| |
Direct Economic Burden of Chronic Hepatitis C Virus in a United States Managed Care Population: 'HCV costs 4 times control group, SVR reduced costs'
|
| |
| |
Download the PDF here
"Adjusted all-cause costs were $20,961 per HCV patient, compared with $5451 per control......patients who achieved SVR incurred approximately half the HCV-related costs per month incurred by those who did not achieve SVR ($717 vs. $1436; P<0.0001)."
"Our study provides current estimates of the direct economic burden of chronic HCV infection to MCOs.....total annual per patient costs of $20,961 estimated in this study. Our study findings thus support earlier predictions that the economic burden of HCV will continue to grow as a result of an aging cohort living with the disease and will be borne increasingly by commercial insurance providers.15,16 Our findings also confirm that the annual per patient cost burden to payers for chronic HCV exceeds that of other more common conditions such as cardiovascular disease ($18,953)27 and type 2 diabetes ($9677)"
"......chronic HCV is an important and potentially costly disease to MCOs and confirms earlier expectations of cost growth associated with the disease. Increased efforts in HCV screening and early treatment, particularly before progression to liver cirrhosis, may therefore, lead to substantial cost savings for MCOs."
Journal of Clinical Gastroenterology: February 2011
Davis, Keith L. MA*; Mitra, Debanjali MA*; Medjedovic, Jasmina MSc; Beam, Cynthia PhD; Rustgi, Vinod MD
*RTI Health Solutions, Research Triangle Park, NC
Human Genome Sciences Inc., Rockville, MD
Georgetown University Medical Center, Washington DC
Novartis Pharma AG, Basel, Switzerland
Funding Source: This study and the writing and preparation of this manuscript were funded in full by Human Genome Sciences Inc. and Novartis Pharma AG. All researchers had complete access to the data that support this manuscript.
"HCV cases had a higher baseline comorbidity burden (mean CCI=0.95) compared with controls (mean CCI=0.24). Finally, among the subset of patients for whom disease severity was assessed (N=2877), 83% were classified has having an APRI score consistent with mild HCV."
"Compared with controls, a higher proportion of patients with HCV had at least 1 claim in every all-cause utilization category examined. The largest differences in the percentage of patients with use were observed for hospitalizations (24% of HCV cases vs. 7% of controls; P<0.0001), emergency room visits (32% vs. 15%; P<0.0001), and laboratory tests (79% vs. 35%; P<0.0001). Among those with HCV, 14% had at least 1 disease-related hospitalization."
C"ompared with controls, HCV patients had substantially higher use and associated costs for all service categories. Adjusted total all-cause costs incurred during the 12-month follow-up period were $20,961 per patient among HCV cases compared with $5451 among controls (P<0.0001). Key cost drivers among HCV patients were prescription drugs ($6191) and hospitalizations ($5892). In contrast, the control group incurred only $1315 and $1159 per patient for pharmacy and hospital services, respectively (P<0.0001). Although HCV patients received, on average, only 2 disease-related prescriptions during follow-up (reflecting a low treatment rate in the population studied), HCV-related drugs accounted for more than half of their prescription drug costs. Overall, disease-related resource use accounted for nearly one-third of all costs incurred by patients with HCV. Disease-related costs in the HCV group were $6864, which exceeded all-cause costs observed for controls by 26% (P<0.0001). This difference was driven primarily by disease-related hospitalization and pharmacy costs in patients with HCV ($2078 and $3433, respectively), which exceeded all-cause hospitalization and pharmacy costs in controls ($1159 and $1315, respectively) by 79% and 161%, respectively (all P<0.0001)."
"........During the minimum 6-month period after the end of treatment and SVR assessment, patients who achieved SVR incurred approximately half the HCV-related costs per month incurred by those who did not achieve SVR ($717 vs. $1436; P<0.0001). Finally, disease severity was assessed for 2877 patients (Table 5). The severity distribution among these patients, as measured by the APRI score, was 83% mild, 13% moderate, and 4% severe. Adjusted disease-related hospitalization costs incurred during 12 months after the index date were 3 times higher in patients with moderate HCV ($4425 per patient) and approximately 5.5 times higher in severe patients ($8077) compared with those with mild disease ($1472; both P<0.001). Total HCV-related costs, inclusive of all service categories, were nearly 2 times higher among patients with severe disease compared with those with mild HCV ($12,481 vs. $6839; P<0.0001). There was no significant difference in total disease-related costs between patients with severe and moderate HCV."
Abstract
Goals and Background: To estimate all-cause and disease-related resource utilization and costs among managed care enrollees with chronic hepatitis C virus (HCV).
Study: A large United States claims database was analyzed (1/1/2002 to 12/31/2006). Inclusion criteria were: diagnosis of chronic HCV; no hepatitis B diagnoses; ≥6 and ≥12 months of continuous plan enrollment prediagnosis and postdiagnosis, respectively. Use and costs of medical services and prescription drugs over a 12-month period postdiagnosis were evaluated. Outcomes were assessed in controls without HCV matched (1:1) on age, sex, and plan enrollment. All cost estimates were generated using multivariate generalized linear models to adjust for additional covariates and skewness common in health care cost data.
Results: Of the 20,662 patients who met all inclusion criteria, mean age was 49 years; 61% were male. Adjusted all-cause costs were $20,961 per HCV patient, compared with $5451 per control (P<0.0001). Hospitalization occurred in 24% of HCV patients compared with 7% of controls (P<0.0001). Mean inpatient costs were $5892 and $1159 per patient, respectively (P<0.0001). Patients with HCV had higher prescription costs compared with controls ($6191 vs. $1315; P<0.0001). At $6864 per patient, disease-related costs were nearly one-third of all costs in patients with HCV, which exceeded all-cause costs among controls by 26% (P<0.0001).
Conclusions: Chronic HCV is a costly disease to managed care organizations. Disease-related costs in HCV exceed all-cause costs in demographically matched controls. Increased efforts in HCV screening and early treatment, particularly before progression to liver cirrhosis, may lead to long-term cost savings in HCV management for managed care systems.
Hepatitis C virus (HCV) is one of the most prevalent blood-borne infections in the world, currently affecting 180 million persons worldwide1 and 3.2 million persons in the United States.2 Each year more than 3 million new cases of HCV are identified worldwide; new cases in the United States alone exceed 25,000 annually.3 Although the true prevalence of HCV infection is unknown, it is likely much higher than what most published estimates suggest. As fewer than 20% of all HCV cases are symptomatic,4,5 many people with HCV have not been diagnosed. Moreover, many populations at increased risk for HCV, including prisoners, intravenous drug users, and the homeless, are typically excluded from national surveys from which most prevalence estimates are drawn.6
Although HCV infection is usually clinically silent during both the acute phase and for decades thereafter,4,7 its eventual clinical consequences can be profound. Approximately 20% of all patients with chronic HCV will develop cirrhosis of the liver.8,9 Chronic HCV is also the leading cause of hepatocellular carcinoma in the United States,6,10 and the most common cause of liver failure and subsequent need for liver transplant.11 Even patients who do not develop cirrhosis and other liver complications may suffer from chronic fatigue and decreased quality of life.6,12,13
Most patients currently infected with HCV are now in their fifth or sixth decade of life, likely having been infected during the 1960s and 1970s, primarily through experimentation with injection drugs.14 As the duration of infection increases in an aging population living with HCV, the incidence of liver disease and related sequelae is expected to rise. One earlier study characterized the current living cohort of HCV-infected patients as constituting an "age wave" of asymptomatic infection that is progressing toward clinical liver disease.15 A later study estimated that by 2020, the proportion of patients with cirrhosis will increase from 16% to 20%, hepatocellular carcinoma will increase by 81%, and liver-related deaths will increase by 180%.16
With the combined aging of the HCV population and the increasing costs for treatment, the economic burden of chronic HCV infection is expected to grow proportionally with its clinical burden over the next 10 to 20 years. Wong et al17 estimated that total direct medical expenditures attributable to chronic HCV in the United States between 2010 and 2019 will be $10.7 billion. Over this same period in the United States, it was also estimated that decompensated cirrhosis and hepatocellular carcinoma will lead to 720,000 life years lost and $21.3 billion in related societal costs, whereas the indirect costs attributable to HCV-related deaths in persons younger than 65 years are projected to be $54.2 billion. Because the age group most frequently diagnosed with HCV (people aged 30 to 49 y) is likely to be employed and enrolled in an employer-based insurance plan, the direct cost burden of HCV in the US may fall largely on managed care payers.18
Despite the potential cost implications of chronic HCV to managed care systems in the United States, few studies have analyzed administrative data to characterize the resource utilization and costs of HCV-infected patients in real-world practice settings. To address this knowledge gap, we analyzed retrospective insurance claims from a large managed care population to estimate the increased direct all-cause health care utilization and costs incurred by persons with chronic HCV relative to comparable individuals without HCV. Our study also provides current estimates of the overall direct economic burden of HCV to third-party payers, in terms of both all-cause costs and costs directly attributable to the disease. For subsets of patients in whom relevant data were available, disease-related costs were furthermore stratified by important clinical factors, including viral genotype, attainment of sustained viral response (SRV), and underlying disease severity.
MATERIALS AND METHODS
Data Source
Data were extracted from the Integrated Health Care Information Services (IHCIS) Managed Care Benchmark Database, a commercially available source of administrative medical and pharmacy claims data, and member enrollment information, from 30 managed care organizations (MCOs) across the United States.
At the time of this study, the IHCIS database included more than 50 million unique lives from 2000 to 2006. The database has a similar distribution of age and sex compared with national managed care enrollment, but geographic representation is skewed toward the United States East Coast. Encrypted patient identifiers allow claims in the database to be linked chronologically for any given patient. Use of the IHCIS database for this study was authorized by a Federal-Wide Assurance institutional review board.
The IHCIS database comprises separate files for inpatient stays, medical claims, pharmacy claims, and health plan enrollment. Laboratory test results also are available for a small subset of patients enrolled in plans that allow submission of laboratory values to IHCIS. The member enrollment file includes start and stop dates of health plan enrollment, type of coverage, and patient demographics. The medical file includes claims for professional services (both inpatient and outpatient) such as physician office visits and laboratory tests. This file includes, among other variables, claim-level revenue codes, Health Care Financing Administration Common Procedure Coding System and Current Procedural Terminology Version 4 procedure codes, International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes, provider type, place of service, and reimbursed costs for each encounter. The inpatient file contains a summarized record for inpatient stays from acute care hospitals or skilled nursing facilities, including admission and discharge date, length of stay, and other variables similar to those contained in the medical claims file. Finally, the pharmacy file contains prescription drug claims for medications dispensed on an outpatient basis. These claims include the National Drug Code, quantity dispensed, days' supply, medication strength, and reimbursed amount for each prescription dispensed.
Study Population
Patients with a diagnosis of chronic HCV (ICD-9-CM codes 070.44, 070.54, 070.70, or 070.71) between January 1, 2002, and December 31, 2006, the most recent years of IHCIS data available for this analysis, were selected for initial study inclusion. Patients were required to have no diagnoses of hepatitis B virus at any point in their claims history. Patients with acute HCV (ICD-9-CM codes 070.41 or 070.51), but without a diagnosis of chronic HCV were excluded.
For each patient, an index date was defined as the date of the first observed HCV diagnosis. To help ensure the availability of an adequate period over which comorbidity history could be evaluated, patients were required to have at least 6 months of continuous health plan enrollment before their index date. To help ensure that any observed lack of health care events and associated costs after the first observed HCV diagnosis was owing to a true lack of medical activity and not cessation of insurance, patients were further required to have at least 12 months of continuous plan enrollment after their index date.
To assess the increased all-cause health care utilization and costs associated with chronic HCV, we also selected a demographically matched control group without a diagnosis of HCV. Patients in the control group met all inclusion criteria applied to the HCV case group. The control group was matched to HCV cases at a 1:1 ratio on age (±3 y), sex, and length of health plan enrollment. Controls were assigned the same index date identified for their matched counterparts in the HCV case group.
Baseline Patient Characteristics
Patient age, sex, geographic region, insurance payer type (eg, commercial, Medicare, Medicaid), and health plan type [eg, health maintenance organization (HMO), preferred provider organization] were measured at the index date. To assess overall comorbidity burden before the index date, we calculated a Charlson Comorbidity Index (CCI) score for each patient. The CCI includes 17 categories of comorbidities, defined by ICD-9-CM diagnosis codes, with associated weights corresponding to the severity of the comorbid condition observed.19 Ranging in value from 0 for patients with no comorbidities to 33 for those with all 17 comorbidities, a higher CCI score represents a higher overall comorbidity burden. A single CCI score was calculated for each patient based on the presence of the corresponding diagnosis during the 6-month period before the index date.
Outcome Measures
All medical claims observed during the 12-month follow-up period were aggregated at the encounter level for each patient to obtain measures of all-cause and disease-related health care utilization and associated costs. All cost data were adjusted to 2007 US dollars and represent reimbursed amounts paid by health plans to providers. Copayments were excluded from all cost data. Medical services were grouped according to the setting in which they were delivered, including hospital admissions, emergency room, physician office, and home health visits, laboratory services, prescription drugs, and a residual category capturing all other outpatient and ancillary services. The subset of all-cause medical services and costs specifically related to HCV were identified as those claims with a primary or secondary ICD-9-CM diagnosis code for chronic HCV. HCV-related pharmacy utilization was identified as the subset of prescription drug claims or Health Care Financing Administration Common Procedure Coding System administration codes for any of these medications used in the treatment of HCV: peginterferon α-2a or α-2b combination therapy with ribavirin; peginterferon α-2a or α-2b monotherapies; interferon α-2a or α-2b combination therapy with ribavirin; interferon α-2a or α-2b monotherapies; interferon alfacon-1 (consensus interferon) regimens with ribavirin; or consensus interferon monotherapies.
Analysis Stratification
All analyses of HCV-related health care costs were furthermore stratified by important clinical factors, including viral genotype, SVR attainment, and underlying disease severity among subsets of patients with valid laboratory test results and other information required for constructing these analysis strata. These patients were identified from a small number of health plans that allow patient-level laboratory findings to be transmitted to IHCIS. Viral genotype was determined for the subset of patients with a valid viral genotype test result, defined by Logical Observation Identifiers, Names, and Codes (LOINC) numbers 32286-7, 20416-4, 48574-8, and 48575-5, and by searching lab result descriptions for combinations of keywords such as "HEPATITIS," "HEP," "GENOTYPE," "GENO," and "GENOTYPING." All lab descriptions obtained for HCV genotype testing were manually reviewed for accuracy. On the basis of these test values, patients were categorized as either genotype 1 or genotype 2/3.
SVR attainment is clinically measured as an undetectable HCV viral ribonucleic acid (RNA) test result at least 6 months after completion of treatment.20 In this study, patients were identified as having attained SVR if an HCV viral RNA test done at any point at least 6 months after completion or discontinuation of an HCV-specific treatment showed the viral load to be undetectable. HCV viral RNA test results were identified using LOINC numbers 10676-5, 1011-4, 20571-6, 11259-9, 20416-4, 29609-5, 34703-9, 34704-7, 38180-6, 42617-1, 47252-2, 48576-3, 5010-4, 5011-2, 5012-0, and 6422-0. If a patient had multiple valid HCV viral RNA tests, the first valid result observed at least 6 months after therapy completion or discontinuation was used. Stratification of disease-related costs by SVR attainment required the application of additional inclusion criteria, including receipt of an HCV treatment, and at least 6 additional months of health plan enrollment after treatment completion or discontinuation. Among patients for whom SVR attainment was evaluated, per patient per month costs were calculated over the period between treatment completion or discontinuation and the first observed qualifying viral RNA test (minimum of 6 mo).
Finally, baseline HCV histologic severity was estimated using the aspartate aminotransferase-to-platelet ratio index (APRI) score. APRI scores were computed for the subset of patients with at least 1 valid laboratory result for both an aspartate aminotransferase and platelet count test (occurring on the same day) during the 6-month period preceding the index date. For patients with multiple valid test results, the most recent result (ie, closest to the index date) was used. Patients were then classified as having mild (APRI≤0.5), moderate (0.5< APRI ≤1.5), or severe (APRI>1.5) HCV fibrosis based on these test results.21
RESULTS
Patient Characteristics
We identified a total of 20,662 unique patients with chronic HCV who met all study inclusion criteria. Table 1 presents descriptive statistics on various baseline characteristics for these patients. Approximately 61% of the study sample was male and mean age was 49 years. Controls, which were matched to HCV cases on sex, age, and health plan enrollment, had a nearly identical sex and age distribution compared with patients with HCV. Geographic distribution varied substantially between cases and controls, with only 55% of cases residing in the Northeast compared with 83% of controls. The case group also had disproportionately lower HMO enrollment (38%) compared with controls (53%). Although reasons for the large discrepancies in geographic and HMO distribution between groups are unclear, these factors are controlled in the multivariate cost analyses. HCV cases had a higher baseline comorbidity burden (mean CCI=0.95) compared with controls (mean CCI=0.24). Finally, among the subset of patients for whom disease severity was assessed (N=2877), 83% were classified has having an APRI score consistent with mild HCV.
Health Care Utilization and Costs-HCV Versus Control
Figure 1 shows the percentage of patients with use of each category of all-cause health care utilization during the 12-month follow-up period. Among HCV cases, the percentage with disease-specific utilization is also presented. Compared with controls, a higher proportion of patients with HCV had at least 1 claim in every all-cause utilization category examined. The largest differences in the percentage of patients with use were observed for hospitalizations (24% of HCV cases vs. 7% of controls; P<0.0001), emergency room visits (32% vs. 15%; P<0.0001), and laboratory tests (79% vs. 35%; P<0.0001). Among those with HCV, 14% had at least 1 disease-related hospitalization.
Table 2 presents the volume of use and associated costs of each service category. Among patients in the HCV case group, all-cause utilization and costs were subset to those directly attributable to HCV. Compared with controls, HCV patients had substantially higher use and associated costs for all service categories. Adjusted total all-cause costs incurred during the 12-month follow-up period were $20,961 per patient among HCV cases compared with $5451 among controls (P<0.0001). Key cost drivers among HCV patients were prescription drugs ($6191) and hospitalizations ($5892). In contrast, the control group incurred only $1315 and $1159 per patient for pharmacy and hospital services, respectively (P<0.0001). Although HCV patients received, on average, only 2 disease-related prescriptions during follow-up (reflecting a low treatment rate in the population studied), HCV-related drugs accounted for more than half of their prescription drug costs. Overall, disease-related resource use accounted for nearly one-third of all costs incurred by patients with HCV. Disease-related costs in the HCV group were $6864, which exceeded all-cause costs observed for controls by 26% (P<0.0001). This difference was driven primarily by disease-related hospitalization and pharmacy costs in patients with HCV ($2078 and $3433, respectively), which exceeded all-cause hospitalization and pharmacy costs in controls ($1159 and $1315, respectively) by 79% and 161%, respectively (all P<0.0001).
HCV-related Costs by Clinical Characteristics
Tables 3 to 5 present adjusted disease-related costs incurred by subsets of HCV patients in whom various clinical characteristics could be evaluated. Among 321 patients with a genotype assessment, 248 (77%) were genotype 1 (Table 3). Patients with viral genotype 2/3, however, incurred approximately $2556 more in total HCV-related costs during follow-up compared with those with genotype 1 (P<0.0001). SVR attainment was assessed for 575 patients, of whom 336 (58%) achieved SVR (Table 4). During the minimum 6-month period after the end of treatment and SVR assessment, patients who achieved SVR incurred approximately half the HCV-related costs per month incurred by those who did not achieve SVR ($717 vs. $1436; P<0.0001). Finally, disease severity was assessed for 2877 patients (Table 5). The severity distribution among these patients, as measured by the APRI score, was 83% mild, 13% moderate, and 4% severe. Adjusted disease-related hospitalization costs incurred during 12 months after the index date were 3 times higher in patients with moderate HCV ($4425 per patient) and approximately 5.5 times higher in severe patients ($8077) compared with those with mild disease ($1472; both P<0.001). Total HCV-related costs, inclusive of all service categories, were nearly 2 times higher among patients with severe disease compared with those with mild HCV ($12,481 vs. $6839; P<0.0001). There was no significant difference in total disease-related costs between patients with severe and moderate HCV.
DISCUSSION
Our study provides current estimates of the direct economic burden of chronic HCV infection to MCOs. Although limited data are available in the literature to compare with the findings of the present study, 1 earlier analysis estimated that employer-based insurance plans incur $10,925 per year in all-cause health care costs for persons with HCV.25 Another study from the same period estimated annual all-cause expenditures by managed care plans for HCV-infected individuals to be approximately $13,000 per patient.26 Adjusted to 2007 dollars using the medical consumer price index, these estimates ($15,107 and $17,977 in 2007 dollars, respectively) are smaller than the total annual per patient costs of $20,961 estimated in this study. Our study findings thus support earlier predictions that the economic burden of HCV will continue to grow as a result of an aging cohort living with the disease and will be borne increasingly by commercial insurance providers.15,16 Our findings also confirm that the annual per patient cost burden to payers for chronic HCV exceeds that of other more common conditions such as cardiovascular disease ($18,953)27 and type 2 diabetes ($9677).28
Our study also documents the proportion of total HCV costs that are directly related to the disease, and the potential impact of various clinical characteristics on disease-related costs. We found that nearly one-third of all costs incurred by HCV patients were attributable to the disease itself; more than one-half of all prescription drug costs were attributable to HCV-specific treatments. Of particular importance was our finding that disease-related costs in the HCV group substantially exceeded all-cause costs observed for controls. This difference was driven primarily by significantly higher disease-related hospitalization and pharmacy costs in patients with HCV compared with all-cause costs for these services in controls. These results suggest that a nontrivial proportion of the additional costs incurred by MCOs for patients with HCV can be attributed to the disease itself.
Stratifying by viral genotype, patients with genotype 2 or 3 seemed to have higher disease-related costs than those with genotype 1. As patients with genotype 1 generally represent cases with more severe HCV fibrosis, higher overall costs in patients with genotype 2 or 3 may seem counterintuitive. However, a substantially higher proportion of participants with genotype 2 or 3 received an HCV-related treatment in our study (62% of patients) compared with patients with genotype 1 (only 35% of patients). Combined with the high unit cost of HCV treatments, the larger proportion of genotype 2 or 3 patients receiving treatment seems to offset the treatment costs incurred by a much smaller proportion of patients with genotype 1 who received treatment. As shown in Table 3, the overall difference in disease-related costs by genotype was driven almost entirely by higher costs for prescriptions and associated office visits.
Stratifying by other clinical characteristics, SVR attainment was associated with reduced posttreatment costs, and persons with APRI scores indicative of severe HCV incurred (as expected) substantially higher disease-related costs than persons with mild chronic disease. Services that were delivered for the management of HCV that did not carry an associated HCV diagnosis were not factored into our calculation of disease-related costs. Our estimates of HCV-related costs are therefore, conservative.
Our study is subject to several limitations. First, our analysis relied on administrative claims submitted solely for purposes of insurance reimbursement and not for purposes of research, with no access to information collected from either the attending physician or the participant. The accuracy with which diagnoses, service categories, and other parameters of the study were identified is therefore, dependent upon the accuracy of record keeping by the MCOs submitting data to IHCIS. The general impact of such misclassification bias stemming from claims data has been described in earlier research.29,30 In this study, case misclassification may have resulted in a smaller study sample than would have otherwise been available. Specifically, the study database contained a nontrivial number of patients with only a diagnosis of acute HCV who met all other inclusion criteria. A total of 6131 acute-only patients were identified, representing 22% of the final study sample before their exclusion from the analysis. As inherent with most administrative claims databases, information was not available on patients' history before health plan enrollment or before the longitudinal period captured by the database. Some patients therefore, may have received and carried a diagnosis of chronic HCV at some point (possibly many years) before study entry. Some patients may also have been identified as having acute HCV owing to simple coding errors on a claim form. For these reasons, and considering that acute HCV is uncommon in real-world practice, it is likely that most of the acute-only patients were in actuality chronic HCV cases. However, it is not possible in our data to determine the extent to which these cases were misclassified because of coding errors or because of a lack of data before the observation period. Despite the likelihood that the acute-only patients were truly chronic HCV cases, a conservative methodology for patient identification, such as that which was implemented, remains the most appropriate and rigorous approach for this type of study. Even with a conservative approach for patient selection, a very large sample of analyzable chronic HCV cases (20,662 patients) was produced. To confirm that the apparent case misclassification did not affect the study results, all analyses were replicated with the acute-only cohort included in the study sample and no changes in the key results or overall study findings were identified (tabular data available upon request).
Second, findings from our stratified analysis of costs by various clinical factors (ie, genotype, SVR attainment, and disease severity) should be interpreted with caution. These clinical measures were available only for a small subset of patients and therefore, results from these analyses may not be generalizeable to the overall study sample or to the national chronic HCV population. Findings from the stratified analysis may be furthermore confounded if patients with and without the relevant lab data systematically differ on key background characteristics (eg, age and sex) or if the categoric distributions of the clinical variables differ in the subpopulation from what is expected in a general chronic HCV population.
Despite these potential sources of confounding, additional data from our study sample and from existing literature suggest that such confounding may be limited. For example, sex distribution and mean age were similar (approximately 60% male and 50 y of age) for both the overall study sample and for the respective subsamples with relevant lab results. Moreover, the genotype distribution and SVR rate among patients for whom these variables were measured were similar to corresponding estimates cited elsewhere in the literature. We found that genotype 1 represented 77% of patients for whom it was measured, whereas genotype 2/3 represented 23% of patients. These estimates are consistent with those from an earlier study31 of genotype distribution in a US population of chronic HCV patients, which found genotype 1 to represent 79% of patients and genotype 2/3 to represent 20%. (Genotype 4 in this study was 1% of the sample). Among treated patients for whom SVR was measured, we found an SVR attainment rate of 58%. This estimate is consistent with the range of SVR rates (56% to 63%) cited in 3 earlier studies of chronic HCV treatment.32-34
Although our data on genotype distribution and SVR rates are consistent with expectations, we found a lower-than-expected rate of severe HCV as measured by the APRI score. Although the APRI score has been shown to have adequate sensitivity for detecting severe disease,35 the test's major strength is the diagnostic exclusion of significant HCV-related fibrosis.21 If healthier patients, such as those found in an employed managed care population, are more likely to receive a noninvasive test such as the APRI (versus an invasive diagnostic measure such as liver biopsy), the population of patients receiving an APRI test may be inherently biased toward less severe disease. Despite this potential bias, the key cost endpoints in our analysis varied as expected when stratified by the APRI severity score, as severe disease was associated with significantly higher costs, particularly for hospitalizations. We additionally estimated the percentage of patients in each APRI severity range who had a diagnosis code for liver disease and nonalcoholic cirrhosis at some point after their index date.
Concordant with expectations, we found that patients with severe HCV at baseline, as measured by the APRI score, developed liver disease at more than 2-fold the rate (46.5% vs. 20.6%) observed in patients with mild HCV at baseline. Patients with severe HCV also developed nonalcoholic cirrhosis during follow-up at nearly 10 times the rate as patients with mild HCV at baseline (15.5% vs. 1.6%). Although these data are suggestive of relatively accurate identification of severe patients using the APRI score, it is not possible in our observational database to confirm disease severity.
Third, our cost analyses represent direct medical costs incurred by MCOs. Although the study population included a small number of patients with supplemental Medicare and Medicaid coverage, our analysis largely ignores costs paid by noncommercial payers and the broader societal costs of HCV, including caregiver burden and lost workplace productivity. Finally, our study sample was derived from a largely employed population enrolled in a commercial managed care health plan. HCV patients selected from this population may therefore, have had better underlying health and more favorable downstream outcomes compared with a national chronic HCV population, which includes other key HCV risk groups such as prisoners and the homeless who are generally excluded from the population from which our study data were extracted.
Despite these limitations, our large study cohort provided a unique opportunity to examine the direct economic burden of HCV to commercial insurers. Findings from this study suggest that chronic HCV is an important and potentially costly disease to MCOs and confirms earlier expectations of cost growth associated with the disease. Increased efforts in HCV screening and early treatment, particularly before progression to liver cirrhosis, may therefore, lead to substantial cost savings for MCOs. Future studies are needed to assess whether the increased costs associated with such efforts would be offset by expected long-term reductions in medical resource utilization. Additional study is also needed to assess the cost implications of HCV in other markets outside of the United States where disease prevalence is higher.
|
|
| |
| |
|
|
|
