Pregnancy Status, Risk Factors, and Opportunities for Referral to Care Among Reproductive-Aged Women With Newly Reported Chronic Hepatitis C Virus Infection in Tennessee

Abstract

Objectives:

From 2006 through 2012, the number of acute hepatitis C virus (HCV) infections increased 364% among persons aged ≤30, including reproductive-aged women, in Central Appalachian states. Outreach to reproductive-aged women with newly diagnosed HCV infection affords a unique opportunity to provide counseling, further testing, and linkage to treatment. We modeled a centrally located statewide effort to reach this population and their health care providers to ascertain pregnancy status, HCV risk factors, and opportunities for linkage to additional services.

Methods:

Using the Tennessee Department of Health’s surveillance database, we identified women aged 18-45 residing in Tennessee with newly reported chronic HCV infection from May through October 2017. We contacted health care providers and patients to request information on pregnancy status and HCV risk factors as well as to provide linkage to treatment services.

Results:

Of 1548 women included in this study, health care provider or patient contact information was available for 1316 (85.0%) women. Of the 1316 women, 806 (61.2%) women had a health care provider or patient response, of whom 242 (30.0%) were pregnant. Of 296 patients contacted, 194 (65.5%) reported intranasal drug use, 193 (65.2%) reported having been incarcerated for more than 24 hours, and 180 (60.8%) reported injection drug use. Ninety-eight (33.1%) patients were referred for confirmatory testing, and 174 (58.8%) were referred to treatment.

Conclusion:

A high proportion of reproductive-aged women with newly diagnosed HCV infection were pregnant. Surveillance-informed outreach to this population was feasible and provided opportunities for counseling and linkage to confirmatory testing and treatment. Future studies should evaluate whether a similar model would enhance testing and linkage to care of HCV-exposed infants.

Hepatitis C virus (HCV) infection is the most common bloodborne infection in the United States, affecting an estimated 2.4 million persons from 2013 to 2016.¹ From 2011 to 2014, national rates of acute HCV infection increased 22%, from 139 to 169 per 100 000 reproductive-aged women. Rates of acute HCV infection among young adults increased 364% from 2006 to 2012 in Kentucky, Tennessee, Virginia, and West Virginia.^2,3 This increase in the incidence of HCV infection is attributed to the opioid crisis.³ With approximately 75% of acutely infected persons progressing to chronic infection, the rates of chronic HCV infection among reproductive-aged women are also rising.⁴ From 2013 to 2017, the prevalence of HCV infection reported on birth certificates among pregnant women in Tennessee increased 94%, from 7.9 to 15.3 per 1000 live births.⁵ With a 6% rate of mother-to-child transmission among pregnant women infected with HCV, an increasing number of infants are being born with HCV infection.⁶

Currently, no US Food and Drug Administration HCV treatment regimens are approved for use during pregnancy; however, outreach to HCV-infected reproductive-aged women offers a unique opportunity to provide valuable services, including posttest counseling, additional testing (mother and infant), and linkage to other prevention and treatment services.⁷ In this study, we modeled a centrally located statewide effort to contact reproductive-aged women with newly diagnosed chronic HCV infection and their health care providers to ascertain pregnancy status, risk factors, and opportunities to link to additional testing and treatment services.

Methods

Data Sources

We identified newly reported cases of chronic HCV infection by using the Tennessee National Electronic Disease Surveillance System Base System (NBS). NBS is an online surveillance system developed by the Centers for Disease Control and Prevention (CDC) for local and state health departments to manage cases, create investigations, and analyze data for reportable diseases, as well as to notify and share data on reportable diseases with CDC.⁸ In addition, we used the Centers for Medicare & Medicaid Services National Plan and Provider Enumeration System National Provider Identifier (NPI) Registry to obtain health care provider contact information unavailable in NBS.⁹ We recorded all data and contact attempts in a Research Electronic Data Capture (REDCap) database developed for this project, which was used to determine response rates and analyze data on patient laboratory reports, pregnancy status, and risk factors.¹⁰

Population

We included data on reproductive-aged women aged 18-45 at the time of HCV testing who were residing in Tennessee and had an NBS investigation for chronic HCV (probable or confirmed) initiated between May 1 and October 31, 2017 (Figure). Per the 2016 CDC/Council of State and Territorial Epidemiologists case definition, a probable case of HCV is defined as a positive test for HCV antibodies, and a confirmed case is defined as a positive HCV nucleic acid test or HCV antigen test.¹¹ We initially drew data on women from NBS who met the inclusion criteria and who were not residents of the Nashville-Davidson region aged <30 and did not have an acute HCV investigation. This starting population included 2253 women (Figure). We additionally excluded women who had hepatitis B virus co-infection or acute HCV infection (n = 34) and had been tested at a local health department (n = 530). We excluded these women because they were already receiving similar posttest counseling and follow-up from other Tennessee Department of Health (TDH) programs, and we wanted to decrease the burden on these women and their health care providers. We also excluded women who were tested at a blood products donation center (ie, blood or plasma donation center; n = 36), a community-based organization (n = 29), or a jail or prison (n = 76); previous surveillance efforts conducted by TDH found that contacting these health care providers was deemed unlikely to yield additional information. Our final analytic sample consisted of 1548 women, most of whom resided in Memphis, Nashville, and Eastern Tennessee.

Figure.

Health care provider and patient response rate during investigation of reproductive-aged women (aged 18-45) with chronic hepatitis C virus infection initiated by the Tennessee Department of Health, May 1 to October 31, 2017, Tennessee. Blood products locations include locations where either blood or plasma is donated.

Flow of Data Collection

The TDH Viral Hepatitis Surveillance team entered laboratory results for newly reported HCV into NBS and created surveillance case investigations. For patients who met inclusion criteria, we determined health care provider fax numbers by using NBS and the NPI Registry. If the fax number was not available in the NPI Registry, we called the health care provider directly with the telephone number listed in either NBS or the NPI Registry to request it. We developed and faxed a project-specific standardized 2-question questionnaire to the health care provider. The questionnaire first asked, “What is the reason for HCV testing?” and provided checkboxes for 4 response options, plus a write-in option for “other.” The 4 response options were (1) symptoms of acute hepatitis, (2) asymptomatic patient with risk factors, (3) patient request, and (4) prenatal screening. More than 1 reason could be indicated. The second question was, “Was patient pregnant at time of testing?” Response options were yes, no, and unknown. If the response was yes, the due date was requested. For nonresponding health care providers, the fax was resent at 8 days and 15 days after the initial fax; the health care provider contact window was closed 30 days after the initial fax.

We waited 30 days after contacting the health care provider to allow the provider ample time to disclose HCV test results to the patient. On day 31, we mailed a letter to the patient at the home address listed on the most recent laboratory report and requested that the patient call TDH about an important health matter. When the patient called TDH, we followed a standardized telephone script that asked 10 yes/no questions about pregnancy status and risk factors for HCV transmission (Box). Respondents could choose more than 1 risk factor. We provided posttest counseling and offered referrals to the patient’s local health department for confirmatory testing (for probable HCV cases) or to a regional viral hepatitis case navigator (for confirmed HCV cases) for linkage to care. The Tennessee viral hepatitis case navigators are located in Tennessee’s public health regions and are responsible for providing viral hepatitis education and linkage to services (HCV, HIV, mental health, substance use disorder, and health insurance enrollment). The patient contact window was closed 60 days after the initial health care provider fax. Both the questionnaire sent to the health care provider and the questions asked of patients were not previously validated in other studies but were piloted and adjusted accordingly in surveillance conducted by TDH in the previous year.

Box.

Standardized telephone script used by Tennessee Department of Health to communicate with reproductive-aged women (aged 18-45) with newly reported chronic hepatitis C virus infection in Tennessee, 2017

1. Are you currently pregnant? □ Yes □ No

   1. If yes,

      1. When is your due date? _____/_____/_____

   2. If no,

      1. Were you pregnant in the last 12 months? □ Yes □ No

      2. Are you planning a pregnancy in the next 12 months? □ Yes □ No

2. Have you been employed in a medical or dental field? □ Yes □ No □ Unknown

3. Been on long-term dialysis for kidney problems? □ Yes □ No □ Unknown

4. Had a needle stick exposure? □ Yes □ No □ Unknown 

5. Received a tattoo? □ Yes □ No □ Unknown

6. Injected drugs, even once? □ Yes □ No □ Unknown

7. Inhaled drugs through your nose, even once? □ Yes □ No □ Unknown

8. Been incarcerated for more than 24 hours? □ Yes □ No □ Unknown

9. Had contact with someone who has hepatitis C? □ Yes □ No □ Unknown

   1. If yes, how would you describe the contact? (Check all that apply.)

      • □–Living in the same household?

      • □–Sexual contact?

      • □–Sharing drugs?

      • □–Other? If other, please specify____________________________

10. Was your mother infected with hepatitis C when you were born? □ Yes □ No □ Unknown

11. Have you spoken with your physician about your results? □ Yes □ No

    • If they have a positive antibody test and a negative RNA test according to newborn screening:

    •  Have you been treated for hepatitis C? □ Yes □ No

Statistical Analysis

We computed response rates for health care providers, patients, and any response from either provider or patient. We determined the percentage of women deemed to be pregnant at the time of testing, defined as confirmation of pregnancy by either health care provider or patient. We stratified pregnancy status by HCV RNA test result upon enrollment into the study (HCV RNA positive, HCV RNA negative, or HCV RNA unknown) to determine the number of women known to be currently infected with HCV and at risk of transmitting HCV infection to their child during pregnancy.

Using responses from the health care provider questionnaire, we tabulated data on the reason(s) for testing. Using responses from the standardized telephone script, we tabulated data on risk factors (eg, employment in the medical or dental field, long-term hemodialysis, needlestick exposure, tattoos, infection drug use, intranasal drug use, history of incarceration, contact with HCV-positive person, or born to an HCV-positive mother) and documented services recommended by TDH during the interview. Institutional review board approval was deemed unnecessary by the Tennessee Department of Health because research was covered under surveillance activities.

Results

Population Outline

Of the 1548 women in our study sample, health care provider and/or patient contact information was available for 1316 (85.0%) women. Of 1316 (85.0%) women with contact information, 856 (65.0%) women had both health care provider and patient contact information, 277 (21.0%) had only provider contact information, and 183 (13.9%) had only patient contact information. Of 856 women with health care provider and patient contact information, 169 (19.7%) had responses from both provider and patient. Of 1133 women with health care provider contact information, 679 (59.9%) responded with a completed form. Of 1039 women with patient contact information, 296 (28.5%) responded to our request for a telephone call and answered our questionnaire in its entirety. Of the 1316 women with health care provider and/or patient contact information, 806 (61.2%) had a response from the provider and/or patient.

Pregnancy Status Based on Health Care Provider and Patient Results

Of the 806 women with responses from the health care provider and/or patient, 242 (30.0%) women were pregnant at the time of HCV testing. Of the 242 women who were pregnant, 133 (55.0%) had a positive HCV RNA test result, 44 (18.2%) were HCV RNA negative, and 65 (26.9%) had unknown HCV RNA status. According to patient response, of 213 women who were not currently pregnant, 8 (3.8%) were planning a pregnancy within the next 12 months, of whom 3 were currently HCV infected (RNA positive).

Of the 169 women with responses from both health care provider and patient, 41 (24.2%) women were pregnant and 115 (68.0%) women were not pregnant at the time of testing, according to both the provider and patient. Therefore, there was agreement on pregnancy status between health care providers and patients for 156 of 169 (92.3%) women; 10 patients stated they were pregnant when the provider did not indicate pregnancy, and 3 providers stated the patient was pregnant when the patient did not indicate pregnancy.

Health Care Provider Results

The most common reason for testing reported by health care providers was that the patient was asymptomatic with risk factors (205 of 679 women; 30.2%; Table 1). Health care providers who wrote in “other” reasons for testing (n = 226; 33.3%) offered reasons such as elevated liver enzymes, patient-reported previous diagnosis or exposure, or that testing was conducted as a part of testing for other sexually transmitted diseases.

Table 1.

Health care provider–reported reason for testing for HCV infection among reproductive-aged women (aged 18-45), Tennessee, 2017 (n = 679)^a

Reason for Testingb	No. (%) of Women (n = 679)
Asymptomatic with risk factorsc	205 (30.2)
Prenatal screening	192 (28.3)
Patient request	126 (18.6)
Symptoms of acute HCV infection	40 (5.9)
Otherd	226 (33.3)

Abbreviation: HCV, hepatitis C virus.

^a Data source: Tennessee Department of Health Viral Hepatitis Surveillance.¹²

^b Testing reasons were not mutually exclusive; numbers total to more than the sample size.

^c Includes patients tested due to physician decision based on risk factors for HCV infection, which could include injection drug use, contact with person with HCV infection, needlestick exposure, employment in the medical or dental field, long-term hemodialysis, needlestick exposure, tattoos, injection drug use, intranasal drug use, history of incarceration, contact with HCV-positive person, or born to an HCV-positive mother.

^d Other testing reasons included elevated liver enzymes, patient-reported previous diagnosis or exposure, or testing that was conducted as part of testing for other sexually transmitted diseases.

Patient Results

Of the 296 patients who participated in the telephone interview, the most common HCV risk factor cited was having received a tattoo (81.8%, n = 242), followed by reporting intranasal drug use (65.5%, n = 194), reporting having been incarcerated for more than 24 hours (65.2%, n = 193), and reporting injection drug use (60.8%, n = 180; Table 2). Of 170 (57.4%) women who reported contact with an HCV-positive person, 89 (52.4%) indicated living in the same household, 78 (45.9%) indicated sexual contact, 73 (42.9%) indicated sharing drugs, and 29 (17.1%) described another type of contact. Only 9 (3.0%) patients did not disclose any risk factors. Of the 169 patients who also had responses from health care providers, only 5 (3.0%) did not disclose any risk factors and 164 (97.0%) disclosed risk factors. In addition, 145 (85.8%) health care providers did not disclose risk factors in their reasoning for testing the patient, whereas 24 (14.2%) patients did. The 5 patients who did not disclose risk factors had agreement from health care providers that the patient had no risk factors, and all 24 providers who disclosed risk factors agreed that the patients disclosed risk factors as well. However, 140 health care providers did not disclose risk factors when the patients did.

Table 2.

Patient-reported risk factors for HCV transmission among reproductive-aged women (aged 18-45), Tennessee, 2017 (n = 296)^a

Risk Factorb	No. (%) of Women (n = 296)
Received a tattooc	242 (81.8)
Inhaled drugs through your nose, even once	194 (65.5)
Been incarcerated for more than 24 h	193 (65.2)
Injected drugs, even once	180 (60.8)
Had contact with HCV-positive persond	170 (57.4)
Been employed in a medical or dental field	80 (27.0)
Had a needlestick exposure	59 (19.9)
Mother infected with hepatitis C when [respondent] was born	12 (4.1)
Been on long-term dialysis for kidney problems	2 (0.7)
None	9 (3.0)

Abbreviation: HCV, hepatitis C virus.

^a Data source: Tennessee Department of Health Viral Hepatitis Surveillance.¹²

^b Risk factors were not mutually exclusive; numbers total to more than the sample size.

^c Did not specify whether the tattoo was professionally done.

^d Any type or frequency of contact.

Of the 296 patients who completed the telephone interview, 50 (16.9%) had not spoken with their physician about their HCV test results, 98 (33.1%) were referred to their local health department for confirmatory testing, and 174 (58.8%) were referred to an HCV navigator for linkage to care. Of the 98 who were referred for confirmatory testing, HCV RNA test results were reported for 13 (13.3%) in NBS by the end of the study.

Discussion

The pregnancy rate of women recently diagnosed with HCV infection in Tennessee in our study sample was 30.0%, far surpassing the pregnancy rate for women of reproductive age in the United States in 2016 (6%).¹³ The pregnancy rate of women in our study could be an overestimation, because health care providers might be more likely to test women for HCV infection and report HCV infection results to TDH for women who are pregnant than for women who are not pregnant, and prenatal screening was the second most common reason for testing (28.3%). That health care providers in our study included HCV testing during routine prenatal screening is interesting, because HCV testing was not recommended by the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America until May 24, 2018.¹⁴ Pregnant women may have also been more concerned than nonpregnant women about the letter they received and therefore more likely to call TDH. Of women who were pregnant, more than half (55%) had a positive HCV RNA test result, demonstrating that more than half of women who were pregnant were currently infected and at risk of transmitting HCV to their child. This high rate of current infection shows that reproductive-aged women are an important group to target for interventions to prevent this possible source of HCV transmission. Furthermore, HCV RNA status was unknown in 65 (27%) women, and the risk of transmission of HCV to their child was also unknown. Increased HCV RNA testing among pregnant women is needed to determine whether there is a risk of transmission to the child so that the child can be properly tested. Of the 169 women who had responses from both health care provider and patient, there was 92% agreement in pregnancy status, demonstrating that the questionnaires used were reliable.

Nearly two-thirds of the women in our study population reported using intranasal drugs or injecting drugs, demonstrating an opportunity to counsel and/or refer patients for substance use disorder treatment when they seek HCV testing. Intranasal drug use was more common than injection drug use in this population. The high percentage of intranasal drug use also supports previous findings from a study conducted in Knoxville, Tennessee, that assessed HCV transmission risk factors among pregnant women and determined that sharing snorting utensils when inhaling drugs intranasally could be an additional risk factor for HCV transmission and was more common among this population than intravenous drug use.¹⁵

We did not ask respondents to specify whether tattoos were received in a licensed facility, which would not have been as risky as receiving tattoos at an unregulated facility or in prison. Because 65.2% of women had been incarcerated, tattoos acquired in jail or prison could be more common in our study population than in the general population. More than half of women (52.4%) who indicated having contact with an HCV-positive person noted it was contact through sharing a household, and almost half specified it was sexual contact (45.9%) and/or sharing drugs (42.9%). Many women stated during the telephone interview that their significant other was the known HCV-positive contact. Approximately one-fourth (27.0%) of patients who responded were employed in a medical or dental field, and 19.9% of patients who responded had a needlestick exposure. The percentage of women who indicated needlestick exposure as a risk could have been overestimated because some women may have interpreted the question about needlestick exposure as a question about injection drug use; however, we recorded the data because the patient answered the question. In comparing risk factors among women for whom both patients and health care providers responded, patients were more likely than providers to disclose risk factors. These differences are likely attributed to the differences in the questionnaires, because we asked patients about their risk factors for HCV infection, whereas we asked health care providers whether risk factors were a reason for testing the patient for HCV infection.

Our response rates were high for health care providers, which was most likely because the questionnaire was short. We were also thorough in our attempt to find health care provider contact information. During this study, we found that NBS and the NPI Registry were largely lacking in accurate health care provider contact information. By calling the health care provider directly, we were able to find correct provider fax information, and we appended accurate provider contact information to REDCap for future reference.

We were able to provide posttest counseling to all 296 women who returned our request for a telephone call. Posttest counseling largely depended on several factors: test results, whether the patient had spoken to her health care provider about her results, and whether she had a positive RNA test and required linkage to care. Of these 296 women, most had already spoken with their physician about their results, but further interpretation of test results was often necessary.

Limitations

Our study had several limitations. First, contact information for both health care providers and patients was missing for some women, which we attempted to overcome by contacting both the provider and the patient to determine pregnancy status. We did not have any variables other than pregnancy that were collected by both questionnaires and were not able to use both sources to assess risk factors and reasons for testing. This disparity could have caused bias in our study if the characteristics of women and health care providers without contact information differed from those with contact information. Second, patient response bias was possible, because we asked women with HCV infection, who are part of an often stigmatized population, to disclose both pregnancy status and drug use history. Pregnancy could also have been overreported because of self-selection bias; health care providers and patients who responded could have been more likely to test and report or respond if the patient was pregnant rather than nonpregnant. This self-selection bias could be especially true for patient response; only 28.5% of patients, compared with 59.9% of health care providers, responded. Lastly, our findings may not be generalizable to the rest of the United States, because we studied only women of reproductive age in Tennessee.

Strengths

Our study also had several strengths. One strength was the large study population (n = 1548) and our use of information from both patients and health care providers, which permitted several opportunities for data collection. A second strength was a decent response rate for health care provider responses (59.9%). Our patient response rate (28.5%) was higher than rates in previous surveillance of this group. With higher response rates, self-selection bias was less likely to occur. Finally, providing posttest counseling and referral to confirmatory testing and viral hepatitis case navigators allowed for an immediate intervention while collecting information from our study population.

Conclusion

Among reproductive-aged women with newly reported chronic HCV infection in Tennessee, 30.0% were pregnant at the time of testing, and nearly two-thirds of women reported a history of intranasal drug use and/or injection drug use, indicating that this population is at high risk of transmitting HCV. All 296 patient who responded received posttest counseling, including referral to a viral hepatitis case navigator for linkage to care and confirmatory testing. This study revealed that a high proportion of reproductive-aged women with newly diagnosed HCV infection were pregnant and that surveillance-informed outreach to this population was feasible and provided opportunities for posttest counseling and linkage to confirmatory testing and treatment. As the opioid crisis leads to an increase in HCV infection, future studies should evaluate whether a similar model would enhance testing and linkage to care of HCV-exposed infants.