Protocol to Evaluate the Implementation of an Enterprise-Wide Initiative to Increase Access to Lung Cancer Screening in the Veterans Health Administration

The Veterans Affairs Partnership to increase Access to Lung Screening (VA-PALS) is an enterprise-wide initiative to implement high quality lung cancer screening programs at VA medical centers (VAMCs). VA-PALS will be using implementation strategies that include program navigators to coordinate screening activities, trainings for navigators and radiologists, an open-source software management system, tools to standardize low-dose computed tomography (LDCT) image quality, and access to a support network. Individual VAMCs can utilize strategies according to their local contextual factors. In this protocol, we describe the planned program evaluation for the initial 10 VAMCs participating in VA-PALS. method of conducting patient identication, eligibility conrmation, smoking cessation counseling, shared decision-making, appointment LDCT appointment, communication of results and scans.

and radiologists, a software management system, low-dose CT image standardization tools, and an expert support network.
Several implementation science frameworks (CFIR, RE-AIM, FRAME) inform the design of the program evaluation. Implementation measures will include evaluation of the inner setting through interviews and assessment of organizational readiness, estimation of the population reached, adoption, implementation of the program, clinical and process outcomes, and long-term maintenance of lung cancer screening programs.
Measurement of contextual factors as well as assessment of work ow processes and adaptations to overcome barriers may generate new knowledge that impacts implementation of lung cancer screening programs more broadly.

Background
In 2020, lung cancer is expected to cause 135,720 deaths in the United States, which is more than colorectal, breast and prostate cancer deaths combined. 1 Most patients (74%) are diagnosed with lung cancer at advanced stages when there is less chance for curative treatment, which contributes to its high mortality. 1 Stimulated by the results of the Early Lung Cancer Action Project (ELCAP), 2-4 two large U.S.
(the National Lung Screening Trial [NLST]) and European (the Dutch-Belgian Lung Cancer Screening Trial [NELSON]) multi-center, randomized-controlled trials demonstrated that early detection with low-dose computed tomography (LDCT) decreased lung cancer-speci c mortality in high-risk individuals. 5,6 Following the results of these trials, many professional organizations developed guideline recommendations [7][8][9][10][11] and the United States Preventive Services Task Force (USPSTF) issued a grade B recommendation for annual lung cancer screening with LDCT in the high-risk population. 12 The USPSTF de nes high-risk individuals as being between the age of 55 to 80, current or former smokers, who have quit within the past 15 years, and have a minimum of a 30 pack-year history of cigarette smoking. 12 Under the Affordable Care Act, LDCT screening is considered a preventive care service with coverage by most private insurance plans. 12 In 2015, the Centers for Medicare & Medicaid Services (CMS) added coverage for LDCT screening high-risk individuals 13 and integrated smoking cessation as part of the screening program. In fact, screening has created a "teachable moment" that can change smoking behavior. [14][15][16][17][18][19][20][21] The newly proposed USPSTF guidelines have suggested even broader entry criteria as does the European Union. 11,22 The Veteran population has higher cigarette smoking rates than non-Veterans 23,24 and lung cancer screening programs may be particularly bene cial in this population. The Veterans Health Administration (VHA) has begun to implement lung cancer screening programs that combine early detection with primary prevention (tobacco cessation). [25][26][27] Between 2013 and 2015, VHA conducted the Lung Cancer Screening Clinical Demonstration Project at eight Veterans Affairs Medical Centers (VAMCs). 28 After two years, participating VAMCs screened approximately 2,100 Veterans and identi ed 31 lung cancers, of which 64.5% were stage I. 26 Signi cant variability in implementation was observed between sites, especially around the processes of identifying eligible Veterans, identifying the staff who engaged in shared decision-making, assigning responsibilities to coordinators, trainings for primary care and radiology, communicating screening results to patients, and assigning responsibility for follow up of abnormalities not related to lung cancer (emphysema, coronary artery disease, etc.). 26 This initial evaluation by VHA was limited in its exploration of local contextual factors as well as associations between processes of care and adaptations over time. 26 Lung cancer screening utilization in the United States is exceedingly low (< 14%), [29][30][31] including in VHA. 32 This makes strategies to improve implementation critically important. Building upon VHA's Lung Cancer Screening Clinical Demonstration Project, the Veterans Affairs Partnership to increase Access to Lung Screening (VA-PALS) initiative was developed by its leadership team with the overall vision to implement lung cancer screening programs that improve Veteran outcomes through evidence-based and high-quality early lung cancer detection and primary prevention. [33][34][35] VA-PALS was designed to consider the unique local environments, distinct culture, and varying levels of resources across VAMCs in disparate geographic locations (Fig. 1). The overarching aims of VA-PALS are to: 1) Implement evidence-based lung cancer screening programs by developing and deploying the opensource VAPALS-ELCAP Management System adapted for the needs of the VA to track Veterans from the initial contact through a long-term screening and diagnostic process that could eventually lead to lung cancer diagnoses, providing training of navigators, radiologists and other team members, assessing screening protocol adherence, continuous quality assurance assessment, and assessing outcomes at the 10 high-volume VAMCs.
2) Perform a formal evaluation of the 10 nascent lung cancer screening programs to measure the impact of increasing Veterans access to lung cancer screening, rates of early detection, and opportunity for mortality reduction.
Lung cancer screening implementation has not been rigorously studied using an implementation science theory-based approach. 26,36 The purpose of this protocol to describe the scienti c frameworks and protocol developed to inform and evaluate the implementation of lung cancer screening programs at VAMCs participating in VA-PALS (Fig. 1). VA-PALS lends itself as a unique opportunity to explore theorybased relationships between types of strategies (the navigators as a team-based strategy, the phantoms as an intervention characteristic-based strategy, and the software management system as a work ow processes-based strategy) and their effects on evaluation constructs (reach, effectiveness, adoption, implementation and maintenance) across multiple settings. In addition, we will also validate an organizational readiness measurement tool 34 in a national cohort of VHA employees, assess reach at multiple points along the screening process, explore barriers of rural Veterans for future outreach interventions, and will report adaptations in screening processes overtime in multiple programs. These data will likely generate new knowledge that will in uence lung cancer screening program implementation more broadly within and outside the VHA.

Aim
The aim of this program evaluation is to develop and report lung cancer screening work ow and clinical outcome measures as well as contextual factors that in uence implementation at sites participating in VA-PALS.

Implementation Strategies
Each VAMC participating in VA-PALS has a designated site champion and will receive the following: (1) a full-time lung cancer screening navigator (NP or PA) to support program development and coordinate screening-related care], (2) a comprehensive, open-source lung cancer screening management software system 33,35 (3) training for navigators and radiologists, (4) expert guidance to standardize screening protocol adherence and LDCT imaging quality, and (5) a support network led by experienced screening leaders.
Decisions regarding individual site's lung cancer screening program structure and work ow will be determined at the local level depending on the environment, resources available, and culture. As such, sites will be able to adapt their processes over time for different reasons (e.g., contextual pressures, leadership support, trial and error, etc.). These changes will be measured to enhance our understanding of how naturalistic implementation in uences processes and the impact these changes may have on overcoming barriers. Logic Model (Figure 2)

Inputs -People, Technology
People. VA-PALS will be overseen by an inter-professional leadership team of radiologists, a pulmonologist, a radiation oncologist, a screening navigator, and technology specialists at Atlanta VA Medical Center, Phoenix VA Health Care System, Early Lung and Cardiac Action Program at Mount Sinai Health System (MS-ELCAP), and Paraxial LLC who have expertise in lung cancer screening, treatment, research and healthcare computer science. 35 The software team is comprised of software developers with experience in VA networks and informatics experts at MS-ELCAP. The VA-PALS program evaluation team includes health services researchers at VA Tennessee Valley Healthcare System Geriatric Research, Education and Clinical Center and Vanderbilt University Medical Center working together with the VA-PALS leadership team and participating VAMCs. Participating VAMCs have or will have local teams that include a site champion, program director(s), screening program navigator, clinical providers, and software engineers.
Technology. The ELCAP Management System was designed to manage and organize lung cancer screening activities and has been in use since 1992. 35 This software system has been used at 82 institutions in 10 countries and is a tool for navigators and radiologists to document, organize, and manage screening-related activities from the initial contact, screening follow up, evaluation of screening ndings, and treatment of those diagnosed with lung cancer. 2,4,35 This software also serves as a database for improvement initiatives. 35 Other technological inputs include the CT scanners for screening examinations and small nodule phantoms of the CT scanning protocols for quality assurance. 35 A phantom is a specially designed object that can be scanned by a CT scanner and provides the information about the performance of the CT scanner and scanning parameters which can be used to optimize images CT screening.
Activities Support Network Activities. The leadership team will provide guidance and support for each VA-PALS program through planned activities and site monitoring. It will provide standardized training for all newly hired navigators and radiologists at each VA-PALS site along with a support network for the navigators, site champions and program directors through regular conference calls, national workshops, and an operational guide to lung cancer screening. The leadership team will also manage a website (www.vapals.org) to disseminate information and shared resources.
Quality Assurance Activities. The software team will translate the ELCAP Management System into the open source VAPALS-ELCAP Management System with guidance by the MS-ELCAP and leadership team. The software documents enrollment, screenings, follow-up, interventions, and treatment if diagnosed with lung cancer. It also provides management and quality assurance reports of screening protocol adherence and follow-up. This VAPALS-ELCAP Management System will then be integrated with the Veterans Health Information Systems and Technology Architecture (VistA) and the VHA's electronic health system electronic medical record, Computerized Patient Record System (CPRS). 35 It will be tested and piloted at the Phoenix VAMC prior to dissemination to other participating VA-PALS VAMCs. The leadership team will provide each VAMC with an imaging phantom and the software team will provide feedback on their scanners' imaging quality. 35 Phantoms will measure imaging sensitivity, dynamic range, contrast/detail detectability, and spatial resolution. This calibration tool will ensure proper parameters are being used for LDCT imaging to assure high-quality assessment of lung nodules. 35 Local teams at each VAMC will use the VAPALS-ELCAP Management System and its quality assurance reports and the CT phantom assurance tools for program activities, adherence to the screening protocol, standardization of image parameters, and improvement initiatives.
Training Activities. Navigators will be trained by the MS-ELCAP and leadership teams on screening eligibility, shared decision-making, use of the management system, and coordination of screening-related activities. Radiologists will also be trained in the work ow and use of the VAPALS-ELCAP Management System.
Local Lung Cancer Screening Activities. Each site director will be tasked with hiring a lung cancer screening navigator (NP or PA), engaging their local stakeholders and administrators, developing a screening work ow, and educating local providers and staff about the program. Implementation of the lung cancer screening program will include the following activities: identify high-risk Veterans eligible for screening, perform shared decision-making, offer resources and/or treatment for smoking cessation, perform the LDCT scan, ensure structured radiology reporting, facilitate any indicated additional workup, and schedule either repeat annual LDCT or work up of ndings suspicious for lung cancer as appropriate.
Data Collection Activities. The primary clinical and implementation outcome measures were developed by the VA-PALS leadership team, as part of their continued overall responsibility, together with the program evaluation team. The combined teams will work together to collect data from the management system, VHA's Clinical Data Warehouse (CDW), surveys, and interviews, will analyze results, and will share the data with the larger VA-PALS network. The program evaluation team will also participate in leadership and program meetings.

Outcomes
Short-term Outcomes: The short-term outcomes consist of VA-PALS's reach, adoption, effectiveness and implementation measures to be used within the rst 24 months of the program, described in detail below and Table 1.
Long-term Outcomes: The long-term outcomes are the reach, implementation, effectiveness and maintenance measures to be used beyond 24 months since start of the program, described in detail below and in Table 1.
Target Population. The target population will include Veterans eligible for lung cancer screening, providers and staff involved in screening activities, and leadership at each VA-PALS site. Veterans eligible for lung cancer screening generally include those meeting the USPSTF criteria 12 ; at least one site is using a more broad eligibility recommendation. 37 Providers involved in screening activities will include primary care providers (physicians, advanced practice providers, and physicians-in-training), radiology providers, and navigators for each program. Staff involved will include primary care nurses, schedulers, clinical application coordinators (CACs) and radiology technicians/technologists. Leadership will include the local program leadership as well as service leaders within primary care, radiology, and specialty services and executive leadership team.
Theoretical Frameworks a. Implementation Framework Implementation of an evidence-based practice into healthcare delivery depends upon multiple internal and external contextual factors that ultimately in uence the processes and outcomes of healthcare delivery. We selected the Consolidated Framework for Implementation Research (CFIR) to guide this program evaluation based on this framework's ability to provide structure to approaching the implementation of a complex, inter-professional program in a pragmatic manner. 38 The CFIR explores the characteristics of an evidence-based practice, the inner and outer settings of where the evidence-based practice is deployed, the characteristics of the individuals interacting with the evidence-based practice, and the process of implementing the evidence-based practice. 38 This mixed-methods program evaluation measures key elements of the ve major CFIR domains: (1) characteristics of the evidence-based practice (image acquisition and CT standardization), (2) inner setting of each VA-PALS site (organizational readiness, resources) and (3) outer setting of VHA as a whole (policies), (4) the characteristics of individuals involved in the screening process (knowledge and beliefs, self-e cacy, and motivation), and the (5) process of implementation at each VA-PALS site (site process maps for implementation initially and adaptations over time) (Table 1). Thoroughly exploring each of these domains will inform future implementation strategies ( Figure 3). 38, 39

b. Evaluation Frameworks
The Reach, Effectiveness, Adoption, Implementation, Maintenance (RE-AIM) framework was used to develop the program evaluation measures. 40 This evaluation framework in the cancer screening setting 41 and provides a comprehensive structure to assess real-world clinical programs. VA-PALS evaluation measures re ect the activities at each VAMC and clinical and process outcomes of each lung cancer screening program. For each site, the date of hire of the navigator will be considered the program start date. Evaluation of all measures of reach, effectiveness and maintenance will begin at the program start date and continue for the next 36 months, as de ned by the navigator hire date. The VA-PALS leadership and evaluation teams will compare measures (i.e. number of Veterans reached, screening utilization, etc.) over time in VA-PALS and non-VA-PALS VAMCs in an interrupted time series analysis. Further, we will compare outcomes in months 1-24 to months 25-36 to evaluate for sustainability in VA-PALS and non-VA-PALS VAMCs (Table 1).

Reach
The management system will capture descriptive information (e.g. age, smoking history, rural status), method of contact for screening (referral from primary care, specialty care or smoking cessation, outreach or self-referral). In order to have a consistent denominator of the eligible population across all sites, we will estimate the number of Veterans potentially eligible for lung cancer screening according to the USPSTF criteria. As used before, 32 we will obtain counts of Veterans between the ages of 55 through 80 seen at each VAMC each year. To calculate the percentage of age-appropriate Veterans with an eligible smoking history, we will multiply the age-appropriate population by 32% for each VAMC in each year; thirty-two percent was the percentage of age-appropriate Veterans who met the smoking eligibility criteria in the VHA's Clinical Lung Cancer Screening Demonstration Project. 26 All of the components of the program's reach will be designed to capture the number of eligible Veterans at each stage of the process including Veterans approached for screening, eligibility as well as agreement and refusal both before and after the shared decision-making process, and unique Veterans completing initial screens (Figure 4).

Effectiveness
The effectiveness of the program will be evaluated using data acquired from VHA's Corporate Data Warehouse (CDW) and the VAPALS-ELCAP Management System. The measures chosen are re ective of the results of screening, the frequency and type of follow-up procedures performed, the number of cancers detected, type and stage of cancers, treatment received, other abnormal ndings not suspicious of lung cancer and their follow up (e.g. cardiovascular disease, COPD, etc.), 42 survival of screened Veterans, the number of smoking cessation quit attempts, 14, 16 the type of smoking cessation services provided, wait times for services provided, and overall experiences of those involved in the screening process (see Qualitative Section) ( Table 1). We will compare different measures of effectiveness (number of LDCT scans, cancers diagnosed, stage at time of diagnosis, etc.) before and after implementation of the navigator and VAPALS-ELCAP Management System using an interrupted time series analysis. Other measures of effectiveness (wait times for services provided, work ow process maps, and overall experiences of those involved in the screening process) will also be tracked over time using the CDW, process mapping and surveys. Adoption Adoption will be assessed by describing the characteristics of the VAMC facilities participating in VA-PALS. Organizational readiness for change will be measured using a validated survey tool to assess change valence, change commitment, and change e cacy 34,43 to understand each VAMC's inner setting (Table 1). Organizational readiness for change is de ned as the "extent to which organizational members are psychologically and behaviorally prepared to implement organizational change." 44 Change valence is de ned as the organization's members belief that pursuing change is bene cial and valuable to the organization. 45 To describe the program implementation process, we will obtain interviews from sites and program leadership: (1) rst date of posting the navigator position for hire by human resources, (2) rst day navigator starts clinical work, (3) date of navigator training, (4) date of installation of the VAPALS-ELCAP Management System, and (5) dates of imaging phantom use and CT scanner standardization.
The number of unique providers and types of providers (specialty; physician, advanced practice provider) referring to the program within the rst 24 months since hiring of the coordinator will be captured in the CDW and VAPALS-ELCAP Management System. Finally, to describe adoption at the site-level, we will capture alignment with VHA priorities through in-depth interviews of program leadership at each site. Implementation Implementation of lung cancer screening programs at the 10 VA-PALS sites will be captured via local program engagement to assess delity of the program (Table 1). Speci cally, delity refers to the degree to which navigators, the management system, trainings, screening protocol adherence, and imaging phantoms are utilized at each site. Measures of delity will include the retention time of navigators, navigator workload, the number of radiologists who use the management system, the degree to which the navigators use the management system for screening related activities (approach/enroll Veterans, write notes, follow-up screening results, manage downstream follow-up, etc.), and the dates of phantom use and image standardization ( Table 1). Measurement of each VAMC's inner setting will assess team characteristics, the processes involved in screening, and barriers and facilitators. We will assess how smoking cessation is incorporated into each local program (who performs it, type of services performed, when it is performed within screening), how shared decision-making discussions are performed, and how and when results are disseminated to patients and referring providers (see section on Process Maps below). These measures will be primarily obtained through interviews with sites and supplemented with data from the VAPALS-ELCAP Management System (Table 1).

Maintenance
We will measure the extent to which VA-PALS is sustainable over time using a combination of data collected from the VA-PALS-ELCAP Management System and interviews with local program leadership. Maintenance measurements will consist of: number of subsequent screenings, screening work ow adaptations during months 25-36, the number of unique Veterans enrolled in each program during months 25-36, the number of unique providers who refer Veterans in each program during months 25-36, navigator retention, and sustained leadership buy-in (Table 1). We will also continue to track radiologists' and lung cancer screening navigators' continued adherence to the VA-PALS ELCAP Management System per the delity measures described in the Implementation Section above.

Process Maps and Adaptations
Each VA-PALS site will describe their initial processes of clinical work ow, which will be used to generate a process map at the start of implementation. These process maps will be generated via telephone interviews with each site's lung cancer screening navigator. Each map documents the site's method of conducting patient identi cation, eligibility con rmation, smoking cessation counseling, shared decisionmaking, appointment scheduling, LDCT appointment, communication of results and follow-up scans. Each site will be contacted every 6 months to review their process map and delineate work ow adaptations as implementation matures. Using the Framework for Reporting Adaptations and Modi cations to Evidence Based Interventions (FRAME), 46 we seek to understand how and why processes are adapted as well as how these adaptations in uence implementation outcomes over time.
Adaptations will be noted across the following FRAME domains: 1) when and how over the course of implementation were modi cations made, 2) were these changes planned and proactive (e.g., intentional adaptation) or unplanned and reactive (e.g., in response to other forces), 3) who decided changes were necessary, 4) what was modi ed, 5) at what level in the system of delivery were changes made, 6) what types of changes were made in content or context, and 7) the reasons modi cations were made and what contextual drivers were at play. Once process maps are developed or revised, they will be reviewed with each site for accuracy. Adaptations will be correlated with implementation outcomes (work ow, barriers and facilitators).

Qualitative Evaluation
Qualitative data will supplement the quantitative measures to recognize elements that were successfully implemented, and elements that posed a challenge at each site ( Table 2). In-depth interviews among providers, staff and leadership at ve sites with different process maps will be used to assess barriers and facilitators around the main factors related to implementation. Key questions asked will be: (1) "Tell me about what role you currently have in the lung cancer screening program at your VA" (CFIR individual/team characteristics); (2) "What is di cult about implementing a lung cancer screening program?" (RE-AIM implementation); (3) "How well does lung cancer screening t with existing processes and practices in your VA?" (inner setting, implementation climate, compatibility); (4) "Are meetings, such as staff meetings, held regularly to discuss work process and practices such as lung cancer screening?" (inner setting -network and communication); (5) "What could help you to continue to perform highquality LDCT screening in your VA?" (RE-AIM maintenance). Targeted participants for interviews include primary care and radiology providers, staff, and leaders. All interviews will be audio-recorded and transcribed for subsequent thematic analysis, a standardized system that produces valid and reliable interpretations. All transcripts will be coded to establish a hierarchical coding system that categorizes comments contextually into themes that occur across multiple interviewees. 47,48 Discussion This report presents a structured and comprehensive evaluation protocol that will assess lung cancer screening program implementation across multiple VAMCs within an integrated healthcare system that has a decentralized leadership structure. 49 This affords an opportunity to evaluate lung cancer screening program implementation in disparate cultures and clinical environments where priorities and resources may vary. The protocol will assess clinical process adaptations over time and examine how these adaptations in uence implementation. The measures described and the overall program evaluation will generate knowledge that may in uence the implementation of lung cancer screening programs more broadly throughout the VHA and beyond. It may help address the current national estimates that demonstrate exceedingly low rates of lung screening across the US. [29][30][31][32] This program evaluation will advance the eld of implementation science primarily through its theorybased approach. We will explore the interaction of strategies that act at the team, intervention and process of care levels (i.e. navigators, phantoms, management system) with constructs within implementation science frameworks, or perhaps, discover new constructs that are important in implementation of complex, imaging-based screening programs. Furthermore, we will be the rst to validate an organizational measurement tool 34 in a large, VHA population. Finally, we will assess the Veteran reach of lung cancer screening at multiple points along the screening pathway and explore barriers of rural Veterans and adaptations to screening programs over time, areas that have not been well studied in the VHA.
There are several unique and important challenges in implementing lung cancer screening compared to other cancer screenings that may be contributing to delayed uptake since publication of the USPSTF recommendations in 2014. First, identi cation of eligible individuals at-risk is di cult because this step depends upon accurate smoking histories, which are often unavailable, recorded incorrectly or in an unstructured format in the medical record. 26, 50 Then, if eligible individuals are identi ed, screening participants and providers are encouraged (and required by CMS for reimbursement) to engage in shared decision-making during which the potential bene ts and harms of lung cancer screening are discussed; this requires time as well as a clinician who is con dent in their ability to perform a shared decisionmaking discussion. 13,51 Furthermore, programs are encouraged to incorporate smoking cessation services for current smokers, which is critically important but adds time and requires clinician familiarity with this service. 12, 13 VA-PALS programs plan to address these challenges by shifting some of these tasks away from busy primary care clinicians, streamlining work ow, and coordinating downstream screening-related activities. This program evaluation can shed more light on this model.
There are several operational considerations in this program evaluation plan. Some measures of program effectiveness (clinical outcomes) rely upon the timing of the installation of the VAPALS-ELCAP Management System at all 10 sites. This is also during a time when the VHA is planning to transition its electronic health record system to Cerner. The VAPALS-ELCAP Management System will be open-source and compatible with Cerner, though such transition may interrupt its functionality. Next, although we describe a comprehensive, theory-based program evaluation, additional limitations may emerge given a reliance on surveys and in-depth interviews to capture patient and provider experiences. As with all survey and qualitative research, self-reported outcomes are limited by recall bias, social desirability bias, and selection bias. We will attempt to overcome these challenges by seeking a high response-rate with local champions and research incentives. We will also conduct a su cient number of interviews to reach thematic saturation. Another consideration is that a few of the data elds within the VAPALS-ELCAP Management System allow for free text in which users can type their individual responses. The study team will need to analyze these data manually, which may lead to potential misclassi cation or misinterpretation. There are also some measures that are not directly captured by the VAPALS-ELCAP Management System, such as how communication of screening results are relayed to patients and providers. The study team will rely on site self-reporting for these measures. Our denominator is imperfect as some sites may use more broad eligibility criteria and smoking histories are not accurately captured in the electronic health record. However, in order to have a consistent denominator across all VAMCs, we will use the same age criteria and estimate the proportion of those who are age appropriate for screening and meet smoking history criteria according to a prior study. 26, 32 Finally, the USPSTF is expected to expand the eligible screening population to include a younger age (starting at age 50) and less smoking history (at least 20 pack-years) in the coming months. 22 This will increase the reach of Veterans, making lung cancer screening a routine service to an even larger population.
In summary, we report a theory-based, comprehensive program evaluation protocol for a multi-site lung cancer screening initiative that describes implementation and clinical outcomes. The results from this program evaluation will generate new knowledge that may in uence the implementation of lung cancer screening programs more broadly as well as advance our scienti c understanding of implementation of image-based screening tools in diverse settings.

Consent for Publication
Not applicable.

Availability of Data and Material
Datasets used in this study will be available to VA researchers in the VA's Corporate Data Warehouse upon appropriate VA ethics and data access approvals. De-identi ed, aggregated data from the VAPALS-ELCAP management and tracking system, interviews, and surveys will be available upon reasonable request to the corresponding author and with VA-PALS leadership approval.

Competing Interests
DFY is a named inventor on a number of patents and patent applications relating to the evaluation of diseases of the chest including measurement of nodules. Some of these, which are owned by Cornell Research Foundation (CRF), are non-exclusively licensed to General Electric. As an inventor of these patents, DFY is entitled to a share of any compensation which CRF may receive from its commercialization of these patents. He is also an equity owner in Accumetra, a privately held technology    Thoroughly exploring each of these domains will inform future implementation strategies Figure 4 unique Veterans completing initial screens