclear to me why WHO Stage IV was used as a reference, rather than WHO Stage I, or similarly why malnutrition was used as a reference, rather than no malnutrition. It would be preferable to use WHO Stage I as the reference, and no malnutrition as the reference. This makes interpreting the findings clearer, particularly comparing WHO stages to Stage I as baseline, rather than comparing stage II or III to a reference of Stage IV. It is also statistically preferable to use the more common state as the reference.

8. In the discussion, would comment on how to interpret the increased LTFU among WHO Stage IV. Might there be unascertained mortality in this group? Overall, I think there needs to be more discussion about LTFU among those with advanced disease and/or TB/HIV. Also, are there data that may clarify if many of those LTFU had actually passed on? And what are the implications for care programs if severe illness is a driver of LTFU (possibly via unascertained mortality)? How could this group be better addressed?

9. Recommend commenting on the increased LTFU among those with TB/HIV. Does this reflect unascertained mortality? Burdens of combined TB/HIV clinic visits and medications? A need to integrate TB/HIV services? Also, I don’t understand the sentence which states, ‘the development of other opportunistic infection among HIV/TB patients might have accelerated LTFU due to death.’ Does this refer to deaths from a different cause and not TB? Would remove this or clarify if something is being misstated here.

10. Can the authors comment on the higher LTFU among female adolescents? What factors might make them more vulnerable? Addressing the earlier point about whether this is predominantly a cohort with horizontal HIV infection would particularly point to the acute vulnerabilities that adolescent females with new HIV diagnoses may be experiencing that present barriers to care engagement.

11. Given the higher LTFU at public facilities, and at dispensaries/health centers, would elaborate more on implications for: is there a need for expanded provision of adolescent-friendly services? Is the public sector under-resourced to provide quality adolescent-friendly services? If there are high workloads on providers in the public sector (recommend revising the phrase ‘patients’ workload’ line 195), might this point toward a need for greater investments in the health sector and in the health workforce?

12. As a minor point, there are multiple places where LTFU is misspelled as LFTU. There are other minor spelling or grammatical errors that would recommend revising. For example, recommend removing the words “contrariwise” (line 163) and “unlikely” (line 188), and revising for clarity.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Neerav Desai

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Review Pone.docx

Click here for additional data file.

11 Mar 2022

Response to editor’s and reviewers' comments

We are grateful for the reviews provided by the editors and external reviewers of this manuscript. The comments and suggestions are thoughtful and have helped much to improve our manuscript. We have taken them into account in the revision. Please see below in blue our detailed response to comments.

Responses to editor’s comments

Editor’s comment No 1: - Explain in methods where the various settings adolescent can access care (dispensary, health center, hospital) and perhaps a line about what that means to someone not familiar with infrastructure of care delivery in Tanzania)

Authors' response: We are thankful for the comment. We have provided information on the various settings that can access care and information on the infrastructure of care delivery in Tanzania and revised version of the Methods section reads;

Methods, Line 74-77; "In Tanzania, adolescents can access ART care in health facilities which are categorized into three levels. These levels include; primary health facilities also known as dispensaries, secondary health facilities (health centers) and tertiary health facilities (hospitals).”

Editor’s comment No 2: - Discussion includes reviews of other references which support or contradict the findings. But what I want to see is thoughtful ideas or references about why each of the factors such as living in the lake zone or getting care at a dispensary vs a hospital is causing diverse outcomes with loss to follow up. In short I want more in the discussion.

Authors' response:

Authors' response: We appreciate the comment. The authors have provided more information on the factors: living in the lake zone and getting care at a dispensary versus hospitals and the revised version reads:

Discussion, Line 288-290; “A study conducted in Lake zone reported that local beliefs that HIV-like illnesses were attributed to witchcraft, thus the majority of HIV patients preferred traditional healers[47]. These beliefs might have resulted in poor attendance of adolescents in ART clinics”.

Discussion, Line 240-245; “In our study, the high risk of LTFU among adolescents at primary health facilities might be attributed to inadequate human resources resulting in long waiting times. Also, non-adherence to ART care might be caused by the poor quality of adolescents’ HIV services at primary health facilities. This calls for greater investment in the healthcare workforce and the establishment of integrated adolescents’ ART clinics at primary health facilities.”

Editor’s comment No 3: - The methodology needs clarification as indicated by reviewer #3. In addition, it is unclear how deaths were separated from LTFU - especially given some of the findings. Can deaths be miss-classified as LTFU?-

Authors' response: Thank you for the comment. We have provided information on how deaths were separated from LTFU and newly added information reads:

Methods, Line 67-68; “Adolescents who were transferred out and died were censored at the time of their last visit.”

Editor’s comment No 4: Also please clearly define all outcomes and variables.

Authors' response: We are grateful for the comment. We have defined all outcomes and variables and the updated paragraphs reads:

Methods, Line 95-120; " The loss to follow up was defined as any of the adolescents who failed to attend at least one clinic visit within 90 days of their scheduled appointment. Retention to care was defined as the state of adolescents being alive, actively attending scheduled appointments at the clinics and receiving ART care at the end of follow up period. Independent variables of the study included; sex which was classified as male and female, age was grouped as 10-14 years and 15-19 years. Marital status was grouped as never married/single and cohabiting/married, types of health facilities were categorized as dispensaries, health centers, hospitals. Also, facility ownership was classified as public and private, WHO clinical stages were grouped as stage I, II, III, and IV. Geographical zones were grouped as coastal, central, lake, northern, southern highland and western zones.

Likewise, nutrition status was categorized into two groups according to BMI scales. BMI was obtained from adolescents’ weight in kilograms divided by the square of height in meters which were assessed at every clinic visit. The BMI scale of <18.5kg/m2 (underweight) and 25.0-29.9kg/m2 (obesity/overweight) were classified as malnutrition, while BMI scale of 18.5-24.9kg/m2 (normal weight) was classified as no malnutrition. ART regimens were categorized as first-line and second-line based on the Tanzania HIV guideline[19]. First-line regimens included; Tenofovir (TDF) 300 mg / Lamivudine (3TC) 300 mg / Efavirenz (EFV) 600mg. Alternative first line regimen used were; Tenofovir (TDF) + Emtricitabine (FTC) + Dolutegravir (DTG), Abacavir (ABC) + Lamivudine (3TC) + Efavirenz (EFV) and Zidovudine (AZT)+Lamivudine(3TC) +Nevirapine (NVP). Also the second-line regimens included; Zidovudine (AZT), Tenofovir (TDF), Abacavir (ABC), Lamivudine (3TC), Emtricitabine (FTC), Atazanavir boosted by Ritonavir (ATV/r), Lopinavir boosted by Ritonavir (LPV/r) and Dolutegravir (DTG). Other independent variables included; prior exposure to ART during enrolment (yes or no) and HIV/TB Co-infection (yes or no).”

Editor’s comment No 5: You indicated that you had ethical approval for your study. In your methods section, please ensure you have also stated whether you obtained consent from parents or guardians of the minors included in the study or whether the research ethics committee or IRB specifically waived the need for their consent.-

Authors' response: Thank you for the comment. We have revised the ethical section under methods and updated the paragraph reads:

Methods, Line 136-138;”This retrospective record analysis utilized anonymized data hence the institutional review board (IRB) waived the need for consent from parents and guardians of the adolescents.”

Editor’s comment No 6: Thank you for stating the following in the Acknowledgments Section of your manuscript: Our special appreciation goes to Tanzania field epidemiology and Laboratory training program for its funding support. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows: We acknowledge the financial support from the Tanzania Field Epidemiology and Laboratory Training Program (TFELTP) during development and preparation’ of this study. TFELTP had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.-

Authors' response: We appreciate the comment. We have revised acknowledgment section and the newly updated paragraph reads:

Acknowledgment, Line 317-319; “We thank the Tanzania National AIDS Control Program (NACP) for letting us access patients’ data for this study as per the Program data access regulations. We acknowledge the valuable support received from Dr Ahmed Abade and Dr Werner in supporting the design of this study.”

Additional Editor’s comment No 7: This is an interesting manuscript addressing an important topic of LTFU among adolescents living with HIV. However, there are several critical weaknesses identified by the reviewers that would need to be addressed before this manuscript could be considered for publication in PLoS One. The methodology needs clarification as indicated by reviewer #3. In addition, it is unclear how deaths were separated from LTFU - especially given some of the findings. Can deaths be miss-classified as LTFU? Also, the definitions of variables and outcomes need to be further clarified.

Authors' response: Thank you for the comment. We have provided information on how deaths were separated from LTFU and define all variables and outcomes this has been answered in comment 3 and 4 above.

Responses to reviewer number 1

Reviewer comment: Explain in methods where the various settings adolescent can access care (dispensary, health center, hospital) and perhaps a line about what that means to someone not familiar with infrastructure of care delivery in Tanzania)

Authors' response: We appreciate the comment. We have provided information on the various settings that can access care and information on the infrastructure of care delivery in Tanzania and revised version reads;

Methods, Line 74-77; "In Tanzania, adolescents can access ART care in health facilities which are categorized into three levels. These levels include; primary health facilities also known as dispensaries, secondary health facilities (health centers) and tertiary health facilities (hospitals).”

Reviewer comment: Discussion includes reviews of other references which support or contradict the findings. But what I want to see is thoughtful ideas or references about why each of the factors such as living in the lake zone or getting care at a dispensary vs a hospital is causing diverse outcomes with loss to follow up. In short, I want more in the discussion.

Authors' response: We appreciate the comment. The authors have provided more information on the factors: living in the lake zone and getting care at a dispensary versus hospitals and the revised version reads:

Discussion, Line 288-290; “A study conducted in Lake zone reported that local beliefs that HIV-like illnesses were attributed to witchcraft, thus the majority of HIV patients preferred traditional healers[47]. These beliefs might have resulted in poor attendance of adolescents in ART clinics”.

Discussion, Line 240-245; “In our study, the high risk of LTFU among adolescents at primary health facilities might be attributed to inadequate human resources resulting in long waiting times. Also, non-adherence to ART care might be caused by the poor quality of adolescents’ HIV services at primary health facilities. This calls for greater investment in the healthcare workforce and the establishment of integrated adolescents’ ART clinics at primary health facilities.”

Responses to reviewer number 2

Reviewer comment: - What is the HIV prevalence in the adolescent age group in Tanzania?

Authors' response: We are thankful for the comment. We have added the information on the HIV prevalence among adolescents in Tanzania in the Methods section and the added sentences reads:

Methods, Line 72-74; " In 2020, population of adolescents aged 10-19 years in Tanzania mainland was projected to be 13,206,921 based on the 2012 national census[17] with an HIV prevalence of 5.8%[18]. "

Reviewer comment: -How was malnutrition defined in this population?

Authors' response: We appreciate the comment. The authors have defined malnutrition and the new version reads;

Methods, Line 106-110; “Likewise, nutrition status was categorized into two groups according to BMI scales. BMI was obtained from adolescents’ weight in kilograms divided by the square of height in meters which were assessed at every clinic visit. Participants with BMI of <18.5kg/m2 (underweight) or 25.0-29.9kg/m2 (obesity/overweight) were classified as “malnutrition”, while those with BMI between 18.5-24.9kg/m2 (normal weight) were classified as “no malnutrition”

Reviewer comment: - What are first and second-line regimens in Tanzania?

Authors' response: Thank you for the comment. We have revised the text and explained first and second-line regimens in Tanzania and updated paragraph reads:

Methods, Line 112-120; “ART regimens were categorized as first-line and second-line based on the Tanzania HIV guideline[19]. First-line regimens included; Tenofovir (TDF) 300 mg / Lamivudine (3TC) 300 mg / Efavirenz (EFV) 600mg. Alternative first line regimen used were; Tenofovir (TDF) + Emtricitabine (FTC) + Dolutegravir (DTG), Abacavir (ABC) + Lamivudine (3TC) + Efavirenz (EFV) and Zidovudine (AZT)+Lamivudine(3TC) +Nevirapine (NVP). Also the second-line regimens included; Zidovudine (AZT), Tenofovir (TDF), Abacavir (ABC), Lamivudine (3TC), Emtricitabine (FTC), Atazanavir boosted by Ritonavir (ATV/r), Lopinavir boosted by Ritonavir (LPV/r) and Dolutegravir (DTG). Other independent variables included; prior exposure to ART during enrolment (yes or no) and HIV/TB Co-infection (yes or no).”

Reviewer comment: Please describe in more detail the different levels of care "dispensaries" vs "clinics

Authors' response: We are grateful for your valuable comments. We have described in detailed the different levels of care “dispensaries vs clinics”, the revised version now reads;

Methods, Line 74-77; "In Tanzania, adolescents can access ART care in health facilities which are categorized into three levels. These levels include; primary health facilities also known as dispensaries, secondary health facilities (health centers) and tertiary health facilities (hospitals).”

Reviewer comment: Please also describe how many times adolescents are seen when starting treatment.

Authors' response: We are grateful for the valuable comments. The authors have described on how many times adolescents are seen when starting treatment. The new sentences now read:

Methods, Line 77-79; “The Tanzania HIV guideline requires adolescents to attend HIV clinics at least once a month for the first three months of the ART treatment and thereafter once every three months depending on their adherence status[19].

Reviewer comment: If follow-up was over two years - surely adolescents were not TB/HIV connected for two years? Was only the baseline variable taken into account or were the variables assessed at every visit? (for malnutrition too?)

Authors' response: We appreciate the comment. The authors have provided information on how TB/HIV co-infection variable and malnutrition variables were assessed, the revised version reads;

Methods, Line 82-84; “. In Tanzania, all people living with HIV (PLHIV) are screened for TB on every clinic visit to prevent them from developing active TB by providing Isonized Preventive Therapy (IPT).”

As already described in the Methods, 106-110; nutrition status is assessed at every clinic visit.

Reviewer comment: Was viral load available if yes why is it not presented?

Authors' response: Thank you very much for your comment. We acknowledge that we did not have comprehensive data on viral load which would have been useful as we are assessing the ultimate outcome of LTFU which is treatment failure. We have described this as a limitation in the following text added into the revised version

Discussion, Line 294-298; “We did not ascertain treatment failure as an ultimate outcome of LTFU from ART care using CD4 counts and viral load counts. However, a strong link between LTFU and treatment failure published in various studies in Tanzania and elsewhere provide justification for addressing LTFU among adolescents[26,48].”

Reviewer comment: Were the authors able to assess if adolescents presented for care at other clinics during the two year time period?-

Authors' response: We are grateful for the comment. The authors were not able to assess if adolescents presented for care at other clinics during the follow up time of two years. Therefore, we have revised the discussion part and explain it as part of limitation of the study.

Discussion, Line 298-299; "Also, we were not able to assess if adolescents presented for care at other clinics during the two years follow-up time.”

Reviewer comment: How many died?

Authors' response: We appreciate the comment. The authors have provided data on how many adolescents died in the results section.

Results, Line 143-145; “About 177(0.7%) records were deceased and 74(1.7%) records were transferred out on their last appointment date.”.”

Reviewer comment: How many transferred out?

Authors' response: We appreciate the comment. The authors have provided data on how many adolescents died in the results section as described in the previous comment.

Responses to reviewer number 3

Reviewer comment: - In the abstract, it would be helpful to specify and delineate bivariate and multivariate associations with LTFU. Consider removing the description of less significant associations (lines 31-32), and removing some redundant descriptions (lines 36-38). Instead, could add more to the implications of the findings in the abstract. For example, that LTFU was high, and that in multivariate analysis, LTFU was most associated with XXXX factors. These findings point to: needed resources? Funding? Interventions? Integration of care? Adolescent-friendly services? See comments below on strengthening the discussion. These can inform how the abstract could be revised to emphasize key conclusions or interpretations. It is stated that ‘novel’ interventions are needed, but perhaps strategies might include those that are not necessarily so novel. See thoughts below on the discussion. Findings may prompt consideration of strategies including: greater investments in healthcare workforce, provision of quality adolescent-friendly services, and targeting support for those with most advanced illness or TB/HIV.

Authors' response: We are thankful for the comment. We have added information on the implications of the findings, the revised now reads:

Abstract, line 28-35 “Predictors associated with LTFU included; adolescents aged 15-19 years (adjusted hazard ratio (aHR): 1.57; 95% Confidence Interval (CI); 1.47-1.69), having HIV/TB co-infection (aHR: 1.58; 95% CI, 1.32-1.89), attending care at a dispensary (aHR: 1.12; 95% CI, 1.07-1.18) or health center (aHR: 1.10; 95% CI, 1.04-1.15), and being malnourished (aHR: 2.27; 95% CI,1.56-3.23). Moreover, residing in lake zone and having advanced HIV disease were associated with LTFU. These findings highlight the high rate of LTFU and the need for intervention targeting older adolescents with advanced disease and strengthening primary public facilities in order to achieve the 2030 goal of ending HIV as a public health threat.”

Reviewer comment: - In the introduction, the list of predictors is confusing because some include e.g. ‘sex’ or ‘age’ and ‘WHO Stage,’ where it’s not actually clear what associations have been seen previously (e.g. older adolescence? Female sex? Advanced WHO stage?). Recommend revising to be more specific about which associations have been noted previously.-

Authors' response: Thank you for your comment. We have revised the list of predictors and shows which associations were noted previously and the newly paragraph reads:

Introduction, Line 46-51; “Predictors associated with increased risk of LTFU among ADLHIV reported in various studies included; ADLHIV aged 15-19 years, female adolescents, those diagnosed with HIV/TB co-infection, those with malnutrition, adolescents who attended clinics at primary facilities and having advanced WHO clinical stage [7-13]. Also increased risk of LTFU was observed among adolescents who had prior exposure to ART[14] and those who attended clinics at public health facilities[15].”

Reviewer comment: - My most important point to revise for clarity relates to the inclusion criteria. It is stated that the study included adolescents “10-19 years initiated and enrolled on ART from 2014 to 2016”. Would add clarity to understand: does this mean this cohort only included those initiating ART at age 10 onwards? Or would it include adolescents who initiated ART before age 10? It’s an important distinction because if the study focuses on those who start ART age 10 or later, the sample may be primarily made up of adolescents with horizontal infection or those with advanced perinatal infection, and this would exclude likely most adolescents with perinatal HIV who would start ART before age 10. This would also have implications for interpreting findings and contrasting with other studies in the discussion section. I may be mistaken in how I am reading this, but having more clarity here would avoid others potentially misinterpreting the criteria as well.-

Authors' response: Thank you for the comment. We have revised the inclusion criteria and exclusion criteria and the new paragraph reads:

Methods, Line 63-65; “This study included adolescents who were enrolled and initiated ART at age 10-19 with either vertical or horizontal HIV/AIDS infection from January 2014 to December 2016.”

Reviewer comment: - Would add specifics regarding “TB history” and “TB/HIV co-infection” definitions. Is this referring to having a diagnosis of TB disease? Does this include others requiring TB preventive therapy? Would add the details of how this was defined, and how it would have been documented in the routine data. Further, would similarly provide a definition for how nutritional status was determined.-

Authors' response: We are grateful for the comments. We have revised the document and added the definition of TB/HIV co-infection and that of nutritional status.

Methods, Line 81-84; “TB/HIV Co-infection refers to HIV patients who were also diagnosed with TB infection during routine TB screening. In Tanzania, all people living with HIV (PLHIV) are screened for TB on every clinic visit and those not infected are provided with Isonized Preventive Therapy (IPT) to prevent them from developing active TB”.

Methods, Line 108-110; “Participants with BMI of <18.5kg/m2 (underweight) or 25.0-29.9kg/m2 (obesity/overweight) were classified as “malnutrition”, while those with BMI between 18.5-24.9kg/m2 (normal weight) were classified as “no malnutrition”.

Reviewer comment: For the section on bivariate analyses (starting line 136) would give the crude hazard risks, not just the p-values.-

Authors' response: We appreciate the comment. We have revised the bivariate analyses and have added the crude hazard risks and newly updated paragraphs reads.

Results, Line 179-184; “In bivariate analysis predictors associated with increased risk of LTFU were; age 15-19 years (crude hazard ratio (cHR): 2.24; 95% CI, 2.13-2.35),, attending clinics at dispensaries (cHR: 1.39; 95% CI, 1.33-1.46)and health centers (cHR: 1.22; 95% CI, 1.16-1.28)and attending clinics at public facilities (cHR: 2.24; 95% CI, 2.13-2.35). Also, residing in central (cHR: 1.22; 95% CI, 1.11-1.34), lake zone (cHR: 1.39; 95% CI, 1.29-1.51) and HIV/TB co-Infection (cHR: 1.29; 95% CI, 1.10-1.51)”.

Reviewer comment: It’s unclear to me why WHO Stage IV was used as a reference, rather than WHO Stage I, or similarly why malnutrition was used as a reference, rather than no malnutrition. It would be preferable to use WHO Stage I as the reference, and no malnutrition as the reference. This makes interpreting the findings clearer, particularly comparing WHO stages to Stage I as baseline, rather than comparing stage II or III to a reference of Stage IV. It is also statistically preferable to use the more common state as the reference.

Authors' response: We are grateful for the valuable comments. We have revised the analysis and kept WHO stage I as the reference, and no malnutrition as the reference. This has been accommodated in table 3“Bivariate and Multivariate analysis of loss to follow up among Adolescents on ART”

Reviewer comment: In the discussion, would comment on how to interpret the increased LTFU among WHO Stage IV. Might there be unascertained mortality in this group? Overall, I think there needs to be more discussion about LTFU among those with advanced disease and/or TB/HIV. Also, are there data that may clarify if many of those LTFU had actually passed on? And what are the implications for care programs if severe illness is a driver of LTFU (possibly via unascertained mortality)? How could this group be better addressed?

Authors' response: We are thankful for the comment. We have revised the discussion about the increased risk of LTFU among WHO stage IV, and the newly paragraph reads;

Discussion, Line 220-227; “No association between LTFU and the WHO stage was observed in the study conducted in Kenya[22]. In our study, an increased risk of LTFU among adolescents who were in WHO stage III and IV might have been caused by unbeknownst death to the health system. Patients with advanced WHO stage might have severe malnutrition or undiagnosed infections such as tuberculosis resulting in an increased mortality rate[28]. Nevertheless, patients with a worse prognosis at baseline are more likely to be loss to follow up[28]. These findings calls for early identification of HIV-infected individuals and early initiation of ART.”

Reviewer comment: Recommend commenting on the increased LTFU among those with TB/HIV. Does this reflect unascertained mortality? Burdens of combined TB/HIV clinic visits and medications? A need to integrate TB/HIV services? Also, I don’t understand the sentence which states, ‘the development of other opportunistic infection among HIV/TB patients might have accelerated LTFU due to death.’ Does this refer to deaths from a different cause and not TB? Would remove this or clarify if something is being misstated here.

Authors' response: We are thankful for the comment. We have revised the discussion about the increased risk of LTFU among those with TB/HIV, and newly added information reads;

Discussion, Line 230-234; “Increased LTFU among HIV/TB adolescents in our study might be due to medication-related issues such as adverse effects, pill burden or complexity of drug regimen. To ensure close follow up among adolescents in ART care it is essential to strengthen TB screening for early diagnosis and the use of home-based care workers.”

Reviewer comment: Can the authors comment on the higher LTFU among female adolescents? What factors might make them more vulnerable? Addressing the earlier point about whether this is predominantly a cohort with horizontal HIV infection would particularly point to the acute vulnerabilities that adolescent females with new HIV diagnoses may be experiencing that present barriers to care engagement.

Authors' response: We are appreciate for the comment. We have revised the discussion and we have commented on the higher LTFU among female adolescents and the new paragraph reads;

Discussion, Line 269-281; “The increased risk of LTFU among female adolescents observed in our study was consistency with the previous findings in Uganda and a study conducted on MTCT-Plus programs in 9 different countries[39,40]. However studies in Tanzania, Ethiopia and Malawi reported a high risk of LTFU on ART among male adolescents[41-43]. Nevertheless, a study conducted in 15 ART programs in Africa, Asia, and South America reported no association between gender and LTFU [44]. A study conducted in Kenya reported that the reasons for high LTFU among females included family commitments, high transport costs and work commitments[45].In our study, the reason for the gender gap is that more females accessed HIV testing services than males, especially during Antenatal Clinic (ANC) services. Therefore, the high LTFU among female adolescents in our study might be attributed to challenges of securing childcare to attend follow-up clinic visits or undocumented transfer to other HIV care clinics. This calls for strengthening the linkage to HIV care and counselling among female adolescents and sustained outreach after delivery.”

Reviewer comment: As a minor point, there are multiple places where LTFU is misspelled as LFTU. There are other minor spelling or grammatical errors that would recommend revising. For example, recommend removing the words “contrariwise” (line 163) and “unlikely” (line 188), and revising for clarity.

Authors' response: We are appreciate for the comment. We have revised the whole manuscript on the grammatical errors.

Submitted filename: Response to reviewers..docx

Click here for additional data file.

10 May 2022

Predictors of loss to follow up from antiretroviral therapy among adolescents with HIV/AIDS in Tanzania

PONE-D-21-31368R1

Dear Dr. Tesha,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Brian C. Zanoni, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The authors have appropriately responded to the comments of the editor and of the reviewers. The manuscript is acceptable for publication. However, I suggest the authors carefully review the manuscript and correct several language, grammar and punctuation errors throughout the manuscript.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: They have addressed the issues I was concerned with. I noticed several grammatical errors in the added sections. Please fix several grammatical mistakes.

Reviewer #2: Thank you for addressing the comments and for clarifications and revising the manuscript

I am satisfied with the responses.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Neerav Desai MD

Reviewer #2: No

29 Jun 2022

PONE-D-21-31368R1

Predictors of loss to follow up from antiretroviral therapy among adolescents with HIV/AIDS in Tanzania

Dear Dr. Tesha:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Brian C. Zanoni

Academic Editor

PLOS ONE