Incidence and prevalence of hepatitis B in patients with diabetes mellitus in the UK: A population-based cohort study using the UK Clinical Practice Research Datalink

1 INTRODUCTION

Hepatitis B virus (HBV) infection is a global public health issue with increasing complication and mortality rates.1 The prevalence of Hep B ranges from 5% to 10% in adults in sub-Saharan Africa and East Asia to <1% of in the population of Western Europe and North America.2 Acute infections progress to chronic infections in 80%-90% of children infected in their first year of life, 30%-50% of children infected before the age of 6 and <5% of people infected as adults. About 20%-30% of chronically infected adults develop cirrhosis and/or liver cancer.2 As of 2015, 185 countries vaccinate infants against Hep B as part of their national immunization programmes and 83% of children received the complete Hep B vaccine course.3 Nonetheless, global disease burden estimates indicate that nearly 240 million people are chronically infected with Hep B,2 suggesting that innovative strategies to address HBV infection are needed.4 Although the prevalence of HBV in the UK is low, an increasing trend has been recently observed.5 As HBV is stable for long periods of time on surfaces and is highly transmissible,6 many Hep B outbreaks have occurred particularly in adults with diabetes who receive assisted blood glucose monitoring.7

The Center for Disease Control and Prevention in the USA has received many reports of Hep B outbreaks in long-term care facilities, including nursing homes and assisted-living facilities since 1996. These outbreaks prompted the Hepatitis Vaccines Work Group of the Advisory Committee on Immunization Practices (ACIP) to evaluate the risk of HBV infection among adults diagnosed with diabetes mellitus (DM).7

Several studies were carried out to understand the burden of Hep B among diabetics. Data from the National Health and Nutrition Examination Survey (NHANES) showed a higher prevalence of HBV infection among noninstitutionalized individuals diagnosed with diabetes as compared to nondiabetics.8, 9 Data from the Emerging Infections Program (EIP) database showed that diabetes was independently associated with an increased risk for acute Hep B among adults without HBV risk behaviours.10 The vaccination of adult diabetics (20-59 years) with Hep B vaccine was found to be cost-effective.11 Therefore, ACIP recommends that all individuals aged 19-59 should be vaccinated with Hep B vaccine “as soon as feasible after a diagnosis of diabetes”.7

To our knowledge, studies investigating the incidence and prevalence of Hep B in DM patients, rather than the general population, are scarce outside the United States. According to WHO estimates, the global prevalence of diabetes was 9% in 2014.12 UK data published in 2013 reported a significant increase in the incidence and prevalence of type 2 DM.5 It has been further projected that by 2030, the total number of adults with diabetes will increase to 9.5% in England.13

This study was therefore undertaken to assess the incidence and prevalence of HBV diagnosis in DM and non-DM cohorts in the UK using the Clinical Practice Research Datalink (CPRD) database and hospitalization records in England.

2 MATERIALS AND METHODS

3 RESULTS

3.1 Study population

From a total of 7 712 043 study eligible subjects, the DM cohort included 160 760 patients. From the eligible 7 551 283 subjects, the non-DM cohort included a random sample of 4 679 010 subjects (Figure 1). The mean ages were 54.4 and 32.0 years, 57.20% and 50.14% were male, and ethnicity data were available for 62% and 50% of subjects in the DM and non-DM cohorts, respectively (Table 1). In the DM cohort, 95.85% of subjects had type 2 diabetes, 4.14% had type 1 diabetes, and 0.01% were nonspecified. For the sensitivity analyses, the DM cohort of 152 376 patients was matched with 1 520 836 subjects in the non-DM cohort.

4 DISCUSSION

This retrospective database study is the first one evaluating the incidence and prevalence of Hep B and Hep B-related hospitalization rates among DM and non-DM populations in the UK using the CPRD data source. Studies from the United States have reported an increased risk for HBV among persons with diabetes regardless of HBV risk behaviours.7, 8, 10 HBV transmission among diabetics, especially in older populations, has resulted from lapses in infection control associated with blood glucose monitoring.18 Another study showed that patients with diabetes were involved in HBV outbreaks as a consequence of the habit of performing capillary blood sampling using nondisposable multipatient devices.19 Furthermore, an increased risk of HBV exposure has been reported in the United States.8, 10 A more severe chronic HBV infection and exacerbation of DM-associated liver disease in the DM population has been shown in several studies.10, 20-23 As a result, ACIP has recommended that individuals aged 20-59 years old should be vaccinated against HBV as soon as possible after the diagnosis of DM.7

In the present study, the risk for new-onset incident Hep B among patients with DM was not different from that in non-DM subjects after adjustment for age at inclusion, gender, geographical region of the GP, year of the inclusion in the cohort and number of consultations during the 12 months prior to inclusion. The IR of Hep B in the DM cohort was highest in the 0-20 years age group. Available literature on laboratory confirmed Hep B has shown that overall the incidence of Hep B in the UK is low.24, 25 In an earlier study conducted in the UK, the IR of Hep B among population ≥15 years of age was reported to be 1.6 per 100 000 person-year.26 However, the authors did not consider diabetes as a risk factor for Hep B infection. The IRR estimate was not adjusted for ethnicity, which is only partially documented in CPRD and is not recorded for 50% of the study subjects. The risk for type 2 diabetes is known to be higher in migrant populations from South-East Asia settling in western countries than in the local population.27, 28 In addition, Hep B infection and transmission are geographically more prevalent in the Asian regions.29, 30 A different proportion of Hep B cases in subjects of South-East Asian ethnicity among diabetics may have impacted the estimation of the IRR, as ethnicity may have been a potential confounder in the comparison between DM and non-DM cohorts.

The observed prevalence of Hep B during the follow-up period was higher in the DM cohort compared to the non-DM cohort in all age groups. Males had a higher prevalence of Hep B than females in both study cohorts. These findings are in line with previous studies where a higher HBV-diabetes prevalence ratio was observed for men.8, 26 Likewise, a study conducted in Turkey revealed higher prevalence of occult HBV infections in diabetic patients.31 The range of prevalence observed in this study is consistent with published prevalence estimates (HPE data: 0.1% and 0.5% of the UK population).32, 33 A higher Hep B-associated hospitalization IR was observed in the DM than that in the non-DM cohort in this study.

A similar study in the United States reported a higher prevalence of Hep B among adults with DM than in those without DM.18 Another study reported a 1.6-times (95% CI 1.3-1.9) higher prevalence of HBV infection among diabetics than nondiabetics.8 Differences in the data reported in our study and in the US studies can be attributed to differences in the study designs. NHANES: National health and nutrition examination survey is a program of studies designed to assess the health and nutritional status of adults and children in the United States and includes a representative sample of the noninstitutionalized US population. The survey is unique in that it combines interviews and physical examinations data; therefore, Hep B is laboratory confirmed. The Emerging Infectious program is a national resource for surveillance, prevention and control of emerging infectious diseases used to study specific infectious diseases and, similarly to NHANES, includes laboratory confirmed Hep B cases while in the CPRD laboratory confirmation of Hep B is not routinely reported.8, 10

The main strength of this study lies in the use of the CPRD database, a large population-based electronic medical record data set which can link primary and hospital data and provide a good representation of the UK population registered with a GP practice.14 Approximately half of the GP practices allowed linkage with the HES inpatient database. However, CPRD does not include data or recorded variables on populations at higher risk of Hep B, for example prisoners, injectable illegal drug users, homeless and temporary migrant workers. The study assumed that these populations would probably be equally distributed between the DM and non-DM cohorts, with minimal potential for differential information or selection bias.

As Hep B infection screening is often performed confidentially during sexually transmitted disease consultations or at walk-in clinics, the GP records may not have complete information on the dates of Hep B infection or the results and serological status. Therefore, the onset date of first diagnosis may be earlier than the first date of recorded diagnosis of Hep B in the GP medical records. We attempted to minimize this bias in the primary design and analyses by considering only new-onset DM during the study period together with new-onset Hep B diagnosis after the DM onset, but this would have reduced the number of eligible participants. Furthermore, Hep B laboratory testing results may not be systematically recorded in the CPRD, and negative Hep B screening results may be less likely to be recorded than positive results. In addition, some of the READ codes included in the case definition may require further validation on their predictive value when recorded for a single event in the patient's GP medical history records. However, we assumed that this was an unlikely differential between DM and non-DM subjects, unless DM subjects were subject to more intensive screening or testing than non-DM patients. HES inpatient subset data linked to CPRD were limited to England; therefore, outcomes involving hospital data were restricted to the English CPRD practices subset.

Hepatitis B vaccination coverage in the UK is very low, as until recently it was only recommended to those UK populations which are at increased risk of infection. As of 2017, Hep B vaccination will be introduced in the infant vaccination schedules through a combination vaccine.34 There is no recommendation for Hep B vaccination of DM patients. The likelihood of Hep B vaccination was assumed to be similarly distributed between DM and non-DM patients and was therefore not considered in the analyses.

In conclusion, we found that the risk of new Hep B diagnosis after DM onset was no different to the risk among patients without DM. However, prevalence of Hep B during the follow-up period was higher in subjects who developed DM in all age groups compared to those subjects without DM. The Hep B-associated hospitalization rate was also higher in the DM cohort than in the non-DM cohort.

AUTHORS CONTRIBUTIONS

CM, LDM, GLCF and YEH were involved in the conception or the design of the study. CM, YF, GLCF and YEH participated in the data collection. GLCF and YEH performed the study/project and participated. TVS contributed with materials or reagents. All authors had access to the data, participated in the analysis and interpretation of the data, were involved with developing this manuscript, gave final approval before submission and are accountable for all aspects of the work.

CONFLICT OF INTEREST

CM and LDM are employees of GSK and hold stock options/restricted shares from the sponsoring company. AD is an employee of GSK. GLCF was an employee of GSK at the time of the study. YF was an employee for GSK when the study started, and then she changed to a statistical consultant for GSK continuing to support this study from statistical point of view since 1 September 2014. YEH has declared no conflict of interest. TVS reports a research grant for the epidemiological study of COPD from GSK and personal fees from NovoNordisk and Sanofi during the conduct of the study.

DISCLOSURE

This study is based in part on data from the CPRD obtained under licence from the UK Medicines and Healthcare Products Regulatory Agency. However, the interpretation and conclusions contained in this report are those of the authors alone.