A study protocol for Project I-Test: a cluster randomized controlled trial of a practice coaching intervention to increase HIV testing in substance use treatment programs

Background People with substance use disorders are vulnerable to acquiring HIV. Testing is fundamental to diagnosis, treatment, and prevention; however, in the past decade, there has been a decline in the number of substance use disorder (SUD) treatment programs offering on-site HIV testing. Fewer than half of SUDs in the United States offer on-site HIV testing. In addition, nearly a quarter of newly diagnosed cases have AIDS at the time of diagnosis. Lack of testing is one of the main reasons that annual HIV incidences have remained constant over time. Integration of HIV testing with testing for HCV, an infection prevalent among persons vulnerable to HIV infection, and in settings where they receive health services, including opioid treatment programs (OTPs), is of great public health importance. Methods/Design In this 3-arm cluster-RCT of opioid use disorders treatment programs, we test the effect of two evidence-based “practice coaching” (PC) interventions on: the provision and sustained implementation of on-site HIV testing, on-site HIV/HCV testing, and linkage to care. Using the National Survey of Substance Abuse Treatment Services data available from SAMHSA, 51 sites are randomly assigned to one of the three conditions: practice coach facilitated structured conversations around implementing change, with provision of resources and documents to support the implementation of (1) HIV testing only, or (2) HIV/HCV testing, and (3) a control condition that provides a package with information only. We collect quantitative (e,g., HIV and HCV testing at six-month-long intervals) and qualitative site data near the time of randomization, and again approximately 7–12 months after randomization. Discussion Innovative and comprehensive approaches that facilitate and promote the adoption and sustainability of HIV and HCV testing in opioid treatment programs are important for addressing and reducing HIV and HCV infection rates. This study is one of the first to test organizational approaches (practice coaching) to increase HIV and HIV/HCV testing and linkage to care among individuals receiving treatment for opioid use disorder. The study may provide valuable insight and knowledge on the multiple levels of intervention that, if integrated, may better position OTPs to improve and sustain testing practices and improve population health.


6
Background 98 In its ongoing recognition of HIV testing as a fundamental component of HIV treatment and 99 uptake of testing by OTP clients, and providing timely linkage to care for persons who test 191 positive. The quaternary objective is to assess the health outcomes, health care utilization, and 192 cost-effectiveness of the PC interventions compared incrementally to one another and to the 193 control condition. This will allow for assessing the budget required to implement (scale up and 194 sustain) the PC interventions nationally. 195

Study Design 197
This protocol manuscript follows the SPIRIT reporting guidelines (27) Sites are randomized into three groups (HIV PC, HIV/HCV PC, and information only 208 control condition) in a ratio of 1:1:1 using a blocked randomization scheme to ensure 209 relative balance across time of entry into the study. The data analyst, who is not 210 involved in the delivery of the intervention, keeps the randomization schedule and 211 sequence secure, and ensures confidentiality and independence of the allocation data. 212 After site personnel complete baseline surveys and interviews, site personnel are 213 notified to which of the three study conditions the site has been assigned. 214 215

Eligibility criteria 216 217
Site eligibility criteria for this study are: 1) OTP site sees at least 150 unduplicated clients per year; 218 2) The site is capable and willing to prospectively collect data on the number of clients who a) are 219 offered any HIV and/or HCV tests; b) completed these tests; c) are referred to care/evaluation (and 220 type of referral) if positive); and d) are linked to care/evaluation within 30 days of diagnosis of 221 HIV and/or HCV; 3) The site is capable and willing to provide aggregate client testing data within 222 demographic categories of gender and race/ethnicity and data on HIV/HCV test reimbursement 223 processes and outcomes; 4) the site is able to select staff willing to consent to participate in study 224 surveys, qualitative interviews, and intervention coaching throughout the study. Sites in which 225 over 50% of clients served in the prior 6 months were HIV or HCV tested are excluded. To be 226 eligible to participate in the study's site surveys, interviews and intervention activities, individuals 227 must be site personnel employed within one of the 51 enrolled sites. 228

Study Settings and Recruitment 229
The sampling frame consists of all opioid treatment programs/sites in the 2017 National Survey of 230 Substance Abuse Treatment Service (N-SSATS), a national census of all US substance use 231 treatment facilities, that have a minimum client census of 150 clients per year. The study draws on 232 a random sampling of 500 eligible sites from this sampling frame, and will draw additional sites as 233 needed. A total of 51 eligible sites will be enrolled in the trial. Recruitment occurs through e-mail 234 and telephone contact. Site leadership (e.g., Chief Executive Officer, Director) are contacted, 235 informed about the study and invited to complete a screening process to determine the OTP's 236 eligibility to participate in the study. If interested in participating in the study, the site leader 237 completes a brief screening by telephone interview (after providing verbal consent) or via self-238 administered survey to determine the site's eligibility to participate. Enrollment consists of 239 obtaining a signed form letter from each participating site, outlining the various study activities in 240 which the site personnel will participate. The site leader must also complete an acknowledgment 241 from noting that participation in the study is voluntary and that there will be no impact to any 242 individual employee of the site for not participating. Participants in this study consists of the 243 professional staff working at eligible treatment programs/sites around the country that treat clients 244 with opioid use disorder. Staff at selected sites that accept the invitation to participate are 245 interviewed and complete brief surveys to confirm that they meet eligibility criteria. We also 246 recruit, via email, directors working at state substance use authorities. We will conduct a survey of 247 state policies and guidelines relevant to HIV and HCV testing. To participate in state surveys, 248 individuals must be directors at state substance use authorities in the participating sites' states. 249

250
Sites complete all surveys and related evaluations according to the study timeline. Participants may 251 retract their consent to participate in the study, and may do so at any time before or during the 252 study. Once a site or staff member participating in the study withdraws from the study during 253 treatment, their data is excluded from our specific analysis, but may inform aggregated analysis of 254 data. 255 256 A number of procedures are in place to promote retention in the study for the duration of the 257 planned intervention. The primary strategies to improve retention in the interventions in this trial 258 are twofold. The first is our incentive structure. Participating sites receive monetary incentives 259 during their 2-year involvement. The payments are given once after completing their initial data 260 collection plan (typically within 2 weeks of randomization) and a second/final time after the site 261 completes the second of four aggregate data transfers. Personnel questionnaires and interviews are 262 compensated at $40 and $50 (respectively); per each site's discretion, these incentives are either 263 issued directly to the personnel completing them or pooled into a single site-wide incentive (e.g., 264 staff luncheon). This is intended to prevent participants from providing partially completed 265 questionnaires, not adhering to treatment as delivered, or withdrawing from the study after 266 enrollment. Secondly, the PCs work with the sites to encourage them to participate and adhere to 267 the intervention sessions/timeline/window. PCs are mindful and respectful of the sites' time and 268 busy schedules and therefore ensure that they meet their scheduling needs. The collaborative 269 nature of the intervention helps as the PC will assist the site in setting goals/action items and help 270 brainstorm and discover ways to implement change. Intervention adherence is part building 271 relationships and part the site staff's time and motivation. Contacting the site to encourage them 272 and move them along is part of the success. 273

Conceptual/theoretical Framework 274
The Consolidated Framework for Implementation Research (CFIR) framework was the basis for 275 identifying essential factors supporting or impeding the adoption of testing.(28) The five CFIR 276 domains we considered in developing the PC interventions are based on contexts that influence the 277 implementation, effectiveness, and sustainability of our approach: inner setting (e.g., networks, 278 climate, readiness), outer setting (e.g., client needs and resources, peer pressure, incentives), 279 intervention characteristics (e.g., evidence strength, adaptability, cost), individual characteristics 280 (e.g., self-efficacy, knowledge, beliefs), and the implementation process (e.g., planning, engaging, 281 executing, evaluating). The implementation of the PC interventions was then guided by the stage 282 theory of organizational change. Change theories guide the implementation of interventions, as 283 well as the evaluation (29-31). Stage theory posits that organizations move through four sequential 284 stages as they change or adopt an innovation: awareness, adoption, implementation, and 285 institutionalization (see Table 1). Each stage involves specific strategies that are matched to that 286 stage, the particular OTP, and factors external to the organization (e.g., how CDC guidelines are 287 implemented in the particular OTP's state). We provide details of the specific steps to be taken 288 within each of the 4 sequential stages of the interventions below. PC is tailored to the context of 289 the OTP, focusing on organizational change. 290

Study Interventions 291
The two PC interventions are manualized and training of Practice Coaches (PCs) emphasizes the 292 importance of adhering to the manual that corresponds to a site's assigned intervention condition 293 (i.e., preventing drift). To ensure consistency of intervention delivery across all PCs, the PCs co-294 facilitated the first few intervention sessions. PCs also co-facilitate some intervention sessions later 295 in the study to ensure that they are still delivering the intervention in the same manner and 296 adhering to the manuals. Additionally, the Intervention Director conducts regularly scheduled 297 "peer to peer" conference calls to discuss difficulties and successes in conducting the PC 298 interventions; to facilitate the PCs learning from and supporting each other; and to facilitate 299 receiving support and feedback from the Intervention Director. 300 All participants are provided with information and resources, per their intervention allocation. 301 Program are discouraged from additional treatments that are not according to the study protocol, 302 during the intervention period. Participants will be required to report all treatments that are not 303 according to the treatment protocol, i.e., an initiative that supports the adoption of HIV tor HCV 304 testing delivered by a coach. 305 Practice Coaching: Skilled PCs serve as a resource for programs. PC's work includes helping the 306 site leader to identify an organizational change agent/champion, who will lead the program's on-307 site testing effort and serve as the primary liaison to the study team. The Champion is supported by 308 a Change Team, who are key staff identified by the Champion, with guidance from the PC, i.e., 309 individuals with high-level of commitment to organizational change and improving testing 310 practices. PC activities will encompass: (1) pre-implementation assessment, feedback and goal 311 setting, (2) information on the provision of HIV or HIV/HCV testing and linkage to care, (3) 312 leveraging existing resources (e.g., staff, space, equipment) to improve the HIV or the HIV/HCV 313 service delivery system and facilitate billing and reimbursement for testing, (4) technical and 314 decision support for reimbursement of testing services, and (5) improved linkages to medical care 315 and city, state, and federal sources for testing resources. PCs support sites by helping them 316 navigate resources, as well as support the site in addressing potential barriers, including, but not 317 limited to, human resources, staff training, and resource allocation. PCs engage OTPs over 6 318 months to guide them through the process of improving the initial and sustained implementation of 319 HIV or HIV/HCV testing services and linkage to care (see Fig. 1). 320

321
The treatments in this study are two active interventions: PC for HIV testing, and PC for HIV/HCV 322 testing. Treatment and Access to Services (ARTAS) counseling). PCs also support sites by helping them 343 navigate resources, focus their use of linkage to care materials, as well as support the site in 344 addressing potential barriers, including, but not limited to, human resources, staff training for 345 linkage, and resource allocation to facilitate linkage to care services. Sites assigned to the 346 HIV/HCV PC intervention condition also receive coaching preventive self-care and protecting 347 liver function from further harm through reducing or eliminating alcohol consumption, and 348 Hepatitis A and B vaccination, as appropriate. PCs also link clients who receive an HCV-positive 349 test result (either antibody or RNA) to follow-up evaluation and/or medical care within 30 days of 350 HCV diagnosis. 351 352

Control Condition 353
Provision of Information: The administrators and/or designated personnel within the OTPs 355 assigned to the information control condition receive the official NIDA/SAMHSA Blending 356 Initiative product, "HIV Rapid Testing in Substance Abuse Treatment Programs," that we will 357 provide to OTPs to educate and motivate them about the importance of offering on-site HIV 358 testing. They will also receive an electronic link and/or hard copy of the ARTAS implementation 359 manual and training information as well as information about PrEP, a daily medication that serves 360 as an HIV prevention tool for individuals who are HIV-negative but at substantial risk of acquiring 361 HIV infection. Resources generated from the HIV rapid testing blending initiative product include 362 a fact sheet, resource guide, marketing materials, and an Excel-based budgeting tool. In addition to 363 the HIV-specific materials, the Website provides opportunities for training, self-study progress, 364 workshops, and distance learning. 365 366 367

Description of Intervention Stages 368 369
Awareness Phase 1 is concerned with raising interest and generating support for the 370 intervention with senior management by defining the problem (i.e., local HIV prevalence, 371 resource allocation for HIV testing), and identifying possible solutions such as establishing a 372 billing and reimbursement system for HIV testing services, training and motivating staff to test 373 clients for HIV, and connecting with a health care center so that procedures are in place to link 374 clients who test HIV-positive to care. 375 376 Phase 1 includes five steps: Step 1 is a teleconference call between the PC and the site's Leader, 377 including advice to select a champion, with appropriate interest, knowledge base, skill set and 378 leadership capacity.
Step 2 is a teleconference call between the PC and the site's designated 379 champion.
Step 3 involves the PC's comprehensive assessment of barriers and facilitators to the 380 provision, client uptake, and reimbursement of HIV testing services. This assessment is based 381 on a structured interview conducted by the PC.
Step 4 is a concentrated in-person or virtual 382 workshop and with the champion and key staff from the site. PCs review the goals and 383 objectives of practice coaching, knowledge-based HIV information, the provision of HIV testing 384 services, quality improvement, monitoring and evaluation tools, billing and reimbursement for 385 HIV testing (including alternatives such as securing free test kits from the local health 386 department and/or establishing Memoranda of Understanding / Agreement (MOU/ MOAs) with 387 the health department and/or other community-based organizations to provide HIV testing 388 services within the site), introduction to evidence-based linkage to care strategies as well as a 389 review of roles/responsibilities and data capture forms. One purpose of the workshop is for the 390 PC to synthesize results of the site's comprehensive barriers/facilitators assessment and pre-391 intervention performance data and present these results to the site's champion(s) and key staff, 392 providing constructive feedback on identified barriers and potential solutions. Another key 393 purpose of the visit is creating an action plan that is tailored to the OTP's context and culture 394 and that addresses identifying/securing resources needed to initiate or increase on-site testing. 395 Step 5 is a debrief phone call with site Champion and Change Team to review and discuss the 396 action plan for testing. This interaction with the PC also presents opportunities for sites to ask 397 additional questions. 398 399 Adoption Phase 2 begins when an organization decides to commit to and initiate an 400 innovation or evidence-based intervention (e.g., on-site testing); this phase includes refining 401 the action plan for on-site testing. The champion and key staff (the "change team") use the 402 plan-do-study-act (PDSA) method, a structure for iteratively guiding goal setting and 403 planning. PCs assist change teams and provide tools to facilitate relationship building with 404 stakeholders for adopting and implementing system/OTP-wide changes, specific strategies to 405 achieve HIV testing goals through appropriate mechanisms. Specific intervention activities in 406 this stage include 1) ongoing video or traditional teleconference call meetings utilizing the 407 PDSA format. Additionally, 2) PCs will guide the champions and the OTP change team in 408 engaging organizational "gatekeepers" to build consensus and negotiate any needed action 409 plan modifications without jeopardizing the integrity of the stated goals. 3) PCs will meet (by 410 phone or video conference) with the change teams biweekly, and as needed, regularly to 411 support any necessary iterations between steps 1 and 2. 412 413 Implementation Phase 3 is the process of integrating an innovation within a setting, involving 414 identification of (and changes to) practice patterns or organizational structures as necessary to 415 overcome identified barriers. This involves the technical aspects of providing HIV testing, 416 including staff training and procurement of materials as well as the support needed for the 417 introduction of change. For linkage to HIV care, it is critical to identify the facilities and teams to 418 which people are linked for follow-up care, and engagement of new sets of stakeholders may be 419 required. Additionally, building staff capacity and motivation for testing and linkage to care is 420 crucial for sustained implementation. PCs provide support on: 1) optimizing workflow (e.g., 421 what type of HIV testing to implement, when to provide testing), 2) application of CDC and 422 state-level HIV testing and linkage to care guidelines, 3) development and maintenance of a 423 20 training and quality assurance program to ensure front-line staff have initial and continued 424 knowledge, support and motivation to provide HIV testing/linkage to care, 4) assistance with the 425 effective use of billing and reimbursements systems (established in Phase 2) for sites with the 426 capacity to bill (e.g., processes to facilitate coding of services, timely submission of claim), and 427 initiation of efforts to translate information and resources for setting-up infrastructure for billing 428 among sites that are not already billing for services, 5) support tools to help sites engage clients 429 (e.g., testing campaigns) and promote the uptake of HIV testing, and 6) increasing utilization of 430 community resources that enhance the site's capacity to provide HIV services.

Study Assessments 489
Three types of data are collected, client data, site data and state data. The assessments used in 490 the study consist of three quantitative surveys with treatment program staff (i.e., treatment 491 program administrators, treatment program clinical staff), and state administrators; and 492 qualitative interviews with treatment program directors and study champions (see Table 2). The 493 treatment program administrator survey measures structure and service setting, client 494 characteristics, staffing characteristics, program guidelines, barriers to care, and perceptions. The 495 treatment program clinician survey measures training, knowledge, experience, barriers, and 496 perceptions. The state administrator survey covers policies/regulations, reimbursement, and 497 prioritization of testing services. The qualitative interviews address in-depth discussion about 498 testing services offered at the site, barriers and facilitators to offering HIV/HCV testing services 499 and linkage to care, attitudes towards services and training at the site, and organizational 500 readiness for change. 501 Outcomes 502

Primary Outcome 503
The primary outcome analysis will compare the PC interventions with the control condition on 504 the initial impact of HIV testing as measured by the proportion of OTP clients tested during the 505 period 7-12 months after randomization ("initial impact", T3), while controlling for HIV 506 testing during the baseline period (T1). 507

Secondary Outcomes 508
The secondary outcome analysis will examine the incremental impact of the HIV/HCV testing 509 intervention condition, compared with the HIV testing condition, on the proportion of OTP 510 clients tested for HIV. Other secondary outcome analyses will examine the impact of the PC 511 interventions on the: sustained impact of HIV testing (proportion of OTP clients tested during 512 T4), initial impact of HCV testing (proportion of OTP clients tested during T3), and sustained 513 impact of HCV testing (proportion of OTP clients tested during T4). The effectiveness of the interventions relative to the control condition will be examined for 518 tertiary outcomes: linkage to HIV care among OTP clients who tested positive for HIV, linkage 519 to HCV care among OTP clients who tested positive for HCV, and change in perceived 520 barriers/facilitators to HIV testing. We will also examine, using mixed methods: the 521 interventions' impact mediated by changes in perceived barriers/facilitators; the impact of the 522 PC interventions on OTPs' organizational and environmental characteristics that serve as 523 facilitators and barriers to the initial and sustained implementation of HIV testing, the uptake 524 of testing by OTP clients, and providing timely linkage to care for persons who test positive. 525 While the intervention emphasizes on-site testing, study outcomes may assess any testing, 526 either on-or off-site, to measure potential spillover effects of the intervention. We will determine health outcomes, health care utilization, and cost-effectiveness of the PC 531 interventions, and compare them incrementally to one another and to the control condition. We 532 will also assess the budget required to implement (scale up and sustain) the PC interventions 533 nationally. 534 535

Data Sources 536 537
We use various approaches to collect data to measure outcomes and covariates (see Table 3). 538 Study sites, upon enrollment in the study, are provided with a spreadsheet which they may use 539 to assist in compiling aggregate de-identified data summaries, including HIV/HCV testing 540 data. These data are transferred from sites at 6-month intervals and are checked for 541 consistency. We use REDCap Survey data capture tools, with automatic range and consistency 542 checks for quantitative survey data collection. PCs track the intervention process and record 543 these data in structured forms, i.e., Practice Coach Interaction Form (PCIF). Some of the 544 intervention process data are collected and managed using REDCap, and other intervention 545 process data are collected using electronic collection forms. Qualitative interview data, 546 including audio recordings and transcriptions, are collected via digital audio recorders. All data 547 are stored securely on an encrypted and password protect server. All personal data of 548 participants, both program and staff, are assigned a unique identifier that is stored on a secure 549 server available only to data analysts and researchers with approved access to the database. 550 Data analysis will only include non-identifiable data. Prior to analyses of the site surveys, 551 factor structure and reliability of scales will be documented and all variables will be assessed 552 for appropriate statistical distributions for analysis. Any missing data will be accommodated 553 using multiple imputation. To ensure monitoring of other study-related participant safety events or incidents, procedures 558 regarding confidentiality and data integrity are continually monitored and regularly audited. 559 Members of the PI team meet regularly (e.g., bi-weekly) during the study period to review trial 560 progress. 561 562 Developed and implemented by the PIs, the data and safety monitoring plan (DSMP) assures 563 minimal risk and data integrity in this study. The plan assures that all data collection procedures 564 concur with all local, state and federal guidelines. To assure integrity of data and safety, all aspects 565 of the program are monitored, including informed consent procedures, data collection and quality 566 (i.e., review for statistical anomalies), fidelity of the practice coaching intervention to the 567 intervention manual and adherence of the qualitative interviews to the interview guide. A 568 designated Data Safety Monitoring Board (DSMB) assists in oversight of the study. The DSMB 569 examines accumulating data to assure protection of participants' safety while the study's scientific 570 goals are being met. The DSMB conducts periodic reviews of accumulating safety and 571 effectiveness data and determines whether there is support for continuation of the study, or 572 evidence that study procedures should be changed, or if the study should be halted for reasons 573 relating to the safety of the study participants, the effectiveness of the treatment under study, or 574 inadequate study progress. In the event that an adverse event or otherwise untoward incident 575 occurs as a direct result of or in the context of the project, we closely follow IRB directives and 576 reporting policies. Specifically, we report to the appropriate IRBs within 10 working days, in 577 writing, all serious adverse or otherwise untoward events associated with procedures. To ensure 578 monitoring of other study-related participant safety events or incidents, procedures regarding 579 confidentiality and data integrity are continually monitored and regularly audited. The PIs 580 promptly inform other Co-Is and NIDA staff of any proposed changes in site enrollment, 581 intervention implementation or in the protocol that are relevant to safety, as well as any actions 582 taken by the IRBs as a result of their continuing review of the project. In the event of any major 583 changes in the status of an ongoing protocol (which occurs only with IRB approval), the PIs inform 584 NIDA's program officer and the DSMB immediately. Such changes would include, but are not 585 limited to protocol amendments, temporary suspension of site initiation/commencement, changes 586 in informed consent or IRB approval status, termination of participation by the site and/or site 587 personnel, or other problems or issues that could affect the human subjects in the study. Statistical Power and sample size were determined using a simulation programed in SAS 9.3. 592 The simulation generated data with a range of intra-class correlations (ICC) from .04 to .08, and 593 an information control condition with a proportion of clients' HIV testing of 20% as found in the 594 control condition in CTN0032, a study assessing the relative effectiveness of three HIV testing 595 strategies on increasing receipt of test results and reducing HIV risk behaviors among patients 596 seen at drug use treatment centers.(32) A sample size of 51 OTPs and an average of about 100 597 clients per site per 6-month period provides over 95% power for the primary outcome and 80% 598 power for the secondary outcome if the proportion of clients' HIV testing in the HIV PC 599 condition is 30% (absolute difference of 10% from control condition) and the proportion of 600 clients' HIV testing in the HIV/HCV PC condition is 41% (absolute difference of 11% from the 601 HIV PC condition) for all expected levels of ICC (.04 to .08). For the quasi-experimental 602 evaluation of the blending product, the study will have over 80% power to uncover an absolute 603 change in proportion testing for HIV of 6% to 7%. For analysis of change in proportion of 604 facilitators/barriers, the study will have over 80% power to uncover a significant difference in 605 change if the difference in change is .5 to .65 of a standard deviation, a medium effect size. 606

Empirical Analysis 607
The primary outcome analysis will test the hypothesis that the two PC interventions will result in 608 significantly higher proportions of clients tested for HIV than the control condition during the 609 "initial impact" period (7-12 months post-randomization or T3), controlling for the proportion of 610 clients tested during the baseline period (T1). We will use a generalized estimating equation (GEE) 611 model with a binary distribution and logit link. The model will include four 6-month periods: T1 612 (months -6 to-1) --prior to randomization, T2 (months 1-6) --during intervention/control period, 613 T3 (months 7-12) --initial impact, and T4 (months 13-18) --sustained impact. Time and 614 participants are both nested within site. However, time is not nested within participants in the 615 primary analysis. Individuals within a site may be more alike (correlated) than are individuals 616 between sites, which will be accounted for in the GEE by inclusion of the working correlation 617 matrix within site and the sandwich estimator for standard errors. The model will include gender 618 and race/ethnicity, and geographic region as control variables. The primary tests of H1 will be 619 done using a contrast of testing differences across conditions in the proportion of clients tested 620 during T3, controlling for the proportion of clients tested pre-randomization (T1). 621 The secondary and tertiary outcome analyses will use similar GEE methods as described for 622 the primary outcome measure. The secondary outcome will test the hypothesis that the 623 HIV/HCV PC intervention will result in significantly higher proportions of clients tested for 624 HIV than the HIV PC intervention during the initial impact period (7-12 months post-625 randomization or T3), controlling for the proportion of clients tested during the baseline 626 period (T1). Other secondary measures will examine, for example, the impact of the PC 627 interventions on the provision and sustainability of HIV testing (T4), the impact of the PC 628 interventions on initial impact of HCV testing (T3). The tertiary outcome analysis related to 629 linkage to care will evaluate the effectiveness of the interventions relative to the control 630 condition on linkage to HIV care among OTP clients who tested positive for HIV, or for 631 HCV, as well as change in perceived barriers/facilitators to HIV testing. 632

633
We will use mixed-methods to evaluate the impact of the PC interventions and the OTPs' 634 organizational and environmental characteristics that serve as facilitators and barriers to the 635 provision and uptake of HIV testing (T3), sustained implementation of HIV testing (T4), and 636 improving timely linkage to care for persons who test positive. We will use a multilevel GEE 637 model to examine whether change in perceived barrier/facilitators mediates intervention impact 638 on HIV testing (T4). Mediation will be assessed by the product of coefficients method. 639 640

Cost Analysis 641 642
The quaternary outcome includes determining health outcomes, associated costs and evaluating 643 the cost-effectiveness of the interventions. We will test the hypothesis that the incremental cost-644 effectiveness ratio (ICER) for the HIV PC intervention will be below a commonly-cited US 645 willingness-to-pay threshold (<$100,000/quality adjusted life years (QALY)) and therefore more 646 economically attractive than the control condition. Our other hypothesis is that the ICER for the 647 HIV/HCV PC intervention will be more economically attractive than the HIV PC intervention. 648 The study will follow a proven model of effective collaboration among the intervention team and 649 computer simulation modelers to evaluate the health outcomes, health care utilization, and cost-650 effectiveness of the PC interventions (11). Established micro-costing techniques will be used to 651 identify the costs of delivering the PC interventions, including personnel and non-personnel costs 652 incurred centrally to deliver the intervention and incurred at the OTPs to participate in the 653 intervention and conduct follow-up activities (excluding time required for research activities). foci, coded as headings. Specific dimensions of the headings are assigned core codes, while 674 dimensions of the core codes are assigned sub codes. We will use ATLAS.ti, a software 675 program for qualitative analysis, to facilitate the analysis. Seven steps will be used to develop 676 the coding scheme: 1) Identify the principal issues discussed by participants; 2) Construct 677 definitions of the primary analytic themes; 3) Develop and apply core codes (themes) and sub 678 codes (sub-themes) to the initial set of interviews; 4) Develop a provisional coding scheme; 5) 679 Test the coding scheme by applying it to a subsample (n=15) of interviews 6) refine the 680 provisional coding scheme; 7) Have two research team members independently apply the 681 coding scheme to a new subsample (n=15) of interviews; 8) Have them meet to reconcile 682 differences in their application of the codes; 9) Refine the coding scheme as needed and finalize 683 it; and 10) Apply the finalized coding scheme to the full data set. Inter-coder reliability will be 684 assessed with kappa statistic. 685 686 After all transcripts have been coded, the study team will extract and examine the content of 687 text linked to specific core codes and sub codes and identify ways in which certain themes are 688 analytically related. Identified relationships among themes may lead to more refined data 689 searches. Once patterns of relationships among themes and issues are established, the study 690 team will try to identify participants' accounts that support or refute these patterns. Identifying 691 and accounting for cases that "deviate" from an interpretative pattern enables us to test and 692 confirm the pattern's validity and robustness. Finally, the study team will attempt to map 693 themes onto the relevant domains of the CFIR framework to assess the framework's adequacy 694 in identifying all the important factors supporting or impeding the adoption of testing. If 695 emergent in these analyses, it will be possible to identify pathways through which adoption (of 696 lack of adoption) of testing evolves in the PC versus the control conditions. 697

Discussion 698
Our PC interventions, if shown to be effective and cost-effective, could be used at multiple 699 levels to provide ongoing support to OTPs in delivering HIV/HCV testing. This promising 700 approach should be adaptable to address HIV testing in other settings, including pharmacies, 701 dental care settings, and community centers. To our knowledge, this study is the first to test   Figure) Page 49, Line 963 (