Magnitude and associated factors of anti-retroviral therapy adherence among children attending HIV care and treatment clinics in Dar es Salaam, Tanzania.

INTRODUCTION: The HIV pandemic continues to contribute significantly towards childhood mortality and morbidity. The up-scaling of the Anti-retroviral therapy (ART) access has seen more children surviving and sanctions great effort be made on ensuring adherence. Adherence is a dynamic process that changes over time and is determined by variable factors. This necessitates the urgency to conduct studies to determine the potential factors affecting adherence in our setting and therefore achieve the 90-90-90 goal of sustainable viral suppression.
OBJECTIVES: To assess the magnitude and associated factors of ART adherence among children (1-14 years) attending HIV care and treatment clinics during the months of July to November 2018 in Dar es Salaam.
METHODS: A cross-sectional clinic-based study, conducted in three selected HIV care and treatment clinics in urban Dar es Salaam; Muhimbili National Hospital (MNH), Temeke Regional Referral Hospital (TRRH), Infectious Disease Centre- DarDar Paediatric Program (IDC-DPP) HIV clinics during the months of July to November 2018. HIV-infected children aged 1-14 years who had been on treatment for at least six months were consecutively enrolled until the sample size was achieved. A structured questionnaire was used for data collection. Four-day self-report, one-month self-recall report and missed clinic appointments were used to assess adherence. Frequencies and percentages were used to describe categorical data. The odds ratio was used to analyse the possible factors affecting ART adherence Logistic regression models were used to determine the factors associated with ART adherence. Analysis was conducted using SPSS version 20.0 and p-value <0.05 were considered statistically significant.
RESULTS: 333 participants were recruited. The overall good adherence (≥95%) was approximated to be 60% (CI-54.3-65.1) when subjected to all three measures. On multivariable logistic regression, factors associated with higher odds of poor adherence were found to be caregivers aged 17-25 years [AOR = 3.5, 95%CI-(1.5-8.4)], children having an inter-current illness [AOR = 10.8, 95%CI-(2.3-50.4)], disbelief in ART effectiveness [AOR = 5.495; 95%CI-(1.669-18.182)] and advanced clinical stage [AOR = 1.972; 95% CI-(1.119-3.484)]. The major reasons reported by caregivers for missing medications included forgetfulness (41%), high pill burden (21%), busy schedule (11%) and long waiting hours at the clinic (9%). CONCLUSION AND RECOMMENDATIONS: In the urban setting of Dar es Salaam, ART adherence among children was found to be relatively low when combined adherence measures were used. Factors associated with poor ART adherence found were younger aged caregivers, and child intercurrent illness, while factors conferring good adherence were belief in ART effectiveness and lower HIV clinical stage. More attention and support should be given to younger aged caregivers, children with concomitant illness and advanced HIV clinical stages. Educating caregivers on ART effectiveness may also aid in improving adherence.

Introduction

HIV (human immunodeficiency virus) is a virus that attacks the immune cells in the human body thus making them more vulnerable to infections and the development of a life-threatening chronic condition called AIDS (acquired immunodeficiency syndrome). To control HIV viral replication and improve the function of the immune system, patients must use ART (antiretroviral therapy). Drug adherence is a key part of ART treatment, it refers to the whole process from choosing, starting, and managing to maintaining a given therapeutic medication regimen to control viral replication and improve the function of the immune system.

The HIV epidemic continues to have devastating consequences globally. According to UNAIDS, in 2018, 37.9 million people were living with HIV/AIDS worldwide, of which 1.7 million (5% of the population) were children aged less than 15 years [1]. Sub-Saharan Africa bears the largest burden of the HIV epidemic, accounting for 1.09 million children living with HIV/AIDS and more than 60% of new infections occur in women, infants or young children [2].

With the recent scale-up of ART access in sub-Saharan countries, the maximum focus needs to be laid on maintaining optimal adherence as it has been proven that, for the greatest success in ART effectiveness [3], adherence should be at 95% or greater [4-6]. The need for near-perfect adherence to a lifelong therapy from an early age has been identified as a major challenge in the administration of ART to HIV-infected children [4, 7, 8]. Poor adherence leads to low therapeutic drug levels, resulting in incomplete suppression of HIV replication.

It leads to resistance to the drugs and moreover cross-resistance within the same class [5, 6], thus compromising future treatment options and increasing mortality. Adherence reports among children taking ARTs in sub-Saharan Africa varies greatly 29–98% [9], with some countries reporting higher adherence over the other [10-13]. This can be attributed to the diversity of methods being used to assess adherence where self-reported measures are the commonest employed amongst low resource settings and thus have a tendency to overestimate adherence compared to the objective methods.

It has been shown in these settings, that a more accurate assessment of ART adherence can be obtained when two or more methods are employed to assess adherence [14]. Studies on adherence amongst children in Tanzanian context are greatly lacking, a study done in 2012 revealed good adherence to ART regimen was only reported in 24.6% of children and adolescents when three assessment measures were employed.

Adherence has been seen to be complex amongst children as compared to adults with different pressing issues affecting it, as they have to rely on the caregivers and their wellbeing to obtain drugs and face multiple challenges in social, health system and drug-related aspects. Poor socio-economic circumstances including age of caregivers, poverty, single marital status and stigma were seen to be associated with poor adherence among countries like Cape Town, Uganda, Ethiopia, Nigeria and Northern Tanzania [10, 12, 15–17]. Various studies have related poor adherence to caregiver factors, as seen with children who are under the care of non-parental guardians and single parents or busy caregivers [18-21].

In some countries, the regimen and health system-related factors have been seen to pose a barrier for good adherence, whereby complexity, tediousness of the pediatric regimen were noted in some places like South India and Ethiopia [10, 22] and drug side effects and nausea were seen to impede adherence in Northern Tanzania and Jamaica [17, 19].

Enhancing adherence and understanding its barriers in the local context is crucial and must be an ongoing process to ensure ART success. Data amongst children and early teenagers is lacking in Tanzania and specifically in Dar es Salaam. This study aims to cater for this gap and assess the magnitude and attributes of ART adherence in this special group in order to design appropriate interventions to improve, or maintain optimal adherence levels.

Materials and methods

Study design and area

This was a hospital based cross-sectional study conducted between July to November 2018 in three sites within the city centre of Dar-es-Salaam; Muhimbili National Hospital (MNH), Temeke Regional Referral Centre (TRRH) and Infectious Disease Centre-DarDar Paediatric Clinic (IDC-DPP) HIV care and treatment clinics catering for children 1–14 years with their caregivers’ and have monthly scheduled follow-ups’ in accordance with the national guidelines for the management of HIV/AIDS. The three sites were chosen as they were more accessible and represented three different levels of health facilities in the country.

Participants and eligibility criteria

The study participants were confirmed HIV positive children (1–14 years) already initiated on anti-retroviral therapy for at least six months and have scheduled follow up clinic visits with their caregivers at the three afore-mentioned CTC sites. Those on presumptive treatment were excluded. Presumptive treatment is the initiation of ARTs in an infant of an HIV positive mother with certain criteria before the availability of HIV PCR results. This is part of the Tanzania’s national HIV guidelines [23].

We screened the child’s clinical hospital records, CTC 1 and 2 cards for confirmation of the diagnosis of HIV and the duration of treatment and follow up at the CTC. Pharmacy records were checked for confirmation of attendance for drug pick up. Participants and their caregivers were consecutively enrolled until the sample size was achieved, if they fulfilled the inclusion criteria and consented.

The sample size for this study was calculated using the Kish Leslie formula using the prevalence of good adherence from the KCMC, a study in Northern Tanzania which employed three methods to assess adherence and found it to be 24.6% when all three methods were combined thus; Where,

Z—Is percentage point corresponding to a significance level of 5% (1.96)

P—Prior judgment of the correct level of ART adherence in children with HIV (%)

Taken as: 24.6% in a study done at KCMC between 2011–2012 [17]

E—Precision of the estimate = 0.05

Thus

Therefore, 285 was the minimum sample size required for the study.

Considering 10% non-response, the required minimum sample size estimated was 316 participants with their caregivers.

To ensure fair distribution from each site, the sampling technique used was Proportion to size sampling, based on the number patients attending/registered at the specific CTC clinic. Then participants were consecutively recruited until the desired site contribution was reached and the final sample size was obtained in order to achieve a representative, yet, unbiased sample. The percent contribution towards the sample size was calculated as follows:

MNH had 255 patients 1–14 years registered thus contributing 35% of the sample size = 111 participants

TRRH had 250 patients 1–14 years registered thus contributing 34% = 107 participants

IDC-DPP clinic had 230 patients 1–14 years registered thus contributing 31% = 98 participants

In our study a total of 330 met the inclusion criteria during the recruitment process as described in Fig 1. The sample can be considered representative of a larger population as it included patients from three different cadres of the health facility.

Fig 1

Flow diagram showing recruitment of participants.

Variables and data measurement

The variables measured in this study included: adherence level, factors associated with ART adherence after being on ART for at least six months. ART adherence being the dependent variable and independent variables were socio-demographic characteristics like child’s age, gender, schooling status, caregiver’s age, relation to the child, marital status, employment status and monthly income. Participant characteristics like HIV status disclosure, stigmatization, child’s illness, availability of adequate food, belief in ART effectiveness and HIV clinical stage, while treatment related factors considered in this study were type of ART regimen, duration of use, pill burden and side effects. Finally, health system related factors included in this study were drug availability, adequate refills, easy access to health facility and counseling on drug adherence.

Measures of adherence

Three measurement tools were used to assess adherences for all participants: four-day self-recall report, one-month self-recall report (using the visual analogue scale) and number of scheduled clinic appointments missed. For each of these measures, greater than or equal to 95% responses were considered as good adherence, while less than that were taken as poor adherence [24].

4-day self-recall report: Caregivers were asked to recall if they had missed any dose over the last 4 days. If they reported missing more than one dose, it was considered as poor adherence <95%. If they reported one or no missed doses it was considered good adherence (≥95%).

1 month-self recall report (Visual Analogue Scale): Caregivers were asked to mark on a calibrated line scaled 1–10, as a reflection of the way they administered the medications to the child over the last one month. Thereafter, the mark was translated into percentage to estimate their drug administration estimate over the entire month. If the mark was below 9.5 (<95%) it was considered as poor adherence. If marked was at or above 9.5 (≥95%) it was taken as good adherence.

The number of missed clinic appointments: Records of the number of visits scheduled over the past 6 months were asked and confirmed from the CTC card and then the number of visits missed/ delayed according to the clinic’s allowed time for delay, were checked from their CTC card and recorded accordingly. Each CTC had a maximum tolerance for delay where they were provided adequate drugs for that delay period. If there was no attendance beyond the allowed delay period, it was considered a missed visit. If the participant missed his/her scheduled appointment but attended it within the maximum allowed by the CTC, it was not considered a delay. This was further counter-checked with the pharmacies’ refill records to ensure the drugs were picked up. If the participant missed more than one scheduled clinic visit it was considered as poor adherence (<95%).

Thus, after documenting all the three adherence measures, overall good adherence was qualified when the participant did well (≥95%) in all three adherence measures. If he/she performed less than 95% in any one adherence measure, he/she was qualified as having poor overall adherence.

Factors affecting adherence

Factors affecting ART adherence taken in this study were: socio-demographic factors, participant-related factors, drug regimen-related factors and health system-related factors, obtained from interviewing the caregiver.

Data collection methods- tools and technique

Data was collected using a researcher-administered, standardized structured questionnaire, which was developed in English and translated into Kiswahili, the national language in Tanzania to enable participants to respond in a language they fully comprehend. The interview was conducted with the caregiver of the participant, and the questionnaire was simultaneously filled during the interview. Data was collected by the principal investigator and a single research assistant who underwent a week of daily training by the principal investigator before the commencement of data collection. The research assistant was a medical officer who had good understanding about the subject of the study. Screening to identify the confirmed HIV positive children was conducted at the HIV CTC. The investigator first established a good rapport with the client and displayed a non-judgmental, empathic attitude before embarking on the interview, which was conducted in a private comfortable room with no interference/on sight listeners. Data collection forms were anonymous and were only identified by a unique study identification number. Some information was obtained from their medical records (CTC cards, case notes and online system) whichever was being used in that facility. For participants with indicators of poor adherence levels, reasons for poor adherence were enquired from the caregivers.

Statistical analysis

Data was analyzed using SPSS software version 20. Quality control checks were assured during data entry process. Measures of central tendency were used to summarize discrete data. Charts, graphs and tables were used to display categorical data. The overall level of ART drug adherence was calculated, and contingency tables were constructed for bivariate analysis, to explore factors associated with levels of ART adherence. Odds ratio was used to determine associations between the dependent and independent variables. Level of significant association was set at p<0.05.

Bivariate analysis and multivariate logistic regression models were used to determine Odds ratios and p values for varied factors associated with adherence. Only those factors whose Odds ratios had p values<0.2 on Bivariate analysis were further adjusted for in grouped manner on multivariate analysis. Adjusted Odds ratios with p values<0.05 on multivariate analysis were considered significant.

Ethical consideration

Ethical approval to conduct the study was obtained from the Ethics Review Committee of Muhimbili University of Health and Allied Sciences and permission was sought from the respective health facilities. A written informed consent was sought from the caregiver of the child after providing them clear information regarding the study, its benefits, and risks of participating in this study.

A formal written assent was obtained from all children seven years and above who had the capacity to understand, after explaining to them in a simple language the aim of the study and the necessary information at a level they could comprehend.

Confidentiality of the participants’ information was ensured throughout, and no identification was used on the data collection tools. A unique case report form (CRF) identification number was used on the questionnaire, such that no information could be traced back to the participants. Information obtained was stored in a password-secured computer database and the hard copies of the questionnaire were secured in a locked cabinet by the principal investigator.

Results

Socio-demographic characteristics of the study participants

Among the 333 participants as shown in Table 1, the median age was 11 years with an interquartile range of 7–13 years. A slightly higher proportion of the participants were male (53.2%) and more of the participants were in the adolescent age group (10-14yrs) (61.3%), and the majority were attending school (90.7%).

Table 1

Participant and caregiver socio-demographic characteristics.

Variable	Frequency (%) n = 333
Sex of Child
Male	177 (53.2)
Female	156 (46.8)
Child Age Group (years)
Pre-Adolescent (1–9)	129 (38.7)
Adolescent(10–14)	204 (61.3)
Child Schooling Status
Attending school	302 (90.7)
Not attending school	31 (9.3)
Parental Status
Single/Both Parents	276 (82.9)
Non-parental caregiver	57 (17.1)
Relation of Caregiver
Biological Parent	245 (73.6)
Non-Biological Parent	88 (26.4)
Caregiver Age Group
17-<25 years	26 (7.8)
25–<35 years	74 (22.2)
> 35 years	233 (70.0)
Caregiver Marital Status
Married/Cohabiting	177 (53.2)
Single/Divorced/Widowed	156 (46.8)
Employment Status
Employed	218 (65.5)
Unemployed	115 (34.5)
Monthly Income
Less than 100,000	250 (75.1)
100000 and above	83 (24.9)

Most of the participants (82.9%), had at least one parent and for most, they were their biological parent (73.6%). About one-third of caregivers (34.5%) were unemployed and the majority of caregivers (75%) had an average gross monthly income of under 100,000 Tzs. Caregivers were aged between 17 and 75 with a median age of 40 years (IQR 34-46yrs), of which 70% were above 35 years of age.

ART Adherence among HIV infected children

Amongst a total of 333 participants that were recruited, approximately 60% (CI- 54.3–65.1) were noted to have good adherence as shown in Fig 2.

Fig 2

Overall level of ART adherence among children 1–14 years.

Fig 3 displays the ART adherence level by each method of adherence assessment.

Fig 3

Adherence level By each method of adherence assessment.

The four-day self-recall and missed clinic visits had an almost similar adherence rate of 93–94% as compared to the one-month self-recall report using VAS which showed a considerably lower level of adherence (61.6%). However, this was concordant with the overall adherence level.

Factors associated with ART adherence

Socio-demographics factors associated with ART adherence

Table 2 shows the univariate analysis of the selected socio-demographic characteristics in relation to ART adherence. The only factor noted to be associated with ART adherence was the caregivers’ age group. Children whose caregivers were 17–25 years had a two times higher odds of having poor adherence [OR = 2.636, 95% CI 1.146–6.066 p = 0.023] when compared to children whose caregivers were over 25 years.

Table 2

Socio-demographic characteristics associated with ART adherence.

Variable	Adherence		Total (n = 333)	OR (95% CI)	P-Value
	Poor (%)	Good (%)
Gender	(n = 134)	(n = 199)
Male	74(41.8)	103(58.2)	177	1.150 (0.741–1.784)	0.534
Female	60(38.5)	96(61.5)	156
Child Age Group
Pre-Adolescent (1-9yrs)	45(34.9)	84(65.1)	129	0.692(0.439–1.092)	0.113
Adolescent (10-14yrs)	89(43.6)	115(56.4)	204
Child Schooling Status
Attending school	121(40.1)	181(59.9)	302	0.926(0.437–1.959)	0.840
Not attending school	13(41.9)	18(58.1)	31
Parental Status
Single/Both Parents	110 (39.0)	166(60.1)	276	0.911 (0.511–1.625)	0.762
Non-parental caregiver	24(42.1)	33(57.9)	57
Relation of Caregiver
Biological Parent	95(38.8)	150(61.2)	245	0.796(0.486–1.303)	0.364
Non-Biological Parent	39(44.3)	49(55.7)	88
Caregiver Age Group
17–<25 years	16(61.5)	10(38.5)	26	2.636 (1.146–6.066)	0.023
25–<35 years	30(40.5)	44(59.5)	74	1.123 (0.658–1.917)	0.669
> 35 years	88(37.8)	145(62.2)	233	1 Reference variable
Caregiver Marital Status
Married/Cohabiting	71(40.1)	106(59.9)	177	0.989 (0.638–1.534)	0.960
Single/Divorced/Widowed	63(40.4)	93(59.6)	156
Employment Status
Employed	86(39.4)	132(60.6)	218	0.909 (0.574–1.440)	0.685
Unemployed	48(41.7)	67(58.3)	115
Monthly Income
Less than 100,000	101(40.4)	149(59.6)	250	1.027(0.619–1.705)	0.918
100000 and above	33(39.8)	50(60.2)	83

On further investigation about these younger aged caregivers, it was noted six amongst the twenty-six were older siblings caring for their younger ones in the absence of their parents and the remaining were young mothers.

Participant factors associated with ART adherence

Participant factors that were associated with poor ART adherence on univariate analysis, were found to be child having an inter-current illness had eighteen times higher odds of having poor adherence (OR = 18.3; 95% CI 4.214–79.513; p<0.001) and those who had adequate access to food, were two and a half times more likely to be associated with poor adherence (OR = 2.507; 95% CI 1.452–4.330; p = 0.001). In addition, it was seen in this study that disbelief in ART effectiveness and advanced clinical stage was associated with higher odds of poor adherence (OR = 8.474; 95%CI 3.145–22.727; p<0.001) (OR = 1.927; 95% CI 1.149–3.23; p = 0.019) respectively and these were found to be statistically significant as seen in Table 3.

Table 3

Participant characteristics associated with ART adherence.

Variable	Adherence		Total (n = 333)	OR (95% CI)	P-Value
	Poor (%)	Good (%)
HIV Status disclosure	(n = 134)	(n = 199)
Yes	66(45.2)	80(54.6)	146	1.444(0.928–2.245)	0.103
No	68(36.4)	119(63.6)	187
Experiencing Stigma
Yes	121(40.5)	178(59.5)	299	1.098(0.530–2.277)	0.801
No	13(38.2)	21(61.8)	34
Child Inter-current illness
Yes	21(91.3)	2(8.7)	23	18.3(4.214–79.513)	<0.001
No	113(36.5)	197(63.5)	310
Availability of Adequate Food
Yes	39(58.2)	28(41.8)	67	2.507(1.452–4.330)	0.001
No	95(35.7)	171(64.3)	266
Believf in ART Effectiveness
Yes	110(36.2)	194(63.8)	304		<0.001
No	24(82.8)	5(17.2)	29	8.474(3.145–22.727)
Clinical Stage
Stage 1–2	94(36.6)	163(63.4)	257		0.012
Stage 3–4	40(52.6)	36(47.4)	76	1.927(1.149–3.23)

Regimen related and health system-related factors associated with ART adherence

Tables 4 and 5 demonstrate the selected regimen related and health facility factors assessed in this study. None of these factors were significantly associated with poor adherence.

Table 4

Drug regimen related characteristics associated with ART adherence.

Variable	Adherence		Total (n = 333)	OR (95% CI)	P-Value
	Poor (%)	Good (%)
Type of ART Regimen	(n = 134)	(n = 199)
Lopinavir/ritonavir based	27(37.5)	45(62.5)	72	1 Reference Variable
Efavirenz based	26(32.5)	54(67.5)	80	0.802(0.411–1.565)	0.519
Nevirapine based	81(44.2)	100(55.2)	181	1.350(0.771–2.363)	0.293
ART Duration Use
Less than 2 years	12(36.4)	21(63.6)	33	0.834(0.396–1.757)	0.633
2 years and more	122(40.7)	178(59.3)	300
Number of pills per dose
1-pill	30(45.5)	36(54.5)	66	1.306 (0.759–2.249)	0.335
More than 1 pill	104(39.0)	163(61.0)	267
Person administering dose
Child him/herself only	10(55.6)	8(44.4)	18	1.846(0.692–4.927)	0.221
Caregiver only	59(38.3)	95(61.7)	154	0.917(0.583–1.442)	0.708
Child and caregiver	65(40.4)	96(59.6)	161	1 Reference Variable
Side effects Experienced
Yes	19(48.7)	20(51.3)	39	1.479(0.757–2.890)	0.253
No	115(39.1)	179(60.9)	294

Table 5

Health system-related characteristics associated with ART adherence.

Variable	Adherence		Total (n = 333)	OR (95% CI)	P-Value
	Poor (%)	Good (%)
Drug Change due to Unavailability	(n = 134)	(n = 199)
Yes	39(39.8)	59(60.2)	98	0.974(0.602–1.576)	0.915
No	95(40.4)	140(59.6)	235
Adequate Drug Refill
Yes	107(41.5)	151(58.5)	258	1.260(0.740–2.146)	0.395
No	27(36.0)	48(64.0)	75
Difficulty Accessing Health Facility
Yes	76(41.8)	106(58.2)	182	1.15(0.740–1.787)	0.535
No	58(38.4)	93(61.6)	151
Clinic Time Convenient
Yes	57(35.0)	106(65.0)	163	0.649(0.418–1.010)	0.055
No	77(45.3)	93(54.7)	170
Counseled on importance of Adherence
Yes	130(39.6)	198(60.4)	328	0.164(0.018–1.485)	0.108
No	4(80.0)	1(20.0)	5

Independent factors associated with poor ART adherence

The possible factors that were seen to be associated with ART adherence on univariate analysis were further analysed for confounder effect. The Hosmer-Lemshow goodness of fit test for logistic regression was used to test the association, the test statistic equalled a p-value of 0.243, which signified that the logistic regression model used to analyse this data was appropriate. Multivariate analysis results for those factors and additionally others with p<0.2 are displayed in Table 6. It was seen that the factors which remained to be significantly associated with a higher odds of poor adherence, were younger age group caregivers 17–25 years (AOR = 3.520; 95% CI 1.471–8.422), children having inter-current illness (AOR = 10.869; 95% CI 2.340–50.489), disbelief in ART effectiveness (AOR-5.495; 95% CI-1.669–18.182) and advanced children clinical stage (AOR-1.972; 95% CI-1.119–3.484).

Table 6

Independent factors associated with poor ART adherence.

Variables	Adherence		cOR (95%CI)	AOR (95% CI)	P-Value
	Poor(%)	Good(%)
Child Age Group
Pre-Adolescent (1-9yrs)	45(34.9)	84(65.1)	0.692(0.439–1.092)	0.848 (0.453–1.587)	0.607
Adolescent (10-14yrs)	89(43.6)	115(56.4)		1
Care Giver Age Group
17-<25 years	16(61.5)	10(38.5)	2.636(1.146–6.066)	3.520 (1.471–8.422)	0.005
25–<35 years	30(40.5)	44(59.5)	1.123(0.658–1.917)	1.431 (0.781–2.619)	0.246
> 35 years	88(37.8)	145(62.2)	1	1
HIV Status Disclosure
Yes	66(45.2)	80(54.6)	1.444(0.928–2.245)	1.303 (0.717–2.367)	0.385
No	68(36.4)	119(63.6)		1
Child inter-current Illness
Yes	21(91.3)	2(8.7)	18.3(4.214–79.513)	10.869 (2.340–50.489)	0.002
No	113(36.5)	197(63.5)		1
Availability of Adequate Food
Yes	39(58.2)	28(41.8)	2.507(1.452–4.330)	1.574 (0.830–2.984)	0.165
No	95(35.7)	171(64.3)		1
Believes on Effectiveness of ART
Yes	110(36.2)	194(63.8)		1
No	24(82.8)	5(17.2)	8.474(3.145–22.727)	5.495(1.669–18.182)	0.005
Clinical Stage
Stage 1–2	94(36.6)	163(63.4)		1
Stage 3–4	40(52.6)	36(47.4)	1.927(1.149–3.23)	1.972(1.119–3.484)	0.019
Clinic Timing Convenient
Yes	57(35.0)	106(65.0)	0.649(0.418–1.010)	0.711 (0.430–1.177)	0.185
No	77(45.3)	93(54.7)		1
Counseled on Importance of Adherence
Yes			0.164(0.018–1.485)	1.448 (0.106–19.706)	0.781
No				1

Availability of adequate food lost statistical significance on multivariate analysis. Moreover, Child age group, HIV disclosure, clinic timings and counselling were not found to be associated with poor ART adherence.

Caregiver reasons for missing doses in HIV infected children

For those caregivers who reported missing even one pill, reasons were explored and some of the main challenges cited by them were; forgetting to take pills daily (41%) and high pill burden (21%) as the predominating factors followed by busy schedule- interrupting drug administration (11%) and long waiting hours at the clinic (9%). These, among the other cited reasons, are shown in Fig 4.

Fig 4

Reasons for missing doses in HIV infected children.

Discussion

ART adherence in children is crucial to ensure treatment success. Poor adherence results in sub-optimal drug levels making it less effective in suppressing viral replication resulting in treatment failure. There is an emerging concern as to what level of adherence can be achieved from children in poor resource settings, since adherence in children and adolescents has been identified as a challenge due to multiple factors, dependence on others to give them the drugs and problems related to the endurance of life-long therapy.

The overall ART adherence of approximately 60%, is in keeping with a study done in Nigeria in 2015, and close to that reported in rural Tanzania, during the same year 65.6% and 70% respectively [16, 20]. A systematic review of ART adherence in low and middle-income countries in 2008 found a range in adherence estimates from 49% to 100% using different adherence measures [25].

In contrast to our findings, adherence was reported to be higher ranging from 72% to 94% among the last two decades in other sub-Saharan countries including; Ethiopia, Tanzania, Malawi and Uganda, with most studies reporting>90% ART adherence by self-recall means [10-13]. Similarly, a single centred study done in Northern Tanzania, subjected to the three methods, reported a significantly lower adherence level of 24.6% among HIV infected children [17]. Our comparatively higher adherence estimates despite multiple methods, could be explained by the considerable improvement in CTC services, free ART access and improved adherence counselling over the past half-decade.

The possible explanation of the exceptionally high contrasting results, obtained by the other sub-Saharan countries, could be due to their use of short-term recall methods to assess adherence. In fact, in this study, using the four-day recall method alone to measure adherence would have shown similar high adherence of 93.4% as these studies. This method, when used alone, is known to over-estimate ART adherence, especially if caregivers are the ones reporting, due to their vulnerability to social desirability and over-reporting. These biases could result in falsely inflated adherence estimates in the actual situation [25]. While the VAS score is subjected to recall bias, it is the only measure which may capture the participant’s own perception of his/her adherence, a measure of self-evaluation and may actually affect the outcome of future adherence and thus is an important measure that should not be omitted.

Multivariate logistic regression analysis indicated that the age of caregiver was an independent factor associated with ART adherence. Children, whose caregivers were 17–25 years, were significantly found to be associated with poor adherence. These findings were similar to those reported from Kano (Nigeria) and Mekelle (Ethiopia) [10, 16]. However, caregivers’ marital status and their economic condition did not seem to have an effect in this study like it did in the aforementioned surveys. Similarly, findings from KCMC in Northern Tanzania, suggested caregivers’ low monthly income was associated with poorer adherence [17]. The caregivers’ relationship to the child had little effect on adherence status of the child, unlike the findings reported from a region in rural part of Tanzania-Ifakara in 2015 by Nyogea et al from Tanzania where step mothers contributed to poor adherence [20].

A child’s optimal adherence strongly relies on the caregiver’s commitment; the younger age group are more likely to be less responsible towards this entrustment than their counterparts, who may have a better understanding of the dynamics of disease and health and thus, deserve more focused attention on adherence counselling and may entail health providers to identify older household members to be involved hand in hand in the care of the child.

Regimen related and health facility related factors affecting adherence

The study took place in centres with trained and motivated staff, where all the facilities provided care which was in line with the national ART scale up policy, by using simple once/twice daily regimens, fixed dose combinations when available and constant supply at all times. Due to the fully funded HIV CTC services, participants no longer faced the regimen related and healthcare related barriers to ART adherence including: cost of drugs, difficulty in accessing healthcare facilities, inadequate refills, unqualified regimen changes and pill burden. Therefore, these factors were not found to significantly influence ART adherence in this study; which was in line with other studies in Sub-Saharan African countries with similar setups in providing ART services [13, 17, 26–28].

This study highlighted that, the most significant barriers to good ART adherence were those related to the participant themselves; Child’s intercurrent illness, availability of adequate food, belief in the effectiveness of ARTs and HIV clinical stage. In a few studies, some but not all these factors were found to be obstacles for ART. A study done in Nigeria in 2010, reported child’s inter-current illness and non-belief in ARTs; while a study in rural Tanzania (Ifakara) found preferring herbal medicine and experiencing stigma as barriers for ART adherence [16, 20]. On the contrary, a study in urban Malawi, Ethiopia and northern Tanzania found no significant relation of ART adherence with a child’s WHO clinical stage or caregivers’ belief in ART effectiveness [13, 17, 29]. Non-disclosure and stigma were not found to affect adherence in this study as it did in earlier studies [12, 30]. The most plausible reason for this difference may be due to improved awareness, acceptance of HIV/AIDS and encouragement of timely child disclosure.

To explain our findings, we speculate; in Tanzania, while the knowledge of HIV/AIDS and its treatment is overall good, beliefs of an existing supernatural power that may have afflicted an HIV infected person still exists, leading to a preference for alternative healing (herbs, spiritual medicine) which may hinder ART use. The findings point to a possible serious gap in a caregiver’s awareness of health implications of non-adherence during illness and advanced stage, coupled with a false belief that taking ARVs with other drugs or during concomitant illness could result in serious adversities. Further studies on caregivers’ belief aspects need to be carried out to confirm this finding.

Surprisingly in this study, the availability of adequate food was found to be associated with poor adherence, similar to findings from Addis Ababa [29], however this lost significance on multivariate analysis. It is difficult to speculate about this association, as several studies have deemed nutrition availability and support as an integral part of successful HIV care and treatment [31-33] losing its significance on further analysis verifying its confounder effect. Moreover, under-nutrition was noted among less than one-fifth of the participants (17%) signifying that it poses less of a problem in recent times than it did around a decade earlier in Tanzania [27] and Ethiopia [26]. The possible explanation for this could be that most of these patients were not ART naïve and had been attending the clinic and receiving counseling, which may have improved their nutritional status. Furthermore, nutrition is not entirely dependent on adherence to ARVs, rather is determined by multiple factors like educational and socioeconomic status, family size, dietary diversity and other comorbid conditions [33-35].

Caregivers’ reasons for missing medications

The main problems cited by the caregivers, who were responsible for a missed dose in this study were: forgetfulness, high pill burden, a busy schedule, and long waiting hours at the clinic. Some, but not all of these factors, were reported in other studies from similar resource limited countries like Ethiopia-Mekelle which found the pill a burdensome and a survey in Addis Ababa found forgetfulness (23.5%) [26] as barriers to observing adherence. A study done in Nigeria also reported similar findings as ours; forgetfulness (59.5%), travel away from home/busy (21.4%) and child inter-current illness (14.3%) [16].

A multi-centre study in Tanzania reported long waiting hours at the clinic to be a hindrance to effective adherence [27] which was similar to findings in this study. This is possible because some HIV clinics operated only once a week for children. In addition to that, with the current provision of free drugs, the demand is higher and more prone to having longer waiting hours, unlike previous times where patients had to pay for their drugs. This can lead to patient dissatisfaction and default from treatment [36]. The pill burden was observed to be high among most of the participants of this study and this could be due to the deficiency of paediatric fixed-dose combinations available for every ART regime, as is available for adults- as a single pill taken once a day.

On the contrary, earlier studies reported explanations of suboptimal adherence being the cost of transportation, difficulty accessing health facilities, cost of medication and lack of adequate counselling [27, 37]. During those times, HIV care was not entirely free and widely distributed and accessible. Patients had to bear some costs of their treatment, including buying their drugs and paying for lab tests. Currently, with support funding from PEPFAR, HIV drugs and services is cost-free, and clinics and transport services are easily accessed and thus not reported to be problematic.

The strengths of this study were that it was conducted in three centres in the region including primary and tertiary level health facilities, serving participants from most districts. Therefore, the data obtained is reflective of the actual situation of adherence in our surroundings. Furthermore, the parameters used to assess adherence were practical and easily employed in any low-resource clinical setting. Three separate measures were in cooperated, to improve the reliability of the assessment, and thus remove the bias of one measure over the other.

This study should be interpreted with the following limitations in mind: Cross-sectional studies of this nature are unable to capture trends in adherence, which are known to change over time, thus are limited in drawing conclusions regarding causal effect relationship of potential factors, a longitudinal component may be the most appropriate for this regard. In addition, classification of different adherence cut-off points, adherence measures and study designs used in different studies may not be perfect to compare and contrast findings meticulously. With regards to the parameters used to assess adherence, they are liable to be affected by recall bias and social desirability bias and these may have over-estimated good adherence. However, efforts were made to mitigate them by employing three different methods, whereby, the 4-day recall method has nearly no recall bias and the clinic attendance method is highly objective and concrete. The one-month recall method is actually the most accurate depiction of the care givers’ own perception of the child’s adherence and has implications on his/her future adherence pattern and is thus an important parameter that should not be excluded.

In conclusion, good adherence was noted only among 60% of the children attending HIV CTC clinics in Dar es Salaam. The factors influencing poor adherence were; younger age of caregivers and a child having inter-current illness whilst, belief in the effectiveness of ARTs and a child in the initial HIV clinical stages fostered good adherence. The most cited caregiver-related reasons for missing drugs were; forgetfulness, high pill burden, busy schedule and long waiting hours at clinic.

Thus, we recommend the following; health workers should assess for ART adherence and routinely do adherence counseling at each encounter with the caregiver, whereby these simple adherence tools should be in-cooperated, in order to identify and detect poor adherence with more focus on child-centered adherence counseling strategies so as to further improve adherence in children in order to achieve 90-90-90 target.

Healthcare providers should also seek to enlist older caregivers above the age of 25 years in the household as treatment supporters for each child. They may team up with the younger caregivers to enhance ART adherence among children. Sick children and those with advanced HIV stages may need to be admitted or have a much closer follow-up with the clinician to improve their clinical status and adherence to treatment, as it may be demotivating to the caregiver when they see their child not improving giving them a reason to default treatment. It is during these periods when children most need their drugs to overcome their illness and improve their clinical stage. Furthermore, health education at point of service, like during the daily morning talks and on social media platforms about the effectiveness of ARVs’ and alleviating the mythical beliefs about the HIV/AIDS and its treatment in order to improve adherence.

The questionnaire.

(DOCX)

Click here for additional data file.

31 Dec 2020

PONE-D-20-25583

MAGNITUDE AND ATTRIBUTES OF ANTI-RETROVIRAL THERAPY ADHERENCE AMONG CHILDREN (1-14 YEARS) ATTENDING HIV CARE AND TREATMENT CLINICS IN DAR ES SALAAM, TANZANIA

PLOS ONE

Dear Dr. Mussa,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by January 29, 2021. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

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: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Satya Surbhi, PhD

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. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified whether consent was informed.

3. Please provide your sample size calculation.

4. In your Methods section, please provide additional information about the participant recruitment method and the demographic details of your participants. Please ensure you have provided sufficient details to replicate the analyses such as:

a) the recruitment date range (month and year),

b) a statement as to whether your sample can be considered representative of a larger population, and

c) a description of how participants were recruited.

5. Please include a copy of the interview guide used in the study to interview caregivers, in both the original language and English, as Supporting Information, or include a citation if it has been published previously.

6. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

7. Thank you for stating the following financial disclosure:

"The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

At this time, please address the following queries:

Please clarify the sources of funding (financial or material support) for your study. List the grants or organizations that supported your study, including funding received from your institution.

State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

If any authors received a salary from any of your funders, please state which authors and which funders.

If you did not receive any funding for this study, please state: “The authors received no specific funding for this work.”

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

8. Please amend either the title on the online submission form (via Edit Submission) or the title in the manuscript so that they are identical.

9. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

10. We note you have included a table to which you do not refer in the text of your manuscript. Please ensure that you refer to Table 1 in your text; if accepted, production will need this reference to link the reader to the Table.

[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: Yes

Reviewer #3: Yes

**********

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

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

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: No

**********

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: Yes

Reviewer #3: No

**********

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: The manuscript needs revision especially on data analysis, methods and discussion section. Besides this the ascent should be stated in ethical clearance for children from their parents or guardian. The factors that are insignificant should be deleted.

Reviewer #2: The manuscript is well-prepared. It is scientifically sound. It will be more clear and precise if the authors include the analysis tools in the methodology section of the abstract, give more background information on the pediatric ART programme in the country like the regimens introduced etc. In addition, even if the number of references are enough better to include more recent related publications. The paper cites references dated back to 2000 to 2018.

Reviewer #3: First, I would like to congratulate the authors for exploring such an important topic. Magnitude and Attributes Of Anti-Retroviral Therapy Adherence Among Children (1-14 Years) Attending HIV Care And Treatment Clinics In Dar Es Salaam, Tanzania. However I have some comments and questions for them, these are found below

Abstract:

Introduction:

Line 1: In Epidemiological disease level classification HIV is a Pandemic disease; use this term instead of Epidemic. In addition at the end of the introduction please add justification/the need to conduct this study in a line or two.

Objectives: Correct the English (Among HIV positive children). Objective must be SMART which includes the “T” which stands for time bound. Therefore specify the time of the study.

Methods: The first statement is not even a full sentence. The method presented is too shallow. The sample size, the method of data analysis, level of significance, the measurement of association, the model fitness assessment, the software used all are lacking.

Result: The magnitude of good adherence reported (60%) is a point estimate, you should provide the confidence interval. Take the type of analysis used in the method section, use the term multivariable logistic regression instead of multivariate (Multivariate is when you have multiple outcome variables). Age is better reported as a range instead of <25 for children of very young (underage) mothers may not have the same association. Belief in ART effectiveness is not clear (Negative or Positive?? And how did you assess it??). After all the associated factors were written the sentence was concluded as “”… were less likely to be associated poor adherence.” Yet all the significant variables are not negatively associated with the outcome variable. Besides, though, the authors reported the magnitude of good adherence, they reported associated factors for poor adherence, you need to be consistent.

Conclusion and recommendations: The authors should re-write this section. Most of it is wrong and the recommendations given are not based on the findings.

Methods and Materials

Study design and area: In the first line authors mentioned “This was hospital based descriptive cross-sectional study” which is totally wrong. You have assessed factors that affecting adherence therefore it is not mere description. (Use hospital based cross-sectional study)

Participants and eligibility criteria

The minimum sample size determination must be clearly presented. There is no such a thing as “quota sampling calculation” Quota sampling is a sampling technique (How you distribute the sample size you determined across quotas).

How was the three study places identified?? Why none probability sampling???

How can you take consent form under 14 years old children?? (This must be clearly described in the Ethical consideration)

Variables and measurements

Clearly put the outcome variable and list of independent variables considered.

Statistical analysis

What method of goodness of fit assessment did you use?? What was its value??

Generally the method part lacks the following: Study period, Sample size determination, Sampling Procedures, Data collection procedure and quality assurance.

Results

The first paragraph must be part of the method section. You should focus on what you have found and not how you did it in the result section.

The order of the result presentation is full of flaws. First report the characteristics of study participants be it socio-demographic characteristics or co-morbidities. Then go to the descriptive findings of your study which the level/magnitude of adherence, finally you can present the factors associated. The level/magnitude of adherence also needs to be reported using confidence intervals.

Factors associated with ART adherence

Why are the authors reporting the associated factors as socio-demographic and participant related and the like all variables must be adjusted in the multivariable analysis and be reported as one multivariable table.

**********

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: Abebe Dilie Afenigus

Reviewer #2: Yes: Yesunesh Teshome

Reviewer #3: Yes: Adhanom Gebreegziabher Baraki

[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: comments for Plos One.docx

Click here for additional data file.

Submitted filename: PONE-D-20-25583_reviewer.pdf

Click here for additional data file.

Submitted filename: Reviewer comment for PONE-D-20-25583.docx

Click here for additional data file.

9 Feb 2021

RESPONSES TO REVIEWERS

REVIEWER 1

Title:

1. Better to say “associated factors” rather than “attributes” in the title.

RESPONSE: CHANGED TO ASSOCIATED FACTORS

2. Originality: what makes your study unique from a similar study conducted in your country which is available with the following link?

https://bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-015-0753-y?

RESPONSE: THIS STUDY WAS CONDUCTED IN A RURAL AREA IN IFAKARA IN 2015, IT WAS A MULTICENTRE STUDY WITH A SMALLER SAMPLE SIZE AND THAT MAY LIMIT GENERALIZABILITY EVEN IN OUR SETTING. A SINGLE MEASURE WAS USED TO ASSESS ADHERENCE. THEIR COHORT WAS A VERY WIDE AGE RANGE FROM 2-19 YEARS AND IT IS KNOWN THAT FACTORS AFFECTING ADHERENCE VARIES GREATLY AMONGST CHILDREN AND ADOLESCENTS.

ADHERENCE AND FACTORS AFFECTING IT, IS A DYNAMIC PROCESS THAT CHANGES OVER TIME AND NEEDS REGULAR ASSESMENT TO IMPROVE AND OVERCOME ITS BARRIERS, THAT WARRANTS US TO REVIEW THE OUTCOMES AS WE CHANGE AND IMPROVE OUR PRACTICES IN THE CARE OF THESE CHILDREN. FURTHERMORE, WE DID A MULTICENTRE STUDY, CHOSING DIFFERENT CADRES OF THE HEALTH SYSTEM FROM THE COUNTRYS NATIONAL REFERAL HOSPITAL TO A REGIONAL HOSPITAL CENTRE, TO A PRIMARY LEVEL OUT PATIENT CLINIC, THUS TRYING TO CAPTURE ADHERENCE AND ITS BARRIERS ACROSS ALL LEVELS IN THE MAIN CITY CENTRE OF THE COUNTRY. WE CHOSE TO TARGET THE YOUNGER AGED CHILDREN UNDER THE CARE OF CAREGIVERS RATHER THAN COMBINING THEM WITH ADOLESCENTS AS THEIR BARRIERS DIFFER GREATLY AND THUS OUR RESULTS ARE QUITE INDICATIVE OF THE SITUATION AFFECTING THIS PARTICULAR AGE GROUP.

Key words:

1. Please use key words rather than phrases like ART adherence, associated factors, children, Magnitude…..

RESPONSE: CHANGED TO KEYWORDS RATHER THAN PHRASES

Abstract

1. Objectives: the objective should be inline with your title i.e. “to assess magnitude and associated factors of ART adherence among children (1-14 years) attending HIV care …

RESPONSE: CHANGED TO THE ABOVE

2. Result: what are all the three measures? Please also approximate the odds ratios and confidence intervals to the nearest number and use only two-digit numbers like 3.520 → 3.5

RESPONSE: THE 3 MEASURES WERE DESCRIBED IN THE METHODS SECTION OF THE ABSTRACT AS FOUR DAY SELF-REPORT, ONE MONTH SELF-RECALL REPORT AND MISSED CLINIC APOINTEMENTS

HAVE APPROXIMATED THEM TO THE NEAREST DECIMAL PLACE

Introduction

1. First begin by defining HIV/AIDS and adherence

RESPONSE: HAVE INCLUDED THAT

2. Please also state the severity and magnitude of the problem (Poor ART adherence) in Tanzanian context as well.

RESPONSE: LIMITED STUDIES AVAILABLE FOR THAT ESPECIALLY AMONGST CHILDREN, HAVE INCLUDED WHAT WAS FOUND.

Methods and materials

Study design and area

1. Your title includes 1-14 years but in study design you included 0-14 years. why?

RESPONSE: MY APOLOGIES, THIS WAS A TYPO, HAVE CORRECTED IT.

2. In the paragraph you included children who had regular follow-up. Since you study adherence you have to include children who didn’t have also regular follow up. If you include clients who had regular follow up alone, it over estimates the magnitude of good adherence.

REPSONSE: THIS WAS A MISINTERPRETATION, PROBABLY A WRONG CHOICE OF WORDS ON MY SIDE; I MEANT CHILDREN WHO ARE ENROLLED IN THAT CLINIC AND SCHEDULED TO ATTEND THERE FOR REGULAR REFILLS AND FOLLOW UP VISITS. NOT THAT I CHOSE PATIENTS WHO REGULARLY ATTENDED. AMONGST THESE PATIENTS WERE THOSE WHO HAD POOR FOLLOW UPS AND THUS DETECTED ON MISSED CLINIC VISIT ADHERENCE MEASURE. HAVE CHANGED IT TO BE MORE CLEAR

Participants and eligibility criteria

1. What is presumptive treatment? Is this recommended by WHO?

PRESUMPTIVE TREATMENT IS PART OF OUR COUNTRIES NATIONAL HIV/AIDS TREATMENT GUIDELINES AND PART OF OUR PRACTICE WHERE INFANTS AND CHILDREN<18MONTHS BORN TO HIV POSITIVE MOTHERS ARE GIVEN A PRESUMPTIVE DIAGNOSIS OF SEVERE HIV INFECTION BASED ON CERTAIN CRITERIA IN ORDER NOT TO DELAY TREATMENT WHILE AWAITING CONFIRMATORY DNA PCR RESULTS. FURTHER DETAILS CAN BE FOUND FROM THE NATIONAL GUIDELINES FOR THE MANAGEMENT OF HIV/AIDS, 2017, CHAPTER 7, PAGE 111 ( Available in the references).

2. Why you used quota sampling? It is better to use probability sampling techniques in order to generalize your findings.

RESPONSE: QUOTA SAMPLING WAS USED AS WE WERE TIME BOUND AND HAD A TIGHT BUDGET.

3. You used clinical hospital records. You have to set some exclusion criteria, if major variables are missed in the chart.

RESPONSE: HOSPITAL RECORDS WERE NOT USED TO COLLECT DATA, AS DATA WAS COLLECTED FROM A STRUCTURED QUESTIONNAIRE DURING AN INTERVIEW PROCESS WITH THE CAREGIVER. THE RECORDS WERE ONLY USED TO CONFIRM THE DIAGNOSIS OF HIV, DRUG REGIMEN THEY ARE ON AND OTHER SUPPORTIVE INFORMATION. THIS ENSURED NO MISSING VARIABLES FOR OUR STUDY.

Variables and data measurement

1. Better to state your variables in categories of dependent (outcome variable) and independent (predictor) variables in paragraph form

RESPONSE: ADDED THAT AS A PARAGRAPH

2. Adherence measurement: adherence can be adherence to care (clinical adherence) meaning regular attendance of the patient according to given appointment, and drug adherence which means taking the drug according to instructions given by the providers. Your focus is adherence to drug (ART). Therefore, why you used missed clinic appointments as measure of adherence?

RESPONSE: AGREED, HOWEVER IN OUR SETUP DRUG REFILLS ARE PROVIDED DURING THESE SCHEDULED MONTHLY VISITS, THUS IF A PARTICIPANT DOESNOT ATTEND A CLINIC VISIT THEY WILL MISS A DRUG REFILL AND BE UNABLE TO TAKE THE DRUGS AS SCHEDULED AND INSTRUCTED BY THE PROVIDER AND THUS CONSIDERED AN OBJECTIVE WAY TO ASSESS DRUG ADHERENCE.

3. It is better to use WHO adherence scale by calculating percent (%) adherence and classify level of adherence as good, fair or poor. You can merge fair and poor based on your operational definition you stated.

RESPONSE: IN THIS STUDY, THE VISUAL ANALOGUE SCALE (VAS) WAS USED TO OBTAIN THE PERCENTAGE ADHERENCE AND CLASSIFIED AS POOR OR GOOD, THIS WAS CHOSEN AS THE CAREGIVER HAD TO OBSERVE THE SCALE AND MARK ON THEIR OWN NOT BEING INFLUENCED OR JUDGED FOR THE MARK THEY PUT THUS REDUCING THE EFFECT OF SOCIAL DESIRABILITY BIAS. FURTHER MORE IT GAVE A DEPICTION OF WHAT THEY PERCEIVED THEIR ADHERENCE WAS, WHICH IS IMPORTANT IN THEIR FUTURE ADHERENCE CHOICES. THIS SCALE HAS BEEN VALIDATED AMONGST STUDIES AND FOUND USEFUL IN RESEARCH AND CLINICAL SETTINGS.

4. In recall report of measuring adherence, how did you handle recall bias?

RESPONSE: IN ORDER TO MINIMISE THE RECALL BIAS WE CAREFULLY SELECTED THE QUESTIONS/ MEANS OF ASKING. USING THE VAS AND RECALLING OVER THE LAST MONTH WHICH WAS NOT TOO LONG AGO, AS IT HAS BEEN SHOWN RECALL BIAS IS PROBLEMATIC WHEN THE EVENT OCCURS LONG TIME AGO. FURTHER MORE WE COMBINED IT WITH SHORTER RECALL METHODS TO REDUCE THE RECALL ELEMENT AND MORE OBJECTIVE MISSED CLINIC VISIT.

Factors affecting adherence

1. Since it is part of variables, it is better to write in variable section before measures of adherence and outcome.

RESPONSE: I HAVE MOVED IT IN THE VARIABLES SECTION, BUT I FEEL ITS BETTER PUT AFTER MEASURES OF ADHERENCE AS ADHERENCE MEASURING IS THE FIRST OBJECTIVE AND FACTORS AFFECTING ADHERENCE IS THE SECOND OBJECTIVE.

Statistical analysis:

1. What is the importance of chi square test if you use odds ratio? Odds ratio shows measure as well as strength of association. Therefore, it is better to use odds ratio and logistic regression model.

RESPONSE: APOLOGIES, THAT WAS A TYPO ON MY SIDE, WE USED ODDS RATIO AND LOGISTIC REGRESSION MODELS. THANKS FOR NOTING THAT. I HAVE RECTIFIED THE TYPO.

2. In the logistic regression model, it is better to use bivariable and multivariable logistic regression rather than univariate regression

RESPONSE: AGREED, BIVARIATE AND MULTIVARIABLE LOGISTIC REGRESSION WAS DONE.

Result:

1. What is the importance of mentioning 830, 476 … if this population doesn’t meet inclusion criteria and not part of your study? Focus only on sample size.

RESPONSE: REMOVED THE FLOW DIAGRAM AND EXPLANATION ON RECRUITMENT.

2. Write sociodemographic characteristics before ART adherence among HIV infected children

RESPONSE: DONE

3. What is the importance of writing table 4 and 5, if all of the variables are insignificant?

RESPONSE: MY TAKE ON THIS WOULD DIFFER, THE STUDY OBJECTIVE WAS TO FIND OUT THE FACTORS AFFECTING ART ADHERENCE AMONG CHILDREN, IN TERMS OF THEIR SOCIO-DEMOGRAPHIC CHARACTERISTICS, HEALTH FACILITY/ DRUG RELATED FACTORS. IT WAS ONLY AFTER MULTIVARIABLE LOGISTIC REGRESSION IT LOST SIGNIFICANCE. I THINK ITS IMPORTANT TO SHOW THEY WERE SOUGHT, HOWEVER LOST SIGNIFICANCE AFTER REGRESSION ANALYSIS.

4. It is better to write the factors affecting adherence in 1 table

RESPONSE: TABLE 2,3,4,5 IS SHOWING BIVARIATE REGRESSION ANALYSIS OF THESE FACTORS. HOWEVER TABLE 6 IS ONLY DEPICTING THOSE FOUND SIGNIFICANT ON BIVARIATE ANALYSIS, BEING FURTHER ANALYSED IN MULTIVARIABLE LOGISTIC REGRESSION MODEL AND THUS COMING UP WITH THE MOST SIGNIFICANT FACTORS IN ONE TABLE.

Discussion

1. Paragraph 1: please discuss only your pertinent findings. Paragraph 1 is not related to your study. Avoid it.

RESPONSE: THANKS, REMOVED IT.

2. Please write similar concepts in 1 paragraph in brief and succinct way.

RESPONSE: THANKS HAVE REMOVED THE INSIGNIFICANT FINDINGS AND DWELLED ON WHAT WAS SIGNIFICANT TO THIS STUDY.

3. No need to write sub themes like “Factors affecting ART adherence in children…” in discussion. Please discuss the factors without subheadings.

REPSONSE : NOTED THANKS

Conclusion:

1. Conclude based on your objective

RESPONSE : THANKYOU, THE FEEDBACK IS NOTED.

Recommendation:

2. Recommend based on your result

RESPONSE: THANKYOU, THE FEEDBACK IS NOTED.

Summary status: Major revision

RESPONSE: THANKYOU FOR ALL YR COMMENTS AND CONSTRUCTIVE FEEDBACK. I APRRECIATE THE TIME YOU TOOK TO READ THROUGH AND ADVISE. HAVE REVISED MOST OF THE RECOMMENDED CHANGES, AND GIVEN AN EXPLANATION FOR THE REST.

REVIEWER 2

First, I would like to congratulate the authors for exploring such an important topic. Magnitude and Attributes Of Anti-Retroviral Therapy Adherence Among Children (1-14 Years) Attending HIV Care And Treatment Clinics In Dar Es Salaam, Tanzania. However I have some comments and questions for them, these are found below

Abstract:

Introduction:

Line 1: In Epidemiological disease level classification HIV is a Pandemic disease; use this term instead of Epidemic. In addition at the end of the introduction please add justification/the need to conduct this study in a line or two.

RESPONSES: CHANGED THE WORD TO PANDEMIC. ADDED A JUSTIFICATION LINE FOR THE STUDY.

Objectives: Correct the English (Among HIV positive children). Objective must be SMART which includes the “T” which stands for time bound. Therefore specify the time of the study.

RESPONSES: CORRECTED THE ENGLISH, SPECIFIED THE TIME

Methods: The first statement is not even a full sentence. The method presented is too shallow. The sample size, the method of data analysis, level of significance, the measurement of association, the model fitness assessment, the software used all are lacking.

RESPONSES: THANKYOU FOR YR CONTRIBUTION; COMPLETED THE SENTENCE. ADDED ALL THE NECESSARY MISSING DETAILS.

Result: The magnitude of good adherence reported (60%) is a point estimate, you should provide the confidence interval. Take the type of analysis used in the method section, use the term multivariable logistic regression instead of multivariate (Multivariate is when you have multiple outcome variables). Age is better reported as a range instead of <25 for children of very young (underage) mothers may not have the same association. Belief in ART effectiveness is not clear (Negative or Positive?? And how did you assess it??). After all the associated factors were written the sentence was concluded as “”… were less likely to be associated poor adherence.” Yet all the significant variables are not negatively associated with the outcome variable. Besides, though, the authors reported the magnitude of good adherence, they reported associated factors for poor adherence, you need to be consistent.

RESPONSES: WE CANNOT DO PREVALENCE AS WE DONOT HAVE THE ENTIRE DENOMINATOR OR TOTAL NUMBER OF CHILDREN ATTENDING THE FACILITY OVER TIME EG. PER YEAR. WE ONLY COUNTED THOSE WHO CAME DURING THE SAID DATES AND CALCULATED A PROPORTION OUT OF THAT FOR GOOD VERSUS BAD ADHERENCE

Conclusion and recommendations: The authors should re-write this section. Most of it is wrong and the recommendations given are not based on the findings.

RESPONSE: THANKYOU FOR YOUR COMMENT. HAVE CONCLUDED AND RECOMMENDED MY STUDY AS I VISIONED IT TO BE AND TO THE BEST OF MY ABILITY.

Methods and Materials

Study design and area: In the first line authors mentioned “This was hospital based descriptive cross-sectional study” which is totally wrong. You have assessed factors that affecting adherence therefore it is not mere description. (Use hospital based cross-sectional study)

RESPONSE: CHANGED AS ADVISED

Participants and eligibility criteria

The minimum sample size determination must be clearly presented. There is no such a thing as “quota sampling calculation” Quota sampling is a sampling technique (How you distribute the sample size you determined across quotas).

How was the three study places identified?? Why none probability sampling???

How can you take consent form under 14 years old children?? (This must be clearly described in the Ethical consideration)

RESPONSES: MY APOLOGIES THAT WAS A TYPO, I MEANT QUOTA SAMPLING WAS DONE. HAVE CLEARLY SHOWN HOW THE SAMPLE SIZE WAS DERIVED AND HOW WERE THEY DISTRIBUTED AMONGST THE 3 SITES.

THE THREE SITES WERE MORE ACCESSIBLE AND REPRESENTED THREE DIFFERENT LEVELS OF HEALTH FACILITIES IN ORDER TO MAKE THE STUDY REPRESENTABLE IN THE CAPACITY WE HAD.

INFORMED CONSENT WAS SOUGHT FROM THE CAREGIVER WHO BROUGHT THE CHILD TO THE CTC CLINIC. A WRITTEN INFORMED ASSENT WAS SOUGHT FROM CHILDREN WHO WERE CAPABLE OF UNDERSTANDING. HAVE CLARIFIED IT IN THE ETHICAL CONSIDERATION SECTION.

Variables and measurements

Clearly put the outcome variable and list of independent variables considered.

RESPONSES: HAVE CLEARLY PUT IT DOWN.

Statistical analysis

What method of goodness of fit assessment did you use?? What was its value??

Generally the method part lacks the following: Study period, Sample size determination, Sampling Procedures, Data collection procedure and quality assurance.

RESPONSE: HAVE ADDED THE DETAILS IN THE METHODOLOGY

Results

The first paragraph must be part of the method section. You should focus on what you have found and not how you did it in the result section.

The order of the result presentation is full of flaws. First report the characteristics of study participants be it socio-demographic characteristics or co-morbidities. Then go to the descriptive findings of your study which the level/magnitude of adherence, finally you can present the factors associated. The level/magnitude of adherence also needs to be reported using confidence intervals.

RESPONSE: HAVE REMOVED THE FIRST PARAGRAPH AND ARRANGED THE ORDER AS ADVISED.

Factors associated with ART adherence

Why are the authors reporting the associated factors as socio-demographic and participant related and the like all variables must be adjusted in the multivariable analysis and be reported as one multivariable table.

RESPONSE: TABLE 2-5 SHOWS THE BIVARIABLE ANALYSIS OF THE FACTORS WE ARE LOOKING INTO CATEGORISED AS SOCIOECONIMIC FACTORS, PARTICIPANT FACTORS, DRUG RELATED FACTORS AND HEALTH SYSTEM RELATED FACTORS. ONLY THOSE WHICH WERE FOUND SIGNIFICANT WERE FURTHER ANALYSED FOR MULTIVARIABLE ANALYSIS AND PUT IN ONE SINGLE TABLE. IT MAY BE DIFFICULT TO PUT ALL THE TABLES IN BIVARIABLE ANALYSIS IN 1 SINGLE TABLE AS WE HAVE CATEGORISED IT. HOWEVER THE MULTIVARIABLE TABLE IS ONE IE: TABLE 6

Submitted filename: RESPONSE TO REVEIWERS.docx

Click here for additional data file.

13 Apr 2021

PONE-D-20-25583R1

MAGNITUDE AND ASSOCIATED FACTORS OF ANTI-RETROVIRAL THERAPY ADHERENCE AMONG CHILDREN ATTENDING HIV CARE AND TREATMENT CLINICS IN DAR ES SALAAM, TANZANIA

PLOS ONE

Dear Dr. Mussa,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by May 12, 2021. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

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:  http://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,

Satya Surbhi, PhD

Academic Editor

PLOS ONE

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

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

Reviewer #3: (No Response)

**********

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

Reviewer #3: Partly

**********

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

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

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

Reviewer #3: 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

Reviewer #3: (No Response)

**********

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: I would like to appreciate the authors for studying a public health problem. Magnitude and associated factors of Anti-Retroviral Therapy Adherence Among Children (1-14 Years) Attending HIV Care and Treatment Clinics in Dar Salaam, Tanzania)

The response and revision given by the authors for my previous concerns is plausible.

# Discussion

In the discussion section, it better to write an introductory paragraph what your study focuses on (magnitude and factors associated with ART adherence) at the beginning and then proceed to discuss pertinent findings as you did already.

Reviewer #2: The authors have addressed the comments from my reviews in an intelligent and scientific way. I accept that the manuscript is worth publishing.

Reviewer #3: Thank you all the authors for making the requested revisions. I believe you have made a significant improvement from the original manuscript. But I have some issues I asked but not well addressed.

Here are my comments and questions

Abstract

1. The English of the manuscript is still poor. For instance in the method section of the abstract I have commented the first sentence is not even a full sentence yet it is not corrected. It has no verb.

2. In the result I requested the authors to put a confidence interval for the point estimate of adherence level which is 60%. But I don’t think they have got the idea of confidence interval. As long as you determine proportion (which actually is a prevalence) you can always determine the confidence interval for it. You just have to check the boostrap in the SPSS while you run to determine the prevalence (Proportion). If you want to use STATA, you have a total of 200 children with good adherence therefore using the command (cii 333 200) you can find the point estimate to be 60% with 95% CI of (54.58, 65.36).

3. In the previous comment I have also asked the authors to put the age range for caregivers/Mothers instead of putting < 25 for the case may not be the same for very young or teen mothers than the others. It will also has implication in terms of showing the effect of teen pregnancy or motherhood. All the associated factors do not have the same direction but the statement in the result section is closed as ” … less likely to be associated” this is not appropriate even if all are negatively associated the phrase “less likely associated does not give sense”. All these comments were previously given but no improvement is made or no explanation is given.

4. The conclusion given is not based on the results. For example WHO stage of the disease was not significantly associated factor, yet Adherence counseling is recommended for children with advanced disease. You do not just give recommendations by your idea of what should be done, it must be based on the evidence/finding you have.

MAIN PART OF THE MANUSCRIPT

5. In the sample size calculation you should put reference for the study you took the prevalence (24.6%) from. The allocation of study participants in each clinic looks proportional. Therefore it is better put as “The sample was proportionally allocated to the three clinics/hospitals” than quota.

6. In the variables of the study the authors write the Dependent variable to be Adherence whereas the associated factors to be independent variables. How could you just say the associated factors are independent variables?? This is just telling the other name for independent variables. What you are asked is to list the independent variables considered in this study. This comment was previously given as well.

7. In the data processing and analysis section the question of goodness of fit was raised but it was not addressed. P-value tells us if the variable is significantly associated with the outcome, it does not tell us if the model used appropriate for the data. We use the Hosmer-Lemshow goodness of fit test to show this in logistic regression. This was not addressed.

8. In the result section: if the sample size is 316 why 333 people are included in the study? The table showing the multivariable outcome (Table 6) should also include the crude odds ratios.

**********

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: No

Reviewer #2: No

Reviewer #3: Yes: Adhanom Gebreegziabher Baraki

[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.

8 Jul 2021

THANKYOU ALL FOR YR VALID COMMENTS.

REVIEWER 1:

1. THANKS FOR YOUR COMMENT, IT WAS PREVISOULSY COMMENTED I SHOULD REMOVE THE INTRODUCTORY COMMENT AND JUMP STRAIGHT TO DISCUSSING MY RESULTS AND THUS I HAD DELETED IT. HOWEVER, I HAVE AGAIN INCLUDED A SHORT INTROCUTORY COMMENT AS SUGGESTED IN THE DISCUSSION SECTION.

REVEIWER 3:

THANKYOU FOR YR COMMENTS AND DETAILED CLARIFICATION, I HAVE TRIED TO CLARIFY AND ANSWER TO MY BEST CAPABILITIES.

1. IT WAS AN OVERSIGHT ON MY PART THAT THE VERB WAS MISSING IN THE METHOD SECTION OF THE MANUSCRIPT, I HAVE NOW INCLUDED IT AND HAVE TRIED MY BEST TO PERFECT THE ENGLISH IN THE BEST OF MY ABILITIES.

2. CONFIDENCE INTERVAL FOR THE GOOD ADHERENCE: I HAVE ADDED IT, CALCULATED IT FROM SPSS AND PUT IT IN THE RESULT SECTION.

3. AGE RANGE<25 CAREGIVERS, WERE 17-25 YEARS, OF WHICH ONLY 7 PARTICIPANTS (2.7%) WERE 17-19 YEARS, THE REST ( 15, ie: 5.4%) WERE 20-24 YEARS.

4. LOWER CLINICAL STAGE (1-2) AS COMPARED TO ADVANCED STAGES (3-4) WERE SEEN TO BE LESS LIKELY ASSOCIATED WITH POOR ADHERENCE. IF YOU RUN THE STATISTICS VICERVERSA IT WILL SHOW THAT ADVANCED STAGES WILL MORE LIKELY BE ASSOCOATED WITH POOR ADHERENCE. THUS, IT SEEMS FAIR TO CONCLUDE THAT CHILDREN WITH ADVANCED STAGES NEED MORE ADHERENCE COUNSELLING TO MY UNDERSTANDING. I WAS TRYING TO KEEP THE LANGUAGE STANDARD WHERE WE COMPARE THE FACTORS AGAINST POOR ADHERENCE RATHER THAN SOME FACTORS AGAINST POOR ADHERENCE AND SOME AGAINST GOOD ADHERENCE.IT WOULD SOUND INAPPROPRIATE TO SAY LOWER CLINICAL STAGE WAS MORE LIKELY ASSOCOATED WITH GOOD ADHERENCE AS THE PREVIOUS FACTORS WERE DESCRIBED IN RELATION TO POOR ADHERENCE.

5. REFERENCE FOR STUDY FOR SAMPLE SIZE- THAT HAS BEEN STATED IN THE MAIN PART OF THE MANUSCRIPT WITH THE REFERENCE GIVEN. IN REGARDS TO THE SAMPLE SIZE ALLOCATION- ALTHOUGH IT SEEMS LIKE THE ALLOCATION OF PARTICIPANTS WERE PROPORTIONAL, IT WAS STILL QUOTA SAMPLING THAT WAS PERFORMED TO GET EXACT NUMBERS AS EXPLAINED. IT WILL RAISE A QUESTION OF WHY THE SAMPLE WAS NOT EXACTLY IDENTICAL IF WE MENTION THE SAMPLE WAS PROPORTIONALLY ALLOCATED. DUE TO THE ALMOST NEAR NUMBER OF PARTICIPANTS REGISTERED FROM EACH CLINIC SITE, IT TURNED OUT TO BE ALMOST PROPORTIONAL.

6. INDEPENDENT VARIABLES: THE INDEPENDENT VARIABLE BY DEFINITION IS THE VARIABLE THE EXPERIMENTER MANIPULATES OR CHANGES AND IS ASSUMED TO HAVE A DIRECT EFFECT ON THE DEPENDANT VARIABLE. IN THIS CASE OUT DEPENDANT VARIABLE OR OUTCOME IS THE ADHERENCE. SO, WHAT INFLUENCES THE ADHERENCE IN THIS CASE ARE THE FACTORS (DERIVED FROM LITERATURE REVIEW) THAT MAY OR MAY NOT CHANGE THE OUTCOME. YES, NOT ALL ASSOCIATED FACTORS ARE INDEPENDENT VARIABLES BUT, IN THIS STUDY ALL THESE FACTORS ARE CAPABLE OF AFFECTING THE OUTCOME VARIABLE. THE AUTHORS REQUEST THE REVIEWER TO GIVE US A FURTHER EXPLANATION CONTRARY TO THIS AND EXPLAIN WHAT HE UNDERSTANDS BY THE TERM INDEPENDENT VARIABLES. THANKYOU.

7. GOODNESS OF FIT

8. SAMPLE SIZE WAS 316, WHICH IS THE MINIMAL SAMPLE SIZE CALCULATED, HOWEVER WE MANAGED TO COLLECT DATA FOR MORE CHILDREN (313) WITHIN THE TIME FRAME AND COULD NOT REFUSE THE PATIENTS TO ENROLL IN THE STUDY. THUS, INCLUDED THEM THE ANALYSIS. IN ADDITION, WE HAVE NOT EXCEEDED THE 10% OF THE ALLOWANCE FOR THE SAMPLE SIZE. HAVING LESS PARTICIPANTS WOULD WEAKEN THE POWER OF THE STUDY, REACHING AND EXCEEDING SAMPLE SIZE INCREASES THE POWER OF THE STUDY.

9. CRUDE ODDS RATIO: INCLUDED AS REQUESTED

Submitted filename: RESPONSE TO REVEIWERS.docx

Click here for additional data file.

1 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,

Satya Surbhi, PhD

Academic Editor

PLOS ONE

Journal Requirements:

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

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 #2: All comments have been addressed

Reviewer #3: (No Response)

**********

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 #2: Yes

Reviewer #3: Partly

**********

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

Reviewer #2: Yes

Reviewer #3: 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 #2: Yes

Reviewer #3: No

**********

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 #2: Yes

Reviewer #3: No

**********

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 #2: (No Response)

Reviewer #3: Thank you, all the authors, for making the requested revisions. I believe you have made improvement from R1. But I have some issues I asked but not well addressed. Here are my comments and questions.

General Comment:

I think it could be better if the authors give the manuscript to a native English speaker or language expert to do proof reading. The overall quality of the language used is still poor.

2. You added the confidence interval in the result section of the main manuscript only put it in the abstract too.

3. The age range 17-25 years was mentioned in the table but not in the other parts of the manuscript like the abstract. When you do corrections in one part try to make it consistent across all parts of the manuscript.

4. In the result section of the abstract. All the associated factors do not have the same direction but the statement in the result section is closed as ” … less likely to be associated” this is not appropriate even if all are negatively associated. The phrase “less likely associated” is not the right way to interpret odds ratio. I recommend reading more on interpretation of Odds ratios.

Main manuscript:

METHOD

5. I am afraid the authors get the philosophy behind quota sampling. The participants are taken from each institution either proportionally or in equal number in this study but not using quota sampling. Quota sampling is a non-probability sampling, and the participants are selected based on specific qualities and traits. There is no such considerations here.

6. The definition of what dependent variable is and what independent variable is not needed here. You just have to mention what the dependent variable is and what independent variables are considered in this study. For instance you can say: The dependent variable in this study was Adherence where as the independent variables were socio-demographic characteristics like age, gender, marital status . . .; Treatment related factors like type of regimen, duration on ART. . .

7. The authors just write the phrase “Goodness of fit” in their response about how the goodness of fit of the model was addressed. In the result section it says 0.243 for the P-value of the model fitness test. I recommend to mention Hosmer-Lemshow goodness of fit was used to test the association.

8. Including sample of study participants above the minimum may not have weaken the study. Actually, it increases the power. The thing is participants are not supposed to be included just because they wanted to participate. This is why we needed the sample size and sampling procedure. 10% is also to replace non-responders and it is considered in the final sample size; It is not in addition to the final sample size you have. The authors need to give scientific justification as this may affect rigor in conducting research and its replicability.

9. The final multi-variable table shall have all the sections provided below

Variable | Adherence (frequency and percentage for both good and poor) | COR(95% CI) | AOR (95% CI)

There is also no need to write P-values for COR and AOR. P-value for AOR is enough or you can use * for P-value less than 0.05 and mention this as a footnote.

**********

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 #2: No

Reviewer #3: Yes: Adhanom Gebreegziabher Baraki

[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.

26 Jan 2022

I am thankful to the entire team for their comments and contributions to this manuscript.

These are the responses to the questions posed by Adhanom Gebreegziabher Baraki.

1) I have sent the work to a native English speaker to proofread. I hope it is satisfactory as it stands.

2/3) I have added the confidence interval for adherence and changed the age range 17-25 in the abstract and throughout the manuscript.

4) Pertaining to the direction to interpretation of the negatively associated odds ratio, we have changed the reference variable and made it a positive odds ratio with a more harmonized interpretation. Thus made a change to those 2 variables ie: belief in ART effectiveness and clinical stage variables.

From my reading, odds ratio may be interpreted as odds of having an event occurring was…….. times more or less likely in the exposed group than in the unexposed.

5) Regarding the sampling procedure I have understood what you meant, I have rectified it to Proportion to size sampling rather than quota sampling and explained the technique used in the manuscript.

6) I have indicated the dependent and independent variables considered in this study as requested.

7) In regards to Goodness of fit, I had specified in the results section the Hosmer-Lemshow goodness of fit was used. However I have rephrased it again as requested.

8) Regarding the sample size, I understand that participants are not supposed to be included because they wanted to participate. I apologize for the over enrollment that took place as we were two interviewers collecting simultaneously at different sites, So that led to over recruitment due to lack of communication. Once we collected the data, we felt it unwise to discard the data. Thankyou for pointing that out.

9) The final multivariate table has been re-visited and all the variables required in cooperated.

Submitted filename: response to reviewers dec 2021.docx

Click here for additional data file.

16 Aug 2022

Submitted filename: Review Report.docx

Click here for additional data file.

5 Sep 2022

Dear editor and reviewer, Thankyou for taking up this review process again. As you may have already known, this work had gone through an extensive review process over the past one and half year, I’m not sure which version of the work you have finally reviewed as some of the comments had already been addressed and edited in the last version. However, I have tried to answer your valid comments in the best of my capacity. Im proud to claim that my work has been validated by 6 reveiwers and 2 editors!!!

Reviewer #2’comments

Comment 1: Better to put AOR which are significant (i.e., p<=0.05) instead of OR as final results in the results section for factors-

Response : have addressed the suggested valid comments in the manuscript.

Comment 2: In the conclusion section, factors contributing to low adherence from Table 6 must be mentioned (like age of care giver, child inter-current illness, believes on effectiveness of ART, and clinical stage).

Response: it was already mentioned in the conclusion of the discussion section but was missing in the abstract, I have now included it.

Comment 3: Usually, sample sizes are denoted by n and population by N. So, better to denote your sample size by n.

Resposnse: Have denoted the sample sizes as n

Reviewer #4 comments

comment 1: Study Design and study area:

Reference was made to National guidelines. It is useful to state the country of origin of the guidelines. Is this the Tanzanian National guidelines? Please clarify.

Response: Yes, I meant the Tanzanian National guidelines for HIV/AIDS. I have clarified it in the document.

comment 2: Methodology is silent on the specific factors affecting adherence only to surface in the results. Unless this is the journal preference, mention should be made in methodology of the specific factors affecting ART adherence.

Response: This comment was already addressed in the revised latest version which I assume you may not have seen, I have highlighted in the document again.

Comment 3: The description of the quota sampling technique was inadequate. This requires more detail to understand the populations of the patients at the CTC sites and how representative the sampled research participants are of the CTC sites. Given the drawbacks of Quota Sampling methodology – Non-probability sampling, researcher’s bias in selection of study participants and limitations with generalization of study findings, a lot more detail would be required. Was the sampling controlled or not controlled for instance.

Response: this also was addressed in the previous review and after discussion, it was concluded that it was not a quota sampling (a misnomer) but rather a proportion to size sampling technique based on the number of patients enrolled in the specific site in order to obtain a representative yet unbiased sample based on the site of recruitment, where participants where then consecutively recruited till the number was reached from each site and eventually the required sample size. I added a better clarification in the manuscript.

Comment 4: Results

Socio-demographic factors affecting adherence: a. Child age grouping of 1-9 years and 10-14 years lumps children with different needs together. The Under-five children age group is entirely dependent on the caregiver for medication administration as opposed to the older age groups. School-age children attending day school would have different challenges when compared with children in boarding school. Combining all these children together into just two groups with different needs may hide information that is inimical to adherence. Consider age groups of U5, 6-10, 11-14.

b. Similarly, parental grouping could be further revised to Single parents, both parents, grandparents and other caregivers. The challenges for each group differ from the other and may impact differently on adherence.

c. Care giver marital status also appears to be lumped together. Married, cohabiting, single, Divorced, widowed caregivers could be assessed separately. Married and cohabiting on one hand and single divorced and widowed caregivers on the other hand may not have the same challenges. It is possible that evaluating each group may be more informative that combining them.

Response: Thank you for this comment, in regards to the child age group we had considered this among many other sub-categorization methods, however during our data collection process we found that under 9 years in our setting were totally dependent on the caregiver for their overall needs and HIV care specifically such that even drugs were being administered solely by the caregiver, even boarding school options were not there for this category. Thus, we deemed it wise to categorize them as we have done due to the similarities. Generally, in this setting children below 10 years are still considered to be totally dependent on their caregivers.

In regards to parental status, we thought it would be wise to look into whether a parental presence would affect the outcome of interest, whether one or both or none as we know a parent would strive hard for their own child as compared to others even if it is a single parent. In our setting, it is quite common to have different partners, married or not at different times, so having another partner instead of the child’s own another parent would not change the outcome and it might be culturally not acceptable to ask if you are still living with the father/mother of the index child, especially if they had no short-term or long-term relationship which is very customary here. Similarly, we opted to group the marital status in a culturally unoffending way as married/cohabiting as we know atleast they have a partner support versus the single/divorced/widowed with no partner support. Cohabiting is culturally accepted here and it may be hostile to ask them whether they’re living together out of wedlock or they have actually married each other.

Conclusion:

Comment 5. Pill burden

Specify the type of pill burden, high or low.

Response: have specified it as high, thankyou.

Recommendation:

comment 6. Adherence counseling: The authors’ results show that there is no association between lack of adherence counseling and poor ART adherence and yet recommended the necessity of counseling. This needs clarification.

Response: ART adherence was still found to be sub-optimal at around 60%, the most important strategy to improve ART adherence is to strengthen counselling, which in our study is reported to be done, however, may still need to be reinforced in the high risk groups like younger aged caregivers, sick children as these children are still dependant of their caregivers. Younger aged caregivers being less mature than their counterparts may need more support and education. Sick children and those with advanced HIV stages may need to be admitted and have closer follow-ups with their clinicians to improve their condition and improve adherence. Health talks during clinic visits and on social media platforms may improve belief and alleviate misconceptions in the ARTs. There’s no single way to predict patients’ adherence or nonadherence status and it has been seen that adherence is always better earlier in the disease in patients who are more motivate and less fatigued, it is ongoing adherence measurements and early identification of faltering adherence and continuing counselling that may be able to bring it back up. Thus these may not be separable and is vital for ART adherence success.

Reviewer #5:

comment 1: Line 4 of the 2nd paragraph of the introduction says 1.7 million children (1-14 years) were living with HIV and 1.9 million of them were in sub-Saharan Africa. I think this is a typo as the number is about 1.09 million (64% of the 1.7 million).

Response: Amends has been made, thanks.

comment 2. In statistical analysis, it is mentioned that measures of central tendency were used to summarize discrete data; the only measure of central tendency that can be applied to discrete data is mode and this should be stated if it was used.

Response: Age in this study was taken in years not considering the months, and median was applied to it to find the median age of children as our age outliers. Mode was used by default for all categorical data.

comment 3. In the results section: for age groups, 25 years appears in 2 strata: 17-25 years and 25-34 years. This should be appropriately revised; either 17 -<25 years and 25-35 years, OR 17-25 years and >25 - 34 years.

This has been addressed and changed accordingly, thank you.

Submitted filename: RESPONSE TO REVEIWERS.docx

Click here for additional data file.

19 Sep 2022

MAGNITUDE AND ASSOCIATED FACTORS OF ANTI-RETROVIRALTHERAPY ADHERENCE AMONG CHILDREN ATTENDING HIV CARE AND TREATMENT CLINICS IN DAR ES SALAAM, TANZANIA

PONE-D-20-25583R4

Dear Dr. Mussa,

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,

Chika Kingsley Onwuamah, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

21 Sep 2022

PONE-D-20-25583R4

Magnitude and associated factors of anti-retroviral therapy adherence among children attending HIV care and treatment clinics in Dar es Salaam, Tanzania

Dear Dr. Mussa:

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. Chika Kingsley Onwuamah

Academic Editor

PLOS ONE