Improving Hepatitis C Screening in the Baby Boomer Birth Cohort in Rural South Dakota: A Medical Student HQIP Project

Abstract

Hepatitis C is a bloodborne viral infection that often leads to liver disease. Individuals born between 1945–1965 (baby boomer birth cohort) are five times more likely to have hepatitis C than other age groups due to blood transfusions and medical procedures performed before the discovery of the virus. The Centers for Disease Control and Prevention and the U.S. Preventive Services Task Force recommend a one-time screening for individuals in the baby boomer birth cohort. Even with these recommendations, national screening rates remain low at around 13 percent, suggesting a need for improvement.

In this study we reviewed the electronic medical record (EMR) data for a rural primary care clinic and determined the percentage of individuals screened in the baby boomer birth cohort in a one-year time period. Interventions (provider/nursing education, community education) were implemented over a four-month period. We compared the EMR data from before, during, and after interventions. Pearson’s chi-squared analysis was used to evaluate differences in proportions.

The results showed no statistical significance between the three timeframes measured (p-value 0.6164). We can conclude that the interventions used in this study were not adequate in producing a statistically significant change in the percentage of baby boomers screened at our local clinic. These results could be due to interventions not being implemented simultaneously, lack of follow-up with staff regarding interventions, and a short time frame for measuring post-intervention changes. Future projects may benefit from modifying interventions and their implementation.

Introduction

Hepatitis C virus (HCV) is the most common chronic bloodborne infection in the U.S.¹ Of those infected, 80 percent progress to chronic infection, often leading to cirrhosis, hepatocellular carcinoma and liver transplant.² An estimated 3.2 million people in the U.S. have chronic hepatitis C, 75 percent of whom were born between 1945–1965. In other words, those born between 1945–1965, a population known as the baby boomers, are five times more likely to have hepatitis C than other age groups.³ This is partially because hepatitis C was not discovered until 1989, and most baby boomers are believed to have been infected between the 1960s and 1980s. Infections were most likely contracted from medical procedures before universal precautions and infection control procedure were adopted, from contaminated blood and/or blood products before screening, or from IV drug use, even if only once in the past.

Unlike hepatitis A and hepatitis B, there is no vaccination for HCV, but there is treatment. Most patients are asymptomatic until the disease has progressed and caused permanent damage. Therefore, it is important to screen patients who have high risk of HCV. In addition, early treatment lowers morbidity and mortality and decreases healthcare costs. The Center for Disease Control and Prevention (CDC) and the U.S. Preventative Services Task Force (USPSTF) both recommend a one-time screening for patients born between the years of 1945 and 1965.^3,4

Preventative screening is estimated to identify 800,000 infections and avert more than 120,000 HCV-related deaths.3 HCV screening is performed by measuring antibodies in the blood. If the sample tests positive, a confirmatory test using RT-PCR is performed. Screening is relatively inexpensive and is covered under the Affordable Care Act, although some additional charges may be applied for blood draw or clinical procedures. It is estimated that screening saves $1.5-$7.1 billion in liver disease related costs in the U.S. Average treatment costs between 2003–2013 for hepatocellular carcinoma was roughly $220,000 per person.⁵

According to the South Dakota Department of Health, new chronic hepatitis C cases in South Dakota have been on the rise since 2010, increasing from 350 to 734 in 2016.⁶ This increase in prevalence is likely the result of increased screening. Most new cases in South Dakota were between the ages of 40–64.

Unfortunately, even with screening recommendations from the CDC and USPSTF, screening rates nationwide remain low at around 13 percent.⁷ This rate displays a need for improved education and screening measures from a public health standpoint. Previous publications were successful in developing interventions that improve HCV screening rates. In one such publication, a nurse practitioner in Oregon showed a significant rise in screening rates after administering health record alerts, speaking with staff, and placing posters in common patient areas.⁸

The goal of this study was to implement a series of interventions to increase the percentage of HCV screened patients born between 1945–1965 who visited a rural family medicine clinic in a 12-month timeframe. Review of electronic medical records (EMR) determined that approximately 45 percent of patients in this age range have previously been screened for HCV. We hypothesized that this percentage would increase during and after intervention implementation. The goal was to increase the percentage of screened individuals by 25 percent post-intervention, resulting in a screened percentage of around 57 percent.

Methods

Interventions

Three types of interventions were implemented at various time points. The first intervention implemented was a discussion with all providers at the rural family medicine clinic, which was comprised of physicians, physician assistants, and nurse practitioners. These discussions between the research team (G.W., E.C.) and individual providers occurred throughout the month of May. The discussions consisted of a brief overview of HCV epidemiology and screening guidelines. Reference materials from the CDC and USPSTF were used to outline the conversation and provide consistent information. The research team also answered questions and discussed potential obstacles to screening. The meetings lasted between 10 and 30 minutes depending on the providers’ familiarity with HCV and screening guidelines. The research team encouraged all providers to counsel HCV screening with all baby boomer patients and to encourage screening as recommended by the CDC and USPSTF.

The second intervention occurred in June and focused on community education. The research team attended the community health fair, a one-day event where the local hospital promotes healthy lifestyles and provides free lab tests to community members. The research team set up a booth and counseled all interested community members on HCV risk factors and HCV screening, stating that people born between the years 1945 and 1965 should have a one-time screening. If a person claimed to meet the criteria, the research team advised them to ask their provider about screening. Informational pamphlets were also handed out to community members (Appendix A, found at sdsma.org).⁹

The final intervention occurred approximately three months later in September. The research team met with the nurses present at the rural clinic all-nursing staff meeting, where they discussed HCV epidemiology, screening recommendations, and the HQIP project goals. The research team proposed the use of rooming templates to identify patients who qualify for screening. A rooming template is a checklist tool used to determine eligibility or risk factors of patients that should be screened. The nursing team felt it was not feasible to implement a new rooming template in the current clinical setting, so rooming templates were not used.

Statistical Methods/EMR Data Review

In order to determine the HCV screening percentages for the clinic, the research team compiled summary data from the rural family medicine clinic’s electronic medical records (EMR) using Slicer Dicer. Slicer Dicer is a self-service reporting tool within the Epic EMR system. This tool was used to identify the total number of patient visits between January and December, aged 53–73, and the percent of these patients who were identified in the EMR as having a hepatitis C screening status “completed.” No individual-level data was used; therefore, IRB approval was not required.

We compared three timeframes: pre-intervention (January-April 2018); ongoing intervention implementation (May-September 2018); and post-intervention (October-December 2018). These time frames were compared to the same months in 2017, which acted as a pre-intervention control in order to look for other seasonal trends. Pearson’s chi-squared analysis was used to evaluate differences in proportions.

Results

The research team had HCV discussions with all seven providers of the rural clinic, consisting of two physicians (MD or DO), four certified nurse practitioners, and one physician assistant. Providers familiarity with the topic varied but all felt their knowledge had improved from the HCV epidemiology and screening guidelines discussion.

At the community health fair, the research team spoke to approximately 50 community members and handed out over 60 pamphlets. A large proportion of community members who participated belonged to the baby boomer cohort; however, many were unaware of the screening guidelines or if they had been screened previously.

Finally, although the use of a rooming template was not feasible for the nurses at this clinic, the research team spoke to approximately 20 nurses about HCV epidemiology and screening guidelines.

A total of 6,896 office visits were recorded for patients aged 53–73 at the rural clinic between January and December of 2018.

Several comparisons were made using a chi-squared test to determine equality of proportions. The pre-intervention, intervention delivery, and post-intervention timeframes of 2018 were compared (Table 1). In the 2018 pre-intervention timeframe 45.0 percent of patients seen in the clinic had completed HCV screening, 46.0 percent of patients were screened during the intervention, and 44.7 percent of patients seen in the months post-intervention completed screening. Comparison of pre-intervention, intervention delivery, and post-intervention screening rates for 2018 indicate no significant difference between the groups (p-value= 0.6164). Similarly, rates were compared between correlating months of 2017, the year before any interventions were started. In 2017, patients seen in the correlating time frames of pre-intervention, intervention delivery, and post-intervention were 45.9 percent, 46.5 percent, and 45.8 percent, respectively. There was no significant difference in screening rates during the different time frames in 2017 (p-value=0.8715).

Table 1.

Comparison of proportion of baby boomers with completed HCV screen at a rural medical clinic between the three intervals of time.

	January-April*	May-September**	October-December***	p-value
2018	0.4497	0.4603	0.4470	0.6164
2017	0.4586	0.4650	0.4576	0.8715

^*Pre-intervention; 2018

^**Intervention delivery; 2018

^***Post-intervention; 2018

Additional analysis compared the three aggregated time frames between 2017 and 2018 (Table 2). There was no statistically significant difference between 2017 and 2018 during pre-intervention (45.9 vs. 45.8 percent, p-value= 0.5618), intervention delivery (46.5 vs. 46.0 percent, p-value=0.7269), and post-intervention (45.8 vs. 44.7 percent, p-value=0.5617).

Table 2.

Comparison of proportion of baby boomers with completed HCV screen at a rural medical clinic between 2017 and 2018 for each interval of time.

	2017	2018	p-value
January-April*	0.4586	0.4497	0.5618
May-September**	0.4650	0.4603	0.7269
October-December***	0.4576	0.4470	0.5617

^*Pre-intervention; 2018

^**Intervention delivery; 2018

^***Post-intervention; 2018

A month to month visual representation of screening rates during 2017 and 2018 can be seen in Figure 1.

Figure 1.

Proportion of baby boomers with completed HCV screen at a rural medical clinic by month during the years of 2017 and 2018.

*shaded region indicates intervention time interval for 2018

Discussion

We saw no significant differences in screening rates between any time frames and did not reach our goal of a 25 percent increase in percentage of patients screened. While the intervention period had the highest percentage of completed screening, this increase was not statistically significant and was also present during 2017, suggesting a possible seasonal variation. These results did not support previous studies that used similar interventions and reported significant results.

Limitations - Data Entry

Due to how the data is represented in the EMR (Epic – Slicer Dicer), we analyzed the monthly number of patients who have completed screening. This measurement is different than the screening rate, which is how many patients get screened divided by the number who qualify for screening during a certain time frame. Although in theory both measurements would rise with improved screening, using screening rate may make it easier to see if improvements were being made in the short term and would be easier to compare to national screening rates.

Our data failed to include patients not being seen during certain months. Looking at the percentage of individuals with a completed screening status is only indicative of that sample size during that time frame and not of the population as a whole. If a patient has been screened previously but has not been to the clinic in 2018, they were not included in our analysis. This results in our data less accurately reflecting the percentage of screened patients.

One last possible limitation involved how screening is logged into the EMR. Patient profiles can list HCV screening status as incomplete, complete, or addressed. Ideally, all patients who have been screened would be listed as complete. However, providers may be inconsistent with logging status. For example, if a provider offers screening and a patient declines, they are still considered incomplete, but a provider may either list them as incomplete or addressed because they were offered screening. On the contrary, if a patient has given blood after the year 1992, they have automatically been screened for HCV. A physician may enter this patient as complete, or they may enter them as addressed with the reasoning that they do not need screening. These discrepancies may alter the data.

Limitations - Timeline

The implementation of the interventions was not simultaneous. Our first intervention of talking to providers had the most flexibility regarding timing and required the least amount of planning. The community health fair is a yearly event scheduled on the first day in June and would not be able to be rescheduled. The nursing meeting was the most difficult to arrange due to scheduling conflicts between the research team and the nursing staff. The first opportunity to meet with the entire nursing staff was not until September. Results may have improved with a more uniform implementation of interventions at a more consolidated time frame. Results may also differ with a longer post-intervention time frame. The time frame selected to measure the post-intervention data was approximately 2 months, which was shorter than the other time frames. It is possible the time frame was not long enough to show an increase in the screening status of patients.

Conclusions and Future Directions

As previous studies have shown, the interventions chosen revealed to be effective in increasing screening in the baby boomer birth cohort. However, our study did not show an increase. Things to be improved upon include uniform and simultaneously implementation of our selected interventions, reinforcement and follow-up of interventions with staff, uniform nursing rooming formats to include screening for hepatitis C, integrating screening in the local community health fair, and increasing community awareness through media (radio, newspaper, etc.) as well as displaying posters in patient areas encouraging screening. Overall, we believe the interventions have potential with improved implementation. The short time frame for data collection and execution of intervention implementation we believe were the reasons for the lack of statistically significance found in the post-intervention phase. Future projects should modify implementation for a better outcome.