Potential conflict of interest: Dr. Younossi consults for Biolex, Vertex, and Tibotee.
Article first published online: 11 FEB 2011
Copyright © 2010 American Association for the Study of Liver Diseases
Volume 53, Issue 3, pages 737–745, March 2011
How to Cite
Stepanova, M., Kanwal, F., El-Serag, H. B. and Younossi, Z. M. (2011), Insurance status and treatment candidacy of hepatitis C patients: Analysis of population-based data from the United States. Hepatology, 53: 737–745. doi: 10.1002/hep.24131
Supported in part by the Liver Outcomes Research Fund of The Center for Liver Diseases, Inova Fairfax Hospital, Falls Church, VA.
- Issue published online: 2 MAR 2011
- Article first published online: 11 FEB 2011
- Accepted manuscript online: 17 DEC 2010 03:44PM EST
- Manuscript Accepted: 9 DEC 2010
- Manuscript Received: 1 SEP 2010
Successful treatment with antiviral therapy could potentially reduce morbidity and mortality in patients with hepatitis C virus (HCV) infection. However, at the population level, these benefits may be offset by a limited number of patients who have access to antiviral treatment. Using data from the National Health and Nutrition Examination Survey conducted in 2005-2008, we analyzed the health insurance status and treatment candidacy of HCV-positive (HCV+) individuals. A total of 10,582 subjects were examined; of those, 1.16% had detectable HCV RNA and were defined as HCV+. The HCV+ patients were less likely to be insured than HCV-negative individuals (61.2% versus 81.2%; P = 0.004). Among those with health insurance, HCV+ patients were less likely to have private insurance, whereas the coverage by Medicare/Medicaid and other government-sponsored plans was similar to the rest of the population. In multivariate analysis, HCV infection was an independent predictor of being uninsured even after adjustment for demographic disparity of the HCV+ cohort (odds ratio, 0.43; 95% confidence interval, 0.24-0.78). Of all HCV+ patients, 66.7% were eligible for anti-HCV treatment. However, only 54.3% of HCV+ treatment candidates had any type of insurance coverage. Finally, only 36.3% of HCV+ patients were potentially eligible for treatment and had health insurance. Conclusion: A high proportion of HCV+ patients are currently uninsured, and many have publicly funded health insurance. Among those who could be candidates for treatment, the rate of insurance coverage is even lower. These findings can have important implications for health insurance coverage of these patients under the new health care reform legislation in the United States. (HEPATOLOGY 2011)
In the United States, hepatitis C virus (HCV) is the most common cause of chronic liver disease, hepatocellular carcinoma, and liver transplantation.1, 2 Most (80%-85%) of individuals infected with HCV (approximately 3.5 million in the United States) develop chronic HCV infection.3, 4 Symptoms of chronic HCV infection are nonspecific, and many patients remain undiagnosed. In fact, in one study, 75% of patients were unaware of their HCV infection.5
The benefits of treating HCV patients and achieving long-term viral eradication have been established.6 Successful treatment with antiviral therapy improves health-related quality of life in patients with HCV and could potentially reduce morbidity and mortality in patients who successfully eradicate the virus (i.e., have sustained virologic response). Although available treatment with pegylated interferon and ribavirin is successful in only half of treated patients,7 recent data demonstrate that addition of direct acting oral protease inhibitors to the current treatment will likely increase the chances of sustained virologic response in the more common genotype 1 patients.8
Whether or not this improved efficacy of the new antiviral treatment demonstrated in clinical trials will translate into a similar increase in the effectiveness at the population level is unclear. The full benefits of treatment may indeed not be realized, largely because a significant proportion of HCV-infected individuals may not even have access to the antiviral treatment or they may be considered ineligible for treatment. Treatment of HCV and its associated monitoring is expensive, with an estimated cost of up to $48,000 per year.7, 9 The cost of treatment and monitoring can be covered by health insurance; however, for uninsured or underinsured individuals, the economic impact can be substantial. With the advent of health care reform in the United States, there is a critical lack of data on the health insurance status of HCV-positive (HCV+) individuals.
In addition to insurance coverage, treatment candidacy is another important factor impacting access and receipt of care in HCV-infected patients. The side effect profile of the current antiviral therapy requires careful selection of treatment candidates and patients with a number of chronic medical conditions (e.g., severe depression, active cardiac disease, or renal failure) cannot be treated because of potentially severe adverse events during treatment.10, 11 The eligibility criteria for initiating antiviral treatment are likely to remain unchanged over the next several years. Furthermore, the increase in the efficacy is likely to come with additional adverse effects and treatment-related costs.12, 13
The aim of this study was to use population-based data to assess the presence and type of health insurance coverage as well as the treatment candidacy of HCV+ individuals in the United States. These data are important, because they may not only explain the existing gap between expected and observed rates of antiviral treatment in HCV,14 but may also estimate the potential impact of universal health insurance coverage for HCV-infected individuals with the advent of the health care reform bill.
Subjects and Methods
We used population-based data from the National Health and Nutrition Examination Survey (NHANES). NHANES is a series of stratified, multistage probability surveys designed to obtain information on the health and nutritional status of the civilian, non-institutionalized United States population. Sampling weights accounting for age, sex, level of education, and race or ethnic group were used to make the distribution of the participants to be representative of that of the United States population. Beginning in 1999, NHANES data have been collected continuously, with every 2 years serving as one analytic cycle. The data are collected by the National Center for Health Statistics of the Centers for Disease Control and Prevention through household interviews, physical examinations and extensive laboratory data from blood samples collected in special examination centers.
The present study included the two most recent publicly available cycles of NHANES (2005-2006 and 2007-2008) for United States adults aged 18 years or older. The subjects included in the two cycles were unique and were not counted twice. All adults with available HCV antibody test and completed insurance questionnaire were included in the study. The presence of HCV antibody was checked using direct solid-phase enzyme immunoassay with the anti-HCV screening enzyme-linked immunosorbent assay and validated with RIBA 3.0 Strip Immunoblot Assay (Chiron Corporation, Emeryville, CA). In those with positive or indeterminate anti-HCV test, the HCV RNA was measured using the COBAS AMPLICOR HCV MONITOR test version 2.0 (Roche, Branchburg, NJ). Patients with positive HCV RNA were defined as having chronic hepatitis C (CHC). Controls were defined as HCV-negative (HCV−) subjects who did not have other causes of chronic liver disease as determined by using the information provided through NHANES' Medical Conditions questionnaires.
Collected Data and Definitions.
The following demographic information were collected for all eligible CHC participants and controls: age at the time of examination, sex, ethnicity (Caucasian, African American, Hispanic, or Other, the latter of which included Aleut, Eskimo, American Indian, Asian, or Pacific Islander), marital status, history of being in military service, citizenship status, being college graduate, household income (calculated as the ratio to the Federal Poverty Level; ratios above 5.0 were not specifically reported).
Possible comorbidities that may affect treatment eligibility for CHC subjects were assessed using the questionnaires completed by NHANES participants. The list of studied medical conditions included history of hypertension, hypercholesterolemia, type 2 diabetes, asthma, arthritis, and ischemic heart disease (which included history of coronary artery disease, angina, or heart attack), congestive heart failure, chronic obstructive pulmonary disease (COPD) (which included chronic bronchitis or emphysema), stroke, kidney failure and history of dialysis in the past 12 months, and history of any cancer. The presence of depression was ascertained using the PHQ-9 questionnaire15: a score 15 or above corresponded to a diagnosis of depression, and a score of 20 or above corresponded to severe depression. Additionally, the Alcohol Use and Drug Use questionnaires were used to collect the history of substance abuse. For the purpose of the study, alcohol use was defined as consumption of ≥20 g alcohol per day during the 12 months prior the survey, and history of drug use was reported as lifetime history of marijuana or hashish use and lifetime history of cocaine, heroin, or methamphetamine. In addition, smoking history was collected using Smoking questionnaire: a participant was classified as having history of smoking if he/she reported current smoking or having 100 or more cigarettes during their lifetime.
The health insurance status of the CHC subjects and controls was studied using the NHANES Health Insurance Questionnaire. Only subjects who completed that questionnaire were included in the study. The questionnaire provides interview data on insurance coverage, type of insurance, and coverage of prescription drugs. The following types of insurance were included in the questionnaire: private insurance, Medicare, Medi-Gap, Medicaid, SCHIP, military (Tricare, VA, or Champ-VA), Indian Health Service, state-sponsored health plan, government insurance, or single service plan. Some individuals might have had two or more types of coverage. For the purpose of the study, two major types of coverage were considered separately: subjects with any private insurance or any military/state/government coverage were included in insurance group 1, whereas those with Medicare/Medicaid coverage were included in insurance group 2.
Selection of Potential Treatment Candidates with CHC.
Stable health insurance coverage allows for consistent access to health care services, which contributes to better health outcomes. However, with the regimen requirements and severity of adverse effects typically accompanying interferon-based anti-HCV therapy, this benefit is limited to HCV-infected individuals who could be candidates for antiviral treatment. To better understand how health insurance status may affect antiviral treatment rates, we further selected only those HCV patients who could potentially be candidates for the current standard of care HCV therapy (pegylated interferon/ribavirin). Eligibility criteria assumed absence of history of important comorbid conditions or active chronic diseases and included history of ischemic heart disease, congestive heart failure, chronic obstructive pulmonary disease, stroke, cancer, or kidney failure. Treatment exclusion criteria also included individuals with severe depression or uncontrolled diabetes (defined as glycohemoglobin ≥9%). The Hepatitis C follow-up questionnaire was completed only by a small portion of HCV+ individuals, hence we did not include the history of previous unsuccessful treatment in our eligibility criteria.
Health insurance coverage as well as medical, demographic, and social variables were compared between HCV+ subjects and HCV− controls without chronic liver disease using. HCV+ individuals with health insurance were further compared with those without health insurance coverage. The proportions of HCV+ subjects who were potential treatment candidates were then calculated, and we compared these proportions between HCV+ subjects with and without health insurance. Finally, insured and uninsured HCV+ individuals who could be treatment candidates were compared with each other, and then the same analysis was also repeated for the HCV+ treatment candidates from insurance group 1 and insurance group 2 separately; these groups were then compared with their uninsured counterparts. We used a logistic regression analysis to identify independent predictors of insurance status in the general United States population, and to study independent predictors of insurability among HCV+ individuals.
Sampling errors were calculated using the Taylor linearization method, and the stratum-specific chi-square test for independence was used for pairwise comparisons. Sampling weights recommended by National Center for Health Statistics guidelines for each questionnaire and laboratory parameter were used to account for nonresponse and unequal selection probabilities for certain categories of population. Stratification and sampling units describing the design stages of the NHANES data collection were also used to account for the complex, multistage probability sample design of these data. According to the 2005 NHANES Analytic and Reporting Guidelines,16 when merging two analytic cycles, a 50% adjustment coefficient was applied to all provided sampling weights. All analyses were run using standalone SUDAAN 10.0 (SAS Institute Inc., Cary, NC). A complete set of all data analyses is provided in (Table 1). The study was approved by the Inova Institutional Review Board.
|No. of subjects||141||10,441|
|Caucasian, %||63.92 ± 6.08||71.03 ± 2.77||0.1753|
|African American, %||24.48 ± 5.01||10.72 ± 1.57||0.0006|
|Hispanic, %||7.75 ± 2.12||12.56 ± 1.74||0.0421|
|Other, %||3.85 ± 2.18||5.69 ± 0.64||0.4391|
|Male sex, %||65.37 ± 5.39||48.08 ± 0.47||0.0072|
|Age, years||47.84 ± 0.89||45.29 ± 0.48||0.7124|
|Married, %||41.44 ± 4.82||57.16 ± 1.16||0.0056|
|United States citizen, %||98.57 ± 0.86||91.31 ± 1.04||<0.0001|
|College degree, %||8.04 ± 3.42||25.34 ± 1.47||<0.0001|
|History of substance use|
|Alcohol, %||26.83 ± 4.87||6.62 ± 0.36||0.0007|
|Smoking, %||85.08 ± 4.70||47.51 ± 1.28||<0.0001|
|Marijuana or hashish, %||87.49 ± 3.52||60.67 ± 1.25||<0.0001|
|Opiates or methamphetamine, %||70.51 ± 5.51||20.05 ± 1.09||<0.0001|
|General health status (self-reported)|
|Excellent, %||4.64 ± 2.11||17.39 ± 0.61||<0.0001|
|Very good, %||25.77 ± 4.75||31.87 ± 0.93||0.1876|
|Good, %||35.30 ± 5.00||33.95 ± 0.69||0.7916|
|Fair, %||24.10 ± 4.31||13.51 ± 0.57||0.0168|
|Poor, %||10.19 ± 2.83||3.23 ± 0.26||0.0116|
|Health care use in last year|
|Number of health care visits||5.37 ± 0.52||4.46 ± 0.07||0.2398|
|Hospitalization, %||13.01 ± 2.95||10.21 ± 0.48||0.3250|
|Mental specialist visit %||21.93 ± 3.99||7.64 ± 0.42||0.0007|
|Primary source of health care|
|Clinic or health center, %||16.07 ± 3.45||14.49 ± 1.11||0.6520|
|Doctor's office or HMO, %||50.20 ± 5.15||66.43 ± 1.55||0.0059|
|Hospital emergency room, %||5.18 ± 1.84||2.06 ± 0.22||0.0825|
|Insurance coverage, %||61.20 ± 5.57||81.20 ± 1.05||0.0043|
|Private insurance, %||34.78 ± 6.34||64.98 ± 1.36||0.0002|
|Military/state/government insurance, %||9.97 ± 2.55||7.04 ± 0.63||0.2561|
|Medicare/Medicaid, %||17.94 ± 3.54||20.34 ± 1.14||0.5058|
Overall, NHANES 2005-2008 included 20,500 subjects. After exclusion of individuals <18 years of age and those without completed Health Insurance questionnaire or hepatitis C antibody test, only 10,582 individuals were considered eligible for the study. Of those, 190 individuals (1.52 ± 0.14% of the population) were positive for hepatitis C antibody. However, only 141 (1.16 ± 0.14% of the population) had detectable HCV RNA. The cohort of HCV+ subjects included 63.9% Caucasians, 65.3% males, and the average age of the HCV+ participants was 47.3 years. In all, 41.4% of HCV+ subjects were married, 98.5% were United States citizens, and 8.0% had a college degree (Table 1).
Compared with HCV− controls, subjects with HCV were more likely to be African American (24.5% versus 10.7%; P = 0.0006), less likely to be Hispanic (7.8% versus 12.6%; P = 0.042), and more likely to be male (65.4% versus 48.1%; P = 0.0072) and unmarried (41.4% versus 57.2%; P = 0.0056). HCV+ patients were more likely to use alcohol, tobacco, or other substances (Table 1).
A number of other medico-social parameters such as education level, presence of certain comorbid conditions and self-reported health status were also different between HCV+ individuals and controls. Most of these data are consistent with previous reports confirming the validity of our analysis. The summary of sociodemographic parameters for HCV+ subjects compared with controls is given in Table 1, and a medical history comparison is given in Table 2.
|Hypertension, %||32.58 ± 4.37||29.43 ± 0.98||0.4717|
|Hypercholesterolemia, %||33.37 ± 7.45||42.27 ± 0.83||0.2559|
|Diabetes, %||7.78 ± 3.20||9.50 ± 0.46||0.5889|
|Depression, %||7.75 ± 2.26||1.75 ± 0.18||0.0112|
|Severe depression, %||2.52 ± 1.41||0.43 ± 0.08||0.1444|
|Kidney failure, %||4.64 ± 2.42||2.15 ± 0.18||0.3257|
|Asthma, %||16.29 ± 3.82||14.05 ± 0.51||0.5624|
|Arthritis, %||35.05 ± 4.13||25.44 ± 1.13||0.0264|
|Ischemic heart disease, %||5.67 ± 2.02||5.50 ± 0.29||0.8077|
|Congestive heart failure, %||3.85 ± 1.76||2.39 ± 0.19||0.4113|
|Stroke, %||5.29 ± 2.27||2.96 ± 0.27||0.3032|
|COPD, %||15.86 ± 3.33||7.34 ± 0.49||0.0246|
|Cancer, %||9.40 ± 2.89||8.52 ± 0.50||0.7647|
|HIV, %||2.09 ± 1.53||0.43 ± 0.11||0.2752|
|Albumin, g/dL||4.07 ± 0.04||4.25 ± 0.01||0.8124|
|ALT, U/L||60.85 ± 4.56||25.29 ± 0.24||0.0174|
|AST, U/L||56.34 ± 4.33||25.25 ± 0.18||0.0233|
|AST-to-platelet ratio index||0.29 ± 0.03||0.10 ± 0.00||0.0492|
|Platelets, 1,000/μL||256.18 ± 8.62||273.60 ± 1.65||0.6362|
Factors Associated with Health Insurance.
HCV+ individuals were more likely to be uninsured (39.7% versus 19.8%; P = 0.0043) than controls without liver disease. Considering different types of insurance plans, fewer HCV-infected individuals had private insurance coverage, whereas the rates of government- and state-sponsored plans as well as military insurance were similar to the rest of the population. HCV+ patients were less likely to be seen in doctors' offices but more likely to seek mental health care (Table 1).
In multivariate logistic regression, HCV infection was found to be an independent predictor of not having health insurance (odds ratio [OR], 0.43; 95% confidence interval [CI], 0.24-0.78). Additionally, being young, male, of non-Caucasian race, unmarried, without a college degree, and with a history of substance abuse were factors independently associated with lack of insurance (Table 3).
|Predictor||OR (95% CI)|
|Greater likelihood of insurability|
|Caucasian race||1.70 (1.36-2.13)|
|United States citizen||3.89 (3.02-5.00)|
|College degree||3.27 (2.51-4.26)|
|No association with insurability|
|In military service||1.31 (0.99-1.74)|
|Alcohol use||0.79 (0.60-1.04)|
|History of marijuana or hashish use||0.94 (0.79-1.11)|
|Ischemic heart disease||0.96 (0.59-1.58)|
|Type 2 diabetes||1.08 (0.81-1.43)|
|Lower likelihood of insurability|
|HCV infection||0.43 (0.24-0.78)|
|History of cocaine, heroin, or methamphetamine use||0.76 (0.62-0.93)|
|Male sex||0.70 (0.59-0.84)|
Health Insurance Coverage of HCV+ Individuals.
There were few significant differences in sociodemographic factors (Supporting Table 2) or comorbid conditions (Supporting Table 3) between HCV+ patients with or without health insurance. Individuals without insurance were less likely to have a college degree (12% versus 1.5%; P = 0.011; national average level, 26.5%17), and less frequently reported any knowledge of their chronic liver disease compared with those with health insurance (24.4% versus 52.2%; P = 0.031).
However, when we considered the types of insurance separately, more differences between HCV+ subjects who were insured and those who were uninsured were found. Specifically, compared with uninsured HCV+ individuals, subjects with Medicare or Medicaid were less likely to be Caucasian (29.7% versus 76.2%; P = 0.0001) and more likely to be African-American (51.8% versus 20.1%; P = 0.0016). Furthermore, they were highly unlikely to have a college degree (no cases), were less likely to be married (17.9% versus 38.8%; P = 0.0073), and had generally poorer health (8.7% reported being in very good health versus 19.7% among uninsured, P = 0.0338; 28.1% versus 8.8% reported poor health, P = 0.465; 29.6% versus 10.4% reported hospitalization last year, P = 0.0878), which is unlikely attributable solely to older age (50.9 versus 46.4 years; P = 0.5621). On the other hand, HCV+ subjects with private or military/state/government-sponsored plans were more likely to be married (53.1% versus 38.8%; P = 0.0393) and less likely to be poor (income/poverty ratio of 2.83 ± 0.23 versus 1.58 ± 0.19; P = 0.0248) or undereducated (16.6% had a college degree versus 1.5%; P = 0.0118) than uninsured (Supporting Table 1).
As expected, uninsured HCV+ individuals were more likely to use a hospital emergency room (9.23 ± 3.51 versus 2.61 ± 1.42; P = 0.0580) and less likely to use any other type of health care (Clinic or Health Center, 9.48 ± 4.21 versus 20.24 ± 4.75, P = 0.1126; doctor's office or HMO, 40.63 ± 6.39 versus 56.27 ± 7.16, P = 0.1221), although the estimates were not statistically significant, perhaps because of power limitations. We indentified an unexpectedly greater proportion of Caucasians among HCV+ subjects without health insurance (Supporting Table 2). The uninsured HCV+ subjects were also less likely to have kidney failure, human immunodeficiency virus (no cases), or cancer (2.9% versus 13.5%; P = 0.0570) (Supporting Table 3). This is most likely due to the eligibility of individuals with conditions for Medicare/Medicaid coverage. Additionally, individuals with Medicare/Medicaid were more likely to have hypertension (54.8% versus 26.0%; P = 0.0183) and arthritis (52.71% versus 37.53% in uninsured subjects [P = 0.1043] and versus 24.64% in privately insured subjects [P = 0.0165]) (Supporting Table 1), both probably being related to the age distributions of the respective populations.
|Demographics||Insured HCV+||Uninsured HCV+||P||Type of Insurance|
|No. of subjects||59||35||42||19|
|Prevalence, %||54.33 ± 7.67||45.67 ± 7.67||40.10 ± 8.15||15.25 ± 4.63|
|Caucasian, %||54.10 ± 8.27||77.97 ± 8.29||0.043||57.64 ± 9.84||41.13 ± 13.48|
|African American, %||28.55 ± 7.17||18.97 ± 7.23||0.332||26.85 ± 7.69||37.81 ± 14.01|
|Hispanic, %||13.67 ± 4.81||3.07 ± 2.23||0.061||15.51 ± 6.16||7.92 ± 5.47|
|Other, %||3.69 ± 2.49||0.00 ± 0.00||0.155||0.00 ± 0.00||13.13 ± 7.19|
|Male sex, %||63.83 ± 8.06||64.18 ± 10.25||0.977||67.15 ± 9.41||54.98 ± 10.61|
|Age, years||47.20 ± 1.21||46.60 ± 1.44||0.723||47.53 ± 1.21||47.20 ± 2.92|
|Married, %||41.34 ± 9.10||38.74 ± 8.24||0.839||51.95 ± 10.73||14.86 ± 9.29*|
|Military service, %||5.15 ± 2.52||11.45 ± 6.62||0.268||6.98 ± 3.52||4.19 ± 4.30|
|United States citizen, %||97.59 ± 1.78||98.17 ± 1.89||0.824||96.74 ± 2.43||100.00 ± 0.00|
|College degree, %||9.47 ± 4.90||1.93 ± 2.00||0.178||12.78 ± 6.46||0.00 ± 0.00*|
|Household income/federal poverty level||2.07 ± 0.27||1.81 ± 0.19||0.623||2.33 ± 0.35||1.38 ± 0.31|
|History of substance use|
|Alcohol, %||23.74 ± 6.60||31.55 ± 8.66||0.484||24.68 ± 8.13||19.17 ± 6.87|
|Smoking, %||86.58 ± 5.96||88.19 ± 7.24||0.852||84.35 ± 7.26||90.77 ± 6.64|
|Marijuana or hashish, %||92.13 ± 3.74||91.12 ± 6.65||0.894||89.02 ± 5.25||100.00 ± 0.00|
|Opiates or methamphetamine, %||71.81 ± 6.92||72.83 ± 9.76||0.927||70.49 ± 8.32||74.24 ± 11.67|
|Excellent, %||3.42 ± 2.22||4.50 ± 2.52||0.765||2.36 ± 1.91||5.98 ± 6.12|
|Very good, %||27.09 ± 7.73||25.15 ± 7.10||0.856||32.95 ± 9.20||9.86 ± 7.28|
|Good, %||46.98 ± 7.52||40.67 ± 8.69||0.588||46.93 ± 8.02||43.94 ± 18.62|
|Fair, %||16.19 ± 4.72||24.20 ± 8.14||0.423||15.88 ± 5.59||22.63 ± 10.34|
|Poor, %||6.32 ± 3.67||5.48 ± 4.40||0.885||1.88 ± 1.46||17.58 ± 12.24|
|Health care use in last year|
|Number of health care visits, %||4.47 ± 0.80||4.41 ± 0.97||0.938||3.02 ± 0.50||8.27 ± 1.24|
|Hospitalization, %||9.65 ± 5.35||11.75 ± 4.96||0.775||3.47 ± 2.51*)||25.26 ± 14.13|
|Mental specialist visit, %||14.89 ± 4.85||21.12 ± 9.25||0.575||14.80 ± 5.93||14.14 ± 8.63|
|Primary source of health care|
|Clinic or health center, %||14.50 ± 5.19||7.03 ± 3.91||0.267||16.42 ± 6.86||8.47 ± 5.59|
|Doctor's office or HMO, %||62.83 ± 7.96||40.94 ± 7.19||0.064||55.70 ± 9.76||79.90 ± 9.31*|
|Hospital ER, %||4.40 ± 2.39||7.68 ± 3.48||0.458||3.14 ± 2.33||11.62 ± 7.63|
|Hypertension, %||32.43 ± 7.33||22.54 ± 7.79||0.396||27.00 ± 7.81||47.05 ± 15.57|
|Hypercholesterolemia, %||16.71 ± 7.29||14.05 ± 10.12||0.835||16.08 ± 8.22||18.32 ± 14.76|
|Diabetes, %||4.02 ± 2.11||6.32 ± 3.55||0.598||1.77 ± 1.17||9.67 ± 7.13|
|Depression, %||3.75 ± 2.48||2.61 ± 2.68||0.762||2.53 ± 2.51||6.69 ± 6.27|
|Asthma, %||7.57 ± 3.18||17.36 ± 8.30||0.264||4.64 ± 3.05||14.74 ± 5.98|
|Arthritis, %||23.52 ± 7.12||31.07 ± 7.70||0.496||13.7 ± 6.7*||51.11 ± 14.50|
|HIV+, %||5.57 ± 4.07||0.00 ± 0.00||0.175||7.87 ± 5.79||0.00 ± 0.00|
After adjusting for these differences simultaneously in the multivariable model, Caucasians (OR, 0.20; 95% CI 0.06-0.64), individuals reporting alcohol use (OR, 0.28; 95% CI, 0.09-0.87), and individuals with diabetes were less likely to be insured than their counterparts. In contrast, HCV-infected individuals with a college degree were more likely to have health insurance than those without a college degree (OR, 3.42; 95% CI, 1.15-8.35).
Insurance Coverage Among Candidates for Treatment with Pegylated Interferon and Ribavirin.
Among all HCV+ subjects, only 66.7% (n = 94) were found to be potential treatment candidates. Of these, 54.3% had any health insurance, which was lower than the coverage rate among those who were potentially ineligible for treatment (75.5%; P = 0.071). The distribution of insurance types among potential treatment candidates was not significantly different from the distribution in the entire HCV+ cohort.
There was little difference in the sociodemographic and health-related characteristics between treatment eligible patients with and without health insurance (Table 4). However, when we considered different types of insurance, HCV+ treatment candidates covered by Medicare or Medicaid were less likely to have a college degree (no cases) and to be married (14.9% versus 40.2% in all HCV treatment candidates) than uninsured. On the other hand, HCV treatment candidates with private or military/state/government plans had lower prevalence of chronic diseases such as asthma, arthritis, and diabetes (4.6%, 13.7%, and 1.8% versus 12.0%, 27.2%, and 5.1% for all HCV treatment candidates, respectively). The patterns of health care use also varied by the type of health insurance: uninsured HCV treatment candidates were significantly less likely to use doctors' offices or HMOs to receive health care compared with those with Medicare/Medicaid and were far more likely to be hospitalized in the year prior to the survey than those with private insurance.
In addition to the described sociodemographic and clinical factors, we also examined the following laboratory parameters: blood creatinine and albumin, ALT, AST, APRI,18 total bilirubin, complete blood count, fasting glucose and insulin, triglycerides, and total cholesterol together with high- and low-density lipoprotein cholesterol, and found no differences between groups based on their insurance coverage or treatment candidacy (Supporting Table 1).
This is a comprehensive study based on recent population-based data that assesses the health insurance coverage and treatment candidacy of HCV-infected individuals in the United States. Our data show that only a third of HCV-infected individuals in the United States can potentially benefit from and have access to antiviral treatment; the remaining individuals are either uninsured or have potential contraindications to antiviral treatment.
We found that approximately two-thirds of HCV-infected individuals in the United States may be potential candidates for treatment. However, only half of these individuals have any form of health insurance coverage. Although treatment exclusions due to absolute contraindications will likely remain an issue, our data show that by removing the insurance-related barrier, twice as many HCV individuals may gain access to potentially effective treatment regimens for hepatitis C.
Our study also shows that, regardless of their treatment candidacy, individuals with chronic hepatitis C have a very low rate of health care insurance coverage. In the United States, HCV+ individuals are twice more likely not to have health insurance than their counterparts without HCV infection. In fact, we found that HCV infection was an independent predictor of lack of insurance, and that this association could not be attributed solely to sociodemographic factors of HCV-infected population. Furthermore, we found that only 24% of HCV+ individuals without insurance had any knowledge of their chronic liver disease (compared with 50% among insured; P = 0.0300). Better insurance coverage may not only improve antiviral treatment rates, but also enhance rates of hepatitis C testing (i.e., screening) and diagnosis, particularly among those who are at high risk for infection. Furthermore, early diagnosis and counseling may enhance patients' knowledge about their liver disease and may increase antiviral treatment rates by identifying patients earlier in the course of their disease.
We also found that HCV+ individuals without health insurance were more likely to report history of alcohol abuse and were less likely to be educated than insured. It is plausible that the high prevalence of social comorbidity and lack of education may still hamper treatment acceptance and initiation among individuals after they are diagnosed with HCV infection. To make any impact on the burden of HCV and to cover the gap between efficacy and effectiveness, not only more individuals need to be screened for and diagnosed with HCV, but more focus is needed on HCV-related social services and education—comprehensive HCV care that may be best delivered through medical homes using the chronic care model approach.3
Our data show that currently, 1.2% of the United States population has active HCV viremia. This rate is lower than the previously reported national prevalence rate of CHC calculated using NHANES III, which was conducted between 1988 and 1994.19, 20 This drop is most likely related to the recent decline in the incidence of new cases of HCV infection coupled with treatment availability.
The strength of our study is the use of contemporary United States population-based data. Although similar data on treatment eligibility are available from the Veterans Administration,20, 21, 22 these are the first large-scale data that may be generalizable to all HCV-infected individuals in the community setting. Furthermore, the NHANES study design provides standardized data collection and follow-up, thus there are no ascertainment or selection biases.
The main limitation of the study is that, though it is based on population-level data, the sample of HCV-infected individuals used for calculations is still relatively small. Another important limitation is that, after applying our study eligibility criteria, a large portion of NHANES participants was excluded primarily due to the age requirement (age >18 years at the time of examination). Furthermore, a proportion of adults was excluded because of incomplete insurance questionnaires and absence of hepatitis C serologic tests. Because the reasons for absence of laboratory tests and having incomplete questionnaires for adult participants are not clear, these exclusions may have caused a bias in our calculations, the direction of which is unclear. Additionally, NHANES participants enrolled by the Centers for Disease Control and Prevention were only United States residents with addresses, and therefore does not include the homeless population, which is expected to have high prevalence of HCV infection without insurance coverage. Given these limitations, our results may underrepresent both HCV prevalence as well as rates of insurance coverage in the HCV population. Finally, we were unable to link insurance coverage to treatment receipt in this analysis. Therefore, we cannot indicate which patients had already received treatment. However, current data suggest that less than 25% of patients diagnosed with HCV have ever been treated. This low treatment rate is especially true for the uninsured, who are even less likely to have received treatment previously. Future studies should look into these issues in more depth.
It is important to note that some of the contraindications to treatment used in this study, such as active congestive heart failure, may be considered as an absolute contraindication by most health care providers. However, other conditions (such as depression and diabetes) could potentially be temporary and reversible. On the other hand, some patients with relative contraindications—who may have been classified as treatment candidates under our study assumption—may develop absolute contraindication or may not be treated by their physicians in the community. Furthermore, there are other reasons that physicians and patients might choose to forego treatment. HCV is a slowly progressing and often asymptomatic condition, and its treatment has significant adverse effects and results in a response in only approximately half of the patients. Thus, a number of patients with early liver disease may be counseled against treatment or elect not to receive treatment. In addition, patients' compliance with their physicians' recommendations and their history of unsuccessful treatment may further reduce these treatment candidacy rates. These determinations (regarding relative treatment contraindications, patient acceptance, and patient compliance) require a comprehensive and HCV-specific evaluation. Therefore, our estimate of treatment-eligible patients, although likely biased upward, may be taken as a best-case scenario with the largest possible denominator that will require a targeted evaluation in order to accurately ascertain treatment eligibility.
In conclusion, a high proportion of HCV+ individuals in the United States are currently uninsured, and many have publicly funded health insurance. Among those who could potentially be candidates for treatment, the rate of insurance coverage is even lower. Although newer treatment regimens with direct acting antivirals may increase efficacy, it will certainly increase the costs of antiviral treatment in HCV—thus further limiting access to treatment for the uninsured/underinsured. This issue of access to care for HCV patients is critical and must be considered by policy makers. We believe that our results can have important implications for health insurance coverage of HCV-infected patients and should be considered under the new health care reform legislation.
- 16National Center for Health Statistics, Centers for Disease Control and Prevention. Analytic and reporting guidelines: The National Health and Nutrition Examination Survey (NHANES). Hyattsville, MD: National Center for Health Statistics, Centers for Disease Control and Prevention; 2005.
- 17U.S. Census Bureau. Percent of people 25 years and over who have completed a bachelor's degree. http://www.census.gov/acs/www/Products/Ranking/2003/R02T040.htm. Published 2003. Accessed August 10, 2010.
|Supplement1Final.doc||258K||Supporting Information data|
|SupplementaryTable2.doc||54K||Supporting Table 2|
|SupplementaryTable3.doc||39K||Supporting Table 3|
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.