MAJOR ARTICLE
Acute Hepatitis Delta Virus Infection in Italy:
Incidence and Risk Factors after the Introduction of
the Universal Anti–Hepatitis B Vaccination Campaign
Alfonso Mele,1 Andrea Mariano,1 Maria Elena Tosti,1 Tommaso Stroffolini,2 Renato Pizzuti,4 Giovanni Gallo,5
Pietro Ragni,6 Carla Zotti,7 Pierluigi Lopalco,8 Filippo Curtale,3 Emanuela Balocchini,9 and Enea Spada1
on behalf of the SEIEVA Collaborating Groupa
1
National Centre of Epidemiology, Surveillance and Health Promotion, Istituto Superiore di Sanità, 2Gastroenterology Unit, San Giacomo Hospital, and
Public Health Agency of the Lazio Region, Rome, 4Regional Health Authority, Regione Campania, Naples, 5Service of Public Health and Hygiene, Unita
Locale Socio Sanitaria 9, Treviso, 6Regional Health Authority, Regione Emilia Romagna, Bologna, 7Department of Public Health and Microbiology,
University of Turin, Turin, 8Institute of Hygiene, University of Bari, Bari, and 9Regional Health Authority, Regione Toscana, Florence, Italy
3
Background. Updates on the incidence of and risk factors for acute hepatitis delta virus infection in Italy, as
well as in other countries, are lacking, and the impact of the mandatory anti–hepatitis B vaccination has not been
evaluated.
Methods. We performed a case-control study within a population-based surveillance for acute viral hepatitis.
Results. During 1993–2004, 344 cases of acute hepatitis delta virus infection were reported. After a peak in
1993 (2.8 cases per 1 million population), the incidence decreased from 1.7 to 0.5 cases per 1 million population.
Coinfections were prevalent. The decrease in incidence particularly affected young adults, and it paralleled the
decrease in incidence of acute hepatitis B. In 1993, being an injection drug user (adjusted odds ratio [ORadj], 67.9;
95% confidence interval [CI], 18.1–254.5) or being a member of a household with a carrier of hepatitis B surface
antigen (ORadj, 14.8; 95% CI, 3.0–72.9) were the only independent predictors of infection. During 1994–2004,
being an injection drug user (ORadj, 36.8; 95% CI, 20.7–65.4), cohabitation with an injection drug user (ORadj,
4.2, 95% CI, 1.7–12.3), hospitalization (ORadj, 3.5; 95% CI, 1.9–6.6), receipt of dental therapy (ORadj, 2.3; 95%
CI, 1.4–3.6), promiscuous sexual activity (ORadj, 2.2; 95% CI, 1.4–3.6), and receipt of beauty treatment (ORadj,
2.0; 95% CI, 1.3–3.2) were independently associated with infection.
Conclusions. Incidence of acute hepatitis delta infection is markedly decreasing in Italy. Undergoing invasive
medical procedures, engaging in promiscuous sexual activity, and receiving beauty treatments are emerging, in
addition to injection drug use, as important risk factors for infection. Further efforts are needed to increase vaccine
coverage in high-risk groups and to implement the safety of invasive procedures performed both inside and outside
health care facilities.
Hepatitis delta virus (HDV) is a defective RNA virus
that requires the helper function of hepatitis B virus
(HBV) for infection [1]. HDV infection can occur either simultaneously with HBV infection (coinfection)
or in chronic carriers of the hepatitis B surface antigen
Received 3 August 2006; accepted 19 September 2006; electronically published
22 December 2006.
a
Members of the Sistema Epidemiologico Integrato dell’Epatite Virale Acuta
(SEIEVA) group are listed at the end of the article.
Reprints or correspondence: Dr. Enea Spada, Istituto Superiore di Sanità,
National Center of Epidemiology, Surveillance and Health Promotion, Clinical
Epidemiology Unit, Via Giano Della Bella, 34- 00162 Rome, Italy ([email protected]).
Clinical Infectious Diseases 2007; 44:e17–24
2006 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2007/4403-00E1$15.00
(HBsAg) (superinfection). The presence of HDV in patients who have chronic HBV disease is associated with
a more rapid progression to cirrhosis and with a higher
risk of liver decompensation and hepatocellular carcinoma [2].
HDV prevalence is high in areas of sub-Saharan Africa, Eastern Europe, the Middle East, central Asia, the
South Pacific islands, the Amazon Basin, and the Mediterranean Basin [3]. In Italy, the virus was endemic in
the 1970s, when it was responsible for a substantial
proportion of cases of juvenile cirrhosis [4]. In the past
decades, HDV prevalence among HBsAg-positive patients has progressively decreased in southern Europe
[5]: in Italy, prevalence decreased from 25% in 1983
Epidemiology of HDV Infection in Italy • CID 2007:44 (1 February) • e17
[4] to 8.3% in 1997 [6]; in Spain, prevalence decreased from
15.1% in 1979–1985 to 7.1% in 1986–1992 [7]; and in Greece,
prevalence decreased from 21.5% in 1970–1974 to 16.5% in
1985–1989 [8]. However, updated data on the incidence of
HDV infection in Italy, as well as in other countries, are lacking.
Data from a national surveillance system for acute viral hepatitis
(Sistema Epidemiologico Integrato dell’Epatite Virale Acuta
[SEIEVA]) [9] have shown a decrease in the incidence of acute
HDV infection in Italy, from 3.1 cases per 1 million population
in 1987 to 1.2 cases per 1 million population in 1992; injection
drug use, household contact with an HBsAg-positive individual,
and promiscuous sexual activity were independent predictors
of infection in that time period [10]. In 1991, a mandatory
universal anti-HBV vaccination campaign of infants and 12year-old adolescents was initiated in Italy. Taking into account
the close bond between HBV and HDV infections, we aim to
update the previous SEIEVA report [10] and to describe the
epidemiology of acute HDV infection after the introduction of
the anti-HBV vaccination campaign.
METHODS
Participants. SEIEVA is coordinated by the Italian National
Institute of Health (Istituto Superiore di Sanità, Rome) and
was established in 1985; since that time, the number of local
health units (LHUs) throughout the country that voluntarily
participate in the system has progressively increased. In the
present study, SEIEVA data for the period 1993–2004 were
analyzed; the percentage of the Italian population under surveillance was ∼40% in 1993–1995, ∼45% in 1996, ∼55% in
1997–1999, and ∼60% in 2000–2004.
Cases of acute viral hepatitis, whether the affected patients
are admitted to the hospital or not, are reported to the surveillance system through the LHUs. Case definition of acute
hepatitis is based on clinical, biochemical, and serological criteria. Clinical and biochemical criteria include the occurrence
of an acute illness that is compatible with hepatitis together
with an increase in serum alanine aminotransferase levels (12.5
times the upper limit of normal). Serological criteria, used to
distinguish the different types of acute viral hepatitis, are as
follows: acute hepatitis A virus (HAV) infection is defined by
IgM anti-HAV–positive serological test results, regardless of
other viral markers; HBV infection is defined by IgM hepatitis
B core antibody (anti-HBc)– and HBsAg–positive and IgM
anti-HAV–negative serological test results; and hepatitis C virus
(HCV) infection is defined by IgM anti-HAV and IgM antiHBc–negative and anti-HCV–positive serological test results.
Cases of acute hepatitis with HBsAg-positive, anti-HDV–positive, and IgM anti-HAV–negative serological test results are
classified as HDV infection. If IgM anti-HBc is also present,
the case is considered to be HBV-HDV coinfection. If IgM antiHBc is absent, the case is considered to be HDV superinfection
e18 • CID 2007:44 (1 February) • Mele et al.
in a chronic HBsAg carrier. Assays to detect viral hepatitis
markers are performed in different local laboratories using standardized methods.
The study was approved by the ethics committee of the Italian National Institute of Health; written informed consent was
obtained from all participants.
Data collection. Patients who received a diagnosis of acute
viral hepatitis were interviewed by either a public health inspector or a physician of the LHU using a standard, 2-page
questionnaire. Information regarding sociodemographic characteristics, exposure to parenteral risk factors within the previous 6 months, and exposure to fecal-oral risk factors within
the 6 weeks before disease onset were collected. The interviewer
was blinded with respect to the type of viral hepatitis, to avoid
bias in the identification of risk factors. The results of assays
to detect hepatitis markers were recorded on the questionnaire
after the interview; the completed questionnaires were forwarded to the coordinating center at the Italian National Institute of Health.
Incidence rates estimation. Incidence rates were estimated
using the new reported cases for each type of hepatitis as the
numerator and the population of LHUs participating in the
surveillance system at various points in time as the denominator. No changes were made in the notification system during
the study period. Most patients with symptomatic acute hepatitis are admitted to hospitals and virtually all cases of hepatitis
in hospitalized patients are recorded by LHUs.
Case-control study. To estimate the association of cases of
HDV infection with potential risk factors, cases of hepatitis A
reported during the same time period as the SEIEVA study were
used as controls. Cases reported in 1993 were analyzed separately, because an outbreak of HDV infection occurred in that
year.
Statistical analysis. Differences in proportion between
coinfections and superinfections were tested using the x2 test
or Fisher’s exact test, when necessary. The difference between
mean age was evaluated using the Student’s t test. A P value
!.05 was considered to be statistically significant.
Crude ORs and associated 95% CIs for the factors under
consideration were calculated in a univariate analysis. Multiple
logistic regression was used to estimate the adjusted OR (ORadj)
for all risk factors. Age, sex, area of residence, and educational
level were also adjusted for during the analysis. All statistical
analyses were conducted using STATA software, version 8.0
(Statacorp).
RESULTS
During the period 1993–2004, 9433 HBsAg-positive, IgM antiHAV–negative cases of acute hepatitis were reported to SEIEVA.
Overall, 46.2% of these case patients were tested for anti-HDV,
although the percentage of of patients tested for anti-HDV
Figure 1. Incidence rates of acute hepatitis delta virus (HDV) infection.
Data were collected by Sistema Epidemiologico Integrato dell’Epatite
Virale Acuta (SEIEVA) during the period 1987–2004.
decreased by approximately two-thirds in the last 2 years of
that period. The anti-HDV test was more frequently performed
in injection drug users (IDUs; 53.9% of whom were tested; P
!.001) and in individuals who engaged in promiscuous sexual
activity (defined as 11 sexual partner in the previous year;
52.5% of whom were tested; P !.001), than in patients reporting
other risk factors. Three hundred forty-four cases were classified
as acute HDV infection, 218 (63.4%) of which were
coinfections.
Figure 1 shows the yearly incidence rates of cases of acute
HDV infection (per 1 million population) since 1987. In figures
2A and 2B, the age-specific yearly incidence rates of cases of
acute HDV infection (per 1 million population) and cases of
acute hepatitis B (per 100,000 population) are reported, respectively. In these figures, an overall decrease in incidence is
evident; in particular, a decrease in incidence—for both HDV
and HBV infections—occurred in the 15–24-year-old age
group. Two peaks are evident in the acute HDV infection incidence curve. A first peak was observed in 1990; this excess
of cases of infection was not linked to any specific geographical
area or reported risk factor. A second peak (64 cases, corresponding to 2.8 cases per 1 million population) occurred in
1993: 56.3% of these cases were from northeastern Italy; 60.7%
of these case patients were IDUs. From 1994 to 2004, 280 acute
cases of HDV infection were reported, and the incidence rate
decreased from 1.7 to 0.5 cases per 1 million population. Coinfections represented 69% of cases of HDV infection in 1993
and 62% in 1994–2004. The characteristics of the acute HDV
infections that were reported to SEIEVA during the period
1987–1992 have been previously reported [9].
Table 1 describes demographic and clinical characteristics of
cases of HDV infection reported in 1993–2004 and compares
coinfected with superinfected cases. Most cases of HDV infection were observed in men 125 years of age who had a mediumhigh educational level and who resided in the northern central
region of the country. Sociodemographic characteristics did not
differ between patients with coinfections and patients with su-
perinfections, except for geographical area: 38.9% of superinfected subjects versus 11.9% of coinfected subjects were reported (see table 1) in the southern region of the country or
in the islands. In additon, the clinical characteristics were similar, but HBV-HDV–coinfected individuals experienced jaundice more frequently than superinfected individuals. Almost
one-half of the patients were anti-HCV positive, and 75% of
these were IDUs. Two patients with HBV-HDV coinfection
died: one was an HCV-positive IDU, and the other was a 65year-old man with cancer.
As mentioned earlier, because an outbreak of HDV infection
occurred in 1993 in northeastern Italy (mainly among IDUs),
risk factors for HDV infection were evaluated separately for the
year 1993 and the period 1994–2004. In 1993, being an IDU
or a member of a household with a chronic carrier of HBsAg
were the only independent predictors of infection (table 2).
In the period 1994–2004, in addition to injection drug use
(which remained strongly associated with HDV infection), hospitalization, receipt of beauty treatments (i.e., tattooing, piercing, manicure/chiropody, and barber shop shaving) or dental
therapy, having had ⭓2 sexual partners in the past year, and
cohabitation with an IDU emerged as other independent risk
factors (table 3).
Figure 2. Yearly age-specific incidence rates of acute hepatitis delta
virus (HDV) infection (A) and acute hepatitis B virus (HBV) infection (B).
Data were collected by Sistema Epidemiologico Integrato dell’Epatite
Virale Acuta (SEIEVA) during the period 1987–2004. Vertical dotted line,
the beginning of the hepatitis B vaccination campaign.
Epidemiology of HDV Infection in Italy • CID 2007:44 (1 February) • e19
Table 1. Characteristics of patients with hepatitis delta virus (HDV) infection reported to
Sistema Epidemiologico Integrato dell’Epatite Virale Acuta (SEIEVA) during the period 1993–
2004.
Characteristic
Percentage of total patients
Sociodemographic characteristics
Sex
Male
Female
Patients with
coinfections
a
(n p 218)
Patients with
superinfections
(n p 126)b
Total patients
(n p 344)
63.4
36.6
100
172 (78.9)
46 (21.1)
103 (82.4)
22 (17.6)
275 (80.2)
68 (19.8)
.434
.069
Pc
Age, years
⭐14
2 (0.9)
6 (4.8)
8 (2.3)
15–24
49 (22.5)
29 (23.0)
78 (22.7)
25–34
97 (44.5)
61 (48.4)
158 (45.9)
70 (32.1)
30 (23.8)
100 (29.1)
29 (18.1)
78 (48.8)
53 (32.1)
22 (24.2)
48 (52.7)
21 (23.1)
51 (20.3)
126 (50.2)
74 (29.5)
North central Italy
192 (88.1)
77 (61.1)
269 (78.2)
Southern Italian islands
Clinical characteristicsd
26 (11.9)
49 (38.9)
75 (21.8)
197 (94.7)
11 (5.3)
106 (88.3)
14 (11.7)
303 (92.4)
25 (7.6)
.036
Yes
202 (95.3)
119 (96.0)
321 (95.4)
.769
No
10 (4.7)
5 (4.0)
15 (4.7)
20.9 12.7
21.6 13.9
21.2 13.1
.653
91 (45.7)
108 (54.3)
56 (49.1)
58 (50.9)
147 (47.0)
166 (53.0)
.563
2 (0.9)
216 (99.1)
0
126 (100.0)
2 (0.6)
342 (99.4)
.534
⭓35
Years of schooling
⭐5
6–8
⭓9
.2
Geographic area
Jaundice
Yes
No
!.001
Hospitalization
Duration, mean days SD
Anti-HCV
Positive
Negative
Death
Yes
No
NOTE. Data are no. (%) of patients, unless otherwise indicated. HCV, hepatitis C virus.
a
b
c
d
Defined as HDV infection that occurs simultaneously with hepatitis B virus infection.
Defined as HDV infection in chronic carriers of the hepatitis B surface antigen.
Comparison between patients with coinfections and patients with superinfections.
Information on selected characteristics may be missing for some patients.
DISCUSSION
With the exception of 2 outbreaks that occurred in 1990 and
1993, the incidence of reported cases of acute HDV infection
in Italy progressively decreased over time, reaching a nadir in
2004. This decrease closely paralleled the decrease in incidence
of acute hepatitis B during that time, and it was already ongoing
before the introduction of the universal vaccination program.
In fact, many factors are thought to have contributed to the
decreasing spread of HBV infection in Italy [11–13], and these
factors probably also explain the similar trend exhibited by
HDV infection: the social changes that occurred in the past
e20 • CID 2007:44 (1 February) • Mele et al.
decades (e.g., general improvement of hygienic standards and
living conditions and a decrease in family size); the increasing
use of disposable syringes; blood screening; the impact of the
anti-AIDS campaign on risky behaviors related to sexual contact and to drug abuse; and the vaccination of high-risk groups
since the 1980s. Most of these factors have probably affected
the risk of acquiring HDV and HBV infections, especially
among younger individuals. The universal vaccination campaign, which was initiated in 1991, further enhanced this trend
by reducing the pool of susceptible children (at high risk for
chronic HBV infection) while having the greatest impact in the
Table 2. Univariate and multivariate analysis (crude and adjusted ORs) of the risk factors reported by patients
with hepatitis delta virus (HDV) infection reported to Sistema Epidemiologico Integrato dell’Epatite Virale Acuta
(SEIEVA) in 1993.
Patients with
HDV infection, %
(n p 64)
Patients with
HAV infection, %
(n p 1148)
Blood transfusion
Surgery
Hospitalization
a
Receipt of beauty treatment
0.0
6.7
12.5
0.1
3.6
5.6
26.7
16.7
1.8 (0.9–3.4)
1.2 (0.4–4.1)
Dental therapy
⭓2 Sexual partners in the past year
20.0
55.6
21.5
11.2
0.9 (0.4–1.8)
9.9 (5.4–18.4)
0.9 (0.3–3.0)
2.2 (0.7–6.9)
IDU
IDU household
60.7
14.0
1.7
1.3
88.6 (41.9–188.3)
12.6 (4.3–34.5)
HBsAg-positive household
20.5
3.0
8.4 (3.2–20.0)
Risk factor
OR (95% CI)
…
1.9 (0.5–5.6)
2.4 (0.8–6.1)
AOR (95% CI)
…
0.7 (0.1–5.9)
2.4 (0.4–14.4)
67.9 (18.1–254.5)
2.6 (0.4–17.4)
14.8 (3.0–72.9)
NOTE. Each patient may have reported 11 risk factor. Cases of HAV infection reported in 1993 were used as a control group. ORs
were adjusted for age, sex, geographical area, educational level, and the risk factors listed in the table. AOR, adjusted OR; HAV, hepatitis
A virus; HBsAg, hepatitis B surface antigen; IDU, injection drug user.
a
Tattooing, piercing, manicure/chiropody, and barber shop shaving.
15–24-year-old age group in terms of the reduced number of
acute symptomatic HBV and HDV infections. At present, Italy
is a country of low HBV endemicity [13]. This translates to a
lower incidence of HDV superinfections, which are at high risk
to result in a chronic infection and to perpetuate HDV endemicity. This is supported by the finding that the proportion
of the total number of cases of acute HDV infection that are
HDV superinfections decreased from 50.6% in 1987–1992 [10]
to 36.6% in 1993–2004; furthermore, this proportion is lower
in northern regions (which have a lower HBV carrier prevalence) than in southern regions. Other epidemiologic data are
consistent with our figures, which reveal a decreasing impact
of HDV infection in Italy. At a general population level, in a
recent survey conducted in a town in southern Italy, the prevalence of chronic carriers of HBsAg was 0.8%, none of whom
were anti-HDV positive [14]. At the hospital level, multicenter
surveys performed in 1983 [4], 1987 [15], 1992 [16], and 1997
[6] have shown that the proportion of HBsAg-positive patients
who were anti-HDV positive was 25%, 23.4%, 14.4%, and
8.3%, respectively. Besides a reduced incidence of HDV infection, it is likely that an increased mortality associated with
chronic HDV-HBV hepatitis [2] also contributed to the progressive decrease in the population of coinfected patients.
Injection drug use played an important role in the outbreak
that occurred in 1993; similar outbreaks have been reported in
other industrialized countries, as well [12, 17, 18]. In 1993,
acute HDV infection was strongly associated, not only with
injection drug use, but also with cohabitation with a chronic
carrier of HBsAg. This condition increases either the risk of
acquiring acute HDV-HBV coinfection or the probability of
being chronically HBV infected. Thus, in periods of high HDV
circulation, the risk of acquiring acute HDV coinfection or
superinfection is increased for households of chronic carriers
of HBsAg.
During the period 1994–2004, household contact with an
IDU, promiscuous sexual activity, and receipt of beauty treatments and dentistry procedures (in addition to injection drug
use) played a role in the transmission of HDV. Hospitalization
was also an independent predictor of HDV infection, whereas
surgery and blood transfusion were not. It is likely that, as has
already been demonstrated for HBV and HCV infections [19],
other invasive diagnostic and therapeutic procedures play a role
in the transmission of HDV in hospitals. In addition, invasive
procedures that are performed outside health care facilities,
such as beauty treatments, have been shown in a previous
SEIEVA report [20] to be risk factors for transmission of HBV
and HCV. It is noteworthy that, in our previous analysis of the
incidence of and risk factors for HDV infection during the
period 1987–1992, only injection drug use, household contact
with an HBsAg-positive carrier, and promiscuous sexual activity
resulted in an independent association with HDV infection.
Thus, an important change in the pattern of HDV transmission
might have occurred in the past 2 decades: besides IDUs and
people engaged in promiscuous sexual activity—who are still
at high risk of infection—there is likely a decreasing impact
from intrafamiliar HDV spread and an increasing impact from
invasive procedures; the latter is probably because of the growing number of people who undergo invasive medical procedures and the recent great diffusion among the general population of fashionable beauty practices, such as piercing.
When interpreting the results of this study, some possible
limitations should be discussed. First, the incidence rates of
acute HDV infection reported in this study are probably far
from the true incidence of infection. In fact, SEIEVA, as well
Epidemiology of HDV Infection in Italy • CID 2007:44 (1 February) • e21
Table 3. Univariate and multivariate analysis (crude and adjusted ORs) of the risk factors reported by patients
with acute hepatitis delta virus (HDV) infection reported to Sistema Epidemiologico Integrato dell’Epatite Virale
Acuta (SEIEVA) during the period 1994–2004.
Patients with
HDV infection, %
(n p 280)
Patients with
HAV infection, %
(n p 16,665)
OR (95% CI)
AOR (95% CI)
Blood transfusion
Surgery
Hospitalization
a
Receipt of beauty treatment
1.2
11.2
18.4
0.3
3.9
4.2
4.5 (0.9–14.2)
3.1 (2.0–4.7)
5.2 (3.6–7.4)
4.7 (0.9–26.3)
1.4 (0.7–2.9)
3.5 (1.9–6.6)
33.2
16.7
2.5 (1.9–3.3)
2.0 (1.3–3.2)
Dental therapy
⭓2 Sexual partners in the past year
28.6
36.5
15.7
10.3
2.2 (1.6–2.9)
5.0 (3.7–6.7)
2.3 (1.4–3.6)
2.2 (1.4–3.6)
IDU
IDU household
44.3
10.0
1.5
0.8
53.8 (4.3–71.7)
14.0 (8.4–22.5)
36.8 (20.7–65.4)
4.2 (1.5–12.3)
HBsAg-positive household
11.7
2.3
5.6 (3.4–8.9)
2.2 (0.8–5.8)
Risk factor
NOTE. Each patient may have reported 11 risk factor. Cases of HAV reported during the period 1994–2004 were used as control
group. ORs were adjusted for age, sex, geographical area, educational level, and the other risk factors listed in the table. AOR,
adjusted OR; HAV, hepatitis A virus; HBsAg, hepatitis B surface antigen; IDU, injection drug user.
a
Tattooing, piercing, manicure/chiropody, and barber shop shaving.
as other surveillance systems that are based on the voluntary
reporting of cases of acute hepatitis, may lack sensitivity because
of underreporting, the occurrence of subclinical infection, and
the lack of HDV testing for all HBsAg-positive, IgM anti-HAV–
negative individuals. However, because the system was unchanged during the study period, its low sensitivity may affect
yearly incidence rates but not the incidence trend of infection
over time. In fact, the decreasing trend of incidence of HDV
infection parallels that of acute hepatitis B in the same time
period, and it is in agreement with the results involving seroepidemiological studies of the general Italian population and
hospitalized individuals [4, 6, 14–16]. Thus, data collected
through a passive surveillance system provide valuable evidence
of a change in secular trend over reasonably long time periods.
Finally, in the study period, a high proportion of Italians were
surveilled through the collaboration of LHUs all over the country. As a result, the data collected are likely representative of
Italy as a whole.
Second, the use of patients with hepatitis A as control subjects to estimate the strength of the association between HDV
infection and various routes of exposure may raise some concern. HAV has different modes of transmission than HDV,
although they can occasionally share the same risk factors, such
as injection drug use or transfusion of blood or blood products.
However, percutaneous transmission of HAV is considered to
be a very rare event in the general population: of the patients
with hepatitis A who participated as control subjects in this
study, only a very low percentage reported blood transfusion
or injection drug use as a risk factor (0.1% and 1.7%, respectively). However, an overinclusion of exposed persons in the
control group would, at most, have led us to underestimate
the related OR. Moreover, it should be considered that, in casee22 • CID 2007:44 (1 February) • Mele et al.
control studies, the comparability between case patients and
control subjects is a crucial factor and is more important than
representativeness. In fact, all individuals in this study who had
acute hepatitis represented reported cases, originated from the
same geographical area, were identified through the same surveillance system, and were interviewed by the same blinded
interviewer; they were, thus, exposed to the same selective factors, if any. The possible confounding effect of some sociodemographic variables was likely removed through the means
of multiple logistic regression analysis. Although patients with
hepatitis A could not be considered to be absolutely the best
control subjects, their inclusion, which also allowed us to save
time and expenses, represents a valid and feasible choice in this
study. Third, because HBsAg-positive, IgM anti-HAV–negative
individuals who reported injection drug abuse or promiscuous
sexual activity were tested more frequently for HDV than were
other risk groups, the role of these risk factors in HDV spread
was likely overemphasized in our study population. Conversely,
we possibly underestimated the risk associated with the other
exposures as a result of underreporting and the relatively low
percentage of HBsAg-positive, IgM anti-HAV–negative individuals with hepatitis who were tested for HDV. This does not
seem to greatly influence the incidence rate estimate, but it
might affect the generalizability of our data concerning the
relative role of the various risk factors.
The present study shows that, although acute HDV infection
is decreasing in Italy, it still exists, and its incidence is probably
underestimated. The lack of HDV testing in HBsAg-positive
individuals represents a culpable negligence of the diagnostic
procedure. Because both HDV superinfection and coinfection
worsen the prognosis of HBV infection [2, 21, 22], and because
nucleoside/nucleotide analogues are ineffective in chronic
HDV-HBV hepatitis [23], clinicians should be aware of the
need for performing an HDV test in every case in which an
HBsAg-positive result is detected. Furthermore, routine performance of this test would allow for a better definition of the
relative role of various risk factors.
Risk factors for HBV and HDV infections are well known
to the scientific and medical communities; however, in every
geographical area, these factors undergo modifications over
time that are often linked with changes in socioeconomic, cultural, and sanitary conditions. Monitoring these modifications
through a local surveillance system is crucial for evaluation of
the impact of the acting preventive measures and to appropriately address the introduction of new preventive measures.
In conclusion, improvements in socioeconomic conditions,
a decrease in the prevalence of HBsAg, and the anti-HBV vaccination campaign are probably the most important determinants of the downtrend of acute HDV infection in Italy. IDUs
are still at high risk of HDV infection and are at increased risk
of mortality because of the high prevalence of underlying
chronic liver diseases, such as hepatitis C and alcohol abuse.
Thanks to the anti-HBV vaccination campaign, however, most
Italians !24 years of age at this time—including young IDUs—
are protected against both HBV and HDV. Further efforts are
needed to increase vaccine coverage in high-risk groups, such
as older IDUs and members of households of HBsAg-positive
carriers. Improvements are also needed in the safety of invasive
procedures that are performed both inside and outside health
care facilities.
THE SISTEMA EPIDEMIOLOGICO INTEGRATO
DELL’EPATITE VIRALE ACUTA (SEIEVA)
COLLABORATING GROUP
F. Marzolini (Isituto Superiore di Sanità, Rome); L. Sudano (Unità Sanitaria Locale Valle d’Aosta, Aosta); J. Simeoni and M.
Fischer (Azienda Sanitaria Locale [ASL], Provincia Autonoma di
Bolzano, Bolzano); V. Carraro and C. Grandi (ASL, Provincia
Autonoma di Trento, Trento); G. Lanci (Università di Torino,
Torino); M. Meda and G. Vizzani (ASL Torino, Torino); A. Gallone and L. Muratori (ASL Collegno, Torino); M.T. Galati and
L. Signorile (ASL Chiasso, Torino); G. Valenza, Scala, and P.
Castagno (ASL Chieri, Torino); M.P. Alibrandi and E. Bresaz
(ASL Ivrea, Torino); A. Incarbona and E. De Luca (ASL Pinerolo, Torino); V. Silano and A. De Simone (ASL, Vercelli);
G. Ara (ASL, Biella); A. Cipelletti (ASL, Novara); S. Iodice (ASL
Verbania, Vercelli); A. Costantino and M. Boscardo (ASL Cuneo, Cuneo); A. Bertorello (ASL Mondovı̀, Cuneo); R. Barberis
(ASL Savigliano, Cuneo); F. Giovanetti (ASL Alba, Cuneo);
M.A. Marchisio (ASL, Asti); B. Rizzi (ASL Tortona, Alessandria); R. Prosperi and F. Fossati (ASL Acqui Terme, Alessandria); M. D’Angelo (ASL Casale Monferrato, Alessandria); S.
Crippa and G. Altomonte (ASL 3 Monza, Milano); F. Zorzut
(ASL, Trieste); L. Donatoni (ASL, Gorizia); T. Gallo and G.
Brianti (ASL, Udine); E. Zamparo (ASL, Pordenone); A. Ferro,
G. Niero, and F. Bressani (Assessorato Sanità, Venezia); R. Mel
(ASL, Belluno); M. Soppelsa (ASL Feltre, Belluno); M. Sforzi
(ASL Bassano del Grappa, Vicenza); F. Russo (ASL Thiene,
Vicenza); E. De Stefani (ASL Arzignano, Vicenza); A. Todescato
(ASL Vicenza, Vicenza); P. Paludetti (ASL Pieve di Soligno,
Treviso); D. Rizzato (ASL Montebellunia, Treviso); M. Forte
(ASL Treviso, Treviso); L. Nicolardi (ASL San Donà di Piave,
Venezia); G. Marchese (ASL Zelarino, Venezia); G. Marchioni
(ASL Dolo, Venezia); M. Boscolonata (ASL Chioggia, Venezia);
G. D’Ettore (ASL Camposampiero, Padova); L. Gottardello
(ASL Padova, Padova); R. Bisi (ASL Conselve, Padova); L. Gallo
(ASL Rovigo, Rovigo); L. Caffara (ASL Adria, Rovigo); G. Zivelonghi (ASL Verona, Verona); S. Soffritti (ASL Legnago, Verona); M. Foroni (ASL Villafranca, Verona); R. Rangoni, B.
Borrini, A. Cappelletti, A. Furini, and C. Ancarani (Assessorato
Sanità, Bologna); A. Capra (Azienda Unità Sanitaria Locale
[AUSL], Piacenza); B.M. Borrini, F. Zilioli, S. Copelli, A. Alessandrini, and A. Morelli (AUSL, Parma); P. Camerlengo, M.
Greci, L. Monici, A. Bertazzoni, M. Bigliardi, and S. Santacroce
(AUSL, Reggio Emilia); L. Gardenghi, G. Casaletti, R. Cagarelli,
F. Errani, A. Lambertini, A. Bulgarelli, D. Vaccina, A. Casolari,
and A.M. Pezzi (AUSL, Modena); R. Todeschini, C. Andalo’,
G. Domenicani, A. Calzolari, E. Dalle Donne, M.L. Dimaggio,
and M. Fabbri (AUSL, Bologna); N. Savoia (AUSL, Imola); F.
Taddia, M. Cova, and A. Califano (ASL, Ferrara); G. Savelli,
G.P. Casadio, and B. Calderoni (AUSL, Ravenna); E. Fiumana
(AUSL Forlı̀, Forlı̀); M. Franceschini (AUSL Cesena, Forlı̀); A.
Pecci (AUSL, Rimini); M. Mela (ASL, Imperia); M.P. Briata
and R. Carloni (ASL, Savona); W. Turello (ASL 3, Genova); G.
Zoppi (ASL 4 Chiavari, Genova); A. Bertone (ASL, La Spezia);
F. Mazzotta and G. Graziani (Assessorato Sanità, Firenze); L.
Barani and G. Ghiselli (AUSL, Massa-Carrara); A. Di Vito (USL
Capannoni, Lucca); F. Mazzoli and W. Wanderlingh (AUSL,
Pistoia); A.C. Epifani (AUSL, Prato); N. Galletti and F. Sbrana
(USL Pontedera, Pisa); M. Battaglini, G. Marinari, A. Lombardi,
and L. Genghi (AUSL di Piombino, Livorno); J. Lezzi (AUSL
di Poggibonsi, Siena); F. Verdelli, R. Bindi, E. De Sanctis, A.
Beltrano, and R. Conti (AUSL, Arezzo); G. Boncompagni, P.
Piacentini, M. Di Cunto, T. Papalini, and L. Incandela (AUSL,
Grosseto); S. Baretti, R. Cecconi, C. Staderini, and G. Ciampi
(AUSL Firenze, Firenze); P. Filidei (AUSL Empoli, Firenze); C.
Raffaelli (AUSL di Viareggio, Lucca); E. Carducci (Assessorato
Sanità, Ancona); A. R. Pelliccioni (ASL, Pesaro); G. Peccerillo
(ASL, Urbino); M. Agostini (ASL Fano, Ancona); R. Rossini
(ASL Senigallia, Ancona); G. Grilli (ASL Jesi, Ancona); N. Burattini (ASL Fabriano, Ancona); P. Marcolini (ASL Ancona, Ancona); R. Passatempo (ASL Civitanova Marche, Macerata); F.
Migliozzi (ASL Macerata, Macerata); G. Moretti (ASL di Camerino, Macerata); M. Nucci (ASL di Fermo, Ascoli Piceno);
Epidemiology of HDV Infection in Italy • CID 2007:44 (1 February) • e23
S. Impulliti (ASL San Benedetto del Tronto, Ascoli Piceno); C.
Angelini (ASL Ascoli Piceno, Ascoli Piceno); G. Taccaliti
(AUSL, Teramo); G. Guaitini and A. Tosti (Assessorato Sanità,
Perugia); A. Pasquale, M. Rossi, and A. Buscosi (ASL Città di
Castello, Perugia); O.C. Penza and C. Ciani (ASL Perugia, Perugia); F. Santocchia (ASL Foligno, Perugia); M.L. Proietti
(ASL, Terni); A. Romizi (ASL Orvieto); A. Corbo, D. Corpolongo, P. Porcelli, A. Ruta, and G. Ciarlo (ASL, Latina); P.
Dionette and R. Guadagnali (ASL, Rietl); A. Ercole, G. Esterni,
A. Piccoli, R. Boggi, P. Russo, P. Napoli, M.G. Catacchio, C.
Cerocchi, D. Linari, P. Grillo, M.T. Tedeschi, C. Gnesivo, M.
Loffredo, V. Labriola, A. Pendenza, E. Tanzariello, E. Santarelli,
P. Bueti, Ferrucci, S. Ursino, Torchia, L. Santucci, F. Alicata, S.
Viafora, R. Cecere, F. Pepe, P. Patti, M. Pesoli, C. Aiello, M.
Ruiu, F. Beato, and M. Mathis (ASL, Roma); P. Vilecco, D. De
Santis, S. Aquilani, A. Passini, and O. Micali (ASL, Viterbo);
V. Pagano (Osservatorio Epidemiologico Regionale, Napoli); N
Minichiello and M.A. Ferrara (ASL Avellino); A. Citarella and
E. Fossi (ASL, Benevento); C. Bove, M. Trabucco, and A.
D’Argenzio (ASL, Caserta); A. Simonetti, A. Parlato, R. Alfieri,
F. Peluso, R. Palombino, and F. Giugliano (ASL, Napoli); A.L.
Chiazzo, M.G. Panico, V. Della Greca, and A.M. Trani (ASL,
Salerno); S. Barbuti and M. Quarto (Università di Bari, Bari);
R. Matera, R. Colamaria, A. Madaro, G. Scalzo, and F. Avella
(ASL, Bari); E. Bellino (ASL, Brindisi); M. Masullo, I. Pagano,
and V. Di Martino (ASL, Foggia); A. Fedele, G. Turco, and S.
Minerba (ASL, Lecce); F.S. Negrone, M. Maldini, M. De Lisa,
E. Polani, and R. La Vecchia (ASL, Potenza); F. Mauri and T.
Russo (ASL Montalbano Ionico, Matera); F. Aloia and S. Giuffrida (ASL Castrovillari, Cosenza); G. Peritore and R. Mangione
(AUSL Licata, Agrigento); M. Cuccia and S. Rinnone (AUSL,
Catania); S. Cau (ASL Sassari, Sassari); A. Saba (ASL Olbia,
Sassari); D. Fracasso and O. Frongia (ASL Macomer, Nuoro);
and M.V. Marras and G. Mereu (ASL, Caligari).
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Acknowledgments
Financial support. Italian Ministry of Health (“Sorveglianze Speciali”).
Potential conflicts of interest. All authors: no conflicts.
References
1. Rizzetto M, Canese MG, Aricò S, et al. Immunofluorescence detection
of new antigen-antibody system (delta/anti-delta) associated to hepatitis B virus in liver and in serum of HBsAg carriers. Gut 1977; 18:
997–1003.
2. Fattowich G, Giustina G, Christensen E, et al. Influence of hepatitis
delta virus infection on morbidity and mortality in compensated cirrhosis type B. Gut 2000; 46:420–6.
3. World Health Organization. Hepatitis delta, 2001. Available at: http:
e24 • CID 2007:44 (1 February) • Mele et al.
20.
21.
22.
23.
//www.who.int/csr/disease/hepatitis/whocdscsrncs20011/en/. Accessed
13 December 2006.
Smedile A, Lavarini C, Farci P, et al. Epidemiologic patterns of infection
with the hepatitis B virus-associated delta agent in Italy. Am J Epidemiol
1983; 117:223–9.
Hadziyannis SJ. Review: hepatitis delta. J Gastroenterol Hepatol
1997; 12:289–98.
Gaeta GB, Stroffolini T, Chiaramonte M, et al. Chronic hepatitis D: a
vanishing disease?—an Italian multicenter study. Hepatology 2000; 32:
824–7.
Navascues CA, Rodriguez M, Sotorrio NG, et al. Epidemiology of
hepatitis D virus infection: changes in the last 14 years. Am J Gastroenterol 1995; 90:1981–4.
Hadziyannis S, Dourakis SP, Papaioannou C, Alexopoulou A, Hadziyannis ES, Gioustozi A. Changing epidemiology and spreading modalities of hepatitis Delta virus infection in Greece. Prog Clin Biol Res
1993; 382:259–66.
Mele A, Rosmini F, Zampieri A, Gill ON. Integrated epidemiological
system for acute viral hepatitis in Italy (SEIEVA): description and
preliminary results. SEIEVA Collaborating Group. Eur J Epidemiol
1986; 2:300–4.
Stroffolini T, Ferrigno L, Cialdea L, et al. Incidence and risk factors of
acute delta hepatitis in Italy: results from a national surveillance system.
J Hepatol 1994; 21:1123–6.
D’Amelio R, Matricardi PM, Biselli R, et al. Changing epidemiology
of hepatitis B in Italy: public health implications. Am J Epidemiol
1992; 135:1012–8.
Mele A, Stroffolini T, Zanetti AR. Hepatitis B in Italy: where we are
ten years after the introduction of mass vaccination. J Med Virol
2002; 67:440–3.
Stroffolini T. The changing pattern of hepatitis B virus infection over
the past three decades in Italy. Dig Liver Dis 2005; 37:622–7.
Pendino GM, Mariano A, Surace P, et al. Prevalence and etiology of
altered liver tests: a population-based survey in a Mediterranean town.
Hepatology 2005; 41:1151–9.
Sagnelli E, Stroffolini T, Ascione A, et al. The epidemiology of hepatitis
delta infection in Italy. J Hepatol 1992; 15:211–5.
Sagnelli E, Stroffolini T, Ascione A, et al. Decrease in HDV endemicity
in Italy. J Hepatol 1997; 26:20–4.
Bialek SR, Bower WA, Mottram K, et al. Risk factors for hepatitis B
in outbreak of hepatitis B and D among injection drug users. J Urban
Health 2005; 82:468–78.
Zhang YY, Hansson BG. Introduction of a new hepatitis agent in retrospect: genetic studies of Swedish hepatitis D virus strains. J Clin
Microbiol 1996; 34:2713–7.
Mele A, Spada E, Sagliocca L, et al. Risk of parenterally transmitted
hepatitis following exposure to surgery or other invasive procedures:
results from the hepatitis surveillance system in Italy. J Hepatol 2001;35:
284–9.
Mariano A, Mele A, Tosti ME, et al. Role of beauty treatment in the
spread of parenterally transmitted hepatitis viruses in Italy. J Med Virol
2004; 74:216–20.
Pasetti G, Calzetti C, Degli Antoni A, Ferrari C, Penna A, Fiaccadori
F. Clinical features of hepatitis delta virus infection in a northern Italian
area. Infection 1988; 16:345–8.
Shattock AG, Irwin FM, Morgan BM, et al. Increased severity and
morbidity of acute hepatitis in drug abusers with simultaneously acquired hepatitis B and hepatitis D virus infections. Br Med J (Clin Res
Ed) 1985; 290:1377–80.
Niro GA, Rosina F, Rizzetto M. Treatment of hepatitis D. J Viral Hepat
2005; 12:2–9.
E R R ATA
In an article published in the 1 February 2007 electronic
issue of the journal (Mele A, Mariano A, Tosti ME, Stroffolini
T, Pizzuti R, Gallo G, Ragni P, Zotti C, Lopalco P, Curale F,
Balocchini E, Spada E, on behalf of the SEIEVA Collaborating
Group. Acute delta hepatitis in Italy: incidence and risk factors
after the introduction of the universal anti–hepatitis B vacci-
nation campaign. Clin Infect Dis 2007; 44:e17–24), an error
appeared in figure 2B. The label for the vertical axis of the
figure is incorrect; the label should read “No. of cases per
100,000 population” (not “No. of cases per million population”). The corrected figure appears below. The authors regret
this error.
Figure 2. Yearly age-specific incidence rates of acute hepatitis delta virus (HDV) infection (A) and acute hepatitis B virus (HBV) infection (B) . Data
were collected by Sistema Epidemiologico Integrato dell’Epatite Virale Acuta (SEIEVA) during the period 1987–2004. Vertical dotted line, the beginning
of the hepatitis B vaccination campaign.
An error appeared in an article published in the 1 January
2007 issue of the journal (Elzi L, Schlegel M, Weber R, Hirschel
B, Cavassini M, Schmid P, Bernasconi E, Rickenbach M, and
Furrer H, for the Swiss HIV Cohort Study. Reducing tuberculosis incidence by tuberculin skin testing, preventative treat-
Clinical Infectious Diseases 2007; 44:1016
2007 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2007/4407-0029$15.00
DOI: 10.1086/516722
1016 • CID 2007:44 (1 April) • ERRATA
ment, and antiretroviral therapy in an area of low tuberculosis
transmission. Clin Infect Dis 2007; 44:94–102). In the last column of table 2, the P value for female sex should be .03 (not
.3). The authors regret this error.