Acute viral hepatitis morbidity and mortality associated with hepatitis E virus infection: Uzbekistan surveillance data.

Abstract

In Uzbekistan, routine serologic testing has not been available to differentiate etiologies of acute viral hepatitis (AVH). To determine the age groups most affected by hepatitis E virus (HEV) during documented AVH epidemics, trends in AVH-associated mortality rate (MR) per 100,000 over a 15-year period and reported incidence of AVH over a 35-year period were examined.
Reported AVH incidence data from 1971 to 2005 and AVH-associated mortality data from 1981 to 1995 were examined. Serologic markers for infection with hepatitis viruses A, B, D, and E were determined from a sample of hospitalized patients with AVH from an epidemic period (1987) and from a sample of pregnant women with AVH from a non-epidemic period (1992).
Two multi-year AVH outbreaks were identified: one during 1975-1976, and one during 1985-1987. During 1985-1987, AVH-associated MRs were 12.3-17.8 per 100,000 for the general population. Highest AVH-associated MRs occurred among children in the first 3 years of life (40-190 per 100,000) and among women aged 20-29 (15-21 per 100,000). During 1988-1995 when reported AVH morbidity was much lower in the general population, AVH-associated MRs were markedly lower among these same age groups. In 1988, AVH-associated MRs were higher in rural (21 per 100,000) than in urban (8 per 100,000) populations (RR 2.6; 95% CI 1.16-5.93; p < 0.05). Serologic evidence of acute HEV infection was found in 280 of 396 (71%) patients with AVH in 1987 and 12 of 99 (12%) pregnant patients with AVH in 1992.
In the absence of the availability of confirmatory testing, inferences regarding probable hepatitis epidemic etiologies can sometimes be made using surveillance data, comparing AVH incidence with AVH-associated mortality with an eye to population-based viral hepatitis control measures. Data presented here implicate HEV as the probable etiology of high mortality observed in pregnant women and in children less than 3 years of age in Uzbekistan during 1985-1987. High mortality among pregnant women but not among children less than 3 years has been observed in previous descriptions of epidemic hepatitis E. The high mortality among younger children observed in an AVH outbreak associated with hepatitis E merits corroboration in future outbreaks.

Full text

Page 1

BioMed Central
Page 1 of 9
(page number not for citation purposes)
BMC Infectious Diseases
Open Access
Research article
Acute viral hepatitis morbidity and mortality associated with
hepatitis E virus infection: Uzbekistan surveillance data
Makhmudkhan B Sharapov1, Michael O Favorov2,3,4, Tatiana L Yashina5,
Matthew S Brown3, Gennady G Onischenko6, Harold S Margolis2,7 and
Terence L Chorba*8
Address: 1Tashkent Pediatric Medical Institute and Central Asia Epidemiology Network, Ministry of Health, Tashkent, Republic of Uzbekistan,
2Division of Viral Hepatitis (World Health Organization Collaborating Center for Research and Reference in Viral Hepatitis), National Center for
HIV/AIDS, Viral Hepatitis, STD and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia, USA, 3Coordinating Office for
Global Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA, 4Division of Translational Research, International Vaccine
Institute, Seoul, Korea, 5Specialty Laboratories, Santa Monica, California, USA, 6Federal Service for Surveillance on Consumer Rights and
Wellbeing, Ministry of Health and Social Development of the Russian Federation, Moscow, Russia, 7Pediatric Dengue Vaccine Initiative Program,
International Vaccine Institute, Seoul, Korea and 8Office of the Director, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention,
Centers for Disease Control and Prevention, Atlanta, Georgia, USA
Email: Makhmudkhan B Sharapov - mbsharapov@yandex.ru; Michael O Favorov - mfavorov@ivi.int;
Tatiana L Yashina - tyashina@specialtylabs.com; Matthew S Brown - zjc5@cdc.gov; Gennady G Onischenko - depart@gsen.ru;
Harold S Margolis - hsmargolis@pdvi.org; Terence L Chorba* - tlc2@cdc.gov
* Corresponding author
Abstract
Background: In Uzbekistan, routine serologic testing has not been available to differentiate
etiologies of acute viral hepatitis (AVH). To determine the age groups most affected by hepatitis E
virus (HEV) during documented AVH epidemics, trends in AVH-associated mortality rate (MR) per
100,000 over a 15-year period and reported incidence of AVH over a 35-year period were
examined.
Methods: Reported AVH incidence data from 1971 to 2005 and AVH-associated mortality data
from 1981 to 1995 were examined. Serologic markers for infection with hepatitis viruses A, B, D,
and E were determined from a sample of hospitalized patients with AVH from an epidemic period
(1987) and from a sample of pregnant women with AVH from a non-epidemic period (1992).
Results: Two multi-year AVH outbreaks were identified: one during 1975–1976, and one during
1985–1987. During 1985–1987, AVH-associated MRs were 12.3–17.8 per 100,000 for the general
population. Highest AVH-associated MRs occurred among children in the first 3 years of life (40–
190 per 100,000) and among women aged 20–29 (15–21 per 100,000). During 1988–1995 when
reported AVH morbidity was much lower in the general population, AVH-associated MRs were
markedly lower among these same age groups. In 1988, AVH-associated MRs were higher in rural
(21 per 100,000) than in urban (8 per 100,000) populations (RR 2.6; 95% CI 1.16–5.93; p < 0.05).
Serologic evidence of acute HEV infection was found in 280 of 396 (71%) patients with AVH in 1987
and 12 of 99 (12%) pregnant patients with AVH in 1992.
Conclusion: In the absence of the availability of confirmatory testing, inferences regarding
probable hepatitis epidemic etiologies can sometimes be made using surveillance data, comparing
Published: 25 March 2009
BMC Infectious Diseases 2009, 9:35doi:10.1186/1471-2334-9-35
Received: 19 July 2008
Accepted: 25 March 2009
This article is available from: http://www.biomedcentral.com/1471-2334/9/35
© 2009 Sharapov et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2

BMC Infectious Diseases 2009, 9:35http://www.biomedcentral.com/1471-2334/9/35
Page 2 of 9
(page number not for citation purposes)
AVH incidence with AVH-associated mortality with an eye to population-based viral hepatitis
control measures. Data presented here implicate HEV as the probable etiology of high mortality
observed in pregnant women and in children less than 3 years of age in Uzbekistan during 1985–
1987. High mortality among pregnant women but not among children less than 3 years has been
observed in previous descriptions of epidemic hepatitis E. The high mortality among younger
children observed in an AVH outbreak associated with hepatitis E merits corroboration in future
outbreaks.
Background
In Uzbekistan, there is a high incidence of infection with
hepatitis viruses including the viruses of hepatitis A
(HAV) [1], hepatitis B (HBV) [2], hepatitis C (HCV) [3],
hepatitis D (HDV) [4], and hepatitis E (HEV) [5]. In
1990–1995, of all reportable infectious processes, the
number of reported cases of acute viral hepatitis (AVH)
was exceeded only by the cumulative number of cases of
acute respiratory disease [6]. AVH is a reportable disease
in Uzbekistan, where periodic epidemics of fecal-orally-
transmitted non-A, non-B hepatitis have been described
[7,8]. Elsewhere, in regions in which epidemics of hepati-
tis E have been observed, increases in mortality among
pregnant women have been reported [9-12].
For reporting purposes in Uzbekistan and other Central
Asian republics of the former Soviet Union, the diagnosis
of hepatitis A has generally been based on signs, symp-
toms and epidemiological data, while the diagnosis of
acute hepatitis B has been based on serologic detection of
hepatitis B surface antigen (HBsAg). However, lack of
serological testing to differentiate hepatitis types has
resulted in official statistics that have generally enumer-
ated cases of AVH with typing of hepatitis A and hepatitis
B only, and reporting of HCV, HDV, and HEV infections
has not been routine.
An indicator of the impact of hepatitis infections is their
associated mortality, most often the result of chronic liver
disease [13]. However, few data exist concerning age- and
gender-specific associations of AVH with mortality. We
analyzed trends in AVH-associated mortality, including
during periods in which hepatitis E epidemics have been
documented [14,15], to determine age- and gender-spe-
cific changes in mortality.
Methods
Incidence Reporting
AVH incidence (morbidity) data from 1971 through 2005
were obtained from reports of the Sanitary-Epidemiologic
Service (SES) of the Uzbekistan Ministry of Health
(MOH). These are standardized data from a surveillance
system that was uniformly used by the republics during
the Soviet era [16,17] and that has continued to function
in Uzbekistan since the dissolution of the Soviet Union.
Uzbekistan is organized into 12 provinces, called viloyats,
and one autonomous republic; each of these is subdivided
into administrative units called raions. At each level
(republic, viloyat, raion), the SES has operated an infec-
tious disease surveillance system in which AVH has been
a reportable condition. The system has evolved under a
mandate that requires all suspected or confirmed cases to
be reported by telephone to the raion SES within 12
hours, followed by a written case report [17,18]. These
reports are forwarded to the viloyat SES. Reporting sources
include physicians and feldshers (primary health care pro-
viders with less formal training than physicians, found
more commonly in rural areas) who evaluate patients in
hospitals, polyclinics, diagnostic centers, ambulatory clin-
ics, or feldsher stations (rural primary health care clinics).
For each case, health care providers report patient's name,
age, date of birth, sex, address, occupation, day care
attendance, last day of work, polyclinic attended, primary
diagnosis, date of symptom onset, date of physician visit,
date of diagnosis, and date of hospitalization.
Telephone case reports received at the raion SES are
recorded and updated as additional clinical, laboratory
and epidemiologic information is received to help raion
SES staff implement disease control measures. The raion
SES obtains additional information on each case within
24 hours using a standardized case investigation form.
Raion SES staff complete a separate case form which sum-
marizes the information collected during the course of the
epidemiologic and clinical work-up, including clinical
presentation, vaccination status, and laboratory results.
SES epidemiologists may request additional laboratory
testing to confirm the healthcare provider's clinical diag-
nosis. However, lack of reagents to perform diagnostic
testing and lack of strict case definitions have had the
result that most reported cases of AVH in the SES database
reflect clinical diagnoses that have not been laboratory
confirmed.
Raions report AVH case totals to the viloyat SES, and raion
cases are tabulated by age group (total, 0 to 14 years; <1
year (infancy); 1 and 2 years (up to but not including the
third birthday); and 3 to 6 years). Annually, aggregated
data are reported to the viloyat SES. Each month, the vilo-

Page 3

BMC Infectious Diseases 2009, 9:35 http://www.biomedcentral.com/1471-2334/9/35
Page 3 of 9
(page number not for citation purposes)
yat SES tabulates confirmed AVH cases by age group, and
reports totals to the republican SES.
Mortality Reporting
AVH-specific mortality data were examined as provided to
the MOH by Goskomstat, the State Committee of Statistics
of the Republic of Uzbekistan, a governmental vital
records agency with its own raion-viloyat-republic report-
ing hierarchy that is separate from the SES. When a patient
dies, the clinician notifies the hospital or polyclinic statis-
tician, who in turn is responsible for reporting the death
and its putative cause(s) to the statistical administrative
units within Goskomstat that are concerned with mortality.
Case Definition
Although there has been no strict case definition of AVH,
MOH guidelines for clinicians have traditionally defined
AVH as an infectious disease characterized by acute onset
of jaundice, hepatomegaly, and (where available) a three-
fold or greater elevation of serum aminotransferase levels.
AVH is noted as a cause of death on a death certificate if a
clinician believes the death was a complication of AVH.
Laboratory Methods
During an identified AVH outbreak in 1987, serum sam-
ples were collected from a convenience sample of hospi-
talized patients with AVH in Eastern and Southern
Uzbekistan (Fergana and Kashkadarya viloyats) on the
first to seventh day after onset of jaundice; criterion for
hospitalization was onset of jaundice. Similarly, during
1992, serum samples were collected from a convenience
sample of hospitalized pregnant women with AVH in
Tashkent viloyat on the first to seventh day after onset of
jaundice; again, criterion for hospitalization was onset of
jaundice. For both groups, case-patients were defined as
those with acute onset of dyspepsia, jaundice, hepatome-
galy, and a three-fold or greater increase in serum concen-
trations of alanine aminotrasferase. Serum specimens
were delinked from personal identifiers at the time of col-
lection and were stored at -20°C until subsequently
tested. As both of these activities predated implementa-
tion of codified institutional accountability of human
research protections, no standardized consenting process
was employed; however, patient participation was limited
to a solitary phlebotomy for minimal serum samples for
diagnostic purposes.
Serologic testing by enzyme immunoassay (Abbott Labo-
ratories, North Chicago, IL) for markers of infection with
hepatitis viruses was performed at the Viral Hepatitis
Diagnostic Department, Ivanovski Institute of Virology,
Moscow. Testing included assays for markers of acute
HAV infection (IgM anti-HAV); markers of acute HBV
infection, HBsAg and IgM antibody to hepatitis B core
antigen (IgM anti-HBc); and markers of acute HDV infec-
tion (IgM anti-HDV and total anti-HDV). Serum speci-
mens which tested negative for all of the above markers
were retained and subsequently tested in the Hepatitis
Reference Laboratory, Division of Viral Hepatitis,
National Center for Infectious Diseases, Centers for Dis-
ease Control and Prevention (CDC), for IgM and IgG anti-
body to HEV (IgM and IgG anti-HEV) using both Western
blot and ELISA [19].
Statistical Methods
Incidence data analyses
To quantify AVH case counts and rates per 100,000 popu-
lation, we analyzed Republican SES reports for the period
1971–2005 and specific viloyat reports for 1985–1995.
To verify reported rates by viloyat, population counts
from viloyat reports were compared with official
Uzbekistan census data.
Mortality data analyses
To ascertain mortality rates per 100,000 (MRs) from AVH,
AVH-associated mortality data for 1981–1995 were
obtained from the Department of Statistics of the Ministry
of Health. AVH-associated MRs for the republic and for
each viloyat for 1985–1995 were calculated by age and sex
on the basis of mortality data analyzed from Goskomstat
reports. Where mortality rates per 100,000 population
were calculated for a specific age and sex group, denomi-
nator data were restricted to population estimates for that
specific age and sex group. These analyses did not include
deaths attributed to chronic viral hepatitis and/or cirrho-
sis.
Statistical analyses
Data were analyzed using Epi Info 6.04 for relative risk
(RR), Student's t-test, correlation analysis (r), Pearson's
conformity test, or Fisher's exact test. Specifically, categor-
ical data were analyzed using Pearson's χ2-test or Fisher's
exact test, as appropriate; Fisher's exact test was used
where categorical data were compared in which any cell
size was <5. Continuous data were analyzed with Stu-
dent's t-test. Statistical significance was determined by cal-
culating a p value (with Yates' correction factor where
needed) and exact 95% confidence intervals (CIs), or Tay-
lor series 95% CIs, as appropriate. The significance level
for all statistical analyses was p = 0.05.
Trends in AVH-associated mortality over the 15 years
(1981–1995) for which aggregate data were available
were compared with trends in reported AVH incidence
over the same period using the Pearson product-moment
correlation coefficient. For the 11-year period (1985–
1995) for which gender- and age-specific AVH-associated
mortality data were available, data were compared by gen-
der, age, and rural/urban designation of the decedent
using Pearson's χ2-test, Fisher's exact test, or Student's t-

Page 4

BMC Infectious Diseases 2009, 9:35http://www.biomedcentral.com/1471-2334/9/35
Page 4 of 9
(page number not for citation purposes)
test as appropriate, depending on whether data were cate-
gorical or continuous.
In performing analyses of serology data, IgG anti-HEV
data for various age groups of children were compared
using Pearson's χ2-test.
Results
Reported Incidence
From 1971 through 2005, national reported AVH inci-
dence in Uzbekistan was consistently greater than 200
cases per 100,000 (Fig. 1). Two multi-year epidemics of
AVH were noted: one spanning the years 1975 and 1976
with annual incidences of 1047 and 1072 cases per
100,000, respectively, and another from 1985 through
1987 with annual incidences of 1259, 1187, and 1431
cases per 100,000. Single-year increases in AVH were also
noted in 1973, 1983, 1995 and 1997, with peak inci-
dences ranging from 600 to 1051 per 100,000.
Verification of reported population counts by viloyat
revealed no discrepancies when viloyat reports for hepati-
tis rates were compared with official Uzbekistan census
data. The geographic distribution of average annual AVH
incidence varied: highest viloyat rates were reported in the
East (Fergana, Namangan, and Andijan) and in the South
(Kashkadarya, Surkhandarya), where incidence in some
viloyats was 1.5- to 1.8-times the national average. In less
densely populated areas, e.g., in the West (Karaka-
lpakstan, Bukhara, and Navoi), incidence tended to be
Acute viral hepatitis incidence (1971–2005) and associated mortality rates (1981–1995), by year, Uzbekistan (per 100,000 pop-ulation)
Figure 1
Acute viral hepatitis incidence (1971–2005) and associated mortality rates (1981–1995), by year, Uzbekistan
(per 100,000 population). Left ordinate (y-axis) and line graph: mortality per 100,000 population. Right ordinate (y-axis) and
bar diagram: reported acute viral hepatitis incidence per population.




0
2
4
6
8
10
12
14
16
18
20
197119731975197719791981198319851987198919911993199519971999200120032005
Year
Mortality
0
200
400
600
800
1000
1200
1400
1600
Incidence
incidence
mortality

Page 5

BMC Infectious Diseases 2009, 9:35http://www.biomedcentral.com/1471-2334/9/35
Page 5 of 9
(page number not for citation purposes)
lower than the national average except in 1985–1987.
Despite higher population density, reported AVH inci-
dence in Tashkent was relatively low and exceeded the
national rate only in 1995 and 1997.
Reported Mortality
AVH-associated mortality varied over 1981 to 1995 and
AVH-associated mortality was closely associated with
reported AVH incidence, r = 0.92; p < 0.001 (Fig. 1). The
highest nationwide AVH-associated mortality occurred in
1987 (17.8 per 100,000) with a corresponding AVH inci-
dence rate of 1,431 per 100,000.
In 1985–1995, no significant difference was observed
between males and females: among males, the MR was
11.8 (95%CI: 10.3–13.3) per 100,000, and among
females, the MR was 10.8 (95%CI: 9.3–12.3) (RR = 1.0;
95%CI: 0.8–1.4; p = not significant (n.s.)). When AVH-
associated mortality was examined by age group for all
ages, highest rates were observed in children aged <3 years
(Fig. 2), especially among those aged <2 years; among
both those aged <1 year and those aged ≥1 to <2 years, the
peak coincided with the 1987 peak observed in reported
AVH incidence. In Fig 2, gender-specific AVH-associated
mortality for 1985–1995 by age group (<1 year, ≥1 to <2
years, ≥2 to <3 years, 3 to 14 years) is presented in a set of
four bar diagrams. Because reported AVH incidence data
were not available in the same age classifications as AVH-
associated mortality, AVH incidence rate for all age groups
combined is presented as a line diagram above each of the
four bar diagrams to give the reader an appreciation of the
temporal distribution of the AVH-associated age-specific
mortality relative to the overall reported AVH morbidity.
From 1987 to 1995, the AVH-associated MR decreased
among infants (children aged <1 year) from 143.7 to 29.3
per 100,000 (p < 0.01), and among children aged ≥1 to <2
years from 166.8 to 30.6 per 100,000 (p < 0.01); markedly
less impressive decreases in rates of AVH-associated mor-
Acute viral hepatitis-associated mortality rates for children in various age groups, and reported acute viral hepatitis incidence in all age groups, per 100,000 population, by year, Uzbekistan, 1985–1995
Figure 2
Acute viral hepatitis-associated mortality rates for children in various age groups, and reported acute viral
hepatitis incidence in all age groups, per 100,000 population, by year, Uzbekistan, 1985–1995. Left ordinate (y-
axis) and bar diagram: age- and gender-specific mortality: "black bars" – mortality among boys; "white bars" – mortality among
girls, per 100,000 population of corresponding age and sex. Right ordinate (y-axis) and line diagram: reported acute viral hepa-
titis incidence in the general population, per 100,000.
0
20
40
60
80
100
120
140
160
180
200
19851986198719881989199019911992199319941995
Year
0
200
400
600
800
1000
1200
1400
1600
Age and gender specific mortality rates
Incidence per 100,000 population
??????????????????????
?
0
20
40
60
80
100
120
140
160
180
200
19851986198719881989199019911992199319941995
Year
0
200
400
600
800
1000
1200
1400
1600
Age and gender specific mortality rates
Incidence per 100,000 population
???????????????
0
20
40
60
80
100
120
140
160
180
200
19851986198719881989199019911992199319941995
Year
0
200
400
600
800
1000
1200
1400
1600
Age and gender specific mortality rates
Incidence per 100,000 population
??????????????????????
0
20
40
60
80
100
120
140
160
180
200
19851986198719881989199019911992199319941995
Year
0
200
400
600
800
1000
1200
1400
1600
Incidence per 100,000 population
Age and gender specific mortality rates
??????????????????????