Hepatitis C in health care professionals: prevalence and association with risk factors

OBJECTIVE: To evaluate the prevalence and risk factors for hepatitis C virus infection among health care professionals. METHODS: The study was carried out at a university hospital in the municipality of São José do Rio Preto, Southeastern Brazil, between January 1994 and December 1999. There were included 1,433 health care professionals, 872 administrative workers, and 2,583 blood donor candidates. Data were collected during admission examinations, periodic screenings, and after occupational accidents. Occupational and non-occupational information was obtained by means of a questionnaire. Data were analyzed using Chi-square and Kruskal-Wallis tests and logistic regression analysis. RESULTS: Prevalence of hepatitis C infection among health care professionals (1.7%) was signifi cantly higher than among administrative workers (0.5%; p=0.007) and blood donor candidates (0.2%; p=0.001). Regarding occupational factors, time in the job was signifi cantly longer (p=0.016) among health care professionals with positive serology than among those with negative serology. Multiple regression analysis showed a 50% increase in risk for every fi ve years of age. There was a signifi cant association between blood transfusion and positive serology among health care workers. CONCLUSIONS: Health care professionals show greater prevalence of hepatitis C than administrative workers and blood donation candidates. Among those with positive serology, occupational and non-occupational factors of greatest risk were age, time in the job, and blood transfusion.


INTRODUCTION
Hepatitis C is an important public health problem worldwide, not only due to the large number of individuals infected with the Hepatitis C virus (HCV).A considerable proportion of these individuals is informed of his or her disease status only when donating blood, which makes them an important link in the viral transmission chain, thereby perpetuating the disease.Over 180 million people globally are chronic carriers of HCV, 2 million of which are in Brazil.
HCV infection is universally distributed, and can be transmitted both sexually and from mother to child. 16High prevalence is associated with the so-called "risk groups" -including hemophiliacs, hemoldiayzed patients, multiply transfused patients, and drug addicts. 4e epidemiological profi le of HCV infection is as complex as the natural history of the disease caused by this viral agent.Present in blood in low titers, the major mechanisms of transmission are infected blood and blood products.
In the United States, prospective studies conducted by the National Institutes of Health 1 show a decrease in HCV infection following the introduction of anti-HCV testing.According to a study by Chiaramonte, 6 the risk of infection by blood transfusion is currently 1:103,000 -i.e., almost null -very different from what was seen between 1960 and 1991.In Brazil, the distribution of incidence and prevalence of HCV is not known.Infection is known to be predominant among young adults, and susceptibility seems to be general.Recent studies have identifi ed that 75% of individuals with HCV were infected by the parenteral route, and that infection may be apparent, unapparent, direct, or indirect. 7HCV transmission by the unapparent direct parenteral route is likely to be related to the familial environment.On the other hand, parenteral unapparent indirect transmission may be related to contact with instruments of personal use or to the use of tools or instruments contaminated with infected blood.* The concern about biological hazards began with the realization that professionals who worked in laboratories where microorganisms were manipulated were at greater risk of infection.However, only with the onset of the AIDS epidemic in the 1980's did security norms for the work environment become fi rmly established. 22On the other hand, the risk of transmission of HCV after a needlestick injury can be up to 10 times greater than that of HIV, and, in contrast to the latter, post-exposure preventive measures are unavailable for HCV. 10,14,20though small in number, cases of infection by percutaneous exposure have been documented, 11 including infection through contamination of the conjunctiva with blood. 21The rate of transmission of Hepatitis C after occupational exposure ranges from 0 to 10%. 18,23sk factors for occupational transmission are still poorly defi ned.There is also, to this date, no information on the survival of HCV virions in any given environment.
The occurrence of HCV among health care professionals ranges from 2% to 10%, and risk of infection is correlated with time in the job, performing invasive procedures, and history of percutaneous accidents.Such variation in incidence may be related to the method used for diagnosis, especially after accidents involving HCV-positive patients. 18e aim of the present study was to explore the magnitude of HCV infection and determine its prevalence among health care professionals, as well as to investigate potential risk factors for infection, including blood transfusions, parenteral substance use, accidents with biological material, and others.

METHODS
The study sample comprised different categories of professionals working in a teaching hospital in São José do Rio Preto, in the state of São Paulo, Southeastern Brazil, and blood donor candidates from the same city.This is a medium-sized hospital, comprising 554 beds.
Between January 1994 and December 1999, 1,444 health professionals worked in the hospital.These professionals either had direct contact with, or manipulated objects used by, patients and worked in areas of disease diagnosis and patient accommodation.Professionals on leave due to disease (N=11) were excluded from the sample, totaling 1,433 subjects.
For comparison of the prevalence of HCV infection, we used two control groups.The fi rst consisted of employees of the hospital's administrative area (N=872), who had no contact with patients, or with material used by them.The second group comprised blood donor candidates (N=2,583) who were tested for the presence HCV antibodies between 1-31 May 1999.The donor candidates had no connection to the hospital.Information regarding the serological status of donor candidates was provided by the blood center.
To determine the prevalence of Hepatitis C, we collected data obtained during admission examinations, yearly follow-up tests, or tests performed following work accidents.These data were available in medical charts compiled by the hospital's occupational medicine sector.Time in the job was defi ned as duration of employment in that hospital.
Sectors were classifi ed according to risk of contamination into maximum risk (hemodialysis, blood center, hemodynamics, blood bank, infectious-contagious diseases, emergency room, laboratory, intensive care unit, transplantation unit); medium risk (clinical, surgical, and pediatric infi rmary, outpatient service, endoscopy, laundry (dirty area), peritoneal dialysis, and surgical center), and minimum risk (operational central, graphic methods, cleaning/hygiene center, radiology center, chemotherapy, and garbage collection).
All health care professionals had provided blood samples for anti-HCV testing.Laboratory diagnosis was carried out by second-generation enzyme-linked immunosorbent assay (ELISA) for serum anti-HCV antibodies.
Work accidents were defi ned as those taken place and notifi ed during the subject's work hours at the hospital.These were characterized as any percutaneous injury with a sharp object, contact with biological materials (blood, liquor, amniotic and pleural fl uid, urine, or feces) on non-integral or mucosal epithelium, and patient bites.
In case of work accidents, professionals were administered HCV testing on three occasions: on the date of the accident and six and twelve months after exposure to biological material from known or unknown source-patients.
We interviewed all health care professionals with notifi ed work accidents and who tested positive for HCV.In this interview we collected specifi c data on the accident, including the serology of the source-patient, type of exposure and biological material, and use of personal protection equipment.We also investigated non-occupational epidemiological variables, such as contact with close relatives with Hepatitis C, hepatitis after age 12, homosexuality, sexual promiscuity, blood transfusion prior to 1993, intravenous drugs, and alcoholism.
In univariate analysis, groups were compared using chi-squared, ANOVA, and Kruskal-Wallis tests.When more than two groups were compared, Bonferroni's correction for multiple comparisons was used.The models used for multivariate analysis by logistic regression included all variables with signifi cant association in univariate analysis.We also included the most important variables from the clinical (sex, age, professional category, sectors according to risk, positive anti-HCV serology) and non-occupational (sexual promiscuity, hepatitis after age 12, alcoholism, blood transfusion prior to 1993) perspectives.We evaluated two models that analyzed dependent variables positive anti-HCV antibodies and work accidents with biological material, using adjusted odds ratios and their respective confidence intervals (IC).The level adopted for statistical signifi cance was 5%.
Table 1 presents the distribution of health care professionals according to the risk of contamination of the sector in which they worked.
The prevalence of HCV among health care professionals, administrative workers, and blood donor candi-dates is presented in Table 2. Health care professionals showed signifi cantly greater prevalence of infection than blood donor candidates (χ 2 =27.52; p=0.001) and administrative workers (χ 2 =7.22; p=0.007).Time working in the hospital among professionals positive for HCV was signifi cantly greater (p=0.016)than among those with negative serology.According to multivariate analysis, the odds of infection increased by 50% for every fi ve years of age among health care  is close to that reported in the literature using second generation tests.Prevalence was signifi cantly greater than that found among administrative workers and blood donor candidates, confi rming that health care professionals are at higher risk of HCV infection. 11,18,23igueredo et al, 8 in an analysis of the literature on the subject, found that the frequency of HCV among health care professionals was relatively low.
Manipulation of material contaminated with blood or secretions is inherent to the fi eld of health care.What is problematic, however, is that professionals often manipulate materials incorrectly, thereby increasing the risk of accidents.The serological follow-up of professionals following work accidents, which lasted for at least one year after exposure, failed to detect any instances of seroconversion.
Health care professionals with positive serology had worked in the institution for longer periods, and had 50% greater chance of being anti-HCV-positive for every fi ve years in the job.This result suggests that daily interaction with patients may contribute to increased prevalence of HCV infection.Moreover, this risk may be even greater if professionals manipulate patients incorrectly.
The presence of work accidents was not signifi cantly associated with positive anti-HCV, corroborating the fi ndings of Polish et al. 17 Also according to these authors, needlestick injuries were an independent risk factor for positive anti-HCV among health care professionals.Mitsui et al also detected such an association. 14sitive anti-HCV was associated with blood transfusion prior to 1993, when specifi c serological testing for HCV-induced antibodies was not yet available.In Brazil, Medeiros et al 13 (2004) found similar results among hemodialyzed patients.Alter, 2 in a recent study of transfusions and transplantation, reports that this risk has been practically eliminated.
Work accidents occurred not only due to the manipulation of material contaminated with blood or bodily fl uids (needles and other sharp instruments), but also due to contact with mucosae, non-integral skin, and patient bites.The transmission of HCV by human bites has been reported, especially in case of deep bites administered by HIV-positive source patients. 2 the present survey, the proportion of occurrence of injuries involving blood or bodily fl uids was high.However, it is known that many of these accidents are not notifi ed.The proportion of health care professionals that had suffered accidents was 12.7% lower than that found in a survey with a random sample of 254 professionals carried out at the beginning of the study to evaluate underreporting.
Of all health care professionals evaluated (N=1,433), 342 (23.9%) were involved in a total of 456 notifi ed work accidents with biological material.Of the source patients for these accidents, 21% had positive serology for HIV, HBsAg, or anti-HCV; 15.6% (N=71) were negative, and status was unknown for 63.4% (N=289).
According to risk of contamination (Table 4), more accidents were notifi ed (52.2%;N=238/456) in maximum risk sectors, followed by medium risk (32.4%;N=148/456) and minimum risk (15.4%;N=70/456) sectors.The awareness of risk in the maximum and medium risk sectors may result in slight underreporting in these areas, which is not likely to have occurred in the minimum risk area.
Accidents most frequently involved blood (70.2%), followed by other bodily fl uids (20%).We were unable to identify the biological material involved in 9.8% of cases.
Regarding the use of personal protection equipment, professionals were wearing gloves in 74.3% of notifi ed accidents.
HCV serology on the date of the accident identifi ed 2.3% of professionals as positive for anti-HCV.Followup tests indicated no cases of serologically demonstrated infection.
Evaluation of work accidents by multivariate analysis showed that males are 28% more protected from work accidents than females (OR=0.724;95% CI: 0.528;0.993;p=0.044).Compared to the sector of minimum risk, risk of accidents is 1.14 times higher in the maximum risk sector (OR=2.14;95% CI: 1.34;3.42;p<0.001) and 0.92 times higher in the medium risk sector (OR=1.92;95%: 1.10;3.33;p=0.020).Regarding time in the job, the chance of suffering an accident increases 4% for each additional year of work (OR=1.04;95% CI: 1.01;1.08;p=0.014).

DISCUSSION
Health care professionals in the studied hospital showed 1.7% prevalence of positive anti-HCV serology, which Regarding age, there was no signifi cant difference in mean age between health care professionals that did or did not suffer work accidents.It was expected that the younger group, considered to be less experienced, could be at higher risk.Underreporting of work accidents may have been higher in this group due to fear of dismissal.
On the other hand, this was not the case for time in the job, with professionals who suffered accidents showing signifi cantly greater permanence.With the increasing experience that results from the years of work in the institution, nursing professionals usually perform tasks of greater risk.This may contribute to greater exposure to risk situations for work accidents.Therefore, it is necessary to carry out recycling programs addressing issues of universal precautions, and continued education aiming at preventing accidents.In addition, multivariate analysis showed that risk of accidents increases 4% for every additional year in the job.
In the present study, males were 28% more protected from work accidents than females.In certain sectors, male health care professionals perform specifi c tasks requiring physical strength as well as the care for male patients (catheterization, bathing, moving, and others).Carrying out such tasks may reduce exposure to sharp objects, suggesting that risk increases along with the frequency of contact with such objects.
In the present study, the hemodialysis unit is among those with greatest frequency of accidents.Procedures in this unit are constant, and a high percentage of dialyzed patients are positive for Hepatitis B and C. Another important sector is the emergency room, especially given that the studied hospital is a reference center for the entire Northeastern Region of Sao Paulo State.The large number of patients and the emotional stress inherent to emergency care contribute enormously to the high risk of accidents observed in this sector.The maximum risk sector, which includes emergency, hemodialysis, intensive care, and other sectors, accounted for 52.2% of accidents.The chance of suffering a work accident in these sectors was 1.14 times higher than in the minimal risk sector (the reference category).
According to Lee et al, 12 emergency and surgery rooms are the sectors with highest risk of accidents, which is justifi ed by the type of procedures and techniques employed in these sectors.
It is important to characterize the types of work accidents suffered by health care professionals, so as to avoid high risk situations.In the present study, we found that 78.9% of accidents involved percutaneous injuries.This result is in agreement with data from the literature, which indicate sharp objects as the major source of work accidents among health care professionals. 15,19his can be explained by the type of service provided by these professionals, especially medical and nursing activities, which show high rates of accidents related to invasive procedures.Therefore, the use of devices that prevent the manipulation or recapping of sharp objects such as needles should be encouraged.According to Garner, 9 such devices can reduce by 50% the occurrence of percutaneous lesions.
Blood was the source of infection in 70.2% of accidents.The risk of work accidents involving large amounts of blood and deep lesions, as in the case of hollow needles of large caliber, also increase the risk of seroconversion. 5e universal precautions recommended by the Centers for Disease Control and Prevention are in use since 1987.These recommendations include the use of gloves for manipulating blood and secretions, of lab coats when there is risk of contamination of the professional's skin or clothing, ant the use of goggles and mask in case of risk of splattering of blood or secretions on oral, nasal, or ocular mucosae.Many accidents can be prevented by the use of such barriers.
The present study showed that the majority of subjects (74.3%) wear gloves, although a proportion of these do so only in maximum risk sectors (emergency and surgery rooms, hemodialysis, etc.), an distinction which does not make sense given the profusion of asymptomatic carriers.Studies in the literature suggest that this frequency could be higher still if educational measures were adopted. 18ere were no cases of HCV seroconversion among the accidents reported in this survey.A similar result was found by Baldo et al. 3 Seroconversion will depend on work accidents involving sharp objects, on the percentage of non-use of safety equipment, and on the amount of accidents involving blood.A total of 78.9% of work accidents were percutaneous injuries, of which 70.2% involved blood, and 25.7% of professionals involved were not using safety equipment, both of which could favor seroconversion.However, seroconversion was not detected.The probability of seroconversion following exposure to biological material from source-patients positive for Hepatitis C ranges from zero to 7%. 18,23ven the reasons presented above, and the documented transmission of pathogens through small lesions, the use of gloves and other safety equipment is recommended as one of the primary components of universal precautions.Also, the use of instruments and needles that include safety devices, especially if within the context of a wide-ranging prevention program, can lead to an important re duction in risk of exposure.A mediumterm cost-benefi t analysis will certainly be able to provide evidence of the relevance of occupational and preventive medicine.

Table 1 -
Distribution of health care professionals according to the risk of contamination of the sectors in which they work.São José do Rio Preto, Southeastern Brazil, January 1994 -December 1999.

Table 2 -
Prevalence of hepatitis C among health care professionals, administrative workers, and blood donor candidates.

Table 3 -
Final multivariate analysis model for health care professionals with positive anti-HCV serology.São José do Rio Preto, Southeastern Brazil, January 1994 -December 1999.

Table 4 -
Occurrence of accidents according to risk of contamination and sector of activity.São José do Rio Preto, Southeastern Brazil, January 1994 -December 1999.