Patient and Provider-level Barriers to Hepatitis C Screening and Linkage to Care

A Mixed-methods Evaluation

Robin T. Higashi; Mamta K. Jain; Lisa Quirk; Nicole E. Rich; Akbar K. Waljee; Barbara J. Turner; Simon Craddock Lee; Amit G. Singal

Disclosures

J Viral Hepat. 2020;27(7):680-689. 

In This Article

Methods

As previously described, we implemented a multi-modal HCV intervention at Parkland, Dallas County's safety-net integrated health system including 12 community-based primary care clinics, outpatient hepatology clinics and a tertiary hospital – all sharing a single EHR system.[3] Concurrent to this intervention, we conducted an evaluation in two phases of data collection: First, we conducted patient surveys, focus groups with clinic providers and staff, and structured observations of clinic flow and processes (Phase 1). Subsequently, we conducted semi-structured interviews with patients and providers to examine domains and trends of interest identified from initial phase data (Phase 2). All participants provided informed consent prior to study enrolment.

Phase 1 Data Collection: Patient Surveys, Provider and Staff Focus Groups, and Structured Observation

Patient Surveys. Our survey assessed (a) attitudes towards HCV and people with HCV, (b) knowledge about HCV and (c) sociodemographic information. Questions were selected because negative attitudes (eg fear or stigma) and lack of knowledge could contribute to reluctance to complete screening, diagnostic testing or treatment evaluation. The first section drew on social discrimination constructs and phrasing from HIV stigma instruments, selecting 10 items to assess HCV-related attitudes and feelings.[4,5] Although stigma and social discrimination should theoretically decrease with the shift to cohort-based screening, we included these constructs because prior research has indicated stigmatizing beliefs are often resistant to change.[6,7] In the second section, eight questions assessed knowledge about HCV infection, transmission, treatment, perceived reasons for testing and sources of information. Thirdly, our survey queried patient age, sex, race and ethnicity, language and insurance status. Our intention in surveying patients was to determine if patient knowledge (about HCV and HCV-related screening and treatment) and stigma would contribute to continued screening process failures after implementation of the HCV intervention. We believed these constructs could contribute to reluctance to complete screening/diagnostic testing as well as impacting patient engagement in treatment evaluation. Such findings would inform future intervention design components to improve patient knowledge and/or reduce other persistent barriers to HCV screening.

We pre-tested survey items, using established cognitive interview techniques with three English-speaking and three Spanish-speaking participants, at which point saturation was achieved (individuals excluded from findings reported below).[8,9] The final survey instrument required approximately 10 minutes to complete.

Surveys were administered to patients sampled from three different points along the screening-to-treatment continuum (Figure 1): (a) patients with positive HCV antibody who had not completed viral load testing ('antibody reactive'); (b) patients with positive HCV viral load who had not completed treatment evaluation ('HCV confirmed'); and (c) patients who completed treatment evaluation at the Hepatology Clinic ('clinic patients'). Exclusion criteria included patient refusal, uncontrolled hepatic encephalopathy or other medical conditions precluding survey completion, no phone number on file and primary language other than English or Spanish. Antibody-reactive and HCV-confirmed patients were identified from the EHR-based patient registry developed as part of the HCV intervention and recruited by telephone; clinic patients were also identified through the registry but recruited from the Parkland Hepatology Clinic waiting room. Research staff administered surveys by telephone for antibody-reactive and HCV-confirmed patients; clinic patients were offered the choice to complete the survey themselves or with the assistance of research staff. We aimed to conduct a sample of at least 50 surveys in each of the three groups, sufficient to inform domains for the second phase of our evaluation.

Figure 1.

Survey and Interview Participant Sampling. Adapted from Infectious Diseases Education & Assessment, 2018.21

Provider and Staff Focus Groups and Observation. We conducted focus groups at 9 of 12 Parkland primary care clinics. We invited providers and clinical staff to participate in a 30-minute open discussion to assess knowledge, attitudes and concerns towards HCV screening and linkage to care and interventions to promote either step in the continuum. Participants at each clinic varied from 8 to 27, depending on clinic staffing volume and volunteer interest. We did not sample from the remaining three clinics due to thematic saturation. We then observed a convenience sample of providers and staff during nonpatient-contact activities to identify potential gaps in the targets of our evaluation, for example communication or documentation not captured in the EHR. We recruited a purposeful sample of key process stakeholders for interviews during the observation process, including primary care nurse managers, primary care providers and HCV Clinic providers and staff.

Phase 2 Data Collection: Semi-structured Interviews

We developed an interview guide to explore themes identified from Phase 1, notably barriers to HCV screening process completion, strategies to enhance screening initiation, follow-up testing and linkage to care, and stigma. We interviewed three sets of individuals 15–18 months after implementation: (a) patients undergoing HCV screening and treatment evaluation, (b) PCPs and staff involved in HCV screening and (c) specialty providers and staff at the HCV Clinic. All interviews lasted 30–45 minutes and explored the following domains: experiences with HCV screening and referral to treatment evaluation processes pre- and post-intervention; communication among providers, staff and patients; knowledge of HCV and treatment; insurance coverage and out-of-pocket costs associated with treatment; and HCV-related stigma.

Patient Interviews. As above, we used the EHR patient registry to identify patients scheduled for appointments across the screening-treatment continuum, as well as opportunistic recruitment in the Hepatology Clinic waiting room. We recruited patients until we reached our target enrolment for the three points of the screening-to-treatment continuum (Figure 1).

Primary Care Providers and Staff Interviews. We interviewed PCPs with the highest and lowest HCV screening rates, defined by the proportion of eligible 'baby boomer' patients for whom HCV screening orders were placed in the prior 6–9 months. We also interviewed two nurse manager staff members, selected by years of experience with clinic processes, at the recommendation of the Director of Nursing.

Hepatitis C Clinic Provider and Staff Interviews. In HCV treatment clinics, we interviewed specialty providers and the clinic nurse navigator to assess perspectives on potential barriers to care.

Analysis

For surveys conducted in Phase 1, we used chi-square tests to compare levels of knowledge and stigma across the three cohorts of patients, using STATA 14 (StataCorp). Focus groups and semi-structured interviews were transcribed and analysed using NVivo 9.0 (QSR). Two researchers trained in qualitative methods (RH, LQ) developed an initial, deductive codebook corresponding to the semi-structured interview guide domains. Researchers jointly coded the first six transcripts ('test set', 13%), sampling from patient and provider transcripts at baseline and post-implementation, to refine codebook definitions and add emergent codes. Researchers double-coded the next 18 transcripts (40%) using the refined codebook, meeting weekly to resolve discrepancies. The lead analyst (RH) single-coded 14 additional transcripts (31%) until reaching thematic saturation at 38 of 45 transcripts (84%). The lead analyst then reviewed resulting thematic node reports and selected representative quotes and findings for discussion with the team.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....