Universal HCV Screening of Baby Boomers is Feasible, but It is Difficult

Abstract

Background/objective

Our objective was to assess the impact of mass mailing and the inclusion of Best Practice Advisory (BPA) “Pop-Up” tool in the electronic medical record (EMR) on HCV screening rates.

Methods

Between June 2015 and March 2020, two interventions were developed for primary care physicians (PCP). An educational letter along with a blood requisition form, signed on behalf of the PCPs, was sent to patients. We also developed a BPA “Pop-Up” screening tool to alert PCPs to order HCV screening tests on patients with no previous screening. Data were collected and analyzed prospectively.

Results

When we started the screening program in June 2015, 33,736 baby boomers were eligible for screening, and the hospital system added an additional 26,027 baby boomers between June 2015 and March 2020. Of the 89 primary care providers employed by the hospital, 75 agreed to participate at different time periods. We screened 23,291 (43.5%) of 53,526 eligible patients during study period. Of these, 399 (1.7%) had HCV antibody, but HCV RNA was positive in only 195 (1%). HCV antibody positivity rates were higher in men, blacks, and in 1951–1960 birth cohorts. Spontaneous clearance rates appeared to be lower in men (OR 0.59, 95% CI 0.39–0.90, P = 0.015) and in blacks (OR 0.31, 95% CI 0.20–0.50, P < 0.001).

Conclusion

Although a formal screening program increased screening rates for HCV among baby boomers, about 50% of baby boomers remained unscreened. In this community screening program, we found that men and blacks are less likely to have spontaneous HCV clearance.

Approximately 2.4 million people in the United States have chronic hepatitis C virus infection, with a higher prevalence reported in western and Appalachian states.^1,2 Unfortunately, more than 50% of these infected patients are unaware of their diagnosis, and many of them remain unknown to the health care system.³ Although, oral direct-acting antivirals (DAAs) provide an opportunity for clinicians to reduce the burden of viral hepatitis C, it is estimated that 320,000 HCV-related deaths are still expected to occur by 2050.⁴ People born between 1945 and 1965, also known as the baby boomer generation account for approximately three-fourths of all chronic HCV infections among adults in the United States.⁵ The Center for Disease and Control (CDC) and US Preventive Services Task Force (USPSTF) recommends a one-time screening for patients suspected to have HCV infection.^6,7 Despite these strong recommendations, screening of the baby boom generation is poorly implemented to date in most parts of the country.^{8, 9, 10} World Health Organization has an ambitious goal to eradicate HCV from this world by 2030, but most countries are behind the target, including the USA.^11,12 To attain the WHO goal of HCV eradication by 2030, multipronged screening programs that target a wider population and a successful linkage of care must be implemented nationally.¹³

At our institution (MMC), we implemented a targeted outpatient (OP) screening of those born between 1945 and 1965 in June 2015. Prior to the program’s initiation in June 2015, only less than 5.6% of the baby boomer population was screened (Figure 1). The primary objective of this paper is to report the impact of the HCV screening program at our institution.

Figure 1.

Flowchart of all patients born between 1945 and 1965 who were screened between June 2015 and March 2020.

Patients and methods

The study targeted patients born between 1945 and 1965 that belonged to the practice of 89 primary care providers (PCP) between seven community physician offices located in Baltimore, MD. Patients who had an office visit within the past 18 months with no prior diagnosis of hepatitis C were eligible for the study.

Before we started the screening program, we recruited one full-time and two part-time HCV screening coordinators. We arranged meetings before initiating the screening with multiple stake holders, including senior hospital administrators, director of PCP, pathology, and information technology (IT) team. The PCP’s participation in the program was voluntary. We had a one-on-one meeting with individual PCP before screening their patients. We provided educational material on HCV to the primary care physicians and instructed providers who agreed to participate to request “Hepatitis C antibody with reflex to RNA” testing, which was hardcoded into the EMR BPA “Pop-Up” (EPIC is used at our institution). We also conducted numerous awareness programs for the general public during the study period via posters, radio, local television, and online interactive sessions. Through regular e-mails and letters, PCPs were updated on CDC Guidelines. Between 2014 and 2015, we also arranged numerous CMEs that focused on the role of the PCP in the management of HCV. Project coordinators were involved in educating providers about the process for contacting and referring patients to HCV treating physicians.

To improve screening rates, we deployed a 2-prong approach. One approach was to send letters with educational material on HCV to baby boomers, with no prior screening, from the primary care physician’s office along with an order for the blood test. The second approach was to develop a Best Practice Advisory “Pop Up” tool that would be triggered when a patient who had not been screened for HCV visited their PCP. Information technology (IT) analysts ensured that this workflow was tested and feasible prior to hospital implementation. PCPs had to choose between 5 key acknowledgments available on the “pop up tool” to screen patients, as shown in Table 1.

Table 1.

BPA Acknowledgments and Definitions.

Has this patient ever been tested for Hepatitis C?
Acknowledgment	Definition
Available	•Patient has results available in Epic for Hepatitis C test either interface or scanned
Unavailable	•Patient has been tested, but the results are currently unavailable
No	•Patient has not been tested for Hepatitis C (with an embedded link to order lab)
Decline	•Patient does not want to be tested for Hepatitis C
Unknown	•Patient status is unknown

HCV antibody and RNA test results triggered by the letter or BPA were forwarded to the EMR inboxes of project coordinators by the laboratory, and project coordinators then notified the PCP of positive results. Patients with positive HCV RNA were then contacted by the PCP and subsequently referred to HCV specialists. The weekly report was forwarded to the Medical Director of Primary Care, who tracked the individual PCP success rates. A standard of the operating protocol was developed prior to June 2015 and entailed guidelines, duties, and screening responsibilities for the OP screening initiative. These guidelines were updated on a monthly basis by research coordinators to ensure program efficiency, high-quality data extraction, collection, and reporting of data in compliance with HIPPA regulations.

We collected all data prospectively, including patient demographics, evidence of previous HCV status, anti-HCV antibody and RNA status, and stimulus for successful screening (BPA, letter, or unsolicited). The patients who were screened by their physician without using the “pop-up” tool or those who did not receive the letter were considered “unsolicited” screening.

For the purpose of this report, we analyzed characteristics of patients who had screening, spontaneous HCV eradication and estimated the success rates of linkage to care. Linkage of care was defined as patients completing an appointment with a physician to discuss the treatment of HCV following a positive HCV RNA test. To determine whether there were differences in HCV status based on the year of birth, we divided the cohorts into three groups (Group 1 - born between 1945 and 1950, Group 2 - 1951 to 1960, and Group 3 1961 to 1965).

The study was funded by Focus Grant (supported by Gilead Sciences).

Statistical Analysis

Descriptive statistics patients’ characteristics are presented as means and standard deviations (SDs) for continuous variables and frequencies for categorical variables. The differences in patients’ characteristics between HCV antibody positive and negative or RNA positive and negative were separately assessed by using Chi-square test for categorical variables, and t-test for continuous variables; normality was checked for all the continuous variables, and nonparametric Wilcoxon test was used when data were not normally distributed. Logistic regression was performed to evaluate the factors associated with HCV and spontaneous eradication separately. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).

Results

When we started the program in June 2015, there were 33,736 eligible patients. Between June 2015 and March 2020, more physicians and patients joined the hospital system and added an additional 26,027 patients, increasing the total number to 59,763 baby boomers. Of those, 2,889 declined to be screened, 2,863 (5.6%) had a previous screening, and 485 had a diagnosis of HCV, leaving 53,526 eligible for screening (Figure 1). Of the 89 PCPs affiliated to the hospital, only 44 agreed to send letters in addition to BPA, 2 sent letters only, 29 opted to use only BPA (no letters), and 14 (15.8%) refused to participate in the entire screening program.

Of those 53,526 patients, 23,291 (43.5%) completed successful screening by HCV antibody patients (Figure 1). We screened 7304 (31.4%) patients by BPA only, 2,543 (10.9%) by letters only, and 3,102 (13.3%) by a combination of letter and BPA. However, 10,342 (44.4%) were screened by their PCP without BPA or letters (“unsolicited”).

Of the screened patients, 456 (1.96%) were HCV antibody positive and of these 195 (1%) were HCV RNA positive; 15 patients had no RNA order or a result after an antibody testing was completed despite repeated efforts (Figure 1). All patients (n = 246) who were HCV antibody positive with negative HCV RNA were contacted by telephone or letters, and of those 57 admitted, they had been previously treated for HCV, and 189 had no previous treatment (Table 2). So the total number of new cases of HCV antibody positivity was 399 (1.7%) and of these, 50.7% (195 of 384) were RNA positive. Of these 195 HCV RNA positive patients, 173 (89%) were successfully linked to care. HCV antibody positivity was low in the “unsolicited” screening group (111 of 10,342, 1.07%) compared to those screened via BPA (242 of 7,304, 3.3%) or letters (58 of 2544, 2.3%).

Table 2.

Patients’ Characteristics Stratified by HCV Antibody Status.

Variable	Response	All (N = 23291)	Negative (N = 22835)	Positive (N = 456)	P-value
Birth cohort	1945–1950	5485 (24%)	5392 (24%)	93 (20%)	<0.001
	1951–1960	11938 (51%)	11665 (51%)	273 (60%)
	1961–1965	5868 (25%)	5778 (25%)	90 (20%)
Age at screening	Mean ± SD	61.3 ± 5.8	61.3 ± 5.8	61.6 ± 5.3	0.31
Gender	Female	13272 (57%)	13042 (57%)	230 (50%)	0.004
	Male	10019 (43%)	9793 (43%)	226 (50%)
Race	White	16636 (71%)	16331 (71%)	305 (67%)	0.008
	Black	5618 (24%)	5482 (24%)	136 (30%)
	Asian	961 (4%)	949 (4%)	12 (3%)
	Others	76 (1%)	73 (1%)	3 (1%)
Median household income by ZIP code	≤$50k	4597 (20%)	4484 (20%)	113 (25%)	0.007
	$50k - $70k	8169 (35%)	8005 (35%)	164 (36%)
	>$70k	10443 (45%)	10266 (45%)	177 (39%)
Approach	BPA	7304 (31%)	7062 (31%)	242 (53%)	<0.001
	Letter	2544 (11%)	2486 (11%)	58 (13%)
	Unsolicited	10342 (44%)	10231 (45%)	111 (24%)
	Combo	3101 (13%)	3056 (13%)	45 (10%)
Location	Baltimore City	6166 (26%)	6079 (27%)	87 (19%)	<0.001
	Anne Arundel County	2856 (12%)	2768 (12%)	88 (19%)
	Baltimore County	14269 (61%)	13988 (61%)	281 (62%)

The characteristics of patients stratified by positive or negative HCV antibody and HCV RNA status are shown in Table 2, Table 3, respectively. The mean age of patients was 61.3 $\pm$ 5.8 years. Among the screened patients, 57% were women, 71% were white, and 45% had a household median income of $70,000 or more (Table 2). There were significant differences (higher in low-income groups) in household incomes between HCV antibody-positive patients and negative patients (P = 0.007). There were also differences in HCV antibody positivity among different geographic locations of the clinic (P < 0.001). HCV antibody and RNA positivity were higher in 1951–1960 birth cohorts compared to two other cohorts.

Table 3.

Patients’ Characteristics Stratified by HCV RNA Status.

Variable	Response	Negative (N = 246)	Positive (N = 195)	P-value
Birth cohort	1945–1950	63 (26%)	27 (14%)	0.009
	1951–1960	141 (57%)	127 (65%)
	1961–1965	42 (17%)	41 (21%)
Age at screening	Mean ± SD	62.1 ± 5.5	61.0 ± 5.03	0.028
Gender	Female	133 (54%)	88 (45%)	0.06
	Male	113 (46%)	107 (55%)
Race	White	184 (75%)	111 (57%)	<0.001
	Black	50 (20%)	81 (42%)
	Asian	10 (4%)	2 (1%)
	Others	2 (1%)	1 (1%)
Median household income in ZIP code	≤$50k	53 (22%)	55 (28%)	0.21
	$50k–$70k	89 (36%)	70 (36%)
	>$70k	103 (42%)	69 (36%)
Approach	BPA	135 (55%)	98 (50%)	0.054
	Letter	35 (14%)	23 (12%)
	Unsolicited	48 (20%)	59 (30%)
	Combo	28 (11%)	15 (8%)
Location	Baltimore City	31 (13%)	54 (28%)	<0.001
	Anne Arundel County	48 (20%)	34 (17%)
	Baltimore County	167 (68%)	107 (55%)

Birth cohort (P = 0.001), gender (P = 0.001), and race (P < 0.001) had significant association with HCV infection (Table 2, Table 3). The probability of having HCV exposure (positive HCV antibody) in patients born between 1951 and 1960 was two times that in patients born between 1945 and 1950 [adjusted Odds Ratio (aOR): 2.13, 95% CI 1.40–3.23]. Males (aOR: 1.80, 95% CI 1.35–2.39) and Blacks (aOR 2.36, 95% CI: 1.76–3.15) had higher odds of having HCV exposure compared to Whites (Figure 2).

Figure 2.

Adjusted Odds Ratios (aOR) with 95% Confidence Intervals (LCL, UCL) of factors associated with having HCV exposure and spontaneous HCV clearance.

Gender and race had a significant association with spontaneous clearance of HCV (positive HCV antibody with negative HCV RNA in the absence of any previous HCV treatment. Men (OR: 0.59, 95% CI: 0.39–0.90, P = 0.015), and Blacks had a lower probability of spontaneous clearance of HCV (OR: 0.31, 95% CI: 0.20–0.50, P < 0.001) when compared to Whites (Figure 2).

Discussion

Our study showed that it is possible to improve HCV screening rates among PCP managed baby boomers within an urban medical center using a combination of targeted mass mailing of HCV education materials, increasing PCP awareness of HCV screening in this population, and by modifying EMR to alert PCPs when previously unscreened patients came for outpatient visits. In order to achieve the WHO goal of HCV elimination by 2030 in the United States, we believe that there has to be a coordinated effort from PCPs to identify HCV-positive subjects and link them to care. Previous studies have reported improved screening rates by the system-wide implementation of HCV screening.^{14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24} Our observations corroborate those observations in an urban medical center taking care of a very diverse population. Nevertheless, about 50% of baby boomers remained unscreened after our concerted effort over a period of more than 4 years.

There were many hurdles for the successful implementation of HCV screening programs in a hospital system. We learned at the onset that the awareness of the higher HCV prevalence among the baby boomer generation was not uniform among PCPs, and moreover, many PCPs felt that HCV screening was not critical for these patients as many of them had multiple other comorbidities. To implement a system-wide HCV screening, we needed the cooperation of multiple stakeholders, including PCPs, pathology service, information technology (IT) team, and more importantly, the hospital administration. After multiple meetings with stakeholders, implementation of the program was made conditional on individual and voluntary PCP participation. Initiation of the program was delayed because of the time required to obtain individual PCP opt-in. The IT team could not make system-wide modifications given the voluntary nature of the program, thereby increasing costs and resulting in delays. Despite our best efforts, only 75 of 89 (84%) PCPs agreed to utilize BPA for HCV screening, and only 46 (52%) allowed the coordinators to mail education letters to their patients on their behalf. Nevertheless, the increased awareness through multiple CME meetings and individual discussions resulted in a significant increase in “unsolicited” screening of patients by many PCPs who had initially declined to participate in the formal program. As a part of the program, HCV project coordinators helped to link the newly diagnosed patients with HCV to appropriate care, thereby reducing the additional workload for PCP, and this was an important “buy-in” incentive for PCPs. Between June 2015 and March 2020, we were able to screen 23,291 baby boomer patients, although this was less than we had planned. Despite the initial reluctance from various stakeholders, increased awareness also led to the implementation of universal screening of medical inpatients, patients coming for initial prenatal visits, and more recently, patients visiting the hospital emergency department.

Several studies that evaluated the efficacy of using a BPA screening program have reported encouraging results with an increase in screening rates ranging from 5 fold to 15 fold.^{14, 15, 16, 17, 18} A recent study conducted in Michigan analyzed the BPA tool and reported a 5-fold increase in screening rates since the implementation of the pop-up function within the EMR.¹⁴ The 3-year screening rate prior to the implementation of the tool was only 28%, but this increased to 72% within the first-year post-BPA in that study. Another study reported a 145% increase in HCV screening after the introduction of a clinical decision support tool in the EMR.¹⁶ In this study they used nonmedically trained HCV care coordinators similar to our study and found this to be cost effective in linkage to care. The proportion of baby boomers screened in our center prior to the implementation of BPA, and targeted mailing strategy was only 5.6%, but after implementation of the screening program, we achieved a 7-fold increase (43.5%) in screening rates. However, to date, more than 50% of baby boomers have not been screened despite our coordinated efforts suggesting that HCV screening should not be an option left to the discretion of PCP but should be mandated at an institutional or government level.

Our study is slightly different from the above-mentioned studies because we incorporated a bimodal strategy to screen patients by sending them individual letters along with the deployment of an EMR screening tool. It has been suggested that personalized text messages to patients who meet screening criteria are another way to improve screening rates.¹⁹ Phone calls, emails, and reminders on patient portal websites are other possible ways to enhance awareness in the community. The targeted mailing campaign is a relatively novel approach for HCV screening through direct patient education. The impact of the letter could not be quantified objectively as many of these patients were also alerted by physicians through a BPA pop-up. We saw a steady increase in “unsolicited” screening during the study period, which could be partly through our direct patient education and also through physician education in the form of CMEs and institutional grand rounds. We speculate that physicians in the community who did not agree with implementing BPA or sending letters also started screening because of this increased awareness and partly because of peer pressure as the progress with HCV screening was often discussed by the director of PCPs in their administrative meetings. The medical director of PCPs encouraged screening by recognizing PCPs (plaques, certificates, etc.) when they achieved significant milestones (such as 100% screening or > 1000 patients screened). By March 2020, among 75 providers who participated in FOCUS, 45 physicians achieved a screening rate of 50% or greater, 11 physicians achieved between a 40–49% screening rate and 5 physicians achieved a screening rate between 30 and 39%, and 14 physicians had a screening rate below 30% (this is by excluding those who refused and those who had the previous testing). PCP’s participation in the decision-making process and the recognition of their challenges are crucial for the successful implementation of HCV screening programs.²⁰

In our study, we found that the prevalence of positive HCV antibodies among baby boomers was only 1.96%, and this could be a reflection of a relatively affluent population who were seen at this institution, especially in its suburban outpatient clinics. Other studies have also reported similarly lower prevalence, which has been attributed to the demographics of the screened patient population.^14,21 Linkage to care in our study was excellent (89%). We defined linkage to care as successful if patients completed an appointment with a PCP or HCV treating physician to discuss treatment of HCV following a positive RNA test. Linkage to care in prior studies was highly variable, ranging anywhere from 38% −85%.^{22, 23, 24} Only one study reported a higher (92%) linkage to care than our study.²⁵ Our relatively higher linkage could be attributed to the higher socioeconomic status of the patient population in the relatively wealthy suburbs, as well as the comprehensive effort by ancillary staff and dedicated HCV screening team who meticulously followed up with RNA-positive patients and encouraged them to follow up with their physicians. Barriers to the linkage of care are well documented, and these include poor socioeconomic background, injection drug use, psychiatric illnesses, homelessness, alcoholism, and coinfection with HBV or HIV. These challenges are significantly higher for emergency room providers and in-patient hospitalists who also test baby boomers and high-risk groups for HCV antibodies in many centers.^{26, 27, 28}

Multivariate analysis in our study revealed that HCV exposure rates were higher in men and blacks (OR of 2.36, P < 0.0001 vs. whites), which is consistent with the national estimate.²⁹ According to modeling data, deaths secondary to chronic hepatitis C are projected to peak between 2030 and 2035, and this can have a significant impact on African Americans and patients from a low-income background, again highlighting the urgent need to develop effective screening programs in the black community.³⁰ In this study, we also examined the rate of spontaneous clearance in those exposed to HCV infection. Following an acute HCV infection, spontaneous clearance may occur in 20–40% of subjects; however, it is very rare in chronic HCV infection.^{31, 32, 33, 34, 35, 36} In our study, a relatively higher proportion of patients (49.3%) with HCV Ab positivity were negative for HCV RNA. If we were to assume that all those patients had spontaneous clearance, it is relatively high, and moreover, the rates were lower in blacks compared to whites (OR: 0.31, 95% CI: 0.20–0.50, P < 0.001) (Figure 2). The ability to spontaneously clear HCV is associated with younger age, female sex, nonblack ethnicity, host genetic polymorphism, and coinfection with hepatitis B.^{31, 32, 33, 34, 35, 36} Our finding of poor HCV clearance among blacks Americans is consistent with other published data.³⁶

The implementation of the best practice advisory “pop-up” in our EMR was the most effective intervention in this study, followed by the targeted mailing campaign. We believe that institution-wide implementation of BPA “Pop-Up” that identifies any eligible, at-risk population for testing is a relatively practical and cost-effective way to improve HCV screening. Although it is not going to be popular, there should be a financial penalty for institutions that do not implement such programs. Additionally, there should be an ongoing education program for patients, care providers, and hospital administrators about the importance of the HCV program. In summary, it was quite sobering to recognize that we had to overcome many hurdles to initiate HCV screening of baby boomers, as recommended or endorsed by multiple organizations, including USPSTF, CDC, Centers of Medicare and Medicaid Services, and so on, in a relatively small urban medical center with 89 PCPs. As suggested in a global mathematical model, HCV elimination is possible, but it will require a coordinated effort from multiple stakeholders.^13,37,38

Credit authorship contribution statement

PJT devised the study, JT collected the data, TZ did the statistical analysis, and all authors contributed to the writing of the manuscript.

Conflicts of interest

The authors have none to declare.

Financial support

Focus Grant (supported via Gilead Sciences, Inc.)