- Email: [email protected]
Routine hepatitis C virus screening in pregnancy: A cost-effectiveness analysis
American Journal of Obstetrics and Gynecology (2005) 192, 1153–61
www.ajog.org
OBSTETRIC INFECTIOUS DISEASE
Routine hepatitis C virus screening in pregnancy: A cost-effectiveness analysis Beth A. Plunkett, MD, MPH, William A. Grobman, MD, MBA Department of Obstetrics and Gynecology, Northwestern University Feinberg School of Medicine, Chicago, Ill Received for publication June 9, 2004; revised September 22, 2004; accepted October 14, 2004
KEY WORDS Hepatitis C Screening Pregnancy Cost-effective
Objective: The purpose of this study was to determine whether routine hepatitis C virus screening in pregnancy is cost-effective. Study design: A decision tree with Markov analysis was developed to compare 3 approaches to asymptomatic hepatitis C virus infection in low-risk pregnant women: (1) no hepatitis C virus screening, (2) hepatitis C virus screening and subsequent treatment for progressive disease, and (3) hepatitis C virus screening, subsequent treatment for progressive disease, and elective cesarean delivery to avert perinatal transmission. Lifetime costs and quality-adjusted life years were evaluated for mother and child. Results: In our base case, hepatitis C virus screening and subsequent treatment of progressive disease was dominated (more costly and less effective) by no screening, with an incremental cost of $108 and a decreased incremental effectiveness of 0.00011 quality-adjusted life years. When compared with no screening, the marginal cost and effectiveness of screening, treatment, and cesarean delivery was $117 and 0.00010 quality-adjusted life years, respectively, which yields a cost-effectiveness ratio of $1,170,000 per quality-adjusted life year. Conclusion: The screening of asymptomatic pregnant women for hepatitis C virus infection is not cost-effective. Ó 2005 Elsevier Inc. All rights reserved.
In the United States, hepatitis C virus (HCV) has become a major public health concern. Nationally, 4 million individuals, most of whom have chronic disease, have been infected with HCV.1-3 Given the high prevalence of HCV infection and the increased risk of the development of cirrhosis or hepatocellular carcinoma in individuals with chronic disease,2,4,5 HCV has Supported by the Institute for Health Services Research and Policy Studies at Northwestern University and National Research Service Award 5 T32HS00078 from the Agency for Healthcare Research and Quality. Reprints not available from the authors. 0002-9378/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.10.600
become the leading indication for liver transplantation in the adult population.2 The greatest prevalence of infection occurs among individuals of reproductive age,1,6 and 1% to 4.3% of pregnant women are infected with HCV.1,7,8 Nevertheless, screening is not recommended currently during pregnancy for asymptomatic women without risk factors for HCV infection,9,10 and only 25% to 30% of women with HCV are aware of their diagnosis.9 Consequently, interventions that potentially could improve both maternal and neonatal long-term health outcomes are not available to most HCV-infected women and their children. These interventions may
1154
Plunkett and Grobman
Figure 1 Simplified decision tree. The encircled M refers to the different Markov models that are specific to disease status (neither mother nor neonate infected, maternal infection without perinatal transmission, maternal infection with perinatal transmission) and knowledge of disease status for both mother and child.
include measures of secondary prevention (early medical treatment for progressive disease for mother and child after the completion of the pregnancy11) and measures of primary prevention (elective cesarean delivery to prevent perinatal HCV infection). Although the benefits11 and cost-effectiveness of medical treatment for HCV12-19 have been analyzed in the nonobstetric population, information regarding cost-effective interventions and, in particular, the benefit of elective cesarean delivery and reduction in perinatal HCV transmission is considerably limited.20 As such, the purpose of this analysis was to determine under what circumstances, if any, routine HCV screening in pregnancy and elective cesarean delivery to avert perinatal transmission would be a cost-effective intervention.
Material and methods A decision tree model that used Markov analysis was developed to compare 3 approaches to HCV infection in pregnancy: (1) no HCV screening in pregnancy (the current standard of care), (2) HCV screening in pregnancy and subsequent treatment for progressive disease, and (3) HCV screening in pregnancy, subsequent treatment for progressive disease, and elective cesarean delivery to avert perinatal transmission (Figure 1). The screened population included all asymptomatic, HIVnegative pregnant women without risk factors for HCV infection who received routine prenatal care in the United States. The time course considered was the lifetime of the mother and her child. The analytic decision model was created with Data 4.0 software (TreeAge Software, Inc, Williamstown, Mass). Costs and quality-adjusted life years (QALYs) were evaluated for both the mother and her child because both individuals potentially are affected by antenatal
screening. As such, Markov models that used data that were specific to both the adult and pediatric populations were developed to determine the lifetime costs and health outcomes of HCV infection for each strategy in the respective populations. The results of the mother and child were then combined to yield the primary outcome, the marginal difference in total lifetime costs, and QALYs of the mother-child dyad. We further analyzed the data to determine the proportion of costs and QALYs that were attributed to the mother as compared with those attributed to the child and calculated a cost-effectiveness ratio for each using our base case estimates. Both costs and QALYs are discounted at a 3% annual rate in our base case analysis (range, 3%-5%). Discounting is a process that is used to convert future costs and QALYs into present value. A strategy was considered cost-effective at a ratio !$50,000/QALY, which is a threshold that is used commonly to define cost-effectiveness.13,17 The analysis is from the perspective of the health care system. With respect to the Markov analysis, we modeled the natural history of HCV for the screened and unscreened adult and pediatric populations. In our model, the initial health state for all chronically infected individuals (adults and neonates) is mild hepatitis, which was defined as a clinically asymptomatic and histologically mild disease. From this stage, individuals may enter remission or progress to histologically moderate hepatitis and onto more advanced disease states (Figure 2). The rate of progression from 1 health state to the next is determined by annual transition probabilities that are derived from the published literature (Table I21-39). Recent data demonstrate slower HCV-disease progression in women,19,24 and these rates were applied for maternal disease progression parameters. The model incorporates death from non–HCV-related causes for both the mother and her child.
Plunkett and Grobman
1155
Figure 2 Markov model of HCV disease progression. Straight arrows denote transition between disease states. Curved arrows denote death from hepatitis C-related or -unrelated causes.
We made simplifying assumptions about the characteristics of the populations that were involved in the model. All pregnant women with chronic HCV (screened and unscreened populations) entered the model at age 30 years. Perinatal HCV-transmission resulted in spontaneous remission or chronic HCV infection in the neonate. Because infected neonates show histologic evidence of mild hepatitis as infants21,38,40 and children tend to have a mild indolent course,40-45 we assumed that infected neonates (screened and unscreened populations) remain in the mild hepatitis health state for a latency period of 20 years,22,45 at which time all HCV-infected offspring enter the Markov model as adults at 20 years of age with mild hepatitis. Importantly, in our model, medical treatment for HCV infection occurs only after individuals reach the moderate hepatitis health state that was defined by histologically moderate disease,46 and the screened and unscreened populations differ only in their likelihood of receiving medical therapy after reaching that health state. In our base case, 70% of screened women and their children receive treatment upon entering into the moderate hepatitis health state.26 For those individuals in the model who are not screened, only a proportion of the cohort will be diagnosed and treated in a timely fashion.1,11 We estimated that 20% of unscreened patients would receive treatment after reaching the moderate hepatitis health state.23,25 With respect to medical therapy, we modeled treatment as a 48-week course of weekly pegylated interferon alfa-2b plus ribavarin.11 All individuals who have a sustained response enter into a long-term state of remission and do not relapse over the course of their lifetime.47-49 The non-responders remain in the moderate hepatitis health state and progress through the model according to the annual transition probabilities. We apply average published response rates and thus do not stratify them by HCV genotype. Additionally, retreatment does not occur after initial treatment failure. In our model, all women in the screening arm received pretest counseling, and only a portion of the
women will accept the test27,50 Women who accept testing are screened initially with a third-generation enzyme immunoassay test that is followed by a confirmatory HCV RNA polymerase chain reaction (PCR). We assumed that HCV RNA PCR is the gold standard for the diagnosis of chronic HCV infection and that HCV RNA positivity denotes infectivity.23,28,29 After completion of testing, posttest counseling is performed on all screened individuals. In the arm of the model that evaluates the impact of elective cesarean delivery, this intervention is offered to all women with positive HCV PCR. However, as demonstrated in previous studies,31 because of the unpredictability of the onset of labor or rupture of membranes, not all women ultimately undergo this intervention. Regarding perinatal HCV transmission, we assume that vertical transmission rates are dependent on the route of delivery. This assumption is based on a single retrospective study of 441 HCV-infected mothers in whom perinatal transmission was 0% (n = 31 women) with elective cesarean delivery, as compared with 7.7% with vaginal delivery and 5.9% with emergent cesarean delivery.20 Consequently, 0% transmission is our best estimate for perinatal HCV transmission with elective cesarean delivery, which is the value that was used in our base case analysis. Because the difference in transmission rates for emergent cesarean and vaginal delivery were not significant, we assumed equal risk of transmission and varied the rate of transmission across the range of reported values.20,32-39 All costs in the model are direct, derived from the published literature (Table II51-54), and presented in 2003 dollars. Costs in the literature from before 2003 were adjusted to 2003 dollars with the use of the medical care component of the Consumer Price Index.55 For neonatal testing, we assumed that infants who are born to women who screen positive for HCV PCR will receive 3 serial HCV PCR tests over the first 18 months of life, and the costs are calculated accordingly.23,29 We assumed a 48-week course of 800 mg of ribavirin in addition to weekly pegylated interferon at 1.5 mg/kg.11,54
1156 Table I
Plunkett and Grobman Probability variables
Variable Annual probability of disease progression Perinatal transmission to spontaneous clearance in first year of life Mild hepatitis to remission Mild hepatitis to moderate hepatitis Mother Child Moderate hepatitis to compensated cirrhosis Mother Child Compensated cirrhosis to decompensated cirrhosis Compensated cirrhosis to hepatocellular cancer Decompensated cirrhosis to liver transplantation Decompensated cirrhosis to hepatocellular cancer Decompensated cirrhosis to death Hepatocellular cancer to death Liver transplantation to death during initial year Liver transplantation to death during subsequent years Probability of receiving HCV treatment Unscreened population Screened population Probability of sustained response to treatment Maternal age at entry into Markov model Probability of accepting HCV screening test Prevalence of HCV infection Prevalence of chronic HCV disease Sensitivity 3rd-generation enzyme immunoassay Specificity 3rd-generation enzyme immunoassay Sensitivity PCR Specificity PCR Unknown or negative HCV status Probability of elective cesarean delivery Probability of emergent cesarean delivery Probability of vaginal delivery Screened positive for HCV Probability of elective cesarean delivery Probability of emergent cesarean delivery Probability of vaginal delivery Probability of perinatal transmission by route of delivery Elective cesarean delivery Emergent cesarean delivery Vaginal delivery
We also added the cost of office visits and laboratory services that were associated with this treatment.17 Finally, we assumed that there is no annual cost that is associated with mild hepatitis in the undiagnosed population. All utilities in our model are assigned a value from 0 to1 (0 is defined as no quality of life [death], and 1 is defined as full quality of life [perfect health]). Utility values by the mode of delivery have not been described previously in the literature. As such, we convened a panel of 5 experts to assign utility values by mode of delivery using the time trade-off technique. These values were averaged and assigned as tolls for a 6-week duration (the
Base case (%)
Range (%)
Reference
10 0.2
0-20.0 0.1-0.4
21, 22 12-18, 23
2.0 3.0
1.0-3.0 2.0-4.0
19, 24 19, 24
2.0 3.0 3.9 1.5 3.1 1.5 12.9 42.7 21 5.7
1.0-3.0 2.0-4.0 2.0-8.3 0.5-2.0 1.0-6.2 1.0-2.0 6.5-19.3 33.0-86.0 6.0-42.0 2.4-11
19,24 19, 24 16,19,24 12-18, 23 12-19, 23,24 12-19, 23,24 12-19, 23 12-19, 23, 24 12-18, 23 12-18, 23
0.2 .7 0.54 30 85 1 74 98.6 99.3 100 98
d 0.2-1.0 0.50-0.58 20-35 85.0-100 1.0-10.0 74-85 97.0-99.9 99.0-99.9 d 97.0-99.0
25 26 11 See text 27 1, 3, 8 1, 3 23, 28-30 23, 28-30 23, 28-30 23, 28-30
12.3 14.5 73.2
d d d
31 31 31
84.3 4.3 11.4
84.3-100 d 0-11.4
31 31 31
0-7.7 5.9-12 5.9-12
20 20, 32-39 20, 32-39
0 7.7 7.7
postpartum period). All other utilities are derived from previously published cost-effectiveness analyses (Table III15-18,23,56). Because mild hepatitis in this analysis, by definition, is asymptomatic, individuals with undiagnosed mild disease were considered to have a quality of life that was undiminished. However, once screening is performed, the disutility of knowledge of the disease and the potential for additional medical attention will reduce an individual’s quality of life. Consequently, individuals who were diagnosed with mild hepatitis had a decrement in their annual utility to 0.98. We validated our Markov model by running 10,000 simulation trials and compared our results with longitu-
Plunkett and Grobman Table II
Cost variables
Variable
2003 Base case ($)
Pretest counseling Posttest counseling for negative test result Posttest counseling for positive test result Enzyme immunoassay, 3rd generation PCR Genotype Delivery cost Elective cesarean delivery Emergent cesarean delivery Vaginal delivery Infant testing Annual cost Mild or moderate hepatitis Compensated cirrhosis Decompensated cirrhosis Hepatocellular cancer Remission Liver transplantation, initial year Liver transplantation, subsequent years Treatment
Table III
1157
34.50 48.60 121.40 47.80 127.70 150.70 6523 8155 3387 383 118.50 177.20 23914 17609 0 118483 23696 14138
Range ($)
Reference
14.70-34.50 8.00-52.00 23.30-121.40 28.40-67.70 99.50-156.00 d
27, 27, 27, 27, 27, 23
51 51 51 51 51
5326-7788 6524-9786 3387-5083 298-468
52, 52, 52, 23,
53 53 53 29
59-391 89-521 11957-38045 10870-29349 0-109 89134-329361 11957-31740 11310-16964
13, 13, 13, 13, 13, 13, 13, 11,
16 16 16 16 16 16 16 17, 54
Utility variables
Variable
Base case
Range
Reference
Remission Mild hepatitis (known disease) Mild hepatitis (not diagnosed) Moderate hepatitis Compensated cirrhosis Decompensated cirrhosis Hepatocellular cancer Liver transplantation, initial year Liver transplantation, subsequent years Treatment Vaginal delivery Elective cesarean delivery Emergent cesarean delivery
1 0.96 1 0.92 0.85 0.6 0.25 0.86 0.95 0.88 (0.0027) (0.0035) (0.0046)
d 0.96-1.0 d 0.82-0.98 0.5-0.90 0.5-0.88 0.1-0.5 0.6-0.9 0.8-0.95 0.82-0.91 (0.0037-0.0017) (0.0045-0.0025) (0.0056-0.0036)
15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23 15-18, 23 See text See text See text
Values in parentheses denote negative numbers (tolls).
dinal data on HCV infection from 5 prospective cohort studies57-61 and recently published empirically calibrated models for HCV.19,24 One-way sensitivity analyses were performed on all model variables, and multivariate analyses were performed on variables of interest.
Results In our 10,000 simulation trials, 18% of the women in the unscreened population experienced cirrhosis over a 15-year time period, 23% experienced cirrhosis over a 30-year time period, and 3% and 4% experienced hepatocellular carcinoma over 15 and 30 years, respec-
tively. These values compare favorably to empirically derived data that reveal that 16% to 29%57-61 and 0.7% to 1.3%58-61 of patients have cirrhosis and hepatocellular carcinoma, respectively in a 15-year period. Our findings also fall within the range of outcomes for HCV-infected women that demonstrated in an empirically calibrated gender-specific model of HCV infection.19,24 In our base case, HCV screening and the subsequent treatment of progressive disease result in an average total lifetime cost for mother and child of $4552 and an incremental cost of $108 relative to the current policy of no screening. The average total effectiveness of screening and treatment for mother and child is 54.48947 QALYs, with a decreased incremental effectiveness of 0.00011
1158
Plunkett and Grobman
Table IV Cost-effectiveness analysis under base case assumptions: Incremental cost-effectiveness ratio calculated with respect to the ‘‘No screen’’ strategy Variable
Cost ($)
Effectiveness (QALY)
Incremental cost-effectiveness ($/QALY)
No screen Screening with treatment Addition of cesarean delivery
4552 4660 4669
54.48958 54.48947 54.48968
(Dominated) 1,170,000
QALY, which indicates that no screening dominates (is a less costly and more effective strategy) the alternate strategy of screening and subsequent treatment for progressive disease (Table IV). With respect to the additional intervention of cesarean delivery, if elective cesarean delivery prevents all perinatal transmission (best case), 30 cases of perinatal HCV infection would be averted for every 100,000 women who are screened. However, despite this benefit, when compared with the current strategy of no screening, the marginal cost and effectiveness of screening, treatment, and cesarean delivery is $117 and 0.00010 QALY, respectively, which yields a cost-effectiveness ratio of $1,170,000/QALY (Table IV). Of the total costs for the current strategy of no screening, $4549 is attributed to the mother, and only $3.00 is attributed to the child. For the mother, the incremental cost of screening with subsequent treatment is $104, with a decrement in effectiveness of 0.0006 QALY. Thus, from the perspective of the mother alone, screening and subsequent treatment of progressive HCV disease is dominated by the current strategy of no screening. Similarly, for the neonate, the incremental cost of screening and subsequent treatment is $4.00, with a decrement in effectiveness of 0.0004. With respect to cesarean delivery, the addition of this intervention only adds to the maternal costs and decreases utility because of the disutility of cesarean delivery itself. However, for the child, prevention of HCV transmission with cesarean delivery adds only $0.54 and improves total effectiveness by 0.00018 QALY, for a cost-effectiveness ratio of $3019/QALY relative to the current strategy of no screening. One-way sensitivity analyses were performed across the full range of values for each variable. The model was found to be robust to all cost, probability, and utility variables and to the discount rate. The current standard of care continued to dominate the strategy of screening with subsequent treatment for progressive disease, even when parameters were adjusted to represent a high prevalence of HCV infection. Although the intervention of screening and treatment was no longer dominated by the current standard of care, it did not become a costeffective intervention even under conditions of improved treatment response rates with a cost-effective ratio of $2,140,000/QALY. Similarly, even when the disutility of knowledge of HCV infection was reduced so that the
utility for known mild hepatitis was set at the maximum of the range (0.99), screening and subsequent treatment was no longer dominated by the current standard; however, it did not become cost-effective at a ratio of $101,727/QALY. The strategy of cesarean delivery in addition to screening and subsequent treatment was not cost-effective under any of these circumstances, when compared with the current strategy of no screening. Even if perinatal transmission rates were high (12%) and elective cesarean deliveries were to eliminate all perinatal transmission, the strategy is not cost-effective relative to the current standard of care. In multivariate analysis, we examined varying both the perinatal HCV transmission rates and the HCV prevalence. When both variables were adjusted to be most favorable to the screening strategy (perinatal transmission at 12%, and HCV prevalence at 10%), neither the strategy of screening and subsequent treatment for progressive disease nor the addition of elective cesarean delivery was a cost-effective intervention. The former strategy continued to be dominated by the current policy of no screening, and the latter yielded a minimum cost-effectiveness ratio of $212,280/ QALY.
Comment Although 1 previous study demonstrated that screening the general population for HCV is not a cost-effective intervention,23 the cost-effectiveness of HCV screening during pregnancy, which is a unique circumstance in which lifelong health implications exist for 2 patients (the mother and her child) has not been explored. We believe that, because a policy of screening for HCV in pregnancy affects both the mother and her child, any decision should weigh equally the cost and utility for each individual. When we analyzed the maternal costs separately from the cost that was attributed to the child, it is clear that the benefit of screening, treatment, and cesarean delivery that may be gained for the child is outweighed by the costs and disutility that are incurred by the mother. When the sum of maternal and child lifetime costs and QALYs are considered, these interventions are not cost-effective interventions. Our analysis demonstrates that screening pregnant women for HCV and subsequent treatment for
Plunkett and Grobman progressive disease are not cost-effective interventions and are, in fact, dominated by the current practice of no screening. Of note, this intervention results in a net decrement in utility, which indicates that the disutility of knowledge of HCV infection for both the mother and her child is not outweighed by the potential benefit of treatment. When the disutility of knowledge of the disease is minimized and the utility value for known HCV is set at the highest end of the range (0.99), screening and subsequent treatment are no longer dominated by the current strategy but still do not nearly breach the cost-effectiveness threshold. Even under circumstances that we might consider to favor screening (such as when the prevalence of the disease is high or the efficacy of treatment is improved), the intervention does not become cost-effective. These results likely reflect the slow, indolent nature of HCV infection, particularly in young women, in combination with the additional annual cost of known HCV infection and the absence of useful intervention for many years. With the additional intervention of elective cesarean delivery to avert perinatal transmission, the benefit of the prevention of neonatal HCV infection results in a net gain in the utility value for the mother and her child. As such, the combined interventions of screening, subsequent treatment, and elective cesarean delivery are not dominated by the current practice of no screening. However, these interventions still do not approach our cost-effectiveness threshold. In fact, even under circumstances that are most favorable to screening, when HCV infection is highly prevalent, perinatal transmission is high, and elective cesarean delivery eliminates all perinatal transmission, these interventions are not costeffective. Of note, our cost-effectiveness threshold of $50,000/ QALY represents a commonly used threshold in research studies. Arguably, some health interventions that currently are used in the United States already exceed this threshold and prompt the question as to what threshold may better reflect the societal priorities that are demonstrated by our health care system. Although the answer to this query may be relevant to the interpretation of our results, this task is well beyond the scope of this study. As such, we chose a costeffectiveness threshold that was consistent with the current body of literature and presented the specific cost-effectiveness ratios for important sensitivity analyses so that readers may interpret the results across a range of cost-effectiveness thresholds. This analysis has several strengths. First, our model was robust to all cost, probability, and utility variables. Each variable was tested across its full range of values in 1-way sensitivity analyses; variables of interest were assessed in multivariate analyses, and our conclusions were not altered under any of these circumstances. Second, we validated our Markov model for HCV
1159 progression in the adult population, and our results fell within the range of outcomes that were expected in empirically calibrated models of HCV disease progression19,24 and compared favorably with prospective cohort analyses.57-61 Finally, our model used a life-long timeline in the analysis of the costs and benefits of screening for HCV infection in pregnancy. We did not limit our analysis to immediate outcomes (ie, prevention of perinatal HCV transmission) but considered the implications of screening over the lifetime of both the mother and her child, the perspective required to make rational health policy decisions. We also demonstrate that a large proportion of the costs are derived from the mother and that the decrement in her QALYs outweighs the small benefit for the child. One limitation of this analysis is that the natural history of HCV infection in the pediatric population is not well described.41-44 To date, cohort studies of HCVinfected infants and children do not include long-term outcomes beyond early adulthood.62 In constructing the Markov analysis, we made conservative assumptions with regard to the rate of HCV disease progression. Although cases of rapidly progressive HCV disease in the pediatric population have been reported, we nevertheless modeled the more typical course of HCV disease. As such, we included a 20-year latency period in which no disease progression occurred in chronically HCVinfected children. As a result, our model of perinatal infection depicts a relatively indolent, slowly progressive disease course. Nevertheless, in sensitivity analyses, in which more aggressive disease transitions were modeled, our conclusions did not change. Similarly, we are limited by the available utility values for HCV-related health states. The values that were used in this study were derived from the published literature and originated from a panel of experts who used standard techniques to derive the values. Expert opinion may not reflect the utility values that would be assigned by a patient or the general population. Furthermore, these utility values may change over the course of a lifetime and may be impacted by HCV transmission to a child or family member. However, we recognized these limitations and tested each utility variable over a wide range of values in our sensitivity analyses, and our conclusions did not change. Another limitation to this analysis is that it is performed from the perspective of the health care system. Thus, it does not take into account indirect costs (such as costs that are attributed to hours lost from work) that are associated with each specific health state. As such, the analysis may underestimate the costs of poor health; however, these costs likely represent only a small portion of total lifetime costs for a minority of the population and should not alter our findings. HCV infection remains a significant public health issue that affects both mothers and their infants alike.
1160 Unfortunately, our ability to treat or cure the disease is relatively limited. Consequently, as our model demonstrates, a screening strategy is not a cost-effective intervention even in the unique circumstances of pregnancy, when 2 individuals potentially could access the benefits of treatment and 1 individual could even avoid the disease altogether. In the future, new treatments may offer the possibility that screening would become a costeffective intervention, but at present, the current standard of care (no screening) is the most cost-effective option.
References 1. Centers for Disease Control and Prevention. Recommendations for prevention and control of hepatitis C virus (HCV) infection and HCV-related disease. MMWR Morb Mortal Wkly Rep 1998;47:1-39. 2. Alter HJ, Seef LB. Recovery, persistence, and sequelae in hepatitis C virus infection. Semin Liver Dis 2000;20:17-35. 3. Alter MJ, Kruszon-Moran D, Nainan OV, McQuillan GM, Gao F, Moyer LA, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. N Engl J Med 1999; 341:556-62. 4. Seeff LB, Buskell-Bales Z, Wright EC, Durako SJ, Alter MJ, Iber FL, et al. Long-term mortality after transfusion-associated non-A, non-B hepatitis: the National Heart, Lung, and Blood Institute Study Group. N Engl J Med 1992;327:1906-11. 5. Alter MJ, Margolis HS, Krawczynski K, Judson FN, Mares A, Alexander WJ, et al. The natural history of community-acquired hepatitis C in the United States: the Sentinel Counties Chronic non-A, non-B Hepatitis Study Team. N Engl J Med 1992;327:1899-905. 6. Wasley AD, Alter MJ. Epidemiology of hepatitis C. Semin Liver Dis 2000;20:1-16. 7. Alter MJ. Epidemiology of hepatitis C in the west. Semin Liver Dis 1995;15:5-14. 8. Silverman NS, Jenkin BK, Wu C, McGillen P, Knee G. Hepatitis C virus in pregnancy: Seroprevalence and risk factors for infection. Am J Obstet Gynecol 1993;169:583-7. 9. National Institutes of Health Consensus Development Conference statement: management of hepatitis C: 2002: June 10-12, 2002. Hepatology 2002;36(suppl):S3-20. 10. American College of Obstetricians and Gynecologists. Viral hepatitis in pregnancy: ACOG educational bulletin, no. 248. Intl J Gynaecol Obstet 1998;63:195-202. 11. Manns M, McHutchison JG, Gordon SC, Rustgi VK, Shiffman M, Reindollar R, et al. Peginterferon alfa-2b plus ribavirin compared with interferon alfa-2b plus ribavirin for initial treatment of chronic hepatitis C. Lancet 2001;358:958-65. 12. Kim WR, Poterucha JJ, Hermans JE, Therneau TM, Dickson ER, Evans RW, et al. Cost-effectiveness of 6 and 12 months of interferon-alpha therapy for chronic hepatitis C. Ann Int Med 1997;127:866-74. 13. Bennett WG, Inoue Y, Beck JR, Wong JB, Pouker SG, Davis GL. Estimates of the cost-effectiveness of a single course of interferonalpha 2b in patients with histologically mild chronic hepatitis. Ann Int Med 1997;127:855-65. 14. Davis GL, Beck JR, Farrell G, Poynard T. Prolonged treatment with interferon in patients with histologically mild chronic hepatitis C. J Viral Hepat 1998;5:313-21.
Plunkett and Grobman 15. Buti M, Casado MA, Fosbrook L, Wong JB, Esteban R. Costeffectiveness of combination therapy for naı¨ ve patients with chronic hepatitis C. J Hepatol 2000;33:651-8. 16. Wong JB, Koff RS. Watchful waiting with periodic liver biopsy versus immediate empirical therapy for histologically mild chronic hepatitis C. Ann Int Med 2000;133:665-75. 17. Wong JB, Poynard T, Ling M-H, Albrecht JK, Pauker SG. Costeffectiveness of 24 or 48 weeks of interferon alpha-2b alone or with ribavirin as initial treatment of chronic hepatitis C. Am J Gastroenterol 2000;95:1524-30. 18. Younossi ZM, Singer ME, McHutchison JG, Shermock CK. Cost effectiveness of interferon alfa-2b combined with ribavirin for the treatment of chronic hepatitis C. Hepatology 1999;30:1318-24. 19. Solomon JA, Weinstein MC, Hammitt JK, Goldie SJ. Empirically calibrated model of hepatitis C virus infection in the United States. Am J Epidemiol 2002;156:761-3. 20. Gibb DM, Goodall RL, Dunn DT, Healy M, Neave P, Cafferkey M, et al. Mother-to-child transmission of hepatitis C virus. Lancet 2000;356:904-7. 21. Resti M, Jara P, Hierro L, Azzari C, Giacchino R, Zuin G, et al. Clinical features and progression of perinatally acquired hepatitis C virus infection. J Med Virol 2003;70:373-7. 22. Sinha M, Das A. Cost effectiveness analysis of different strategies of management of chronic hepatitis C infection in children. Pediatr Infect Dis J 2000;19:23-30. 23. Singer ME, Younossi ZM. Cost effectiveness of screening for hepatitis C virus in asymptomatic, average-risk adults. Am J Med 2001;111:614-21. 24. Solomon JA, Weinstein MC, Hammitt JK, Goldie SJ. Costeffectiveness of treatment for chronic hepatitis C infection in an evolving patient population. JAMA 2003;290:228-37. 25. Piton A, Poynard T, Imbert-Bismut F, Khalil L, Delattre J, Pelissier E, et al. Factors associated with serum alanine transaminase activity in healthy subjects. Hepatology 1998;27:1213-9. 26. McHutchinson JG, Gordon SC, Schiff ER, Shiffman ML, Lee WM, Rustgi VK, et al. Interferon alfa-2b alone or in combination with ribavirin as initial treatment for chronic hepatitis C. N Engl J Med 1998;339:1485-92. 27. Grobman WA, Garcia PM. The cost-effectiveness of voluntary intrapartum rapid human immunodeficiency virus testing for women without adequate prenatal care. Am J Obstet Gynecol 1999;181:1062-71. 28. Cheney CP, Chopra S, Graham C. Infections of the liver. Infect Dis Clin North Am 2000;14:633-7. 29. Carithers RL, Marquardt A, Gretch DR. Diagnostic testing for hepatitis C. Semin Liver Dis 2000;20:159-71. 30. Pawlotsky JM, Lonjon I, Hezode C, Bruno R, Darthuy F, Remire J, et al. What strategy should be used for diagnosis of hepatitis C virus infection in clinical laboratories? Hepatology 1998;27:1700-2. 31. The European Mode of Delivery Collaboration. Elective caesarean-section versus vaginal delivery in prevention of vertical HIV-1 transmission. Lancet 1999;353:1035-9. 32. La Torre A, Biadaioli R, Capobianco T, Coloa MG, Monti M, Pulli F, et al. Vertical transmission of HCV. Acta Obstet Gynecol Scand 1998;77:889-92. 33. Spencer JD, Latt N, Beeby PJ, Collins E, Saunders JB, McCaughan GW, et al. Transmission of hepatitis C virus to infants of human immunodeficiency virus-negative intravenous drug-using mothers: rate of infection and assessment of risk factors for transmission. J Viral Hepat 1997;4:395-409. 34. Ohto H, Terazawa S, Sasaki N, Sasaki N, Hino K, Ishiwata C, et al. Transmission of hepatitis C virus from mothers to infants. N Engl J Med 1994;330:744-50. 35. Giacchino R, Picciotto A, Tasso L, Timitilli A, Sinelli N. Vertical transmission of hepatitis C. Lancet 1995;345:1122-3. 36. Okamoto M, Nagata I, Murakami J, Shunsaku K, Iitsuka T, Hoshika T, et al. Prospective reevaluation of risk factors in
Plunkett and Grobman
37.
38.
39.
40.
41.
42. 43. 44. 45. 46.
47.
48.
49.
mother-to-child transmission of hepatitis C virus: high virus load, vaginal delivery, and negative anti-NS4 antibody. J Infect Dis 2000;182:1511-4. Conte D, Fraquelli M, Prati D, Colucci A, Minola E. Prevalence and clinical course of chronic hepatitis C virus (HCV) infection and rate of HCV vertical transmission in a cohort of 15,250 pregnant women. Hepatology 2000;31:751-5. Molin GD, D’Agaro P, Ansaldi F, Ciana G, Fertz C, Alberica S, et al. Mother-to-infant transmission of hepatitis C virus: rate of infection and assessment of viral load and IgM anti-HCV as risk factors. J Med Virol 2002;67:137-42. Pembrey LJ, Tovo PA, Newell ML. Effects of mode of delivery and infant feeding on the risk of mother-to-child transmission of hepatitis C virus. BJOG 2001;108:371-7. Bortolotti F, Resti M, Giacchino R, Azzari C, Gussetti N, Crivellaro C, et al. Hepatitis C virus infection and related liver disease in children of mothers with antibodies to the virus. J Ped 1997;130:990-3. Garcia-Monzon C, Jara P, Fernandez-Bermejo M, Hierro L, Frauca E, Camarena C, et al. Chronic hepatitis C in children: a clinical and immunohistochemical comparative study with adult patients. Hepatology 1998;28:1696-701. Birnbaum AH, Shneider BL, Moy L. Hepatitis C in children. N Engl J Med 2000;342:290-2. Ruiz-Moreno M, Leal-Orozco A, Millan A. Hepatitis C virus infection in children. J Hepatol 1999;31(suppl):124-9. Aach RD, Yomtovian RA, Hack M. Neonatal and pediatric posttransfusion hepatitis C. Pediatr 2000;105:836-42. Jonas MM. Treatment of chronic hepatitis C in pediatric patients. Clin Liver Dis 1999;3:855-67. Desmet V, Gerber M, Hoofnagle JH, Manns M, Scheuer PJ. Classification of chronic hepatitis: diagnosis, grading and staging. Hepatology 1994;19:1513-20. Lau DT-Y, Kleiner DE, Ghany MG, Park Y, Schmid P, Hoofnagle JH. 10-Year follow-up after interferon-alpha therapy for chronic hepatitis C. Hepatology 1998;28:1121-7. Shindo M, Hamada K, Oda Y, Okuno T. Long-term follow-up study of sustained biochemical responders with interferon therapy. Hepatology 2001;33:1299-302. Marcellin P, Boyer N, Gervais A, Martinot M, Pouteau M, Castelnau C, et al. Long-term histologic improvement and loss of detectable intrahepatic HCV RNA in patients with chronic
1161
50.
51.
52.
53.
54. 55. 56. 57.
58.
59.
60.
61.
62.
hepatitis C and sustained response to interferon-alpha therapy. Ann Int Med 1997;127:875-81. Lindsay M, Feng TI, Peterson HB, Slade BA, Willis S, Klein L. Routine human immunodeficiency virus infection screening in unregistered and registered inner-city parturients. Obstet Gynecol 1991;77:599-603. Mauskopf JA, Paul JE, Wichman DS, White AD, Tilson HH. Economic impact of treatment of HIV-positive pregnant women and their newborns with zidovudine. JAMA 1996;276:132-8. Traynor JD, Peaceman AM. Maternal hospital charges associated with trial of labor versus elective repeat cesarean section. Birth 1998;25:81-4. Rouse DJ, Owen J, Goldenberg RI, Cliver SP. The effectiveness and costs of elective cesarean delivery for fetal macrosomia diagnosed by ultrasound. JAMA 1996;276:1480-6. Cohen HE, editor. Drug topics red book. Montvale (NJ): Medical Economics; 2002. p. 473 (pegylated interferon alfa-2b), p. 521 (ribavirin). US Department of Labor, Bureau of Labor Statistics. Available at www.bls.gov/cpi/. Wong JB, Bennett WG, Koff RS, Pauker SG. Pretreatment evaluation of chronic hepatitis C. JAMA 1998;280:2088-93. Di Bisceglie AM, Goodman ZD, Ishak KG, Hoofnagle JM, Melpolder JJ, Alter MJ. Long-term clinical and histopathological follow-up of chronic posttransfusion hepatitis. Hepatology 1991;14:969-74. Hopf U, Moller B, Kuther D, Stemerowicz R, Lobeck H, LudtkeHandjery A, et al. Long-term follow-up of posttransfusion and sporadic chronic hepatitis non-A, non-B and frequency of circulating antibodies to hepatitis C virus (HCV). J Hepatol 1990;10:69-76. Koretz RL, Abbey H, Coleman E, Gitnick G. Non-A, non-B posttransfusion hepatitis: looking back in the second decade. Ann Int Med 1993;119:110-5. Mattson L, Sonnerborg A, Weiland O. Outcome of acute symptomatic non-A, non-B hepatitis: a 13-year follow-up study of hepatitis C virus markers. Liver 1993;13:274-8. Tremolada F, Casarin C, Alberti A, Drago C, Tagger A, Ribero ML, et al. Long-term follow-up from non-A, non-B (type C) posttransfusion hepatitis. J Hepatol 1992;16:273-81. Vogt M, Lang T, Frosner G, Klingler C, Sendl A, Zeller A, et al. Prevalence and clinical outcome of hepatitis C infection in children who underwent cardiac surgery before the implementation of blood-donor screening. N Eng J Med 1999;341:866-70.
www.ajog.org
OBSTETRIC INFECTIOUS DISEASE
Routine hepatitis C virus screening in pregnancy: A cost-effectiveness analysis Beth A. Plunkett, MD, MPH, William A. Grobman, MD, MBA Department of Obstetrics and Gynecology, Northwestern University Feinberg School of Medicine, Chicago, Ill Received for publication June 9, 2004; revised September 22, 2004; accepted October 14, 2004
KEY WORDS Hepatitis C Screening Pregnancy Cost-effective
Objective: The purpose of this study was to determine whether routine hepatitis C virus screening in pregnancy is cost-effective. Study design: A decision tree with Markov analysis was developed to compare 3 approaches to asymptomatic hepatitis C virus infection in low-risk pregnant women: (1) no hepatitis C virus screening, (2) hepatitis C virus screening and subsequent treatment for progressive disease, and (3) hepatitis C virus screening, subsequent treatment for progressive disease, and elective cesarean delivery to avert perinatal transmission. Lifetime costs and quality-adjusted life years were evaluated for mother and child. Results: In our base case, hepatitis C virus screening and subsequent treatment of progressive disease was dominated (more costly and less effective) by no screening, with an incremental cost of $108 and a decreased incremental effectiveness of 0.00011 quality-adjusted life years. When compared with no screening, the marginal cost and effectiveness of screening, treatment, and cesarean delivery was $117 and 0.00010 quality-adjusted life years, respectively, which yields a cost-effectiveness ratio of $1,170,000 per quality-adjusted life year. Conclusion: The screening of asymptomatic pregnant women for hepatitis C virus infection is not cost-effective. Ó 2005 Elsevier Inc. All rights reserved.
In the United States, hepatitis C virus (HCV) has become a major public health concern. Nationally, 4 million individuals, most of whom have chronic disease, have been infected with HCV.1-3 Given the high prevalence of HCV infection and the increased risk of the development of cirrhosis or hepatocellular carcinoma in individuals with chronic disease,2,4,5 HCV has Supported by the Institute for Health Services Research and Policy Studies at Northwestern University and National Research Service Award 5 T32HS00078 from the Agency for Healthcare Research and Quality. Reprints not available from the authors. 0002-9378/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ajog.2004.10.600
become the leading indication for liver transplantation in the adult population.2 The greatest prevalence of infection occurs among individuals of reproductive age,1,6 and 1% to 4.3% of pregnant women are infected with HCV.1,7,8 Nevertheless, screening is not recommended currently during pregnancy for asymptomatic women without risk factors for HCV infection,9,10 and only 25% to 30% of women with HCV are aware of their diagnosis.9 Consequently, interventions that potentially could improve both maternal and neonatal long-term health outcomes are not available to most HCV-infected women and their children. These interventions may
1154
Plunkett and Grobman
Figure 1 Simplified decision tree. The encircled M refers to the different Markov models that are specific to disease status (neither mother nor neonate infected, maternal infection without perinatal transmission, maternal infection with perinatal transmission) and knowledge of disease status for both mother and child.
include measures of secondary prevention (early medical treatment for progressive disease for mother and child after the completion of the pregnancy11) and measures of primary prevention (elective cesarean delivery to prevent perinatal HCV infection). Although the benefits11 and cost-effectiveness of medical treatment for HCV12-19 have been analyzed in the nonobstetric population, information regarding cost-effective interventions and, in particular, the benefit of elective cesarean delivery and reduction in perinatal HCV transmission is considerably limited.20 As such, the purpose of this analysis was to determine under what circumstances, if any, routine HCV screening in pregnancy and elective cesarean delivery to avert perinatal transmission would be a cost-effective intervention.
Material and methods A decision tree model that used Markov analysis was developed to compare 3 approaches to HCV infection in pregnancy: (1) no HCV screening in pregnancy (the current standard of care), (2) HCV screening in pregnancy and subsequent treatment for progressive disease, and (3) HCV screening in pregnancy, subsequent treatment for progressive disease, and elective cesarean delivery to avert perinatal transmission (Figure 1). The screened population included all asymptomatic, HIVnegative pregnant women without risk factors for HCV infection who received routine prenatal care in the United States. The time course considered was the lifetime of the mother and her child. The analytic decision model was created with Data 4.0 software (TreeAge Software, Inc, Williamstown, Mass). Costs and quality-adjusted life years (QALYs) were evaluated for both the mother and her child because both individuals potentially are affected by antenatal
screening. As such, Markov models that used data that were specific to both the adult and pediatric populations were developed to determine the lifetime costs and health outcomes of HCV infection for each strategy in the respective populations. The results of the mother and child were then combined to yield the primary outcome, the marginal difference in total lifetime costs, and QALYs of the mother-child dyad. We further analyzed the data to determine the proportion of costs and QALYs that were attributed to the mother as compared with those attributed to the child and calculated a cost-effectiveness ratio for each using our base case estimates. Both costs and QALYs are discounted at a 3% annual rate in our base case analysis (range, 3%-5%). Discounting is a process that is used to convert future costs and QALYs into present value. A strategy was considered cost-effective at a ratio !$50,000/QALY, which is a threshold that is used commonly to define cost-effectiveness.13,17 The analysis is from the perspective of the health care system. With respect to the Markov analysis, we modeled the natural history of HCV for the screened and unscreened adult and pediatric populations. In our model, the initial health state for all chronically infected individuals (adults and neonates) is mild hepatitis, which was defined as a clinically asymptomatic and histologically mild disease. From this stage, individuals may enter remission or progress to histologically moderate hepatitis and onto more advanced disease states (Figure 2). The rate of progression from 1 health state to the next is determined by annual transition probabilities that are derived from the published literature (Table I21-39). Recent data demonstrate slower HCV-disease progression in women,19,24 and these rates were applied for maternal disease progression parameters. The model incorporates death from non–HCV-related causes for both the mother and her child.
Plunkett and Grobman
1155
Figure 2 Markov model of HCV disease progression. Straight arrows denote transition between disease states. Curved arrows denote death from hepatitis C-related or -unrelated causes.
We made simplifying assumptions about the characteristics of the populations that were involved in the model. All pregnant women with chronic HCV (screened and unscreened populations) entered the model at age 30 years. Perinatal HCV-transmission resulted in spontaneous remission or chronic HCV infection in the neonate. Because infected neonates show histologic evidence of mild hepatitis as infants21,38,40 and children tend to have a mild indolent course,40-45 we assumed that infected neonates (screened and unscreened populations) remain in the mild hepatitis health state for a latency period of 20 years,22,45 at which time all HCV-infected offspring enter the Markov model as adults at 20 years of age with mild hepatitis. Importantly, in our model, medical treatment for HCV infection occurs only after individuals reach the moderate hepatitis health state that was defined by histologically moderate disease,46 and the screened and unscreened populations differ only in their likelihood of receiving medical therapy after reaching that health state. In our base case, 70% of screened women and their children receive treatment upon entering into the moderate hepatitis health state.26 For those individuals in the model who are not screened, only a proportion of the cohort will be diagnosed and treated in a timely fashion.1,11 We estimated that 20% of unscreened patients would receive treatment after reaching the moderate hepatitis health state.23,25 With respect to medical therapy, we modeled treatment as a 48-week course of weekly pegylated interferon alfa-2b plus ribavarin.11 All individuals who have a sustained response enter into a long-term state of remission and do not relapse over the course of their lifetime.47-49 The non-responders remain in the moderate hepatitis health state and progress through the model according to the annual transition probabilities. We apply average published response rates and thus do not stratify them by HCV genotype. Additionally, retreatment does not occur after initial treatment failure. In our model, all women in the screening arm received pretest counseling, and only a portion of the
women will accept the test27,50 Women who accept testing are screened initially with a third-generation enzyme immunoassay test that is followed by a confirmatory HCV RNA polymerase chain reaction (PCR). We assumed that HCV RNA PCR is the gold standard for the diagnosis of chronic HCV infection and that HCV RNA positivity denotes infectivity.23,28,29 After completion of testing, posttest counseling is performed on all screened individuals. In the arm of the model that evaluates the impact of elective cesarean delivery, this intervention is offered to all women with positive HCV PCR. However, as demonstrated in previous studies,31 because of the unpredictability of the onset of labor or rupture of membranes, not all women ultimately undergo this intervention. Regarding perinatal HCV transmission, we assume that vertical transmission rates are dependent on the route of delivery. This assumption is based on a single retrospective study of 441 HCV-infected mothers in whom perinatal transmission was 0% (n = 31 women) with elective cesarean delivery, as compared with 7.7% with vaginal delivery and 5.9% with emergent cesarean delivery.20 Consequently, 0% transmission is our best estimate for perinatal HCV transmission with elective cesarean delivery, which is the value that was used in our base case analysis. Because the difference in transmission rates for emergent cesarean and vaginal delivery were not significant, we assumed equal risk of transmission and varied the rate of transmission across the range of reported values.20,32-39 All costs in the model are direct, derived from the published literature (Table II51-54), and presented in 2003 dollars. Costs in the literature from before 2003 were adjusted to 2003 dollars with the use of the medical care component of the Consumer Price Index.55 For neonatal testing, we assumed that infants who are born to women who screen positive for HCV PCR will receive 3 serial HCV PCR tests over the first 18 months of life, and the costs are calculated accordingly.23,29 We assumed a 48-week course of 800 mg of ribavirin in addition to weekly pegylated interferon at 1.5 mg/kg.11,54
1156 Table I
Plunkett and Grobman Probability variables
Variable Annual probability of disease progression Perinatal transmission to spontaneous clearance in first year of life Mild hepatitis to remission Mild hepatitis to moderate hepatitis Mother Child Moderate hepatitis to compensated cirrhosis Mother Child Compensated cirrhosis to decompensated cirrhosis Compensated cirrhosis to hepatocellular cancer Decompensated cirrhosis to liver transplantation Decompensated cirrhosis to hepatocellular cancer Decompensated cirrhosis to death Hepatocellular cancer to death Liver transplantation to death during initial year Liver transplantation to death during subsequent years Probability of receiving HCV treatment Unscreened population Screened population Probability of sustained response to treatment Maternal age at entry into Markov model Probability of accepting HCV screening test Prevalence of HCV infection Prevalence of chronic HCV disease Sensitivity 3rd-generation enzyme immunoassay Specificity 3rd-generation enzyme immunoassay Sensitivity PCR Specificity PCR Unknown or negative HCV status Probability of elective cesarean delivery Probability of emergent cesarean delivery Probability of vaginal delivery Screened positive for HCV Probability of elective cesarean delivery Probability of emergent cesarean delivery Probability of vaginal delivery Probability of perinatal transmission by route of delivery Elective cesarean delivery Emergent cesarean delivery Vaginal delivery
We also added the cost of office visits and laboratory services that were associated with this treatment.17 Finally, we assumed that there is no annual cost that is associated with mild hepatitis in the undiagnosed population. All utilities in our model are assigned a value from 0 to1 (0 is defined as no quality of life [death], and 1 is defined as full quality of life [perfect health]). Utility values by the mode of delivery have not been described previously in the literature. As such, we convened a panel of 5 experts to assign utility values by mode of delivery using the time trade-off technique. These values were averaged and assigned as tolls for a 6-week duration (the
Base case (%)
Range (%)
Reference
10 0.2
0-20.0 0.1-0.4
21, 22 12-18, 23
2.0 3.0
1.0-3.0 2.0-4.0
19, 24 19, 24
2.0 3.0 3.9 1.5 3.1 1.5 12.9 42.7 21 5.7
1.0-3.0 2.0-4.0 2.0-8.3 0.5-2.0 1.0-6.2 1.0-2.0 6.5-19.3 33.0-86.0 6.0-42.0 2.4-11
19,24 19, 24 16,19,24 12-18, 23 12-19, 23,24 12-19, 23,24 12-19, 23 12-19, 23, 24 12-18, 23 12-18, 23
0.2 .7 0.54 30 85 1 74 98.6 99.3 100 98
d 0.2-1.0 0.50-0.58 20-35 85.0-100 1.0-10.0 74-85 97.0-99.9 99.0-99.9 d 97.0-99.0
25 26 11 See text 27 1, 3, 8 1, 3 23, 28-30 23, 28-30 23, 28-30 23, 28-30
12.3 14.5 73.2
d d d
31 31 31
84.3 4.3 11.4
84.3-100 d 0-11.4
31 31 31
0-7.7 5.9-12 5.9-12
20 20, 32-39 20, 32-39
0 7.7 7.7
postpartum period). All other utilities are derived from previously published cost-effectiveness analyses (Table III15-18,23,56). Because mild hepatitis in this analysis, by definition, is asymptomatic, individuals with undiagnosed mild disease were considered to have a quality of life that was undiminished. However, once screening is performed, the disutility of knowledge of the disease and the potential for additional medical attention will reduce an individual’s quality of life. Consequently, individuals who were diagnosed with mild hepatitis had a decrement in their annual utility to 0.98. We validated our Markov model by running 10,000 simulation trials and compared our results with longitu-
Plunkett and Grobman Table II
Cost variables
Variable
2003 Base case ($)
Pretest counseling Posttest counseling for negative test result Posttest counseling for positive test result Enzyme immunoassay, 3rd generation PCR Genotype Delivery cost Elective cesarean delivery Emergent cesarean delivery Vaginal delivery Infant testing Annual cost Mild or moderate hepatitis Compensated cirrhosis Decompensated cirrhosis Hepatocellular cancer Remission Liver transplantation, initial year Liver transplantation, subsequent years Treatment
Table III
1157
34.50 48.60 121.40 47.80 127.70 150.70 6523 8155 3387 383 118.50 177.20 23914 17609 0 118483 23696 14138
Range ($)
Reference
14.70-34.50 8.00-52.00 23.30-121.40 28.40-67.70 99.50-156.00 d
27, 27, 27, 27, 27, 23
51 51 51 51 51
5326-7788 6524-9786 3387-5083 298-468
52, 52, 52, 23,
53 53 53 29
59-391 89-521 11957-38045 10870-29349 0-109 89134-329361 11957-31740 11310-16964
13, 13, 13, 13, 13, 13, 13, 11,
16 16 16 16 16 16 16 17, 54
Utility variables
Variable
Base case
Range
Reference
Remission Mild hepatitis (known disease) Mild hepatitis (not diagnosed) Moderate hepatitis Compensated cirrhosis Decompensated cirrhosis Hepatocellular cancer Liver transplantation, initial year Liver transplantation, subsequent years Treatment Vaginal delivery Elective cesarean delivery Emergent cesarean delivery
1 0.96 1 0.92 0.85 0.6 0.25 0.86 0.95 0.88 (0.0027) (0.0035) (0.0046)
d 0.96-1.0 d 0.82-0.98 0.5-0.90 0.5-0.88 0.1-0.5 0.6-0.9 0.8-0.95 0.82-0.91 (0.0037-0.0017) (0.0045-0.0025) (0.0056-0.0036)
15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23 15-18, 23, 56 15-18, 23 15-18, 23 15-18, 23 See text See text See text
Values in parentheses denote negative numbers (tolls).
dinal data on HCV infection from 5 prospective cohort studies57-61 and recently published empirically calibrated models for HCV.19,24 One-way sensitivity analyses were performed on all model variables, and multivariate analyses were performed on variables of interest.
Results In our 10,000 simulation trials, 18% of the women in the unscreened population experienced cirrhosis over a 15-year time period, 23% experienced cirrhosis over a 30-year time period, and 3% and 4% experienced hepatocellular carcinoma over 15 and 30 years, respec-
tively. These values compare favorably to empirically derived data that reveal that 16% to 29%57-61 and 0.7% to 1.3%58-61 of patients have cirrhosis and hepatocellular carcinoma, respectively in a 15-year period. Our findings also fall within the range of outcomes for HCV-infected women that demonstrated in an empirically calibrated gender-specific model of HCV infection.19,24 In our base case, HCV screening and the subsequent treatment of progressive disease result in an average total lifetime cost for mother and child of $4552 and an incremental cost of $108 relative to the current policy of no screening. The average total effectiveness of screening and treatment for mother and child is 54.48947 QALYs, with a decreased incremental effectiveness of 0.00011
1158
Plunkett and Grobman
Table IV Cost-effectiveness analysis under base case assumptions: Incremental cost-effectiveness ratio calculated with respect to the ‘‘No screen’’ strategy Variable
Cost ($)
Effectiveness (QALY)
Incremental cost-effectiveness ($/QALY)
No screen Screening with treatment Addition of cesarean delivery
4552 4660 4669
54.48958 54.48947 54.48968
(Dominated) 1,170,000
QALY, which indicates that no screening dominates (is a less costly and more effective strategy) the alternate strategy of screening and subsequent treatment for progressive disease (Table IV). With respect to the additional intervention of cesarean delivery, if elective cesarean delivery prevents all perinatal transmission (best case), 30 cases of perinatal HCV infection would be averted for every 100,000 women who are screened. However, despite this benefit, when compared with the current strategy of no screening, the marginal cost and effectiveness of screening, treatment, and cesarean delivery is $117 and 0.00010 QALY, respectively, which yields a cost-effectiveness ratio of $1,170,000/QALY (Table IV). Of the total costs for the current strategy of no screening, $4549 is attributed to the mother, and only $3.00 is attributed to the child. For the mother, the incremental cost of screening with subsequent treatment is $104, with a decrement in effectiveness of 0.0006 QALY. Thus, from the perspective of the mother alone, screening and subsequent treatment of progressive HCV disease is dominated by the current strategy of no screening. Similarly, for the neonate, the incremental cost of screening and subsequent treatment is $4.00, with a decrement in effectiveness of 0.0004. With respect to cesarean delivery, the addition of this intervention only adds to the maternal costs and decreases utility because of the disutility of cesarean delivery itself. However, for the child, prevention of HCV transmission with cesarean delivery adds only $0.54 and improves total effectiveness by 0.00018 QALY, for a cost-effectiveness ratio of $3019/QALY relative to the current strategy of no screening. One-way sensitivity analyses were performed across the full range of values for each variable. The model was found to be robust to all cost, probability, and utility variables and to the discount rate. The current standard of care continued to dominate the strategy of screening with subsequent treatment for progressive disease, even when parameters were adjusted to represent a high prevalence of HCV infection. Although the intervention of screening and treatment was no longer dominated by the current standard of care, it did not become a costeffective intervention even under conditions of improved treatment response rates with a cost-effective ratio of $2,140,000/QALY. Similarly, even when the disutility of knowledge of HCV infection was reduced so that the
utility for known mild hepatitis was set at the maximum of the range (0.99), screening and subsequent treatment was no longer dominated by the current standard; however, it did not become cost-effective at a ratio of $101,727/QALY. The strategy of cesarean delivery in addition to screening and subsequent treatment was not cost-effective under any of these circumstances, when compared with the current strategy of no screening. Even if perinatal transmission rates were high (12%) and elective cesarean deliveries were to eliminate all perinatal transmission, the strategy is not cost-effective relative to the current standard of care. In multivariate analysis, we examined varying both the perinatal HCV transmission rates and the HCV prevalence. When both variables were adjusted to be most favorable to the screening strategy (perinatal transmission at 12%, and HCV prevalence at 10%), neither the strategy of screening and subsequent treatment for progressive disease nor the addition of elective cesarean delivery was a cost-effective intervention. The former strategy continued to be dominated by the current policy of no screening, and the latter yielded a minimum cost-effectiveness ratio of $212,280/ QALY.
Comment Although 1 previous study demonstrated that screening the general population for HCV is not a cost-effective intervention,23 the cost-effectiveness of HCV screening during pregnancy, which is a unique circumstance in which lifelong health implications exist for 2 patients (the mother and her child) has not been explored. We believe that, because a policy of screening for HCV in pregnancy affects both the mother and her child, any decision should weigh equally the cost and utility for each individual. When we analyzed the maternal costs separately from the cost that was attributed to the child, it is clear that the benefit of screening, treatment, and cesarean delivery that may be gained for the child is outweighed by the costs and disutility that are incurred by the mother. When the sum of maternal and child lifetime costs and QALYs are considered, these interventions are not cost-effective interventions. Our analysis demonstrates that screening pregnant women for HCV and subsequent treatment for
Plunkett and Grobman progressive disease are not cost-effective interventions and are, in fact, dominated by the current practice of no screening. Of note, this intervention results in a net decrement in utility, which indicates that the disutility of knowledge of HCV infection for both the mother and her child is not outweighed by the potential benefit of treatment. When the disutility of knowledge of the disease is minimized and the utility value for known HCV is set at the highest end of the range (0.99), screening and subsequent treatment are no longer dominated by the current strategy but still do not nearly breach the cost-effectiveness threshold. Even under circumstances that we might consider to favor screening (such as when the prevalence of the disease is high or the efficacy of treatment is improved), the intervention does not become cost-effective. These results likely reflect the slow, indolent nature of HCV infection, particularly in young women, in combination with the additional annual cost of known HCV infection and the absence of useful intervention for many years. With the additional intervention of elective cesarean delivery to avert perinatal transmission, the benefit of the prevention of neonatal HCV infection results in a net gain in the utility value for the mother and her child. As such, the combined interventions of screening, subsequent treatment, and elective cesarean delivery are not dominated by the current practice of no screening. However, these interventions still do not approach our cost-effectiveness threshold. In fact, even under circumstances that are most favorable to screening, when HCV infection is highly prevalent, perinatal transmission is high, and elective cesarean delivery eliminates all perinatal transmission, these interventions are not costeffective. Of note, our cost-effectiveness threshold of $50,000/ QALY represents a commonly used threshold in research studies. Arguably, some health interventions that currently are used in the United States already exceed this threshold and prompt the question as to what threshold may better reflect the societal priorities that are demonstrated by our health care system. Although the answer to this query may be relevant to the interpretation of our results, this task is well beyond the scope of this study. As such, we chose a costeffectiveness threshold that was consistent with the current body of literature and presented the specific cost-effectiveness ratios for important sensitivity analyses so that readers may interpret the results across a range of cost-effectiveness thresholds. This analysis has several strengths. First, our model was robust to all cost, probability, and utility variables. Each variable was tested across its full range of values in 1-way sensitivity analyses; variables of interest were assessed in multivariate analyses, and our conclusions were not altered under any of these circumstances. Second, we validated our Markov model for HCV
1159 progression in the adult population, and our results fell within the range of outcomes that were expected in empirically calibrated models of HCV disease progression19,24 and compared favorably with prospective cohort analyses.57-61 Finally, our model used a life-long timeline in the analysis of the costs and benefits of screening for HCV infection in pregnancy. We did not limit our analysis to immediate outcomes (ie, prevention of perinatal HCV transmission) but considered the implications of screening over the lifetime of both the mother and her child, the perspective required to make rational health policy decisions. We also demonstrate that a large proportion of the costs are derived from the mother and that the decrement in her QALYs outweighs the small benefit for the child. One limitation of this analysis is that the natural history of HCV infection in the pediatric population is not well described.41-44 To date, cohort studies of HCVinfected infants and children do not include long-term outcomes beyond early adulthood.62 In constructing the Markov analysis, we made conservative assumptions with regard to the rate of HCV disease progression. Although cases of rapidly progressive HCV disease in the pediatric population have been reported, we nevertheless modeled the more typical course of HCV disease. As such, we included a 20-year latency period in which no disease progression occurred in chronically HCVinfected children. As a result, our model of perinatal infection depicts a relatively indolent, slowly progressive disease course. Nevertheless, in sensitivity analyses, in which more aggressive disease transitions were modeled, our conclusions did not change. Similarly, we are limited by the available utility values for HCV-related health states. The values that were used in this study were derived from the published literature and originated from a panel of experts who used standard techniques to derive the values. Expert opinion may not reflect the utility values that would be assigned by a patient or the general population. Furthermore, these utility values may change over the course of a lifetime and may be impacted by HCV transmission to a child or family member. However, we recognized these limitations and tested each utility variable over a wide range of values in our sensitivity analyses, and our conclusions did not change. Another limitation to this analysis is that it is performed from the perspective of the health care system. Thus, it does not take into account indirect costs (such as costs that are attributed to hours lost from work) that are associated with each specific health state. As such, the analysis may underestimate the costs of poor health; however, these costs likely represent only a small portion of total lifetime costs for a minority of the population and should not alter our findings. HCV infection remains a significant public health issue that affects both mothers and their infants alike.
1160 Unfortunately, our ability to treat or cure the disease is relatively limited. Consequently, as our model demonstrates, a screening strategy is not a cost-effective intervention even in the unique circumstances of pregnancy, when 2 individuals potentially could access the benefits of treatment and 1 individual could even avoid the disease altogether. In the future, new treatments may offer the possibility that screening would become a costeffective intervention, but at present, the current standard of care (no screening) is the most cost-effective option.
References 1. Centers for Disease Control and Prevention. Recommendations for prevention and control of hepatitis C virus (HCV) infection and HCV-related disease. MMWR Morb Mortal Wkly Rep 1998;47:1-39. 2. Alter HJ, Seef LB. Recovery, persistence, and sequelae in hepatitis C virus infection. Semin Liver Dis 2000;20:17-35. 3. Alter MJ, Kruszon-Moran D, Nainan OV, McQuillan GM, Gao F, Moyer LA, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. N Engl J Med 1999; 341:556-62. 4. Seeff LB, Buskell-Bales Z, Wright EC, Durako SJ, Alter MJ, Iber FL, et al. Long-term mortality after transfusion-associated non-A, non-B hepatitis: the National Heart, Lung, and Blood Institute Study Group. N Engl J Med 1992;327:1906-11. 5. Alter MJ, Margolis HS, Krawczynski K, Judson FN, Mares A, Alexander WJ, et al. The natural history of community-acquired hepatitis C in the United States: the Sentinel Counties Chronic non-A, non-B Hepatitis Study Team. N Engl J Med 1992;327:1899-905. 6. Wasley AD, Alter MJ. Epidemiology of hepatitis C. Semin Liver Dis 2000;20:1-16. 7. Alter MJ. Epidemiology of hepatitis C in the west. Semin Liver Dis 1995;15:5-14. 8. Silverman NS, Jenkin BK, Wu C, McGillen P, Knee G. Hepatitis C virus in pregnancy: Seroprevalence and risk factors for infection. Am J Obstet Gynecol 1993;169:583-7. 9. National Institutes of Health Consensus Development Conference statement: management of hepatitis C: 2002: June 10-12, 2002. Hepatology 2002;36(suppl):S3-20. 10. American College of Obstetricians and Gynecologists. Viral hepatitis in pregnancy: ACOG educational bulletin, no. 248. Intl J Gynaecol Obstet 1998;63:195-202. 11. Manns M, McHutchison JG, Gordon SC, Rustgi VK, Shiffman M, Reindollar R, et al. Peginterferon alfa-2b plus ribavirin compared with interferon alfa-2b plus ribavirin for initial treatment of chronic hepatitis C. Lancet 2001;358:958-65. 12. Kim WR, Poterucha JJ, Hermans JE, Therneau TM, Dickson ER, Evans RW, et al. Cost-effectiveness of 6 and 12 months of interferon-alpha therapy for chronic hepatitis C. Ann Int Med 1997;127:866-74. 13. Bennett WG, Inoue Y, Beck JR, Wong JB, Pouker SG, Davis GL. Estimates of the cost-effectiveness of a single course of interferonalpha 2b in patients with histologically mild chronic hepatitis. Ann Int Med 1997;127:855-65. 14. Davis GL, Beck JR, Farrell G, Poynard T. Prolonged treatment with interferon in patients with histologically mild chronic hepatitis C. J Viral Hepat 1998;5:313-21.
Plunkett and Grobman 15. Buti M, Casado MA, Fosbrook L, Wong JB, Esteban R. Costeffectiveness of combination therapy for naı¨ ve patients with chronic hepatitis C. J Hepatol 2000;33:651-8. 16. Wong JB, Koff RS. Watchful waiting with periodic liver biopsy versus immediate empirical therapy for histologically mild chronic hepatitis C. Ann Int Med 2000;133:665-75. 17. Wong JB, Poynard T, Ling M-H, Albrecht JK, Pauker SG. Costeffectiveness of 24 or 48 weeks of interferon alpha-2b alone or with ribavirin as initial treatment of chronic hepatitis C. Am J Gastroenterol 2000;95:1524-30. 18. Younossi ZM, Singer ME, McHutchison JG, Shermock CK. Cost effectiveness of interferon alfa-2b combined with ribavirin for the treatment of chronic hepatitis C. Hepatology 1999;30:1318-24. 19. Solomon JA, Weinstein MC, Hammitt JK, Goldie SJ. Empirically calibrated model of hepatitis C virus infection in the United States. Am J Epidemiol 2002;156:761-3. 20. Gibb DM, Goodall RL, Dunn DT, Healy M, Neave P, Cafferkey M, et al. Mother-to-child transmission of hepatitis C virus. Lancet 2000;356:904-7. 21. Resti M, Jara P, Hierro L, Azzari C, Giacchino R, Zuin G, et al. Clinical features and progression of perinatally acquired hepatitis C virus infection. J Med Virol 2003;70:373-7. 22. Sinha M, Das A. Cost effectiveness analysis of different strategies of management of chronic hepatitis C infection in children. Pediatr Infect Dis J 2000;19:23-30. 23. Singer ME, Younossi ZM. Cost effectiveness of screening for hepatitis C virus in asymptomatic, average-risk adults. Am J Med 2001;111:614-21. 24. Solomon JA, Weinstein MC, Hammitt JK, Goldie SJ. Costeffectiveness of treatment for chronic hepatitis C infection in an evolving patient population. JAMA 2003;290:228-37. 25. Piton A, Poynard T, Imbert-Bismut F, Khalil L, Delattre J, Pelissier E, et al. Factors associated with serum alanine transaminase activity in healthy subjects. Hepatology 1998;27:1213-9. 26. McHutchinson JG, Gordon SC, Schiff ER, Shiffman ML, Lee WM, Rustgi VK, et al. Interferon alfa-2b alone or in combination with ribavirin as initial treatment for chronic hepatitis C. N Engl J Med 1998;339:1485-92. 27. Grobman WA, Garcia PM. The cost-effectiveness of voluntary intrapartum rapid human immunodeficiency virus testing for women without adequate prenatal care. Am J Obstet Gynecol 1999;181:1062-71. 28. Cheney CP, Chopra S, Graham C. Infections of the liver. Infect Dis Clin North Am 2000;14:633-7. 29. Carithers RL, Marquardt A, Gretch DR. Diagnostic testing for hepatitis C. Semin Liver Dis 2000;20:159-71. 30. Pawlotsky JM, Lonjon I, Hezode C, Bruno R, Darthuy F, Remire J, et al. What strategy should be used for diagnosis of hepatitis C virus infection in clinical laboratories? Hepatology 1998;27:1700-2. 31. The European Mode of Delivery Collaboration. Elective caesarean-section versus vaginal delivery in prevention of vertical HIV-1 transmission. Lancet 1999;353:1035-9. 32. La Torre A, Biadaioli R, Capobianco T, Coloa MG, Monti M, Pulli F, et al. Vertical transmission of HCV. Acta Obstet Gynecol Scand 1998;77:889-92. 33. Spencer JD, Latt N, Beeby PJ, Collins E, Saunders JB, McCaughan GW, et al. Transmission of hepatitis C virus to infants of human immunodeficiency virus-negative intravenous drug-using mothers: rate of infection and assessment of risk factors for transmission. J Viral Hepat 1997;4:395-409. 34. Ohto H, Terazawa S, Sasaki N, Sasaki N, Hino K, Ishiwata C, et al. Transmission of hepatitis C virus from mothers to infants. N Engl J Med 1994;330:744-50. 35. Giacchino R, Picciotto A, Tasso L, Timitilli A, Sinelli N. Vertical transmission of hepatitis C. Lancet 1995;345:1122-3. 36. Okamoto M, Nagata I, Murakami J, Shunsaku K, Iitsuka T, Hoshika T, et al. Prospective reevaluation of risk factors in
Plunkett and Grobman
37.
38.
39.
40.
41.
42. 43. 44. 45. 46.
47.
48.
49.
mother-to-child transmission of hepatitis C virus: high virus load, vaginal delivery, and negative anti-NS4 antibody. J Infect Dis 2000;182:1511-4. Conte D, Fraquelli M, Prati D, Colucci A, Minola E. Prevalence and clinical course of chronic hepatitis C virus (HCV) infection and rate of HCV vertical transmission in a cohort of 15,250 pregnant women. Hepatology 2000;31:751-5. Molin GD, D’Agaro P, Ansaldi F, Ciana G, Fertz C, Alberica S, et al. Mother-to-infant transmission of hepatitis C virus: rate of infection and assessment of viral load and IgM anti-HCV as risk factors. J Med Virol 2002;67:137-42. Pembrey LJ, Tovo PA, Newell ML. Effects of mode of delivery and infant feeding on the risk of mother-to-child transmission of hepatitis C virus. BJOG 2001;108:371-7. Bortolotti F, Resti M, Giacchino R, Azzari C, Gussetti N, Crivellaro C, et al. Hepatitis C virus infection and related liver disease in children of mothers with antibodies to the virus. J Ped 1997;130:990-3. Garcia-Monzon C, Jara P, Fernandez-Bermejo M, Hierro L, Frauca E, Camarena C, et al. Chronic hepatitis C in children: a clinical and immunohistochemical comparative study with adult patients. Hepatology 1998;28:1696-701. Birnbaum AH, Shneider BL, Moy L. Hepatitis C in children. N Engl J Med 2000;342:290-2. Ruiz-Moreno M, Leal-Orozco A, Millan A. Hepatitis C virus infection in children. J Hepatol 1999;31(suppl):124-9. Aach RD, Yomtovian RA, Hack M. Neonatal and pediatric posttransfusion hepatitis C. Pediatr 2000;105:836-42. Jonas MM. Treatment of chronic hepatitis C in pediatric patients. Clin Liver Dis 1999;3:855-67. Desmet V, Gerber M, Hoofnagle JH, Manns M, Scheuer PJ. Classification of chronic hepatitis: diagnosis, grading and staging. Hepatology 1994;19:1513-20. Lau DT-Y, Kleiner DE, Ghany MG, Park Y, Schmid P, Hoofnagle JH. 10-Year follow-up after interferon-alpha therapy for chronic hepatitis C. Hepatology 1998;28:1121-7. Shindo M, Hamada K, Oda Y, Okuno T. Long-term follow-up study of sustained biochemical responders with interferon therapy. Hepatology 2001;33:1299-302. Marcellin P, Boyer N, Gervais A, Martinot M, Pouteau M, Castelnau C, et al. Long-term histologic improvement and loss of detectable intrahepatic HCV RNA in patients with chronic
1161
50.
51.
52.
53.
54. 55. 56. 57.
58.
59.
60.
61.
62.
hepatitis C and sustained response to interferon-alpha therapy. Ann Int Med 1997;127:875-81. Lindsay M, Feng TI, Peterson HB, Slade BA, Willis S, Klein L. Routine human immunodeficiency virus infection screening in unregistered and registered inner-city parturients. Obstet Gynecol 1991;77:599-603. Mauskopf JA, Paul JE, Wichman DS, White AD, Tilson HH. Economic impact of treatment of HIV-positive pregnant women and their newborns with zidovudine. JAMA 1996;276:132-8. Traynor JD, Peaceman AM. Maternal hospital charges associated with trial of labor versus elective repeat cesarean section. Birth 1998;25:81-4. Rouse DJ, Owen J, Goldenberg RI, Cliver SP. The effectiveness and costs of elective cesarean delivery for fetal macrosomia diagnosed by ultrasound. JAMA 1996;276:1480-6. Cohen HE, editor. Drug topics red book. Montvale (NJ): Medical Economics; 2002. p. 473 (pegylated interferon alfa-2b), p. 521 (ribavirin). US Department of Labor, Bureau of Labor Statistics. Available at www.bls.gov/cpi/. Wong JB, Bennett WG, Koff RS, Pauker SG. Pretreatment evaluation of chronic hepatitis C. JAMA 1998;280:2088-93. Di Bisceglie AM, Goodman ZD, Ishak KG, Hoofnagle JM, Melpolder JJ, Alter MJ. Long-term clinical and histopathological follow-up of chronic posttransfusion hepatitis. Hepatology 1991;14:969-74. Hopf U, Moller B, Kuther D, Stemerowicz R, Lobeck H, LudtkeHandjery A, et al. Long-term follow-up of posttransfusion and sporadic chronic hepatitis non-A, non-B and frequency of circulating antibodies to hepatitis C virus (HCV). J Hepatol 1990;10:69-76. Koretz RL, Abbey H, Coleman E, Gitnick G. Non-A, non-B posttransfusion hepatitis: looking back in the second decade. Ann Int Med 1993;119:110-5. Mattson L, Sonnerborg A, Weiland O. Outcome of acute symptomatic non-A, non-B hepatitis: a 13-year follow-up study of hepatitis C virus markers. Liver 1993;13:274-8. Tremolada F, Casarin C, Alberti A, Drago C, Tagger A, Ribero ML, et al. Long-term follow-up from non-A, non-B (type C) posttransfusion hepatitis. J Hepatol 1992;16:273-81. Vogt M, Lang T, Frosner G, Klingler C, Sendl A, Zeller A, et al. Prevalence and clinical outcome of hepatitis C infection in children who underwent cardiac surgery before the implementation of blood-donor screening. N Eng J Med 1999;341:866-70.