Published Online: April 17, 2014
Vinutha Vijayadeva, PhD, MPH; Philip R. Spradling, MD; Anne C. Moorman, BSN, MPH; Loralee B. Rupp, MS; Mei Lu, PhD; Stuart C. Gordon, MD; Emily Henkle, PhD, MPH; Joseph A. Boscarino, PhD, MPH; Eyasu H. Teshale, MD; and Cynthia Nakasato, MD
Objectives: Asian and Pacific Islanders (APIs) constitute less than 6% of the US population, but account for more than half of Americans with chronic hepatitis B virus (HBV) infection. We sought to examine the effect of country of origin on HBV testing and chronic HBV infection prevalence among APIs.
Methods: We analyzed demographic and clinical data collected for adults from Kaiser Permanente Hawaii with 1 or more healthcare encounters during 2006 to 2008, 12 months or more of follow-up before 2009, and no HBV-related diagnosis within 6 months of enrollment. Persons who received a test and a positive test result for HBV surface antigen or HBV DNA were classified “tested” and with “chronic HBV infection,” respectively.
Results: Of 92,687 eligible APIs, 53,573 (58%) had country-of-origin data available. Among those, 41,263 were US born; 28.3% were tested; and 1.8% of those tested had chronic HBV infection. Of 12,310 foreign-born APIs, 30.5% were tested and 7.4% of those tested had chronic HBV infection. Foreignborn APIs had higher odds of being tested (odds ratio [OR] = 1.15) and testing positive (OR = 4.18) compared with US-born APIs. Persons with 2 or more abnormal tests for alanine aminotransferase (ALT) levels had higher odds of getting tested (OR = 6.12) and of testing positive (OR = 1.86) compared with persons with other ALT levels.
Conclusions: Less than one-third of this managed care API population (29% of 53,573) was tested, yet the prevalence of chronic HBV infection (3.2%) was 12 times higher than that of the general US population. These findings underscore the importance of adherence to HBV testing guidelines to identify persons with infection so they may be linked to care.
Am J Manag Care. 2014;20(4):e98-e104
In the United States, between 800,000 and 2 million people live with chronic hepatitis B virus (HBV) infection and 2000 to 4000 deaths are attributed to HBV infection annually.1-4 Cirrhosis and hepatocellular carcinoma (HCC) are 2 major long-term complications of chronic HBV infection that significantly increase morbidity and mortality. Asian and Pacific Islanders (APIs) make up less than 6% of the US population but account for more than half of Americans living with chronic HBV infection5,6; hepatocellular carcinoma (HCC) rates are highest among APIs, and HCC is a leading cause of cancer deaths in this population.7-9 Most APIs with chronic HBV infection acquire their infection at birth through mother-to-child transmission or during childhood.10
In 2009, an estimated 1.32 million foreign-born individuals with chronic HBV infection were thought to be living in the United States, and 58% of those persons migrated from Asia, where HBV infection is highly endemic.2 The population makeup of Hawaii, the 50th US state, includes 47% persons of Asian descent and 29% persons of Native Hawaiian/Pacific Islander descent.5,11,12 Among Hawaii’s 212,229 foreign-born residents in 2000, 73% were born in countries with endemic HBV infection, including China, Vietnam, the Philippines, and elsewhere among the South Pacific Islands.13
In Hawaii, Kaiser Permanente (KPHI) has approximately 220,000 members and is the largest health maintenance organization (HMO) providing care in the state.14 Commensurate with the high representation of APIs living in Hawaii, APIs constitute approximately half of KPHI’s membership. Given the unique demographic conditions in Hawaii relative to other US states, and our access to demographic and clinical data collected from patient encounters, we sought to examine the effect of country of origin on HBV infection testing and prevalence among APIs at KPHI.
Persons eligible for the study included those who: (1) were aged 18 years or older; (2) had at least 1 KPHI clinical encounter at any time during January 1, 2006, to December 31, 2008; and (3) completed at least 12 months of continuous follow-up any time before 2009. To examine HBV infection testing and prevalence among persons without previously diagnosed infection, we excluded those with any HBV-related International Classification of Diseases, Ninth Revision (ICD-9) diagnosis code within 6 months of their first KPHI clinical encounter. Retrospective electronic data were available through January 1, 1998.
Patient data were collected from electronic medical records and included age (as of last encounter before December 31, 2008), gender, race/ethnicity, annual income (derived from census tract data based on geocode), and serum alanine aminotransferase (ALT) level. Abnormal ALT was defined as a level higher than the lab-specific upper limit of normal value (range: 3-63 IU/L). Data were collected from the date of the patient’s earliest health plan enrollment through the last health plan encounter, ending December 31, 2008. Methods were identical to those used in an earlier analysis of hepatitis testing and prevalence among 1.2 million persons in 4 healthcare organizations including KPHI, but which lacked country-of-origin data.15 Briefly, persons were classified as “tested” if they had at least 1 test performed for HBV surface antigen (HBsAg) or a qualitative or quantitative test for HBV DNA; those classified with “HBV infection” had at least 1 positive result for any of these tests. The number of persons with infection divided by the number of persons tested was calculated to determine HBV infection prevalence. APIs were classified as foreign born (birth in Asian or Pacific Island countries) or US born (birth in the United States or Canada), based on the self-reported country of origin.
This study was reviewed and approved by the KPHI Institutional Review Board and followed guidelines of the US Department of Health and Human Services for protection of human subjects.
Statistical analysis was performed using SAS, version 9.2 (SAS Institute Inc, Cary, North Carolina). Chi-squared test (χ²) was used to ascertain differences in characteristics between US-born and foreign-born APIs.
After exclusion of 198 persons with an HBV-related ICD-9 diagnosis code within 6 months of their first KPHI clinical encounter, a total of 189,550 KPHI adults met the study inclusion criteria. Among eligible adults, 92,687 (48.9%) were APIs. Information on country of origin was available for 53,573 persons: 12,310 (23%) were born in Asian and Pacific Island countries and 41,263 (77%) were born in the United States (n = 41,235) or Canada (n = 28). Among all APIs (N = 53,575), 29% (n = 15,439) were tested for HBV infection and, among those tested, 3.2% (n = 494) were infected.
The principal demographic and clinical characteristics of the US and foreign-born APIs are presented in Table 1. Compared with foreign-born APIs, US-born APIs were more likely to be younger than age 30 years or older than age 69 years. Foreign-born APIs were more likely to be female, to have lower annual incomes, and to have had at least 2 abnormal ALT test results before undergoing HBV infection testing. A greater proportion of foreign-born APIs were tested and a greater proportion had HBV infection compared with US-born APIs (tested: 3752 [30.5%] vs 11,687 [28.3%]; tested positive: 279 [7.4%] vs 215 [1.8%]). Among APIs with at least 2 abnormal ALT test results, 46% of foreign-born and 41% of US-born were tested for HBV infection, 8.6% of foreign born and 1.9% of US-born were infected.
Table 2 show the important factors associated with odds of getting tested and testing positive for HBV infection. Individuals aged 30 to 39 years had higher odds of getting tested and of testing positive than other age groups. Females had higher odds of getting tested than males. Repeating the analysis excluding the pregnant women who were ever tested for HBsAg (n = 7092), the odds ratio (OR) for females attenuated from 3.26 to 1.13 (data not shown). Foreign-born APIs had higher odds of being tested (OR = 1.15) and testing positive (OR = 4.18) compared with USborn APIs. Persons with at least 2 or more abnormal ALT tests had higher odds of getting tested (OR = 6.12) and testing positive (OR = 1.86) compared with other ALT levels. Repeating the analysis with pregnant women excluding OR for foreign-born APIs and ALT levels did not produce results that differed much.
Table 3 shows the proportion of foreign-born APIs who were tested and infected with HBV by country of birth. Among the 12,310 foreign-born APIs, 11,058 (89.8%) were from countries with high HBV endemicity (ie, HBsAg prevalence ≥8%)16; of these, 29.9% were tested for HBV infection and 7.9% of those were positive for HBV infection. The remaining 10.2% of foreign-born APIs were from countries with intermediate HBV endemicity (ie, HBsAg prevalence 2%-7%)15; of these, 29.8% were tested and 2.7% were positive for HBV infection. Among all foreign-born APIs tested, the proportion of those infected ranged from 2.7% (Japan) to 12.3% (China). Approximately 50% of the foreign-born APIs were from the Philippines; of these, 29.9% were tested and 5.5% were positive for HBV infection.
In this study of HBV testing and infection prevalence among APIs in a large HMO, we found that less than onethird of persons without a HBV diagnosis code at enrollment were tested for HBV infection. Among APIs who were tested, the prevalence of HBV infection (3.2%) was almost 12 times higher than that of the general US population (HBsAg prevalence 0.27%). Among foreign-born APIs who were tested, 90% of whom were born in countries with high HBV endemicity, infection prevalence was largely commensurate with that of their country of origin (7.5% overall). These findings support US Centers for Disease Control and Prevention (CDC) guidelines that recommend HBV infection screening for pregnant women, persons with high-risk behavior, and those born in Asia, Africa, and other geographic areas where HBsAg prevalence exceeds 2%.3
Even among US-born APIs, the prevalence of HBV infection was high (1.8%) relative to the general US population. Maternal country of birth information among USborn APIs might have permitted differentiation of first generation “API Americans” from those whose families had resided in the United States for multiple generations, and may have permitted further differentiation of infection prevalence; however, we did not have access to such information.
PDF is available on the last page.