Evaluation of the Cherokee Nation Hepatitis C Virus Elimination Program in the First 22 Months of Implementation

This cohort study evaluates the first 22 months of implementation of the Cherokee Nation Heath Services community-based program designed to eliminate hepatitis C virus (HCV) infection.


Details of CNHS HCV screening strategies and treatment activities implemented by period
Universal screening was initiated on November 1, 2015 as part of the CNHS elimination program. Universal screening included a one-time screen for all individuals aged 20-69 years who accessed care at CNHS. Various screening strategies were implemented at WWH hospital and in the satellite clinics to achieve universal screening during the elimination program period, November 1, 2015-August 31, 2017. These strategies, presented the main manuscript, are described in more detail below and in eTable 1.
WWH hospital implemented a laboratory-triggered screening (LTS) process in December of 2015. The LTS process involved collecting an additional tube of blood for HCV antibody testing from all individuals aged 20-69 years who were undergoing phlebotomy for any other reason. On March 31, 2016 LTS was discontinued because there were concerns that it did not allow for an "opt out process". LTS was resumed on May 19, 2016 with the condition that a patient had an up-to-date consent for treatment form on file in the CNHS EHR system. This form is a general medical care consent document that all individuals who access the CNHS have to sign to receive medical care and is renewed annually. This form states that the individual will be screened for HCV and HIV at the medical providers' discretion with an option to opt out of these tests. In July 2016, an EHR HCV screening reminder in the new EHR system replaced the LTS with consent to treat on file approach. Under the default EHR settings, this HCV screening reminder was not automatically visible to providers during routine EHR system use; therefore, training was offered throughout CNHS in July and August 2016 to show providers where to find the reminders in the new EHR system. Point-of-care HCV rapid-test screening was implemented using OraQuick rapid HCV tests kits (OraSure Technologies, Inc) in March 2016 at the hospital dental clinic, while point of care testing was initiated at the WWH hospital Behavioral Health clinic in August 2016 but discontinued after one month due to logistical issues.
On-site HCV screening with OraQuick rapid HCV tests kits replaced send-out tests for universal HCV screening at all satellite clinics in January 2016. Rapid test kits were used in the clinic laboratory on blood samples collected through phlebotomy. Hence, the test results were not available during individuals' medical visit. If a test was positive, the individual was contacted and scheduled for RNA testing. Point-ofcare HCV rapid screening using a finger prick was expanded to a second dental clinic located in a satellite clinic in June 2016. Pre-elimination Program Period: Risk-based plus birth-cohort period. Some HCV activities occurred at WWH hospital and in the satellite clinics during the pre-elimination program period, October 1, 2012-October 31, 2015. These activities are described in more detail below and in eTable 2.
An EHR HCV screening reminder was installed in July of 2013 and discontinued in July of 2015 when the EHR system was replaced. Installation of a HCV screening reminder in the new EHR was requested but it did not occur until July 19, 2016 (see above). The HCV ECHO program used telehealth technology 2 and online modules 3 to train and support CNHS primary care providers and pharmacists to develop knowledge in HCV management and treatment and enable them to deliver best practice HCV care. The CNHS HCV ECHO program initially consisted of seven HCV care providers. The program had a hub, physically located at WWH hospital, and four spokes located in the outlying clinics. The hub was composed of a lead clinician (Infectious Disease Specialist), a behavioral health psychologist, a case manager, an ECHO coordinator, and onsite Information Technology support. The spokes were staffed by a total of six clinicians, including two physicians, two nurse practitioners, and two clinical pharmacists, working at the four outlying clinics.
As part of the CNHS HCV ECHO, the spokes' providers participated in weekly HCV clinics via videoconference, called "Knowledge Networks", where they presented their cases by sharing individual medical histories, laboratory results, treatment plans, and questions about best practices and individual challenges. The hub team provided advice and clinical mentoring during these sessions. Additionally, these case-based discussions were supplemented with short didactic presentations to improve content knowledge. Have certain medical conditions, including persons: who received clotting factor concentrates produced before 1987, were ever on long-term hemodialysis, with persistently abnormal alanine aminotransferase levels (ALT), or who have HIV infection; 3. Were recipients of transfusions or organ transplants before July 1992, including persons who: were notified that they received blood from a donor who later tested positive for HCV infection, or received a transfusion of blood, blood components, or an organ transplant; 4. Had a recognized exposure such as: healthcare, emergency medical, and public safety workers after needle sticks, sharps, or mucosal exposures to HCV-positive blood; and 5. Children born to HCV-positive women.

Statistical analyses
For the CNHS HCV Cascade of Care, the number of people with current HCV infection was estimate for the whole CNHS population who had accessed care at least once between October, 1 2012-August 31, 2017 by multiplying the gender-specific cumulative incidence of current HCV infection calculated using the EHR data by the number of males and females accessing the CNHS from October 1, 2012 to August 31, 2017. The data gathered from the EHR on individuals with newly detected HCV RNA was used to calculate the percentages for each cascade of care element for each period. For these estimates, the date of first identification of current infection (i.e., first RNA test), the first date of evaluation, the date treatment was initiated, and the date SVR12 was reached were used to estimate the number of individuals in each cascade of care element. There were some individuals diagnosed in a previous period, but evaluated in a later period. Individuals evaluated during the elimination program period, regardless of their diagnosis date, were included in the evaluation, treatment, and cure elements of the cascade of care. Using the Infectious Disease clinic database, the proportion of individuals cured in the elimination program period was estimated according to the intent-to-treat (ITT) and a Per Protocol Treatment (PPT) approach analyses. The ITT cure proportion was estimated by dividing the number of individuals initiating treatment who attained SVR12 by all those who initiating treatment, irrespective of SVR12 data availability. The PPT proportion was estimated by dividing the number of individuals who attained SVR12 by the number of those who initiated treatment with SVR12 data available. Only individuals who had accessed the CNHS between October 1, 2012 and October 31, 2017 and had a documented HCV antibody or RNA in the EHR were included in these elements of the cascade of care. Cure proportions were evaluated up to April 15, 2018.
First-time screening coverage was calculated by dividing the number of newly HCV-screened individuals by the number of individuals eligible for HCV screening, defined as those aged 20-69 and accessing care at least once in CNHS during each period and not previously screened for HCV antibodies. In addition, the total proportion screened was calculated by dividing the total number of individuals screened during each period by all individuals accessing CNHS during that period, regardless of previous screening.
Bayesian analyses used a Beta distribution to estimate the number and cumulative incidence of anti-HCV seropositive individuals among those newly screened within the period. The cumulative incidence of newly detected anti-HCV cases was calculated by dividing the number of newly identified anti-HCV seropositive individuals among the total number of newly screened individuals during the period. The number of individuals with newly detected current HCV infection among seropositive individuals with missing HCV RNA test results was estimated by multiplying the proportion of newly detected current HCV infection among anti-HCV seropositive individuals tested for HCV RNA by the number of anti-HCV seropositive individuals without an HCV RNA test result. This estimate was added to the number of newly detected HCV RNA-positive individuals to obtain an estimated total number of individuals with current HCV infection among all newly screened. The cumulative incidence of current HCV infections for each period was assumed to follow a beta distribution with alpha parameter equal to the estimated total number of current HCV individuals +1 and beta parameter equal to the total number of individuals screened overall minus the estimated total number of current HCV individuals +1. The CNHS population estimates for the frequency and cumulative incidence were estimated by period and birth cohort. Additionally, since the two periods differed by number of months, the annual mean number of newly detected HCV infections were estimated within each period by summing the number of newly detected HCV infections for each month in that period, then dividing the resulting total by the number of months within that period and multiplying the result by 12.
Chi-square tests were used to compare estimates by birth-cohort and period. Chi-square tests estimates and 95% confidence intervals were generated in SAS 9.4. Cumulative incidence of anti-HCV and current HCV infection and corresponding 95% Bayesian Credible Intervals (95% BCI) were estimated using Logistic regression analyses were used to estimate the odds of being cured from HCV by demographic group, gender and age, HCV genotype, and severity of liver fibrosis, during the elimination program period. Logistic regression analyses were conducted using SPSS statistical software, version 19.

Awareness Campaign
The multi-media public awareness campaign included out-of-home advertising, such as billboards and gas station pump toppers, and ran for eight weeks throughout the CN reservation. Indoor posters were continuously displayed in CNHS hospital and clinics. Radio advertising included 1,261 spots on local stations that ran for 21 days. CN website digital advertisements and social media marketing via the CN Public Health Facebook page ran for 8 days. The multi-media campaign had an estimated impact that included 12,255,000 impressions by out-of-home modalities, 1,300,000 impressions by radio spots, and 8,282,459 impressions by digital display. Social media marketing reached 64,815 individuals within the 10-targeted zip codes, with an average of 5.99 exposures per person and a 0.30% click-through rate. Among the 1,170 individuals making website clicks, 576 were 50-65 years old and 594 were 20-49 years old.

Expansion of the CNHS HCV ECHO Program
During the elimination program period the ECHO spokes increased from four to seven sites, covering all except one of the satellite clinics. The hub team expanded to include a clinical pharmacist, a nurse practitioner, and a dedicated DAA procurement case manager. During this process, 11 primary care providers and 6 pharmacists were recruited and trained in HCV care.

Screening coverage by site, screening strategy and birth cohort
The number of individuals accessing care at CNHS without prior screening at CNHS during the elimination program period and the percentage newly screened among the older and younger birth cohorts are displayed in eFigure 2 for those seen at the hospital (eFigures 2A and 2B) and at the satellite clinics (eFigures 2C and 2D). Among older and younger birth cohorts seen at the hospital, more individuals were screened during the LTS period (28.9% and 27.3%, respectively), followed by LTS with consent to treat on file (17.5% and 17.2%, respectively), than during the time period when LTS was discontinued (10.4% and 10.0%, respectively) or when the EHR reminder was re-established (11.2% and 11.9%, respectively). Among both birth cohorts seen at satellite clinics, the proportions screened were almost ten times higher after the satellite clinics obtained rapid test kits for the universal screening in January 2016. eFigure 2. Individuals accessing care at CNHS without previous HCV screening and the percentage of those screened for HCV with universal screening* during the elimination program period by screening strategy, location, and birth cohort at WWH hospital (eFigures 2A and 2B) and CNHS satellite clinics (eFigures 2C and 2D).

Cure Proportions Stratified by Birth Cohort, demographics and Liver Fibrosis Severity
Of those initiating treatment, 61% were male, with a median age of 47. The percentage of women initiating treatment was slightly higher among the younger birth cohort than the older birth cohort (41% vs 36%, eTable3). The majority of individuals had a HCV RNA genotype 1 (61%), followed by genotype 2 (22%), and genotype 3 (17%). Moderate and advanced liver fibrosis were detected in 29% and 11% of individuals, respectively. Among individuals with SVR12 available, PPT cure proportions were 97% (331/343) (eTables 3-4). In addition, no association was found between HCV cure and gender, genotype, or fibrosis severity (eTable 5). Age at treatment initiation, range, years