Abstract
Context: Approximately 2.7 million people in the United States currently live with chronic hepatitis C virus (HCV) infection, and many are unaware that they have the disease. Community health centers (CHCs) serve as the primary care safety net for more than 22 million patients who are at risk for health inequities and represent an important frontline resource for early screening and treatment for HCV infection.
Objective: To understand HCV infection screening rates among CHC patients, and to quantify the screening gap by demographic characteristics.
Methods: The authors analyzed a deidentified dataset obtained through electronic health records from a large national network of CHCs. All adults at risk for HCV infection, according to the US Preventive Services Task Force (USPSTF) birth cohort screening guidelines for HCV infection, were considered eligible if they had a patient office visit between January 1, 2013, and December 31, 2013. Data were reviewed to determine the documentation of HCV infection screening from January 1, 2010, to December 31, 2013, and HCV infection screening rates were analyzed by age, race/ethnicity, and sex.
Results: Among 60,722 eligible patients, 5033 (8.3%) had an HCV infection screen in accordance with USPSTF birth cohort screening guidelines. Women were less likely to be screened than men in every race/ethnic group, including white Hispanic (9.3% in women vs 5.4% in men), black Hispanic (15.1% in women vs 9.0% in men), white non-Hispanic (13.6% in women vs 8.1% in men), black non-Hispanic (14.9% in women vs 8.9% in men), Caribbean Islander or Haitian (6.5% in women vs 3.7% in men), and other races/ethnicities (6.3% in women vs 3.6% in men).
Conclusion: To the authors’ knowledge, this is the first large-scale study among CHCs to assess the screening gap of the USPSTF birth cohort screening guidelines for HCV infection. This study suggests that CHCs should consider opportunities to improve HCV infection screening, thereby contributing to the reduction of health inequities resulting from untreated HCV infection.
Approximately 2.7 million people in the United States currently have chronic hepatitis C virus (HCV) infection,
1 and approximately 17,000 new cases of HCV infection each year are unreported.
2 Often termed the “silent epidemic,” HCV infection frequently goes unnoticed because many people with the disease exhibit no obvious symptoms.
3
In persons with HCV infection, cirrhosis will develop in 20% to 25% (approximately 800,000 people) within 20 to 30 years; some cases will progress to hepatocellular carcinoma, end-stage liver disease, and death. Costs linked to HCV-related liver morbidity and mortality during the next 20 years are expected to increase by $55 billion.
4 Efforts to improve the identification and early treatment of patients with HCV infection in the United States are supported by the Healthy People 2020 national health objectives,
5 the US Department of Health and Human Services’ Action Plan for the Prevention, Care & Treatment of Viral Hepatitis,
6 and the birth cohort screening guidelines issued by the US Preventive Services Task Force (USPSTF) in June 2013.
7 Before these guidelines were issued, HCV infection screening was recommended for patients at high-risk, including those who had large or repeated percutaneous exposures, such as injection drug users and persons who received donated blood before blood bank screening for HCV infection was implemented. The 2013 release of the USPSTF birth cohort screening guidelines for HCV infection recommends a 1-time screening for HCV infection for all adults born between 1945 and 1965.
7
Community health centers (CHC) provide affordable and accessible health care services (including screenings and preventive care to reduce health disparities), improve management of chronic diseases, and decrease unnecessary hospitalizations.
8,9 Nationally, CHCs provide primary care to more than 22 million people annually, including the working poor, uninsured, unemployed, and homeless populations,
8 and therefore present a frontline opportunity for early identification and treatment of patients with chronic diseases such as HCV infection.
The purpose of the present study was to determine the prevalence of HCV infection screening according to the new USPSTF birth cohort screening guidelines among a large CHC population and to describe screening disparities across demographic characteristics.
Fourteen CHCs were included in this study. Inclusion criteria were use of an electronic health record (EHR) system for at least 4 full years and the availability of complete laboratory data received electronically. Of the 14 CHCs included in the study, 8 were from Florida, 3 from Utah, 1 from California, 1 from Kansas, and 1 from New Mexico. Inclusion criteria for patients in the study were birth between January 1, 1945, and December 31, 1965, and a documented established patient visit at 1 of the 14 CHCs between January 1, 2013, and December 31, 2013. Exclusion criteria were previous diagnosis of HCV infection based on the International Classification of Diseases, Ninth Revision, Clinical Modification, inactive medical record, and death.
We used SPSS software (version 21; IBM) and R-3.1.2 for statistical analysis. Descriptive statistics were used to estimate the prevalence of birth cohort screening guidelines for HCV infection among eligible patients who had an established visit between January 1, 2013, and December 31, 2013, by age, race/ethnicity, and sex groups. To determine whether the number of HCV infection screenings differed by race/ethnicity and sex, χ2 tests of independence (P<.05) were conducted. A nested logistic regression model was created and tested using age, race/ethnicity, and sex as predictors of HCV infection screening. The dependent variable, which measures whether a patient was screened for HCV infection, was equal to 1 if the respondent was screened and 0 otherwise. Various models were compared using the Akaike information criterion, and the model with the lowest Akaike information criterion value was retained. Multicollinearity and outliers were assessed using the variance inflation factor and an examination of the residual plot. No offending variables were found. The final model included age, race/ethnicity, and sex and controlled for the correlation within individual CHCs. The le Cessie-van Houwelingen-Copas test supported the final model (P=.11), and the model correctly predicted screening 92% of the time.
A number of reasons exist for low screening rates that are suggested in the literature but were not assessed directly in this study.
Limited awareness of USPSTF birth cohort screening guidelines for HCV infection, prioritization of birth cohort HCV infection screening vs screening on the basis of specific risk factors such as injection drug use, communication barriers, and limited treatment options for low-income patients are some of the limiting factors.
13 Other limitations of this study include possible misclassification bias of patients who received screening outside the CHC that was either not reported in their CHC EHR or was captured in the EHR in a field not extracted for this study (eg, using a free-text note), which would result in higher screening rates among the population than reported in the current study.
The CHCs included in this study were all early adopters of EHR systems. Although EHRs have been shown to be a valuable tool for conducting surveillance among underserved populations and hold promise for ongoing study among underserved patients who are typically underrepresented in research, data quality is an ongoing improvement initiative.
17 Misclassification bias may have been introduced by patients who were classified as not having been screened but who received screening before the 3-year “look-back” period used in the current study or among patients with previous diagnosis of and treatment for HCV infection. Although initiatives to improve interoperability of EHR data are still in their infancy, health care professionals can help to reduce misclassification bias and support understanding of HCV infection screening and the effectiveness of treatment modalities in future studies.
Another limitation of the current study is that it sought to determine HCV infection screening among all patients who had a primary care visit from January 1, 2013, to December 31, 2013, according to the new USPSTF birth cohort screening guidelines for HCV infection, which were released in June 2013. Before the USPSTF guidelines, high-risk populations were eligible for screening. Therefore, most patients included in the current study who were seen before June 2013 were not recommended for screening according to the USPSTF guidelines in place at that time. Although changes in screening practices may have been notable after the release of the USPSTF birth cohort screening guidelines, we could not assess HCV infection screening rates separately before and after the USPSTF guidelines change because of the timeframe of the study. Future studies should take this point into consideration and, in addition to understanding the screening gap, should assess the change in screening practices as a result of the 2013 USPSTF birth cohort screening guidelines.
We gratefully acknowledge the support of Health Choice Network, Inc, and the valuable input of the Business Intelligence and Health Solutions team, including Raydel J. Monterde, in conducting this research. The authors also gratefully acknowledge the support of John Turse, MD, and Sara Turse, MD, in preparing the manuscript.