Comparing Strategies for Implementing Primary HPV Screening

Cervical cancer screening is an important component of women's health care. Most adult women are familiar with the conventional screening modality, Pap test, which has successfully reduced the burden of cervical cancer in industrialized countries. However, Pap test has limited accuracy and can miss a progressing disease. Advancement in knowledge and technology has led to changes in national recommendations to focus on the testing of high-risk human papillomavirus (HPV) infection, the cause of cervical cancer. Screening with primary HPV testing detects more diseases compared with Pap test alone or co-testing, and requires less number of tests. However, despite the advantages of primary HPV screening over conventional approaches, the switch to primary HPV screening is limited in the United States. The scientific literature reports barriers at both the provider and women level, which include lack of knowledge, resistance, and attachment to Pap test. We currently have insufficient guidance on how to select and deploy implementation strategies most likely to facilitate use of newly recommended cancer screening modality. This project seeks to generate evidence regarding effective strategies to achieve successful implementation of the primary HPV testing for routine cervical cancer screening in a large community-based health care system. A successful implementation will be defined by uptake of the primary HPV screening, adequate knowledge of the HPV test for both patients and providers, and patient/provider satisfaction during the transition. This project is important to most adult women, as a timely adoption of the best evidence-based cancer screening approach means better patient outcomes. Further, the proposed project will not only inform about cervical cancer screening, but other clinical conditions when a physician practice change is recommended by professional societies and/or national guideline body. By engaging patients, providers and other professional stakeholders in this project, we ensure that successful project outcomes are those most important for women and their doctors. Further, the stakeholder partners will help ensure generalizability of our findings to other health care systems, design strategies that maximize completeness in data collection, and lead the dissemination effort for wide application of the knowledge to be gained in this project.

Definition: Extended description of the protocol, including more technical information (as compared to the Brief Summary), if desired. Do not include the entire protocol; do not duplicate information recorded in other data elements, such as Eligibility Criteria or outcome measures. Limit: 32,000 characters. Background and Significance: Strong evidence supports testing for the high risk human papillomavirus (HPV) infection, the etiology agent of cervical cancer, in routine cervical cancer screening. In 2014, the FDA approved the first test for primary HPV screening. Between 2015 and 2017, professional societies and national guideline bodies released practice recommendation for primary HPV screening. However, none of the large health care systems in the U.S. have systematically adopted this new screening strategy. Kaiser Permanente Southern California (KPSC) is preparing to transition to primary HPV screening in 2019, with considerable expected impact and barriers at multiple levels. There is a critical gap of knowledge on effective strategies to guide implementation of uptake of new evidence-based practice, especially around cancer screening where changes to clinical practice guidelines often created confusion among clinicians and patients. The overarching goals of this application is to generate insights and evidence regarding barriers and facilitators and effective strategies to achieve clinical practice substitution. Study Aims: SA1) Compare a local-tailored vs. a centralized approach for facilitating adoption of primary HPV testing for cervical cancer screening on (a) implementation outcomes including uptake of primary HPV screening, acceptability, appropriateness and feasibility; and (b) stakeholder-centered outcomes including knowledge, experience, behavior and satisfaction; SA2) Explain variations in implementation strategy effectiveness on study outcomes by multi-level factors; and SA3) Develop guidance for use of the effective implementation strategies in additional settings and for additional implementation problems. Study Descriptions: We will conduct a prospective, cluster randomized programmatic trial to compare and evaluate a local-tailored versus a centralized implementation approach. The local tailoring strategy will be guided by a structured process, using a menu of core functions and forms with evidence-based barrier assessment and intervention options. The centralized implementation will be based on the prevalent KPSC regional approach to new practice implementation, involving the design of a multi-component approach that is delivered in a relatively consistent manner. Twelve of the 14 KPSC medical centers will be randomized to receive one of the two implementation approaches, with two medical centers serving as pilot sites. The study subject will include screening age women, primary care and obstetrics & gynecology physicians, as well as clinic staff, administrators and operational leaders. The primary outcome of interest is uptake of primary HPV screening at the provider level. The secondary outcomes include stakeholder-centered outcomes such as knowledge and satisfaction, and additional implementation and system outcomes as well as implementation process evaluation. Data collection will be via electronic medical record extraction, patient and provider surveys, and semi-structured key-informant interviews. Multi-level models and generalized estimating equations will be used to evaluate the effect of the local-tailored approach on each outcomes of interest. Effect heterogeneity by multi-level factors will be examined by interaction terms. Content analysis will be used to evaluate qualitative data collected for Aims 1 and 3. We will use the Consolidated Framework for Implementation Research (CFIR) and the Theoretical Domains Framework (TDF) to inform our overall study approach and provide rigor and structure to our analyses.

Cervical Cancer, Human Papillomavirus, Health Knowledge, Attitudes, Practice, Health Care Utilization

This prospective study will use a cluster randomization design to compare two system-level implementation strategies for primary HPV screening, with outcomes measured at multiple levels. KPSC has 13 service areas linked to over 200 ambulatory clinics. Randomization will be performed at the service area level, since these are system-level implementation strategies. Service areas will first be matched in pairs on key attributes that may be associated with implementation success prior to randomization, such as the size and the number of clinics, access, and current screening rates. Thus, the two service areas in a matched pair will be most like each other in those attributes. Within each pair, the two service areas will then be randomized to determine which receives the local-tailored vs. hybrid usual care approach. The intervention period for both arms will take place immediately after the roll out of primary HPV screening.

The intervention arm will consist of six KPSC service areas randomly assigned to the intervention arm. Immediately after primary HPV screening opens at KPSC, the intervention arm will receive the local tailoring interventions. All providers (physicians, nurses and medical assistants) and department administrator from the primary care departments (family medicine and internal medicine) and the department of obstetrics and gynecology at these six service areas randomized to this arm will be included in the study, as well as female patients at these service areas between age 30-65 who received cervical cancer screening during the data collection period.

The hybrid-usual care arm will consist of six KPSC service areas randomly assigned to this arm. The hybrid usual care arm will receive regional educational activities for the transition (as will the intervention arm) before the roll out of primary HPV testing. However, they will not receive any research-led intervention or adaptation guidance after primary HPV screening opens at KPSC. All providers (physicians, nurses and medical assistants) and department administrator from the primary care departments (family medicine and internal medicine) and the department of obstetrics and gynecology at these six service areas randomized to this arm will be included in the study, as well as female patients at these service areas between age 30-65 who received cervical cancer screening during the data collection period.

The "guided local tailoring" approach will employ a standard structured process, including (1) convening a project team, (2) conducting a local diagnostic process to identify likely barriers using provider/patient survey and interviews with providers/administrators, (3) selecting from a pre-developed menu of implementation strategies categorized by core function (form and function menu), and (4) deploying the selected strategies in collaboration with local implementation and improvement consultants.

Proportion of primary HPV testing of all cervical cancer screening performed for women age 30-65 at the provider level.

Knowledge about the efficacy of the new test. We will assess provider knowledge via provider surveys which elicits information on provider knowledge and practice on cervical cancer screening. Provider delivery of education during patient screening visit will be measured in provider survey, including whether education is delivered and average time spent on patient education/counseling. Provider/staff satisfaction of the transition process and provider resistance to the new screening modality will be ascertained via survey questions.

Knowledge about the new test and why the change is made. To measure this outcome, we will use and modify validated survey questions on HPV test knowledge. Emotional reactions (stigma and shame) of a positive HPV test result will be measured using survey questions developed by Waller et al. Patient satisfaction during the transition may be measured with survey questions such as "Were you able to discuss all your concerns/questions about the new screening approach with your provider?" "Where these concerns/questions adequately addressed?" "Did you get your test result back in a timely manner?" "Did someone explain to you what the test result mean?" and "How satisfy are you with the screening experience?"

Time to colposcopy after a positive test result represents the combination of correct and timely referral, patient compliance, and availability of colposcopy appointments and will be measured using laboratory and utilization data collected in KPSC's electronic medical records.

For patients, only female patients will be included in this study because male patients will not go through cervical cancer screening.

Inclusion Criteria: This is a cluster randomized trial. All KPSC's 12 service areas except Orange Country will be randomized and included in this study. All providers (physicians, nurses and medical assistants) and department administrator from the primary care departments (family medicine and internal medicine) and the department of obstetrics and gynecology at these 12 service areas randomized to this arm will be included in the study, as well as female patients at these service areas between age 30-65 who received cervical cancer screening during the data collection period. Exclusion Criteria: Patients who are younger than 30 years old Providers working for departments other than Ob/Gyn, family or internal medicine

Hahn EE, Munoz-Plaza C, Altman DE, Hsu C, Cannizzaro NT, Ngo-Metzger Q, Wride P, Gould MK, Mittman BS, Hodeib M, Tewari KS, Ajamian LH, Eskander RN, Tewari D, Chao CR. De-implementation and substitution of clinical care processes: stakeholder perspectives on the transition to primary human papillomavirus (HPV) testing for cervical cancer screening. Implement Sci Commun. 2021 Sep 23;2(1):108. doi: 10.1186/s43058-021-00211-z.