Digestive Diseases and Sciences

, Volume 53, Issue 6, pp 1693–1698

Hepatitis B and C Among Veterans on a Psychiatric Ward


    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Donna A. Wirshing
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Joseph M. Pierre
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Lisa H. Guzik
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Michael D. Kisicki
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Itai Danovitch
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • Shirley J. Mena
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
  • William C. Wirshing
    • Department of Psychiatry, West Los Angeles Veteran’s Affair’s Hospital, David Geffen School of MedicineUniversity of California
Original Paper

DOI: 10.1007/s10620-007-0045-5

Cite this article as:
Tabibian, J.H., Wirshing, D.A., Pierre, J.M. et al. Dig Dis Sci (2008) 53: 1693. doi:10.1007/s10620-007-0045-5


Hepatitis B and C are public health problems. Psychiatric patients may be at risk of hepatitis B and C exposure due to lifestyle and inadequate health care. We aimed to determine prevalence of hepatitis B and C virus exposure and associated risk factors in acutely hospitalized psychiatric veterans. A total of 234 individuals consecutively admitted to the psychiatric wards at the West Los Angeles Veterans Affairs Hospital were asked to participate. A total of 129 patients consented and were screened for viral hepatitis risk factors, hepatitis B surface antigen, hepatitis B surface and core antibodies, and hepatitis C antibodies. About 31 and 38% of the patients had been exposed to hepatitis B and C viruses, respectively. Several risk factors were associated with exposure. Inpatient psychiatric veterans seem to have increased rates of hepatitis B and C exposure. This highlights the need for prevention of risk behavior in this vulnerable population.


HepatitisEpidemiologyPsychiatricVeteransRisk factors


Hepatitis B and C viruses (HBV and HCV) represent major causes of severe liver disease such as cirrhosis, liver failure, and hepatocellular carcinoma. The Center for Disease Control estimates 0.4 and 1.2% of the general US population to be chronically infected with HBV and HCV, respectively [1, 2]. While these viruses are a great public health concern in the general population, the sub-population of individuals with severe psychiatric illness is at particular risk.

Studies from various countries have found an increased prevalence of two- to three-fold with HBV and two- to eight-fold with HCV in patients with psychiatric illness compared to those without [310]. The ubiquity of exposure to risk-conferring lifestyles and behavior in those with severe psychiatric illness is likely related to this increased prevalence. Homelessness, poverty, substance abuse, unprotected sex, sex while intoxicated, sex bartering, sex with intravenous drug users, anal sex, sex with strangers, and coerced sex have all been found to be widespread amongst patients with major psychiatric illness [1117]. This explanation is not unique to HBV and HCV, but rather it may be applied to other, similarly-transmitted infections such as HIV [1822].

In addition to the issue of heightened risk of viral exposure, there also exists the challenge of treatment of psychiatric patients who develop chronic HBV or HCV infection. Interferon alpha is part of HBV and HCV infection treatment but has been reported to have side effects including depression, anxiety, psychosis, mania, and delirium. Overall, these adverse effects occur approximately twice as often in patients with psychiatric illness than those without, and thus they have deterred some providers from treating psychiatric patients [23, 24]. Treatment is further withheld in this sub-population as a result of the severity of psychiatric illness or due to persistent substance abuse [24].

The purpose of this study was to serologically assess the prevalence of HBV and HCV exposure amongst veterans in psychiatric inpatient facilities in the Greater Los Angeles Veterans Affairs (VA) Healthcare System. Behavioral risk factors and diagnoses were also surveyed to find particular associations with seropositivity in order to categorize groups at risk. Ultimately, such information could help guide preventative practices, treatment, and policy.


Study participants

Institutional Review Board approval was obtained to ask individuals who were voluntarily and acutely admitted to psychiatric inpatient facilities at the Greater Los Angeles VA Healthcare System to enroll in our observational study. Between December 2002 and March 2003, 234 individuals were admitted to one of two adult acute psychiatric wards: one ward cared predominantly for patients with combined psychiatric and substance abuse disorders, while the other was a general acute psychiatric ward. All 234 patients were approached upon admission for participation. A total of 46 were unable to consent due to involuntary admission or conservator status and 59 declined participation. Ultimately, 129 patients consented and completed the study.

All but three of the 129 patients were male, and the average age was 48.9 years with a standard deviation of 8.9. The average education level was 13.4 years with a standard deviation of 2.1. The most represented racial group was Caucasian (50.3%), followed by African American (33.3%), Hispanic (9.3%), and combined Asian American, Native American, and “other” (7.0%). A psychotic type diagnosis was given to 40.3% of the patients, 57.4% had an affect or anxiety disorder, and 66.6% of the sample had a substance abuse disorder (some patients had dual diagnoses). All diagnoses were in accordance with the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, as conferred by the treating physician.


Experienced interviewers, who were either physicians or medical students, performed research assessments. After informed consent was given, the 129 participants were interviewed using a standardized questionnaire regarding socio-demographic data known to be risks for HBV and HCV exposure. No information was recorded that compromised patient anonymity, and researchers were protected from outside infringement of said confidence by a Certificate of Confidentiality from the National Institute of Health. The risk behaviors assessed were according to a Veteran’s Affairs screening tool used by physicians to evaluate patient’s risk and included consideration of sexual behavior, intravenous drug use (IVDU), intranasal drug use, tattoos, piercing, acupuncture, organ transplantation, health care employment, being stuck with bloody objects, close contact with a HBV or HCV patient, previous incarceration longer than 3 days, and history of homelessness.

Each consenting subject’s medical record was checked for prior HBV and HCV serological screening. Due to study design, immunization history was not available. If the subject had not been screened for HBV or HCV in the past 6 months, a trained phlebotomist collected a blood specimen, which was screened for HBV surface antigen (HBsAg), HBV surface antibody (HBsAb), HBV total core antibodies (HBcAb), and HCV antibody (HCAb). Patients found to have positive HBsAg or HCAb as part of this study were referred for appropriate care.

The observed prevalence of exposure to HBV and HCV was calculated for the entire sample. A positive HBsAg, HBsAb, or HBcAb result were all considered markers for HBV exposure. A positive HCAb was considered the marker for HCV exposure.

Bivariate relationships between viral exposure and risk behaviors, as assessed by the standard interview items, were examined using chi-square analysis for comparison of proportions and Student’s t test for comparison of means. Similarly, such relationships were also examined between diagnostic (psychiatric or substance-related) categories and viral exposure.

All sample data, including patient demographics, diagnoses, and risk factors may be seen in Table 1. Significant differences between exposed and unexposed groups were defined as a two-tailed P-value less than 0.05 and are indicated in Table 1 by an asterisk.
Table 1

Sample demographics table









Age (years)





   Standard deviation




Male (%)




Ethnicity (%)





   African American












Education (years)





   Standard deviation




Risk factors (%)

   Sex with intravenous drug user




   Sex with prostitute




   Sex bartering








   Intranasal drug use
















   Organ transplantation




   Health care worker




   Bloody object contact




   HBV/HCV personal contact




   Previous incarceration








Diagnosis (%)

   Affect and substance abuse




   Psychosis and substance abuse




   Psychosis and affect




   Affect only




   Psychosis only




   Substance abuse only




* Significant difference between exposed and unexposed group, two-tailed P < 0.05

IVDU intravenous drug use

Processing of specimens and laboratory procedures

Hepatitis testing was performed locally at the West Los Angeles VA Healthcare facility. All specimens were collected in a marble top tube, spun down, and refrigerated at 2–8°C. No sample was held for longer than 3 weeks prior to testing in the immunology laboratory at the West Los Angeles VA Hospital. HBsAg, HBsAb, total (IgM + IgG) HBcAb, and HCAb were measured using Abbott EIA in which samples are plated using bead sandwich technology and run through FPC (flexible pipetting center) or PPC (parallel processing center) instruments.


Of the 129 patients, 40 (31%) were found to be exposed to HBV. Of these 40 patients, 21 (16%) were positive for HBcAb, of which one patient was positive for HBsAg (indicating active infection), ten were HBsAb positive, and 19 (14.7%) were solely HBsAb positive. Four of the 40 patients were unaware of being at risk for prior or present HBV exposure. Patients exposed to HBV were equally distributed amongst the ethnic groups (P = 0.27). The average age of HBV-exposed patients was 50 years, which was not significantly different from the rest of the sample (P = 0.38). The average number of years of education completed by HBV-exposed patients was 13.3, which also was not significantly different from the rest of the sample (P = 0.75).

With regard to HCV, 49 of 129 patients (38%) had blood samples positive for HCAb, thus indicating exposure to HCV. Eight of these 49 patients were unaware of being at risk for prior or present HCV exposure. Individuals testing positive for HCAb were equally distributed amongst the ethnic groups, demonstrating that, similar to HBV-exposed patients, there is no relationship between ethnicity and exposure (P = 0.93). The average age of HCV-exposed patients was 49 years, which was not significantly different from the rest of the sample (P = 0.20). The average number of years of education completed by HCV-exposed patients in this sample was 13.1, which also was not significantly different from the rest of the sample (P = 0.22).

Coexposure was also assessed in our sample, yielding 26 patients (20%) with positive serology for both HCAb and HBsAg, HBsAb, or HBcAb.

Among patients exposed to HBV, the most common diagnostic groups were psychotic disorder with comorbid substance abuse and mood or anxiety disorder with substance abuse (Fig. 1). Each of these diagnostic groups was found in 37.5% of HBV-exposed patients, compared to 15.7 and 44.9%, respectively, in the unexposed patients. Of these two diagnostic groups, only psychotic disorder with comorbid substance abuse was significantly more associated with HBV exposure (P = 0.006 vs. 0.4).
Fig. 1

Risk factors significantly associated with HBV exposure (P < 0.05). IVDU intravenous drug use

The most common diagnostic group among HCV-exposed patients was a mood or anxiety disorder with a comorbid substance abuse disorder, seen in 38.8% of those patients (Fig. 2). However, this percentage was not significantly different than the 33.8% seen in the unexposed group (P = 0.56). The next most common diagnostic group among HCV-exposed patients was psychosis with a comorbid substance abuse disorder, seen in 28.6%, although this too was not a significant difference from the 18.8% seen in the unexposed patients (P = 0.19). No diagnostic group was significantly more common in the HCV-exposed patients than in the unexposed.
Fig. 2

Risk factors significantly associated with HCV exposure (P < 0.05). IVDU intravenous drug use

To better understand the route of transmission and the presence of particular risk factors in patients with severe psychiatric illness, we analyzed a battery of 14 potential risk factors. IVDU (P = 0.0006) and previous incarceration (P = 0.04) were both significantly more associated as risk factors in HBV-exposed patients compared to unexposed patients (Fig. 3).
Fig. 3

Diagnoses in HBV-exposed versus unexposed

In HCV-exposed patients, IVDU (P < 0.0001), intranasal drug use (P = 0.0002), tattoos (P = 0.0002), close contact with a HBV or HCV infected individuals (P = 0.004), exposure to a bloody object (P = 0.01), intercourse with an intravenous drug user (P = 0.01), healthcare employment (P = 0.03), and previous incarceration (P = 0.03) were all significant risk factors compared to unexposed patients (Fig. 4). From this analysis, IVDU was the risk behavior most clearly associated with both exposures.
Fig. 4

Diagnoses in HCV-exposed versus unexposed


This study demonstrates that HBV and HCV exposure are approximately 6 and 20 times more common, respectively, in our sample of inpatient psychiatric VA patients than in estimates of the general US population. To further objectively assess the effect of psychiatric illness, we also compared our sample to VA patients overall since the latter has a greater frequency of risk factors for blood-borne infection than the general population. Despite the already increased prevalence of HBV and HCV exposure in general VA patients, our sample had an increased rate of exposure compared to VA patients in general [8, 10]. The increased HBV and HCV exposure found in our study is comparable to that in other studies of patients with psychiatric illness.

In a study of 535 psychiatric patients chronically hospitalized at Oregon State Hospital, Meyer et al. found 20.3% seropositivity for HCAb and 24.7% for HbcAb [3]. A large NIMH/VA multi-state survey of 931 patients with severe mental illness treated in outpatient or inpatient facilities presented similar exposure rates of 23.4% with HBcAb and 19.6% with HCAb [4]. A study by Dinwiddie et al. of 1,556 psychiatric patients in the Chicago area found 27.8% with HBsAb and 8.5% with HCV RNA [5]. Lastly, The Veterans Healthcare Administration Northwest Network conducted a retrospective study in psychiatric patients and found that 21.6% of 25,080 individuals were HCAb positive [6]. In Japan, Sawayama et al. studied 196 neuropsychiatric patients and found that 10.2% of the patients were HCAb positive compared to 1.5% of controls and 44.4% were HBcAb positive compared to 20.5% of controls [7]. Nakamura et al. studied 1,193 psychiatric patients in Japan and found 9.1% positive for HCAb compared to an estimated 1.2% HCAb positivity rate in the general Japanese population [9]. Despite differences in serologic testing parameters and specific psychiatric diagnoses, all of these studies suggest that patients with psychiatric disease are at increased risk of HBV and HCV exposure. In our study, several risk factors were significantly associated with HBV or HCV exposure. Substance abuse, particularly IVDU, accounted for the greatest proportion of individuals who were found to have either HBV or HCV exposure. Another significantly associated risk factor for HBV exposure included previous incarceration, whereas for HCV exposure, intranasal drug use, tattoos, close contact with HBV or HCV infected individuals, exposure to a bloody object, intercourse with an intravenous drug user, healthcare employment, and previous incarceration were all also significantly associated. The majority of these risk factors have been verified as such in other studies of blood-borne viral infections [6, 7, 1217].

Interestingly, patients with schizophrenia alone (without other psychiatric or drug abuse diagnosis) appeared to constitute a smaller percentage of the HBV or HCV-exposed group than they did of the total sample (P = 0.1 and P = 0.01, respectively). This protective effect is likely due to the fact that many psychotic patients live marginalized from society and thus protected from certain high risk behaviors. Furthermore, psychotic patients would be expected to be at lower risk for sexually-transmitted viral infection than the general VA population because of their relatively low rate of sexual activity. Patients with combined schizophrenia and substance abuse diagnoses, however, had relatively high rates of HBV and HCV exposure (P = 0.006 and P = 0.34, respectively). This is consistent with the observation that patients with schizophrenia have poor judgment and are often neglectful of self-care, which may lead to relatively greater probability of exposure when immersed in risk factors or individuals with risk factors such as IV drug abuse as compared to those with psychiatric illness or substance abuse alone.

A noteworthy problem with HBV and HCV in psychiatric patients, aside from the increased prevalence of exposure, is the issue of treatment of patients who proceed to chronic viral hepatitis. Unfortunately, interferon, the standard first-line treatment for HCV and part of the treatment for chronic HBV infection, is documented to have significant psychiatric side effects, which has been and is an obstacle to receiving treatment in many patients [23, 24]. However, use of interferon is generally safe and approved by the NIH and the VA in patients with psychiatric illness, unless their psychiatric illness is uncontrolled or if they have suicidal or homicidal ideations [2528]. In addition to adequate management of the psychiatric illness, cessation of exposure risk-conferring substance abuse is important, as it has been shown that patients may otherwise not be offered treatment by clinicians [24]. Another treatment issue of interest in psychiatric patients is if hepatically-cleared psychotropic medications can be used safely in patients with hepatitis. In HBV and HCV infection, any medication that is indicated can be given; however, for drugs that have hepatic clearance, serum levels should be monitored. A 2003 study suggested that patients with HCV infection can be safely prescribed valproic acid—a commonly prescribed hepatically-cleared medication for agitation and manic behavior as well as an augmentation strategy for psychosis—but liver function tests and valproic acid levels need to be followed more rigorously [29]. In this context, medications that are renally excreted, such as lithium or neurontin, may be another or better choice as mood stabilizing agents, particularly for patients with known compromised liver function [30].

Given the increased risk of spread of infection and the possible need for medications with adverse psychotropic effects or hepatic clearance, routine screening for HBV and HCV in severely mentally ill patients is advocated, and HBV as well as hepatitis A immunization is recommended. Although inpatient stays are short, it is possible to prescribe the first immunization during a hospitalization.

Three limitations of our study include: (1) the VA population is distinct from the general US population, including a greater baseline prevalence of risk behaviors and lifestyles, demographic differences, and male predominance, and thus results from our sample may not be accurately extrapolated to the general population with psychiatric illness; (2) our sample size was small; thus, a larger study could reduce error and increase reliability, and further, it may permit an analysis of the exposure prevalence in each diagnostic group individually; and (3) there is no perfect method of determining HBV exposure; our criteria for HBV exposure were based on our prior experience of infrequent immunization in psychiatric patients, the potential for HBcAb to fall to undetectable levels over time, and the methods from similar studies [35, 7]. Had we not included the HBsAb only positive patients, our rate of HBV exposure prevalence would have been slightly lower than that reported in other psychiatric populations (16%), whereas the prevalence calculated by the criteria we ultimately used (31%) was slightly higher than these reported values; thus, it is possible that the true prevalence of HBV exposure in our sample may have been between 16.3 and 31%.

Our future work will focus on further exploring the association between specific psychiatric illnesses and HBV and HCV exposure, as well as the particular risk factors in each diagnostic group. Our aim is to delineate ways in which we can educate our patients and prevent further spread of these infectious agents.


We thank Dr. Neshan Tabibian for his help and critiques in writing this paper from a gastroenterological perspective.

Copyright information

© Springer Science+Business Media, LLC 2007