Nosocomial transmission of bloodborne viruses

OCD

CANADIAN JOURNAL OF INFECTIOUS DISEASES
July/August 2003, Volume 14, Number 4: 192


Nosocomial transmission of bloodborne viruses from infected health care
workers to patients

B Lynn Johnston MD1, John M Conly MD2

The risks to health care workers (HCWs) of occupationally-acquired infection
with hepatitis B virus (HBV), human immunodeficiency virus (HIV) and, to a
lesser extent, hepatitis C virus (HCV) have been reasonably well quantified
(1). Evidence from the HIV and HBV experience suggests that the risk of
infection is increased where the level of viremia is high, as manifested by
high HIV viral load or the presence of hepatitis B e antigen (HBeAg) (1). It
has also been recognized that patients may acquire one of these viruses
following significant exposure to the blood of an infected HCW (2-4). While
the magnitude of this risk to patients is considerably less than that to
HCWs, the 1990 report by the Centers for Disease Control and Prevention
(CDC) that a Florida dentist had transmitted HIV to patients in the course
of dental care triggered widespread public concern about the risk of
infection from HCWs. In 1991 CDC published recommendations for preventing
HIV and HBV transmission to patients, which included the recommendation that
HCWs who are infected with HIV or HBV (and HBeAg positive, a marker of
higher infectivity) should not perform exposure-prone procedures unless they
have sought counsel from an expert review panel (5). In 1998, Health Canada
published guidelines for the management of HCWs infected with HBV, HCV,
and/or HIV (6). Both these documents generated controversy at the time of
their publication. Since that time, however, several provincial regulatory
bodies have formed committees to advise physicians infected with these
bloodborne pathogens (BBPs) regarding their practice. This article reviews
what we know about the transmission of HBV, HCV and HIV from infected HCWs
to patients in medical and dental settings.

Hepatitis B

Several studies have examined the prevalence of HBV markers in HCWs. A 1985
study in Italy found that prevaccination immunological markers of current or
past HBV infection were found in 23.3% of HCWs tested, with 1.8% being HBV
surface antigen positive (7). The prevalence of serological markers for HBV
in Chicago HCWs in 1983 was 15.2%, with 1% HBV surface antigen positive (8).
These results are similar to a previous study conducted between June 1977
and October 1978 of 624 HCWs in Boston (9). A more recent prevaccination
study of HCWs in San Francisco (1984 to 1992) found that 21.7% had markers
of HBV infection (10). While it is expected that the prevalence of HBV in
HCWs will go down with community and occupational HBV vaccination programs,
in the immediate future it is anticipated that there will continue to be
chronically infected HCWs who have come from endemic areas or previously had
other non-occupational exposures.

The first reported transmission of HBV from HCW to patient was in 1972 (11).
In that study, cases of icteric hepatitis in 11 patients were linked to a
nurse with HBV, but the mode of transmission was not determined (11). Since
that time, there have been a number of isolated transmissions and outbreaks
of HBV related to patient exposures to infected HCWs (Table 1). The
methodology used in these studies has differed considerably and many of the
early studies are flawed by several factors: the vast majority were
retrospective; case finding was confined to symptomatic patients; only a
minority of exposed patients had serological testing; and it was often
impossible to distinguish infection related to HCW exposure from that due to
other exposures. While more recent studies have been rigorous in their
follow-up of potentially exposed patients, there still remains the
difficulty of distinguishing the source of hepatitis B infection in those
who no longer have detectable viremia for molecular typing. To date,
therefore, it has not been possible to precisely quantify the risk of
transmission to patients from a significant exposure to an infected HCW.
However, it is apparent that a number of infected HCWs have been associated
with nosocomial transmissions to patients since 1972 (2,11-34).

In an effort to develop a better understanding of the risk of nosocomial HBV
transmission, investigators in the 1970s and 1980s did serological testing
on patients cared for by HBV carrier HCWs who had not been associated with
the transmission of clinical infection (35-37). None of these studies
demonstrated transmission of HBV from infected HCWs to patients. However,
these studies had limitations that could have affected their results. The
first study involved no surgical personnel (35). The second study had
complete serological evaluation on only 28 of 49 patients operated on by an
orthopedic resident with acute hepatitis B (36). In the third study, of 213
patients exposed to six chronic carrier physicians (two surgeons), it could
not be determined how many were exposed to the surgeon who was HBeAg
positive (37).

Some important observations can be made by examining the studies that have
looked at transmission of HBV from infected HCWs to patients. The limited
data suggest that HBV-infected HCWs who are not involved in invasive
procedures do not transmit infection (35,37) unless there are breaches of
infection control (32). It would appear that not all HBV- infected HCWs who
perform invasive procedures transmit infection to patients (36). However,
HBV-infected HCWs have been associated with nosocomial transmissions
(2,12-34). Physicians associated with transmissions have been at all levels
of experience including trainees, clinical assistants and senior surgeons.
Technical expertise or competence has not usually been identified as a
contributing factor in the transmissions.

In the literature on nosocomial HBV transmissions, a few specialties appear
to be over-represented, including dentistry and oral surgery, obstetrics and
gynecology, and cardiothoracic/cardiovascular surgery. It has been suggested
that these are specialties where surgery involves one or more of the
following: operating in a confined and/or poorly visualized field, palpation
of needles or other sharp objects, or frequent exposure to sharp objects,
all of which increase the potential for the HCW's blood to come into contact
with the patient's blood or open tissues. This has led to the terminology
"exposure-prone procedures," first coined by the CDC in 1991 (5).
Characteristics of exposure-prone procedures include digital palpation of a
needle tip in a body cavity or the simultaneous presence of the HCW's
fingers and a needle or other sharp instrument or object in a poorly
visualized or highly confined anatomical site (5). There have been no
reported HBV transmissions related to dental or oral surgery since the
implementation of universal precautions in 1987. Additionally, four of the
dentists/oral surgeons associated with outbreaks returned to practice with
their practices modified to include the routine use of gloves (13,15,17,18).
These observations highlight the importance and effectiveness of education
of the HCWs and the use of gloves in preventing the transmission of BBPs in
the dental setting.

Until 1997, all HCWs associated with HBV transmissions who were tested were
HBeAg positive, which is associated with higher levels of circulating virus
and greater infectivity (2,15,16,19-24,26,28-30,32,33). This was identified
as a marker for HCWs who should refrain from performing exposure-prone
procedures (5,25). Since that time, there have been several reported
transmissions involving physicians who were HBeAg negative, documenting that
they may also transmit infection (25,31,34). Detailed investigations of the
surgeons in these transmissions revealed that all carried a precore HBV
mutant (25,31,34). This mutation, located in the genome just before the gene
that codes for the core protein, prevents expression of the e antigen (34).
The mutation does not interfere with the replication of infectious virus and
some individuals with this precore mutant will have high levels of
circulating HBV DNA (25). It is not known what proportion of HBeAg-negative
carriers have the precore mutant, nor whether this mutation is more or less
likely to be associated with nosocomial transmissions (25).

The ability to accurately measure the serum HBV DNA level might allow a more
precise determination of the degree of infectivity of carriers. In-house
assays of blood from five HBeAg-negative surgeons found the lowest level of
HBV DNA to cause infection was 2.5 x 105 copies per ml (25,34). Based on
this, it has been suggested that the cutoff HBV DNA level below which an
HBeAg-negative carrier would become non-infectious would probably be around
104 HBV DNA copies per ml (38). It should be noted that the performance
characteristics of some commercial assays may not detect this level of
viremia (38).

More recent look-back surveys of patients exposed to infected surgeons have
provided an estimate of transmission risk ranging from 0.5% to 13%
(2,25,31,34). Risk factors for infection have not been identified, but can
be speculated to include the host's health, immune system and vaccine
status, as well as the route of infection, volume of inoculum, HBV DNA
level, and strain pathogenicity (38).