Costs of implementing universal test and treat in three correctional facilities in South Africa and Zambia.

INTRODUCTION: Universal test and treat (UTT) is a population-based strategy that aims to ensure widespread HIV testing and rapid antiretroviral therapy (ART) for all who have tested positive regardless of CD4 count to decrease HIV incidence and improve health outcomes. Little is known about the specific resources required to implement UTT in correctional facilities for incarcerated people. The primary aim of this study was to describe the resources used to implement UTT and to provide detailed costing to inform UTT scale-up in similar settings.
METHODS: The costing study was a cross-sectional descriptive study conducted in three correctional complexes, Johannesburg Correctional Facility in Johannesburg (>4000 inmates) South Africa, and Brandvlei (~3000 inmates), South Africa and Lusaka Central (~1400 inmates), Zambia. Costing was determined through a survey conducted between September and December 2017 that identified materials and labour used for three separate components of UTT: HIV testing services (HTS), ART initiation, and ART maintenance. Our study participants were staff working in the correctional facilities involved in any activity related to UTT implementation. Unit costs were reported as cost per client served while total costs were reported for all clients seen over a 12-month period.
RESULTS: The cost of HIV testing services (HTS) per client was $ 92.12 at Brandvlei, $ 73.82 at Johannesburg, and $ 65.15 at Lusaka. The largest cost driver for HIV testing at Brandvlei were staff costs at 55.6% of the total cost, while at Johannesburg (56.5%) and Lusaka (86.6%) supplies were the largest contributor. The cost per client initiated on ART was $917 for Brandvlei, $421.8 for Johannesburg, and $252.1 for Lusaka. The activity cost drivers were adherence counselling at Brandvlei (59%), and at Johannesburg and Lusaka it was the actual ART initiation at 75.6% and 75.8%, respectively. The annual unit cost for ART maintenance was $2,640.6 for Brandvlei, $710 for Johannesburg, and $385.5 for Lusaka. The activity cost drivers for all three facilities were side effect monitoring, and initiation of isoniazid preventive treatment (IPT), cotrimoxazole, and fluconazole, with this comprising 44.7% of the total cost at Brandvlei, 88.9% at Johannesburg, and 50.5% at Lusaka.
CONCLUSION: Given the needs of this population, the opportunity to reach inmates at high risk for HIV, and overall national and global 95-95-95 goals, the UTT policies for incarcerated individuals are of vital importance. Our findings provide comparator costing data and highlight key drivers of UTT cost by facility.

Introduction

Universal test and treat (UTT) for HIV is a population-based strategy that aims to ensure widespread HIV testing and rapid antiretroviral therapy (ART) for all who have tested HIV-positive regardless of CD4 count to decrease HIV incidence and improve health outcomes. UTT has been tested in several settings and adopted as part of national HIV programmes in many countries, including both South Africa and Zambia. UTT was introduced in South Africa in September 2016 and recommended same-day ART initiation from October 2017 [1, 2]. Prior to UTT, one retrospective study done using programmatic data in South Africa described viral load suppression at six and 12 months after ART initiation at 94.7% and 92.5% respectively for incarcerated people with HIV treated in an on-site ART programme [3]. A cross sectional study done in Malawi also prior to UTT, found 95% viral suppression among a sample of incarcerated people with HIV who received ART for at least 6 months in a prison clinic [4]. The TasP study conducted in South Africa and Zambia showed that it is feasible to implement UTT in diverse southern African correctional settings [5]. The study observed 86% ART uptake among incarcerated participants, many of whom initiated ART on the same day as HIV testing, and retention in care and viral load suppression exceeding 90% at 6 months post-ART initiation for individuals who remained incarcerated [5].

Current studies examining cost for HIV care typically include general populations; little is known about the specific resources required to implement UTT in correctional facilities [6-8]. A modelling study based on general population data from South Africa (SA) in 2014 reported that universal treatment averted more HIV infections, but had a higher total cost of about $320 per quality adjusted life year (QALY) gained compared to the status quo (providing ART based on an initiation threshold of CD4 count ≤350 cells/mm3), which had a total cost of around $290 per QALY gained [9]. Another modelling study (2014) based on data from SA, Zambia, Vietnam and India reported that UTT met cost-effectiveness thresholds [10]. In SA, the cost per Disability Adjusted Life Year (DALY) averted for changing eligibility to all HIV-positive adults compared with eligibility for those with CD4 counts of ≤350 cells per μL ranged from $438 to $3790 in seven models used while in Zambia results showed $790 per DALY in four mathematical models [10].

The primary aim of this study was to describe the resources used to implement UTT in correctional settings as part of the Treatment as Prevention (TasP) study [5]. TasP was an implementation research study designed to assess the feasibility of UTT delivery in three correctional complexes, two in SA and one in Zambia. Through this study, we aimed to provide costing to inform UTT scale-up in similar settings.

Methods

Details on how inmates at each level of the HIV cascade engaged with HIV self-services have been detailed elsewhere [5]. The TasP study provided voluntary universal HCT at facility entry, and through HCT campaigns to reach all inmates at least once annually and offered ART to all inmates who tested HIV-positive. The TasP study encouraged increased HIV services uptake and assured universal HCT and ART access. The TasP study augmented ART care, offering ART initiation to all inmates (after screening for TB and kidney disease) regardless of CD4 count.

Study design and settings

TasP study findings on clinical and select implementation outcomes have been reported previously [5, 11]. The costing component of the TasP study was a cross-sectional descriptive study to describe the resources needed to implement UTT in two correctional facilities in South Africa: Johannesburg Correctional Facility in Johannesburg (>4000 inmates), and Brandvlei Correctional Facility in Western Cape (~3000 inmates) and Lusaka Central in Lusaka, Zambia (~1400 inmates). The three correctional complexes collectively encompassed ten correctional units, which included six male units (two maximum security), three female units, and one youth unit (for people aged 18–22 years) [5]. TasP sites were selected purposively to reflect a range of security levels and sociodemographic characteristics of incarcerated people in southern Africa [5]. Each site had routine health services available on site, including HIV testing and treatment [5].

In all three facilities there were specific structured management systems in place prior to initiation of the UTT. In Brandvlei and Johannesburg, unit managers managed doctors, nurses, and peer educators providing services in their clinical area. Each unit manager reported to a HIV/AIDS co-ordinator who reported to the health service manager. All health service managers reported to a central area co-ordinator who had responsibility for a management area that included several correctional facilities. The rationale for the staffing in SA was based on the national policy regarding staff per facility rather than on need.

Lusaka had fewer managerial layers but had a clinic manager providing line management to doctors, nurses, and peer educators (Table 1).

Table 1

Staff involved in HTS at correctional facilities.

Brandvlei Correctional		Johannesburg Correctional		Lusaka Central
Cadre	Number	Cadre	Number	Cadre	Number
Operations manager	5	Health manager	1	Offender manager	1
Head of Centre	1	Clinic in charge	2	Clinic in charge	1
HIV coordinator	3	HIV coordinator	1	HIV coordinator	1
Medical doctor	1	Medical doctor	2a	Clinical officer	4
Professional nurse	7	Professional nurse	18	Professional nurse	1
				Enrolled nurse	6
Professional nurse counsellor	1	Adherence counsellor	2	Adherence counsellor	2
		Pharmacist	5	Pharmacist	2
HTS counsellor/Health screener	4	HTS counsellor	10	HTS counsellor/Health screener	5
Officer in charge	1			Officer in charge	1
Case officer	4
		Peer educator	10	Peer educator	10
Security officer	10	Security officer	7	Duty officer	5
Warden	2			Cell captain	32
		Project manager	1	Project manager	1
Project coordinator	1	Project coordinator	1	Project coordinator	1
Data capturers	2	Data manager	1	Data associate	2
		Data monitor	1	Xray/Xpert technician	3
Total staff	42		62		78
Total inmates	3,000		4,000		1,400
Staff: Inmates ratio	1:71		1:65		1:18
Total correctional units	2	Total correctional units	3	Total correctional units	3
Male and Female units co-joined (medium security)	2	Male (maximum security)	1	Male (maximum security)	1
		Male (medium security)	1	Male (medium security)	1
		Female units	1	Youth units	1
Estimated HIV Prevalence	5%		15%		15%
Clients initiated on ART	35		233		229

a First medical doctor at 100% FTE (Full Time Equivalent) and second medical doctor at 60% FTE

Study participants

We included staff working in the correctional facilities involved in any activity related to UTT implementation. Staff cadres included managers, HIV co-ordinators, counsellors, adherence counsellors, medical doctors, professional nurses, peer educators and security officials including those employed by the correctional services, the health departments, and non-governmental organizations (NGOs). We purposively selected all staff within the cadre where there were two or less staff in the role. In cadres where three or more staff performed the same duties, we used convenience sampling where every 2nd person was selected.

Costing methods

Data collection

Costing was determined through a survey conducted between September and December 2017 that we used to identify materials and labour used for three separate components of UTT: HIV testing services (HTS), ART initiation, and ART maintenance. This approach considered resource-use in terms of cost of each item: salaries, equipment, laboratory tests and medications. The other consideration was the overall infrastructure present at the time of program implementation. National and facility health management were in place in both South Africa and Zambia, allowing the UTT program to be placed within an existing health infrastructure. In addition, logistics for supplies, off-site laboratory testing, and testing for incarcerated people were in place. This program augmented staff and supplies for service delivery without a need to develop a management structure, clinical infrastructure, or logistics capabilities.

A standardized interview tool developed by investigators was used to assess four main inputs across the three facilities: staff, equipment, supplies and medical tests. The paper-based tool was administered by trained research assistants in a private setting within the correctional facility and the interview lasted between 30–60 minutes.

Costs were determined through record reviews of invoices and price lists for both South Africa and Zambia. We used a combination of top-down and bottom-up micro-costing approaches to estimate the resources/inputs that were used for UTT. The inputs were costed over a 12-month period. The outcomes of interest were unit and total costs per activity for each facility and were presented from a provider perspective.

Data analysis

For all cadres interviewed, the main activity categories included HIV testing, ART (antiretroviral) initiation and ART maintenance. HIV testing included the following sub-activities, namely: group counselling, pre-test counselling, testing/screening for HIV and counselling after result. ART initiation included: adherence counselling, phlebotomy and initial ART prescription. ART maintenance involved: phlebotomy, educational messages on HIV treatment and adherence and ART maintenance (assessing for side effects of ART, isoniazid preventive therapy, cotrimoxazole, and fluconazole and considering preventive therapies). Allocation of costs for phlebotomy between ART initiation and ART maintenance was based on the proportions of clients seen for each visit type. All time spent on research activities was excluded from the analysis.

Data were entered into REDCap software version 7.6.9 (Vanderbilt University) and analysis was completed using MS Excel (Microsoft Corp, USA, 2003). We calculated unit costs as cost per client served. All costs were converted into the US dollar (USD) using exchange rates on 1 January 2017: 1USD = 13.78 South African Rand (ZAR) for the South African sites and 1USD = 9.90 Zambian Kwacha (ZMW) for the Zambian site. Costing results are reported separately for each facility and as overall cost for care provision and as cost per client served (unit cost). Notably, these costs are not related to a specific payer as the budget for various HIV testing, HIV care, and program costs came from a variety of sources including the grant supporting this project, national health and/or corrections budgets, and donor funding.

Staff. Study team members administered a costing instrument to staff to describe activities and time spent over the preceding work week (Monday to Friday). The survey included questions regarding the amount of the workday spent on an activity, the time spent per client to complete the activity (estimated and self-reported), and all equipment and supplies used during the activity.

To calculate costs over a 12-month period, we used the proportion of time spent on implementation activities and the annual salary (base salary plus fringe benefits) for NGO staff and the median annual salary for corrections staff. Salaries in SA were obtained from the Department of Public Service and Administration and the NGO Finance department for staff employed by the NGO. Salaries in Zambia were obtained from the correctional facility. Staff unit costs were computed by dividing the annual total costs with the number of clients seen over a 12-month period which we obtained from the TasP study (S1 Table) for a period of 17 months, though we based our costs on the average number of clients seen over a 12-month period. For ART maintenance, we excluded incarcerated persons who were released, transferred or died during the study period and costed only those who remained incarcerated and on ART.

Equipment. All staff involved in UTT were asked to state the quantities of all the equipment they used and the activities they performed and to specify whether the equipment was shared and with whom and for what other tasks. Costs were then based on the proportion of time the equipment was used for UTT and the market price or invoices. The resulting costs for the equipment were annuitized using a discount rate of 10.5% for both South Africa and Zambia [12]. The same useful life was used in both countries, with 10 years for examination tables, filing cabinets and weighing scales, and 5 years for computers, chairs, desks and blood pressure machines. To avoid double counting the costs of any shared equipment, cost per minute estimates were derived using the time spent data and applied to the annuitized cost of the equipment. We considered this approach bottom-up since cost per minute estimates were used to calculate the annual cost of the equipment used [13].

Supplies. Costs for supplies used for UTT implementation were determined using the most complete package of items mentioned by the participants as needed for the activity. This was a top-down approach that relied on interviews and item cost abstraction and did not use direct observation. Supplies were categorised as: HIV testing, ART initiation, and ART maintenance. The cost of antiretroviral therapy drugs over the 12-month period was also included. For ART initiation, a one month’s supply of ART was costed since the incarcerated individuals returned after one month for their next supply. For ART maintenance, we costed over the remaining 11 months. ART maintenance costs did not include any additional acute or chronic care medical services clients received, but included HIV lab monitoring costs such as viral load testing.

Medical tests. A standard package of medical tests was included in the costing for both South Africa and Zambia. During ART initiation, clients receive three baseline tests: CD4, Creatinine and rapid plasma reagin (RPR) test. During ART maintenance, all clients on ART are meant to receive the following tests: creatinine (3, 6 & 12 months), HIV RNA Viral Load (6 & 12 months), while an estimated 5% (based on empiric data from these facilities) of clients receive haemoglobin, alanine transferase (ALT) and Xpert MTB/RIF testing for specific HIV care related indications. Unit prices for each test were obtained from local laboratories.

Ethics approval

This study was approved by the Human Research Ethics Committees/Institutional Review Boards of: University of Witwatersrand (South Africa); the University of the Western Cape (South Africa); the University of Zambia (Zambia); the University of North Carolina at Chapel Hill, the University of Alabama at Birmingham, and Johns Hopkins University all in the USA; the London School of Hygiene and Tropical Medicine (UK); and James Cook University (Australia). The study was also approved by the Department of Correctional Services in South Africa and the Zambia Correctional Service. Written informed consent was obtained from all study participants in the study approved languages namely English or Afrikaans for South Africa and English, Nyanja and Bemba for Zambia. No reimbursements or incentives were offered to participants.

Results

For each staff cadre we interviewed 1–5 individuals at each site. A total of 107 staff were interviewed with the following breakdown: in Johannesburg, 37 interviews (enrolment target 37); in Brandvlei, 31 interviews were conducted (target 31); and in Lusaka, 39 interviews were conducted (target 40).

Costing

The cost of HIV testing services (HTS) for one client was $ 92.12 at Brandvlei, $ 73.82 at Johannesburg, and $ 65.15 at Lusaka (Table 2). The cost per client initiated on ART was $917 for Brandvlei, $421.8 for Johannesburg, and $252.1 for Lusaka (Table 3). The annual unit cost for ART maintenance was $2,640.6 for Brandvlei, $710 for Johannesburg, and $385.5 for Lusaka (Table 4). Fig 1 shows a summary of the cost per client for HTS, ART initiation and ART maintenance. In addition, the unit cost of gathering incarcerated people to be tested and escorting them for each clinic visit costs $6.56 in Brandvlei, $1.18 in Johannesburg and $0.77 in Lusaka.

Table 2

Unita and total costsb ($US) of HIV testing services (HTS) at correctional facilities in 2017.

		Brandvlei Correctional				Johannesburg Correctional				Lusaka Central
		Staff	Equipmentc	Suppliesd	Total	Staff	Equipmentc	Suppliesd	Total	Staff	Equipmentc	Suppliesd	Total
Group counselling	Unit cost	11.41	0.03	0.37	11.81	2.78	0.13	0.82	3.73	2.55	0.09	0.52	3.16
		(22.3%)	(8.1%)	(0.9%)	(12.8%)	(8.7%)	(50.0%)	(2.0%)	(5.1%)	(32.2%)	(10.8%)	(0.9%)	(4.9%)
	Total cost	21 685	48	712	22 445	8 507	385	2 496	11 388	6 325	230	1 279	7 834
Pre-test counselling	Unit cost	8.47	0.07	0.04	8.58	8.31	0.04	0.42	8.77	1.58	0.21	7.08	8.87
		(16.5%)	(18.9%)	(0.1%)	(9.3%)	(26.1%)	(15.4%)	(1.0%)	(11.9%)	(20.0%)	(25.3%)	(12.6%)	(13.6%)
	Total cost	16 110	124	85	16 319	25 409	130	1 283	26 822	3 928	523	17 699	22 150
Testing for HIV	Unit cost	25.95	0.26	40.07	66.28	12.46	0.04	40.19	52.69	2.12	0.37	48.79	51.28
		(50.7%)	(70.3%)	(98.9%)	(71.9%)	(39.1%)	(15.4%)	(96.4%)	(71.4%)	(26.8%)	(44.6%)	(86.5%)	(78.7%)
	Total cost	49 338	492	27 627	77 458	38 113	128	54 765	93 006	5 263	914	114 066	120 244
Counselling after result	Unit cost	5.40	0.01	0.04	5.45	8.31	0.05	0.27	8.63	1.66	0.16	0.02	1.84
		(10.5%)	(2.7%)	(0.1%)	(5.9%)	(26.1%)	(19.2%)	(0.6%)	(11.7%)	(21.0%)	(19.3%)	(0.0%)	(2.8%)
	Total cost	10 264	23	69	10 355(	25 409	143	813	26 364	4 119	389	48	4 557
Total	Unit cost	51.23	0.37	40.52	92.12	31.86	0.26	41.70	73.82	7.91	0.83	56.41	65.15
		(55.6%)	(0.4%)	(44.0%)	(100%)	(43.2%)	(0.4%)	(56.5%)	(100%)	(12.1%)	(1.3%)	(86.6%)	(100%)
	Total cost	97 397	686	28 493	126 576	97 437	786	59 357	157 580	19 635	2 057	133 092	154 784

Equipment and

supplies lists are detailed in supporting information S2 and S3 Tables

Table 3

Unit and total costs ($US) of ART initiation at correctional facilities in 2017.

Brandvlei Correctional						Johannesburg Correctional					Lusaka Central
	Staff	Equipmentc	Suppliesd	Medical Tests	Total	Staff	Equipmentc	Suppliesd	Medical Tests	Total	Staff	Equipmentc	Suppliesd	Medical Tests	Total
Adherence counselling	541.9	2.4	0.1		544.4	62.8	0.6	0.5		63.9	11.4	0.4	0.0		11.8
	(68.7%)	(20.7%)	(0.2%)		(59.4%)	(20.7%)	(58.3%)	(2.0%)		(15.1%)	(20.5%)	(30.2%)	(0.00%)		(4.7%)
	22 951	103	3		23 058	28 166	270	204		28 640	4 965	169	0		5 134
Phlebotomy	222.1	2.0	3.4		227.4	39.1	0.4	3.6		43.1	10.0	0.4	38.7		49.1
	(28.2%)	(16.8%)	(58.7%)		(24.8%)	(12.9%)	(35.9%)	(15.8%)		(10.2%)	(18.0%)	(28.7%)	(81.9%)		(19.5%)
	9 405	83	1 138		10 626	17 541	164	1 617		19 322	2 284	162	16 842		19 288
ART initiation	24.8	7.4	18.8	94.2	145.2	201.9	0.1	18.7	94.2	314.8	34.2	0.5	8.6	147.9	191.2
	(3.1%)	(62.5%)	(41.1%)	(100%)	(15.8%)	(66.4%)	(5.8)	(82.1%)	(100%)	(74.6%)	(61.5%)	(41.1%)	(18.1%)	(100%)	(75.8%)
	877	259	665	1 031	2 832	47 027	15	4 358	6 805	58 204	7 808	121	1 956	9 561	19 446
Total	788.8	11.8	22.3	94.2	917.0	303.9	1.0	22.8	94.2	421.8	55.6	1.3	47.3	147.9	252.1
	(86.0%)	(1.5%)	(5.8%)	(11.8%)	(100%)	(72.0%)	(0.2%)	(5.4%)	(22.3%)	(100%)	(22.0%)	(0.5%)	(18.8%)	(58.7%)	(100%)
	33 232	446	1 807 1 031	36 516	92 734	449	6 179	6 805	106 167	15 056	452	18 798	9 561	43 868

Unit costs are showed on the first row of each activity while

total costs are shown on the second row

Equipment and

supplies lists are detailed in supporting information S2 and S3 Tables

Table 4

Unit and total costs ($US) of ART maintenance at correctional facilities in 2017.

Brandvlei Correctional						Johannesburg Correctional					Lusaka Central
	Staff	Equipmentc	Suppliesd	Medical Tests	Total	Staff	Equipmentc	Suppliesd	Medical Tests	Total	Staff	Equipmentc	Suppliesd	Medical Tests	Total
Phlebotomy	137.7	1.2	13.4		152.3	24.2	0.6	14.4		39.2	3.6	0.1	155.0		158.7
	(5.9%)	(9.9%)	(5.5%)		(5.8%)	(6.4%)	(7.6%)	(5.1%)		(5.5%)	(6.1%)	(4.0%)	(62.1%)		(41.2%)
	5 830.8	51.7	294.1		6 176.6	10 843.7	101.3	2 078.1		13 023.0	817.5	58.0	12 679.3		13 554.8
Educational messages	1 301.1	4.8	2.7		1 308.6	36.7	1.9	0.8		39.3	30.9	0.8	0.5		32.2
	(55.4%)	(39.1%)	(1.1%)		(49.6%)	(9.6%)	(24.0%)	(0.3%)		(5.5%)	(52.4%)	(23.5%)	(0.2%)		(8.4%)
	45 921.2	221.3	59.0		46 201.4	8 537.7.1	432.2	117.5		9 087.4	7 072.9	175.9	42.2		7 291.0
ART maintenance	908.1	6.3	228.3	36.9	1 179.6	319.8	5.3	269.4	36.9	631.4	24.6	2.4	94.1	73.5	194.6
	(38.7%)	(51.0%)	(93.4%)	(100%)	(44.7%)	(84.0%)	(68.4%)	(94.6%)	(100%)	(88.9%)	(41.6%)	(72.6%)	(37.7%)	(100%)	(50.5%)
	19 871.6	425.3	4 996.4	2 038.8	27 332.0	46 056.6	762.6	38 788.6	13 455.8	99 063.6	2 011.6	194.6	7 700.8	21 223.0	31 130.0
Total	2 346.9	12.3	244.5	36.9	2 640.6	380.7	7.8	284.6	36.9	710.0	59.1	3.3	249.6	73.5	385.5
	(88.9%)	(0.5%)	(9.3%)	(1.4%)	(100%)	(53.6%)	(1.1%)	(40.1%)	(5.2%)	(100%)	(15.3%)	(0.9%)	(64.8%)	(19.1%)	(100%)
	71 623.6	698.2	5 349.5	2 038.8	79 710.0	65 438.0	1 296.0	40 984.2	13 455.8	121 174.0	9 901.9	428.5	20 422.2	21 223.0	51 975.7

Unit costs are showed on the first row of each activity while

total costs are shown on the second row

Equipment and

supplies lists are detailed in supporting information S2 and S3 Tables

Fig 1

Summary of overall cost per patient for HIV testing services, ART initiation and ART maintenance.

The largest cost driver for HIV testing at Brandvlei were staff costs at 55.6% of the total cost, while at Johannesburg and Lusaka, supplies were the largest contributor at 56.5% and 86.6% of the total cost, respectively, due to the greater numbers of clients served (S1 Table). The process of conducting the two HIV tests per inmate.

For ART initiation, the main input cost drivers were staff costs at both Brandvlei (86%) and Johannesburg (72%), while baseline medical laboratory testing were the largest at Lusaka (58.7%). The activity cost driver was adherence counselling at Brandvlei (59%), whereas at Johannesburg and Lusaka it was the initial ART prescription at 75.6% and 75.8% respectively.

For ART maintenance, the main input cost drivers were staff costs at Brandvlei (88.9%) and Johannesburg (53.6%) and supplies at Lusaka (64.8%). The activity cost drivers at all three facilities were side effect monitoring and initiation of Isoniazid Preventive Therapy (IPT), cotrimoxazole, and fluconazole, with this comprising 44.7% of the total cost at Brandvlei, 88.9% at Johannesburg, and 50.5% at Lusaka.

Discussion

In this descriptive costing study, we provide cost estimates for delivering HTS, ART initiation, and ART maintenance for UTT implementation in three diverse correctional settings in Zambia and South Africa. Notably, we observed substantial variation in cost across these three correctional complexes which was largely related to the ratio of health staff to the inmate population, and, to a lesser extent, to variation in staff salary between South Africa and Zambia. We believe that these findings can help guide considerations for resource allocation for implementing UTT in similar correctional settings.

The costs that we estimated are higher than some prior reports for HIV care in the community [6, 7]. For example, HTS during PMTCT in South Africa has been estimated to cost between $31 and $38 and, in Zambia, $19 per client tested [6, 7]. This is in contrast to the unit costs we found for HTS in correctional settings, which ranged from $65.15 in Zambia to $73.82 in Johannesburg and $92.12 in Brandvlei in South Africa. There are several reasons why these costs are likely higher than community health facility HTS costs. First, additional resources are required for security within a correctional facility environment due to the complexities of transferring incarcerated persons from cell blocks to the clinic setting. Correctional wardens and other corrections cadres were involved in HTS to provide security while HIV testing or counselling was delivered by nurses and counsellors. Second, prior HTS costing has generally focused on the clinic setting, where personnel providing HTS may be multi-tasking leading to only the time spent on an HTS session contributing to the HTS cost. This contrasts with HTS screening campaigns or HTS among newly arriving corrections inmates. In the correctional facility setting, HTS personnel may need to wait for additional inmates to be brought to queue or spend set-up time and wait time to provide HTS to a small number of newly incarcerated individuals. Similar decreased efficiency has been reported from community-based and other non-facility HTS settings [14]. The advantage of HTS in the correctional setting is reaching a greater number of people living with HIV, and particularly men and key populations disproportionately affected by HIV/AIDS, many of whom may not visit a health facility when not incarcerated [15].

The cost of ART was also higher in the correctional facility-setting for Brandvlei and Lusaka than a prior report from community clinics in those countries. The annual cost of ART care (not specifying initiation or maintenance) was estimated to be $682 in South Africa and $278 in Zambia [8]. This contrasts to $2,640.6 for Brandvlei, $710 for Johannesburg, and $385.5 for Lusaka. The higher cost at Brandvlei is related to a small volume of ART clients and higher cadres of staff involved in the various components of HIV testing and ART delivery (nurses rather than counsellors). Security requirements were an additional cost as corrections wardens were used to escort clients to and from the clinic and to monitor incarcerated persons while at the clinic. In contrast, standard clinics in the community have minimal or no security and security personnel do not accompany clients to and from places of residence. The cost per client could plausibly decline as the HTS and treatment program matures and ways to achieve greater efficiency are identified such as task shifting to peer educators. However, multiple elements increase costs in the correctional facility, including limited available hours of contact with incarcerated individuals due to a schedule of when cellblocks are locked down, the need for security escorts, and the staffing numbers based on overall correctional services policies for each facility rather than the workload.

This study had several limitations. Many of the corrections staff involved in UTT have other responsibilities outside of HIV service delivery making it difficult to estimate time spent on HIV service delivery alone. Furthermore, estimation of time spent on UTT activities depended on self-report. Some of the material resources required for UTT were shared with other health services, so we relied on self-report by staff members to assign the proportion of the resource that went to UTT. Unit cost (cost per client) depended on an accurate denominator of the number of clients who received the service for those inputs that were costed top-down. Lastly, the findings need to be considered in context of the correctional facility size, number of people living with HIV, and staffing level of health personnel in the facility. We believe that when considering these factors, reasonable assumptions can be made to apply these findings to other facilities in South Africa or Zambia. However, the recruitment of participants from high and low volume facilities with varied human resource capabilities and HIV prevalence was likely generally representative of correctional facilities in southern Africa.

Conclusion

As a rule of thumb, reaching the last 20% of individuals for a service is estimated to cost as much as reaching the first 80% [16, 17]. Efforts to reach those who are unreached with HIV testing and ART should include correctional facilities and incarcerated people, despite higher costs. Given the needs of the incarcerated population, the opportunity to reach men at high risk for HIV, and overall national and global 95-95-95 goals, we believe that the cost of providing scaled-up universal test and treatment is worth the potential returns for the health of incarcerated people and the broader community.

Service statistics over a 12-month period at correctional facilities in 2017.

(DOCX)

Click here for additional data file.

Equipment lists for Brandvlei, Johannesburg and Lusaka correctional facilities.

(DOCX)

Click here for additional data file.

Supplies lists for Brandvlei, Johannesburg and Lusaka correctional facilities.

(DOCX)

Click here for additional data file.

Datasets for Brandvlei, Johannesburg and Lusaka correctional facilities.

(ZIP)

Click here for additional data file.

11 Apr 2022

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,

Elizabeth S. Mayne, M.D.

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

Additional Editor Comments:

This paper shows potential given the limited data on health costings in HIV treatment. The reviewers had some concerns however regarding the failure to acknowledge that universal test and treat is a mandate currently. The second reviewer has also recommended that the authors adhere to the CHEERS checklist for health economics studies and also that the tables are clarified.

[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

**********

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

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

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

Reviewer #1: No

Reviewer #2: Yes

**********

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

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

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

Reviewer #1: The authors report results from a study that explored the costs of implementing UTT in three correctional facilities, 2 in South Africa and 1 in Zambia. This is an interesting paper given the dearth of costing information particularly related to HIV health-care provision in correctional facilities. The paper is well written and provides a useful perspective although I feel there are some areas that need further expansion. In addition, UTT is now widely rolled out across South Africa and the paper would benefit from further information on how these data will help given that UTT is now standard of care in most health-facilities and whether this is also the case in Zambia and across correctional facilities. I would recommend that references to some of the more recent work on UTT implementation be reviewed and included as I am certain that more has been published in the last five years.

Comments and suggestions:

Line 78 – Introduction – it should be made clear that UTT has now been adopted in both South Africa and Zambia. In fact UTT was introduced in South Africa in September 2016 and recommended same-day ART initiation from October 2017.

• SA-NDOH circular – Implementation of Universal test and treat strategy for HIV positive patients and differentiated care for stable patients. South Africa National Department of Health 2016

• SA NDoH Fast tracking implementation of the 90-90-90 strategy for HIV through implementation of the test and treat (TT) poicy and the sam-day ART initiation for positive patients, 2017

Lines 80-82 – can the authors include stats or what ART uptake, viral load suppression and retention were prior to UTT?

Lines 93-95 – can the authors clarify that these data are from that same reference number 6 and that in this context they are referring to mathematical models that are used not models of care.

Line 104 and paragraph – Methods – one piece of information that would be helpful to include here is a brief description of how inmates at each level of the HIV cascade engage with HIV self-services

Line 127 – Table 1 – it would be helpful if this table could be expanded a bit further to include additional information about each faciity e.g. to incorporate the number of correctional units in those facilities and how these break down into male and female units, maximum security and youth. It might also be helpful to know some of the HIV data for each facility as well if it is known. I had missed the total inmates and staff: inmate ratio rows and wonder if those should not be taken to the top or highlighted in some way

I am not sure if I missed it but it would be helpful if there was supplementary material that also present the various staff costs which might give more context to the differences between facilities and countries. This would provide important context in the results as currently it is difficult to understand Table 2, 3 and 4 where despite the units being quite similar the costs are radically different.

Line 163 – it is not clear what the “(i.e., Correctional services”) is referring to in this line?

Line 170 – I think some punctuation is missing between ART prescription and ART maintenance.

Line 179 – unit costs were calculated as cost per client served. Where did information about the number of clients served come from and could that be incorporate in the tables?

Line 180 – please just clarify that the currencies are ZAR – South African Rand and Zambian Kwacha.

Line 190 – ‘the time spent to complete the activity’ please clarify how this was determined and recorded. Was it recorded for each client i.e. observed or was it estimated and self-reported? This will then also give better context for line 211.

Line 197-198 – where was the number of patients seen over a 12-month period determined from and what was the specific period this was collected for.

Line 207 – please give a reference for the interest rates that were used for South Africa and Zambia.

Line 244 – the restriction to approved study languages in South Africa to English and Afrikaans seems like a limitation. Whilst these languages might have been appropriate for the Western Cape I would expect that for Johannesburg Zulu and Sotho should have been included. Were any staff excluded because they did not speak one of the study languages?

Line 249 – Results how was the number of each cadre of staff at each site determined and how were they selected. This is particularly important to understand how the smallest facility had the largest number of interviews.

Line 263 – I think by including more information in table 1 this might help understand how Brandvlei came to have significantly more staff support HTS than Johannesburg and Lusaka

Line 265 – this point again speaks to there needing to be more information in the text and tables, one assumes that the ART costs should be the same in terms of cost per patient. It is therefore difficult to understand why Jo’burg and Lusaka had much higher supply costs unless they had greater numbers.

Tables 3, 4 and 5 – can the authors just clarify that the first row is unit and the second row is costs. It might also be helpful to include footnotes as to what exactly was included under equipment and supplies.

Was there any thought of disaggregating these costs by gender (or age) e.g. higher or lowers costs for treating male prisoners vs. female, younger vs. older, maximum security vs. general population prisoners?

Line 311 – Discussion - the authors conclude that these data can guide resource allocation for implementing ITT in similar correctional settings. Given that UTT is now implemented as standard of care is this still useful? Is there evidence of slower uptake/implementation in correctional facilities vs primary health care facilities? Would we need more information about these correctional facilities in order to generalize these to other facilities?

Line 344 – is there any new or current evidence that the cost per client did or did not decline 5 years later in the UTT programme?

Line 358 – this relates to a point that I made earlier – if these findings cannot be generalized to all correctional facilities how can they then be used to guide future implementation? If we had more information about these facilities maybe that might help determine where these findings could be used?

Line 365 – Conclusion – please add a reference for your first statement.

Line 370 – again it would be good to know what the level of UTT implementation now is in correctional facilities in SA and Zambia.

Reviewer #2: This is valuable study describing costs of delivering HIV services in a correctional services setting.

Major issues:

Please indicate whether this manuscript complies with the CHEERS checklist for health economic studies (https://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-021-07460-7). In the methods section, some aspects are not adequately described such as outcome of interest, discount rate, etc. From reading the submission, the reference to unit costs appears to be the cost per patient for HST, ART initiation and maintenance. Therefore, it would be more accurate to refer to the total cost and cost per patient across the manuscript. Unit costs refers to cost of a Creatinine medical test for example.

The survey was conducted for the September and December 2017 period. However, the time horizon for the costing analysis is not clearly defined in methods: Is data reported for the January to December 2017 period?

Tables 2 to 4 are difficult to read and could be amended to provide more details. It would easier to read if the total costs, number of patients seen and then the cost per patient are reported. Perhaps the percentage contribution should only be reported for the cost per patient. This would make the costing analysis much easier to follow.

There is a need for a summary figure report the overall cost per patient for HST, ART initiation and ART maintenance for the three correctional facilities.

For the costing data reported for Table 4, please indicate why staffing costs for educational messages was so high at Brandvlei. This is almost 40-fold more expensive than the other correctional facilities. Is this not a calculation error?

Minor issues:

Abstract introduction: HIV is repeated thrice in one sentence – please rephrase (Lines 38-40)

‘Universal test and treat (UTT) for HIV is a population-based strategy that aims to

ensure widespread HIV testing and rapid antiretroviral therapy (ART) for all who have

tested HIV-positive regardless of CD4 count to decrease HIV incidence and improve

health outcomes.’

Abstract conclusion: Abbreviate ‘universal test and treatment’ to UTT (Line 70)

Line 170. Please add a full stop after ‘initial ART prescription’

Line 207: Please provide more details on how equipment costs were annualized.

Table 4: Staff heading missing for the Johannesburg correctional facility.

**********

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: No

Reviewer #2: No

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

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

25 May 2022

May 23, 2022

Elizabeth S. Mayne

Academic Editor

PLOS ONE

RE: Revision and resubmission of article titled, “Costs of Implementing Universal Test and Treat in Three Correctional Facilities in South Africa and Zambia”

Dear Editor

I would like to thank you for the careful review and invitation to submit a revised version of the manuscript that addresses the points raised during the review process. Please see here the responses to reviewer comments where the lines in the comments below refer to the “Revised Manuscript with Track Changes (all markup)” version of the manuscript. We have also submitted the following documents:

• 'Revised Manuscript with Track Changes'

• An unmarked version of your revised paper without tracked changes labelled 'Manuscript'

• Supporting information labelled ‘S1_Table.docx’ and ‘S2 and S3_Table.docx’

Comments to the Author

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

Reviewer #1: No

We have extended Table 1 to include the additional information namely number of correctional units, break down into male and female units, maximum security and youth. We have also included some of the HIV data for each facility namely the estimated HIV prevalence and clients initiated on ART.

We have included information with the service statistics on the clients served as supporting information. Please see the document labelled ‘S1 _Table.docx’. We have indicated this on line 219.

Reviewer #2: Yes

5. Review Comments to the Author

Reviewer #1:

Comments and suggestions:

1) Line 78 – Introduction – it should be made clear that UTT has now been adopted in both South Africa and Zambia. In fact UTT was introduced in South Africa in September 2016 and recommended same-day ART initiation from October 2017.

Thank you for the suggestion. We have now added a sentence on lines 80-82 that makes this clear.

“UTT has been tested in several settings and adopted as part of national HIV programmes in many countries, including both South Africa and Zambia. In fact, UTT was introduced in South Africa in September 2016 and recommended same-day ART initiation from October 2017 (1,2).”

We have also added the suggested articles below to our references:

• SA-NDOH circular – Implementation of Universal test and treat strategy for HIV positive patients and differentiated care for stable patients. South Africa National Department of Health 2016

• SA NDoH Fast tracking implementation of the 90-90-90 strategy for HIV through implementation of the test and treat (TT) policy and the same-day ART initiation for positive patients, 2017

2) Lines 80-82 – can the authors include stats or what ART uptake, viral load suppression and retention were prior to UTT?

We have now included stats on lines 82-87. “Prior to UTT, one retrospective study done using programmatic data in South Africa described viral load suppression at six and 12 months after ART initiation at 94.7% and 92.5% respectively for incarcerated people with HIV treated in an on-site ART programme(3). A cross sectional study done in Malawi also prior to UTT, found 95% viral suppression among a sample of incarcerated people with HIV who received ART for at least 6 months in a prison clinic(4).”

There is a lack of data on overall ART coverage in correctional facilities prior to UTT.

3) Lines 93-95 – can the authors clarify that these data are from that same reference number 6 and that in this context they are referring to mathematical models that are used not models of care.

We apologise for this oversight. We have provided the reference number 10 (originally number 6) on line 105 and we have also clarified that we are referring to mathematical models.

4) Line 104 and paragraph – Methods – one piece of information that would be helpful to include here is a brief description of how inmates at each level of the HIV cascade engage with HIV self-services

On lines 114-120, we have included a brief description on how inmates at each level of the HIV cascade engaged with HIV self-services.

5) Line 127 – Table 1 – it would be helpful if this table could be expanded a bit further to include additional information about each facility e.g. to incorporate the number of correctional units in those facilities and how these break down into male and female units, maximum security and youth. It might also be helpful to know some of the HIV data for each facility as well if it is known. I had missed the total inmates and staff: inmate ratio rows and wonder if those should not be taken to the top or highlighted in some way.

We have extended Table 1 to include the additional information namely number of correctional units, break down into male and female units, maximum security and youth. We have also included some of the HIV data for each facility namely the estimated HIV prevalence and clients initiated on ART.

6) I am not sure if I missed it but it would be helpful if there was supplementary material that also present the various staff costs which might give more context to the differences between facilities and countries. This would provide important context in the results as currently it is difficult to understand Table 2, 3 and 4 where despite the units being quite similar the costs are radically different.

We have included information with the service statistics on clients served as supporting information. Please see the document labelled ‘S1 _Table.docx’. We have indicated this on line 219.

7) Line 163 – it is not clear what the “(i.e., Correctional services”) is referring to in this line?

We have deleted “(i.e., Correctional services”) on lines 181-182 to avoid any confusion.

8) Line 170 – I think some punctuation is missing between ART prescription and ART maintenance.

We have included a full stop between ART prescription and ART maintenance on line 189.

9) Line 179 – unit costs were calculated as cost per client served. Where did information about the number of clients served come from and could that be incorporated in the tables?

The information on the clients served came from the TasP study. We have included information with the service statistics on clients served as supporting information. Please see the document labelled ‘S1 _Table.docx’. We have indicated this on line 219.

10) Line 180 – please just clarify that the currencies are ZAR – South African Rand and Zambian Kwacha.

We have provided clarity on lines 198-200.

“All costs were converted into the US dollar (USD) using exchange rates on 1 January 2017: 1USD = 13.78 South African Rand (ZAR) for the South African sites and 1USD = 9.90 Zambian Kwacha (ZMW) for the Zambian site.”

11) Line 190 – ‘the time spent to complete the activity’ please clarify how this was determined and recorded. Was it recorded for each client i.e. observed or was it estimated and self-reported? This will then also give better context for line 211.

The time spent per client to complete the activity was estimated and self-reported. I have specified this on line 210.

12) Line 197-198 – where was the number of patients seen over a 12-month period determined from and what was the specific period this was collected for.

The number of clients seen over a 12-month period was determined from the service statistics obtained from the TasP study. I have included this on line 219. They were collected over a 17-month period (September 2016 – March 2018) so we worked out the average number of clients seen over a 12-month period.

13) Line 207 – please give a reference for the interest rates that were used for South Africa and Zambia.

I have added a reference for the interest rate on lines 230 -231 and edited the text for clarity.

14) Line 244 – the restriction to approved study languages in South Africa to English and Afrikaans seems like a limitation. Whilst these languages might have been appropriate for the Western Cape I would expect that for Johannesburg Zulu and Sotho should have been included. Were any staff excluded because they did not speak one of the study languages?

Since the study on the costs was only conducted to staff, we did not believe that this would lead to exclusion of any staff members. In fact, no staff in Johannesburg or Brandvlei were excluded from the study due to language barriers as all of them spoke English.

15) Line 249 – Results how was the number of each cadre of staff at each site determined and how were they selected. This is particularly important to understand how the smallest facility had the largest number of interviews.

On lines 154-156 we have included the following methods that were used to determine and select the number of each cadre of staff at each site.

“We purposively selected all staff within the cadre where there were two or less staff in the role. In cadres where three or more staff performed the same duties, we used convenience sampling where every 2nd person was selected.”

16) Line 263 – I think by including more information in table 1 this might help understand how Brandvlei came to have significantly more staff support HTS than Johannesburg and Lusaka.

We apologise for any lack of clarity but we would like to clarify that at the bottom of table 1 we included 42 staff at Brandvlei with a staff to inmate ratio of 1:71, Johannesburg had 62 staff and a staff to inmate ratio of 1:65 while Lusaka had 78 staff and a staff to inmate ratio of 1:18. Based on these numbers, it appears that Brandvlei did not have more staff support HTS but instead Lusaka did.

We have specified the rationale for the staffing on lines 140-141. “The rationale for the staffing in SA was based on the national policy regarding staff per facility rather than on need.”

17) Line 265 – this point again speaks to there needing to be more information in the text and tables, one assumes that the ART costs should be the same in terms of cost per patient. It is therefore difficult to understand why Jo’burg and Lusaka had much higher supply costs unless they had greater numbers.

Yes, the total costs of supplies for HTS are higher at Jo’burg and Lusaka due to the greater numbers of clients served. I have included information with the service statistics on the number of clients served as supporting information. Please see the document labelled ‘S1 _Table.docx’.

On lines 291-294, I have also included additional text in the manuscript to clarify this. “The largest cost driver for HIV testing at Brandvlei were staff costs at 55.6% of the total cost, while at Johannesburg and Lusaka, supplies were the largest contributor at 56.5% and 86.6% of the total cost, respectively, due to the greater numbers of clients served (supplementary information ‘S1 _Table.docx’).”

18) Tables 3, 4 and 5 – can the authors just clarify that the first row is unit and the second row is costs. It might also be helpful to include footnotes as to what exactly was included under equipment and supplies.

We have clarified that the first row is aunit costs and the second row is btotal costs using footnotes.

We have included detailed lists of cequipment and dsupplies in supporting information and we have indicated this within the footnotes. Please see the document labelled ‘S2 and S3 _Table.docx’.

19) Was there any thought of disaggregating these costs by gender (or age) e.g. higher or lower costs for treating male prisoners vs. female, younger vs. older, maximum security vs. general population prisoners?

Thank you for this question. Unfortunately, we did not anticipate that the costs would vary by gender, age or maximum vs general population prisoners so our survey questions did not specify these characteristics.

20) Line 311 – Discussion - the authors conclude that these data can guide resource allocation for implementing UTT in similar correctional settings. Given that UTT is now implemented as standard of care is this still useful? Is there evidence of slower uptake/implementation in correctional facilities vs primary health care facilities? Would we need more information about these correctional facilities in order to generalize these to other facilities?

Given that UTT is now policy this evidence is still useful for resource allocation for implementation. Adding responsibilities without appropriate resources can lead to limited implementation of policy. Our study also contributes to early evidence that can be used as a comparator for future studies.

To our knowledge, there are no studies comparing UTT implementation in correctional facilities vs primary health care facilities. It is plausible that implementation is less complete in some correctional facilities due to resources and staff priorities.

We do not need any additional information about these correctional facilities in order to generalize these findings to other facilities.

21) Line 344 – is there any new or current evidence that the cost per client did or did not decline 5 years later in the UTT programme?

To our knowledge there is no current evidence that the cost per client did or did not decline 5 years later in the UTT programme.

Our results do provide a point of comparison for future studies.

22) Line 358 – this relates to a point that I made earlier – if these findings cannot be generalized to all correctional facilities how can they then be used to guide future implementation? If we had more information about these facilities maybe that might help determine where these findings could be used?

On lines 389-393, we have revised this statement: “The findings need to be considered in context of the correctional facility size, number of people living with HIV, and staffing level of health personnel in the facility.” We believe that when considering these factors, reasonable assumptions can be made to apply these findings to other facilities in South Africa or Zambia.

23) Line 365 – Conclusion – please add a reference for your first statement.

We have added the following references to the statement on line 399-400.

Hill LG, Maucione K, Hood BK. A focused approach to assessing program fidelity. Prev Sci. 2007 Mar;8(1):25-34. doi: 10.1007/s11121-006-0051-4. Epub 2006 Sep 13. PMID: 16967341.

Harolds, Jay MD Quality and Safety in Health Care, Part I, Clinical Nuclear Medicine: August 2015 - Volume 40 - Issue 8 - p 660-662 doi: 10.1097/RLU.0000000000000877

24) Line 370 – again it would be good to know what the level of UTT implementation now is in correctional facilities in SA and Zambia.

We agree that a follow-up study or program evaluation would be valuable. We don’t have any recent assessments to include or report for this manuscript.

Reviewer #2:

Major issues:

1) Please indicate whether this manuscript complies with the CHEERS checklist for health economic studies (https://bmchealthservres.biomedcentral.com/articles/10.1186/s12913-021-07460-7). In the methods section, some aspects are not adequately described such as outcome of interest, discount rate, etc. From reading the submission, the reference to unit costs appears to be the cost per patient for HST, ART initiation and maintenance. Therefore, it would be more accurate to refer to the total cost and cost per patient across the manuscript. Unit costs refers to cost of a Creatinine medical test for example.

We apologise if some aspects of the CHEERS checklist may not have been adequately described in our manuscript. We have now clarified our outcomes of interest (lines 180 - 181). The discount rate was initially included in the original version and has now been edited slightly for clarity (lines 230 - 231).

Thank you for the second suggestion. We would like to clarify that we defined the ‘units’ in the unit costs as the ‘clients served’ which we mentioned on line 202. “Costing results are reported separately for each facility and as overall cost for care provision and as cost per client served (unit cost).”

To guide us, we used the following definition from the Reference case written by the Global Health Costing Consortium (GHCC).

Intervention ‘unit’ cost Average cost of an intervention (or strategy) (e.g., unit cost per person or episode of expanding TB treatment, or costs of peer education per person reached)

2) The survey was conducted for the September and December 2017 period. However, the time horizon for the costing analysis is not clearly defined in methods: Is data reported for the January to December 2017 period?

The costing analysis is reported over a 12-month period with no time horizon specified because the number of clients seen were collected over a 17-month period (September 2016 – March 2018) so we worked out our costs based on the average number of clients seen over a 12-month period.

We have included the following statement on lines 218 – 221 to make this clear.

“Staff unit costs were computed by dividing the annual total costs with the number of clients seen over a 12-month period which we obtained from the TasP study (S1_Table.docx) for a period of 17 months, though we based our costs on the average number of clients seen over a 12-month period.”

3) Tables 2 to 4 are difficult to read and could be amended to provide more details. It would easier to read if the total costs, number of patients seen and then the cost per patient are reported. Perhaps the percentage contribution should only be reported for the cost per patient. This would make the costing analysis much easier to follow.

On tables 2 to 4 we have now reported the total costs and cost per patient. The clients seen are reported as supporting information. Please see the document labelled ‘S1 _Table.docx’. The percentage contribution is only reported for cost per patient.

4) There is a need for a summary figure report the overall cost per patient for HST, ART initiation and ART maintenance for the three correctional facilities.

We have included a summary figure 1 reporting the overall cost per patient for HST, ART initiation and ART maintenance in Fig 1 (lines 287 - 289).

5) For the costing data reported for Table 4, please indicate why staffing costs for educational messages was so high at Brandvlei. This is almost 40-fold more expensive than the other correctional facilities. Is this not a calculation error?

The high staffing costs for educational messages at Brandvlei was due to one of our study limitations because of using self-reported time data. We have mentioned this on line 384-385. ‘Furthermore, estimation of time spent on UTT activities depended on self-report’.

Minor issues:

6) Abstract introduction: HIV is repeated thrice in one sentence – please rephrase (Lines 38-40)

‘Universal test and treat (UTT) for HIV is a population-based strategy that aims to

ensure widespread HIV testing and rapid antiretroviral therapy (ART) for all who have

tested HIV-positive regardless of CD4 count to decrease HIV incidence and improve

health outcomes.’

We have edited the sentence on lines 38-40 and reduced the repetition of the word HIV.

7) Abstract conclusion: Abbreviate ‘universal test and treatment’ to UTT (Line 70)

We have abbreviated ‘universal test and treatment’ to UTT (Line 69). We have also revised the statement on lines 68 – 71 to the following:

“Given the needs of this population, the opportunity to reach inmates at high risk for HIV, and overall national and global 95-95-95 goals, the UTT policies for incarcerated individuals are of vital importance. Our findings provide comparator costing data and highlight key drivers of UTT cost by facility.”

8) Line 170. Please add a full stop after ‘initial ART prescription’

We have added a full stop after ‘initial ART prescription’ on line 189.

9) Line 207: Please provide more details on how equipment costs were annualized.

On line 230-231 we have provided a reference for the discount rate used to annualize the equipment costs.

10) Table 4: Staff heading missing for the Johannesburg correctional facility.

We have now included it.

Additional revisions

1) Formatted spacing on line 130

2) Edited footnote beneath table 1 on line 147 for clarity

3) Formatted manuscript to PLOS ONE requirements namely heading levels 1,2 and 3

4) Edited the title page according to PLOS ONE's style requirements

5) We have edited the term “patients” to “clients” throughout the manuscript

Submitted filename: Response to Reviewers_PLOS ONE.docx

Click here for additional data file.

22 Jul 2022

Costs of implementing Universal Test and Treat in three correctional facilities in South Africa and Zambia

PONE-D-21-36716R1

Dear Dr. Mukora,

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,

Elizabeth S. Mayne, M.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

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

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

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

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

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

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

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

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

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

Reviewer #1: The authors have done well address the reviewer comments that were made and have added clarity where needed. A couple of minor points that might be addressed.

Table 1 might need an updated title given the inclusion of the additional data. Perhaps also to check the HIV prevalence for the Lusaka correctional facility as it is given as 15% yet 16% of the inmates are currently on ART suggesting that HIV prevalence might be higher. Likewise it might be worth mention that in the other correctional facilities based on the proportion that are on ART (2% in Brandvlei and 6% in Johannesburg) it suggests that there are large numbers that are not on ART and this must be in part what is driving up the cost as there are a large number of people serving a very few people on ART. I think it is worthy of discussion that not only is the variation in costs the result of differences in the staff:inmate ratio and variation of salary but must also be driven by the variation in the inmate population (Brandvlei looks to have a higher proportion of female inmates and no maximum security units compared to the other two) which in turn might explain the difference in HIV prevalence and low yield of HIV positive.

The only other point I would make - where the authors discuss supplies as a cost driver they link this to 'the number of clients served' - this to me indicates the number of client who receive care or services whereas I think that often the authors are specifically meaning the number of clients tested or the number supported on ART. I think it is worthwhile being specific about these and what is meant in each case.

Reviewer #2: All comments that were raised have been addressed and there no further changes required for this manuscript.

**********

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: Yes: Naseem Cassim

**********

16 Aug 2022

PONE-D-21-36716R1

Costs of implementing Universal Test and Treat in three correctional facilities in South Africa and Zambia

Dear Dr. Mukora:

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. Elizabeth S. Mayne

Academic Editor

PLOS ONE