Trends in HCV RNA Testing Among HCV Antibody–Positive Persons in Care, 2003–2010

Abstract

Background.

A test for hepatitis C virus (HCV) RNA is essential to identify persons with active, or current, HCV infection. We assessed trends in HCV RNA testing following a positive HCV antibody result among persons in 4 large healthcare organizations.

Methods.

Data collected from adults with ≥2 clinical encounters during 2003–2010 were analyzed to determine the frequency of, interval between, and factors associated with having an RNA test after a first positive HCV antibody test.

Results.

From 2003–2010, 5860 persons had a positive antibody test, of whom 3570 (60.9%) had a follow-up RNA test. During this period, the annual frequency of persons with an eventual RNA test did not change significantly; however, the fraction of persons who had the follow-up RNA test within 6 months improved significantly, from 45% in 2003 to 57% in 2010 (P < .001, for trend). Persons born during 1945–1965, men, and those with annual income <$30 000 (by census geocode) were less likely to have had a follow-up RNA test done within 6 months of a positive antibody test.

Conclusions.

Less than two-thirds of persons with a positive HCV antibody test had a follow-up RNA test. Rapid ascertainment of HCV infection status with reflex testing to RNA is critical to identify persons eligible for treatment.

An estimated 3.6 million persons in the United States have been infected with hepatitis C virus (HCV), of whom approximately 2.7 million are chronically infected [1]. In 1998, the Centers for Disease Control and Prevention (CDC) issued guidelines that focused on testing persons with identified risk factors for HCV infection [2]. In 2003, guidelines for laboratory reporting of HCV antibody test results were published, and recommendations for one-time HCV antibody testing of all persons born during 1945–1965, regardless of other risk factors, were published in 2012 [3, 4].

Positive test results for HCV antibody (eg, enzyme immunoassay, chemiluminescent immunoassay, and HCV recombinant immunoassay) do not distinguish past, resolved infection from “active” (or current) infection. In the absence of symptomatic acute hepatitis or evidence of recent seroconversion from a known HCV exposure, identification of HCV RNA in the serum typically represents chronic infection [5]. Identification of HCV RNA in the serum is also necessary to determine the magnitude of viremia and HCV genotype. With the impending approval of more options for curative pharmacotherapy for most patients [6] and evidence that many persons with a positive antibody test were not subsequently tested for active infection [7, 8], testing guidelines published in 2013 emphasized the identification of active HCV infection [5]. In a previous study, we found that 12.7% of persons from 4 healthcare organizations in the United States received either an HCV antibody or HCV RNA test and, of those tested, 5.5% had a positive test result [9]. To determine baseline trends in testing for active HCV infection before the updated 2013 guidelines, we examined the frequency of and factors associated with the performance of an HCV RNA test subsequent to a positive antibody test among persons in these same 4 large integrated healthcare organizations from 2003 through 2010.

METHODS

We included in the analysis adults aged ≥18 years from 4 healthcare organizations (Geisinger Health System, Danville, Pennsylvania; Henry Ford Health System, Detroit, Michigan; Kaiser Permanente Northwest, Portland, Oregon; Kaiser Permanente, Honolulu, Hawaii) of the HMO Research Network with at least 2 clinical services (ie, hospital admission or outpatient provider, laboratory, or emergency department visit), 1 of which was during 1 January 2006 through 31 December 2010. Data were collected via electronic health records and chart abstraction. For each person, the follow-up period originated from the date of the first (which could have preceded 2006) until the last encounter in their respective healthcare organization. To allow persons in the latter years of the study period sufficient follow-up time for the performance of an HCV RNA test, we excluded positive antibody test results after 31 December 2010, but data on the performance of a follow-up HCV RNA were collected through 2011. We excluded persons with documented chronic hepatitis B virus infection (ie, the Chronic Hepatitis Cohort Study [CHeCS] Hepatitis B cohort [10]), those with a first encounter or HCV antibody test performed before 1 January 2003, those with no observation time (ie, had only a single service encounter during their entire follow-up), and persons who had HCV RNA test results but were missing antibody results. We chose 2003 as the earliest cohort year in the study period because HCV RNA testing was not US Food and Drug Administration (FDA)–approved until 2001.

We determined the proportion of persons with a positive HCV antibody test result who had a subsequent quantitative or qualitative HCV RNA test, regardless of the time interval between the 2, by year, from 2003 through 2010. The calendar year of the positive antibody test was regarded as the cohort year associated with the first subsequent HCV RNA test, irrespective of the year the first follow-up HCV RNA was performed. For example, persons with a first positive HCV antibody test in 2003 and the first follow-up HCV RNA in 2006 were counted among those in 2003 who had a follow-up HCV RNA test. Among persons with a positive antibody test in each cohort year, we determined time interval between the dates of the first positive antibody test result to the dates of the first subsequent HCV RNA test result. Multivariate analysis was used to determine demographic factors associated with having an HCV RNA test within 6 months of the first positive HCV antibody result. Although arbitrary, we used 6 months because it was an interval that balanced sufficient time for HCV RNA testing to occur and a limited, finite period of follow-up to apply equally among each of the cohort years during the study. The adjusted model included the following covariates: age, sex, race, median household income, site, year of first positive HCV antibody test, and duration of follow-up. Adjusted odds ratios (ORs) were obtained along with 95% confidence intervals (CIs). A P value <.05 was considered statistical significant. SAS version 9.3 (SAS Institute, Cary, North Carolina) was used for all statistical analyses.

The investigation followed the guidelines of the US Department of Health and Human Services regarding protection of human subjects. The study protocol was approved and renewed annually by each participating institution’s institutional review board.

RESULTS

Figure 1 displays a schematic derivation of the study population. Of a total of 2 318 780 adults from the 4 healthcare organization sites during 2003–2010, 1 373 382 were excluded from the analysis because of entry into their respective healthcare system before 2003, had an HCV antibody test before 2003, or because they were lacking an antibody test result that preceded an HCV RNA test. Of the 945 398 persons eligible for analysis during 2003–2011, 87 431 (9.2%) had an antibody test, of whom 5860 (6.7%) had a positive result. Of these 5860 persons, 3570 (60.9%) had an HCV RNA test performed, of whom 3047 (85.3%) had a positive result. Among those with a positive RNA result, 2431 (80%) had had their RNA test performed within 6 months of their positive HCV antibody test.

Figure 1.

Schematic representation of persons excluded to achieve study population. Abbreviations: anti-HCV, antibody to hepatitis C virus; HBV, hepatitis B virus; HCV, hepatitis C virus.

Table 1 shows the frequency of test results, by year, during 2003–2010. During the interval, the number of positive antibody tests in the cohort increased, from a low of 296 in 2003 to 887 in 2010 (maximum 978, in 2009); the proportion of persons who had a follow-up HCV RNA test ranged from 58.2% (in 2007) to 66.2% (in 2003). Consequently, the proportion of persons who had a follow-up HCV RNA performed within 6 months increased significantly during the interval, from 45.3% (134/296) in 2003 to 57.0% (506/887) in 2010 (P < .001, trend). The median number of days of observation after the first positive HCV antibody result to the date of the last clinical service delivered was significantly greater for those who received an HCV RNA test compared with persons who did not (1033 [range, 1–3700] vs 615 [range, 1–3571] days [P < .001]).

Table 1.

Proportion of Hepatitis C Virus (HCV) Antibody–Positive Persons Who Received HCV RNA Test and Tested Positive for RNA, and the Days Between First HCV Antibody–Positive Test and First HCV RNA Test, by Year of Antibody-Positive Test, 4 US Healthcare Organizations, 2003–2010

Year of First Positive HCV Antibody Test	No. of Patients With First Positive HCV Antibody Test	From Previous Column, No. (%) Patients Ever Tested for HCV RNAa,b	From Previous Column, No. (%) Patients With Positive HCV RNA	No. (%) of Persons With HCV RNA Test Done in 6 Monthsc
2003	296	196 (66.2)	166 (84.7)	134 (45.3)
2004	452	283 (62.6)	244 (86.2)	195 (43.1)
2005	604	387 (64.1)	324 (83.7)	284 (47.0)
2006	861	502 (58.3)	434 (86.4)	387 (44.9)
2007	827	481 (58.2)	417 (86.7)	378 (45.7)
2008	955	578 (60.5)	470 (81.3)	482 (50.5)
2009	978	587 (60.0)	505 (86.0)	504 (51.5)
2010	887	556 (62.7)	487 (87.6)	506 (57.0)
Total	5860	3570 (60.9)	3047 (85.3)	2870 (49.0)

Abbreviation: HCV, hepatitis C virus.

^aIncludes persons tested during 2003–2011.

^bP value for trend = .32.

^cP value for trend <.001.

Table 2 shows the results of multivariate analysis to determine factors associated with having an HCV RNA test within 6 months of the first positive antibody test, controlling for age, sex, race, median household income, year of first positive HCV antibody test, duration of follow-up, and study site. Compared with persons born after 1965, those born during 1955–1965 and during 1945–1954 were less likely to have an HCV RNA test performed within 6 months of a positive antibody test (52.4% vs 48.8%, adjusted odds ratio [aOR], 0.76 [95% CI, .58–.99] and 52.4% vs 46.2%, aOR, 0.69 [95% CI, .51–.94], respectively). Compared with men, women were more likely to have an HCV RNA test within 6 months (49.8% vs 48.5%, aOR, 1.33 [95% CI, 1.06–1.66]). Compared with 49.0% of persons with an annual income <$30 000 (by census tract geocode), those with higher incomes were more likely have an HCV RNA test within 6 months ($30 000–$49 999: 53.9% [aOR, 1.32; 95% CI, 1.02–1.71]; $50 000–$74 999: 52.4% [aOR, 1.45; 95% CI, 1.00–2.11]; ≥$75 000: 55.3% [aOR, 2.68; 95% CI, 1.44–4.97]).

Table 2.

Characteristics of Persons Tested for Current Hepatitis C Virus (HCV) Infection Within 6 Months of a Positive HCV Antibody Test, 4 US Healthcare Organizations, 2003–2010

			HCV RNA Done in 6 mo
Characteristic	Anti-HCV+	HCV RNA Done, No. (%)	No. (%)	Adjusted Odds Ratio (95% CI)a
Total	5860	3570 (60.9)	2870 (49.0)
Birth year
>1965	1572	980 (62.3)	823 (52.4)	Reference
1955–1965	2105	1279 (60.8)	1027 (48.8)	0.76 (.58–.99)
1945–1954	1831	1093 (59.7)	846 (46.2)	0.69 (.51–.94)
≤1944	346	217 (62.7)	173 (50.0)	1.36 (.85–2.18)
Sex
Male	3584	2185 (61.0)	1737 (48.5)	Reference
Female	2273	1384 (60.9)	1132 (49.8)	1.33 (1.06–1.66)
Race
Non-Hispanic white	3448	2234 (64.8)	1808 (52.4)	Reference
Non-Hispanic black	1329	711 (53.5)	537 (40.4)	1.02 (.73–1.44)
Others	1083	625 (57.7)	525 (48.5)	1.27 (.90–1.78)
Median household incomeb
Missing	987	385 (39.0)	325 (32.9)
<$30 000	1455	907 (62.3)	713 (49.0)	Reference
≥$30 000–$49 999	2425	1612 (66.5)	1307 (53.9)	1.32 (1.02–1.71)
≥$50 000–$74 999	834	563 (67.5)	437 (52.4)	1.45 (1.00–2.11)
≥$75 000	159	103 (64.8)	88 (55.3)	2.68 (1.44–4.97)

Abbreviations: anti-HCV⁺, hepatitis C virus antibody positive; CI, confidence interval; HCV, hepatitis C virus.

^aThe adjusted model included the following covariates: age, sex, race, median household income, site, year of first positive HCV antibody test, and duration of follow-up.

^bBy census tract geocode.

DISCUSSION

Testing for HCV RNA after a positive test for HCV antibody is necessary to determine the presence of current HCV infection and is imperative for making clinical decisions involving treatment, to counsel those infected to minimize additional hepatic injury (eg, restricting alcohol intake), and to prevent HCV transmission. In our study from 4 integrated healthcare organizations, we found that approximately 60% of persons with an initial positive antibody test result subsequently had a follow-up HCV RNA test during 2003–2010, and that the proportion of persons tested for HCV RNA following a positive antibody test did not vary appreciably by year for the duration of the study. However, the proportion of persons with a follow-up HCV RNA test performed within 6 months of a positive antibody test improved significantly.

All persons whose HCV infection status is unknown and who test positive for HCV antibody should promptly undergo HCV RNA testing. The low frequency of follow-up with HCV RNA testing in our cohort might be explained by patient, provider, laboratory, and guideline-related factors. Patients, for whatever reason, might not return to have an RNA test performed despite it being ordered. Some providers might be unaware of the need to order an HCV RNA to determine HCV infection status, and even if alerted to the necessity, still may not pursue additional testing [11]. Assays for HCV RNA require that serum or plasma samples be collected, processed, and stored in a manner to reduce the likelihood of a false-negative result. Such samples can be used only if blood was collected in tubes with no additives or by using ethylenediaminetetraacetic acid (EDTA), if serum or EDTA plasma was separated from cellular components within 2–6 hours after collection, if storage of serum or EDTA plasma at 2°C–5°C was limited to 72 hours (for longer storage, at −20°C or −70°C), and, if shipping was required, frozen samples were protected from thawing [3]. If a specimen obtained for HCV serologic testing is not handled properly, then the patient must be recalled to obtain a new specimen. Previous HCV testing guidelines, written when the specificity of serologic tests and sensitivity of HCV RNA tests were poorer than assays available today, emphasized the detection and confirmation of the presence of HCV antibody [3]. In the past, options for HCV antiviral therapy were limited and, when utilized, were difficult to tolerate and frequently ineffective; thus, some patients and providers may have harbored negative attitudes about the prospects for successful treatment, opting against confirmation of chronic infection, or simply assuming a positive antibody test was already reflective of chronic infection. Another possibility is that some patients had the antibody test ordered because it was required for purposes other than for consideration of treatment (eg, application for life insurance).

A recent study from the New York City Department of Health and Mental Hygiene examined the “gap” in testing for HCV RNA and found that two-thirds of persons, which included persons screened in detox facilities and jails, had an HCV RNA ordered subsequent to a positive antibody test [11]. Their results, which were more favorable than ours, might have been a consequence of active follow-up with providers of persons with positive antibody tests. A Veterans Affairs (VA) study found that 76% of 13 257 antibody-positive patients received HCV RNA testing, although patients aged >65 years and illicit drug users were significantly less likely to receive HCV RNA testing [12]. Another group of investigators using surveillance data for HCV infection from 8 US sites estimated that only 50% of persons newly reported with HCV infection had an HCV RNA test performed [7].

In our multivariate analysis of factors associated with having an HCV RNA test within 6 months of a positive antibody test result, we found that the aOR for having an HCV RNA test within 6 months increased incrementally for persons of each higher income strata (by census tract geocode), compared with persons with an annual income <$30 000. Whether this finding was related to type of provider or insurance coverage we could not determine, and income data were missing for approximately one-sixth of persons. More difficult to explain was the finding that persons born during 1945–1965 were less likely to have a follow-up HCV RNA test within 6 months compared with younger and older age cohorts, and men were less likely to have an HCV RNA follow-up within 6 months compared with women, despite controlling for age, sex, race, household income, year of first positive HCV antibody test, duration of follow-up, and study site. However, we found with further analysis (results not shown) that men and persons born during 1945–1965 constituted a higher fraction of persons who tested HCV antibody positive during the early years of the study period, when the median time interval between antibody testing and HCV RNA was longer. Since 2012, the CDC has recommended that all persons born during 1945–1965 receive a one-time HCV antibody test; persons with a positive antibody test result should have an HCV RNA test performed to determine HCV infection status [4].

One manner of ensuring an expedient determination of HCV infection status is to perform reflex HCV RNA testing on a routine basis (ie, done automatically if the initial antibody test is positive). Such a process requires that a portion of the initial specimen be retained and obviates the need to separately order HCV RNA testing and require an additional patient visit to obtain a new specimen. Research from the VA system determined that reflex testing was cost-effective, and the practice was initiated in the VA system in 2009 [13]. Many commercial laboratories now offer reflex testing [14], but such testing must be ordered up front by the provider, and HCV antibody and RNA tests may still be ordered distinctly. Thus, it is incumbent on providers to pursue testing in a manner that ensures determination of HCV infection status and that positive antibody test results are immediately followed with an HCV RNA test. In this respect, hospital laboratory directors could also help to ensure that RNA tests are ordered by reviewing laboratory procedures and messages to clinicians that accompany positive HCV antibody results.

We informally investigated among the 4 healthcare organizations involved in this study the impediments to the introduction of reflex HCV RNA testing. Primary among them was that the quantitative HCV RNA assays commonly used by their in-house laboratories are FDA-approved only for use in monitoring RNA levels during antiviral therapy, rather than for diagnostic use [15, 16]. Accordingly, in-house laboratories either need to locally validate the RNA assay for diagnostic use (a long process per accreditation standards) or send all reflex testing out to a reference laboratory that has performed the required validation process. We expect that this obstacle is typical in the United States. In response, the National Laboratory Training Network and HIV, Hepatitis, STD and TB Programs, Association of Public Health Laboratories, in collaboration with the Division of Viral Hepatitis, sponsored a 2-day seminar on 15–16 May 2014. A major portion of this meeting focused on the procedures necessary for clinical validation of quantitative assays, which were initially approved by FDA exclusively for monitoring antiviral therapy, for diagnostic purposes.

Our analysis was subject to a number of limitations. Although inclusion criteria necessitated that persons have at least 2 clinical encounters during the study period, we did not stipulate a necessary time interval between the 2. Thus, it is conceivable that some patients with a short interval between the 2 visits may not have had sufficient opportunity to have an HCV RNA test performed. As mentioned, the median number of days of observation after the first positive antibody test to last follow-up encounter was significantly higher for persons who had an HCV RNA test performed than for those who did not; however, even among persons who did not have an HCV RNA test, the median follow-up was approximately 20 months. Another potential limitation was that some testing could have been performed on patients in a setting other than the 4 healthcare organizations involved in the study, for which the records were not available to our abstractors, or that testing conducted within the 4 organizations was not recorded in the electronic health record. Finally, as our study was dependent on the presence of recorded HCV RNA test results, we cannot exclude the possibility that tests may have been ordered but not performed.

In summary, in our study population we found that only 60% of persons with an initial positive test for HCV antibody had a follow-up HCV RNA test performed to determine HCV infection status, but the fraction of persons who had a follow-up HCV RNA performed within 6 months improved significantly over time. With curative pharmacotherapy potentially available now for most persons, application of reflex testing to HCV RNA following a positive HCV antibody test is critical to the prompt determination of HCV infection status. As efforts to expand the use of reflex testing are planned, in part through the clinical validation of quantitative assays, a reassessment of HCV RNA testing trends in the future will be useful to determine the degree to which reflex testing has been implemented.