Historically, the highest risk for occupationally acquired hepatitis among HCPs has been associated with exposure to hepatitis B virus (HBV); in fact, before the advent of the hepatitis B vaccine, HBV infection was the major occupational risk to HCP.⁵ In the 1980s, the annual incidence of HBV infection among HCPs in the United States was staggering. The Centers for Disease Control and Prevention (CDC) estimated that in the mid-1980s, ˜12 000 HBV infections occurred annually in HCP who had frequent occupational exposure to blood or other potentially infectious materials, with an annual rate of infection between 4.89 and 6.63 per 1000 exposed susceptible providers.⁶ Of these 12 000 occupationally infected HCPs each year, CDC scientists estimated that 3000 developed symptomatic clinical illnesses, more than 600 were hospitalized, and more than 250 of these HCPs died. The CDC estimated that between 600 and 1200 of these healthcare providers became chronic hepatitis B carriers. Since the HBV vaccine was developed and aggressive hepatitis B vaccination of HCP in the United States has been promoted, HBV infections among HCPs have decreased dramatically to an estimated 100 annually by 2009.⁷

Numerous studies have documented that HCPs exposed to blood are at high risk for acquiring HBV infection. In one of the earliest studies, Williams and associates⁸ investigated a large epidemic of hepatitis B infections among HCPs and found that clinical hepatitis attack rates and HBV antibody prevalence rates correlated with occupational exposure to blood from patients being treated with hemodialysis. Transmission was thought to occur by both accidental parenteral and so-called “inapparent parenteral” routes of inoculation of contaminated blood. Pattison and co-workers⁹ studied providers in a large community hospital between 1972 and 1974 and found a significant association between frequency of blood contact and prevalence of HBV, but no association between frequency of patient contact and HBV prevalence. The first nationwide, cross-sectional seroepidemiological survey of occupationally acquired HBV infection among physicians was conducted by Denes and colleagues in 1975-1976.¹⁰ These investigators found that infection rates were higher among those practicing in urban settings, that the risk for infection increased with the number of years in practice, and that infection rates were highest among pathologists and surgeons. Dienstag and Ryan¹¹ studied workers at a large urban hospital and found that the prevalence of HBV serologic markers increased as a function of contact with blood, years in a healthcare occupation, and age, but not as a function of contact with patients, years of education, and previous needlestick, transfusion, or globulin
injection. The highest seroprevalences were found among emergency room nurses, pathology staff members, blood bank staff members, laboratory technicians, intravenous teams, and surgical house officers. Similar high-risk occupations (emergency room, medical and surgical intensive care units, and dentistry-oral surgery) were identified by Jovanovich and colleagues in a study conducted in an urban hospital.¹²

Snydman and associates¹³ conducted a multiinstitutional seroepidemiological survey of hospital employees in 1980 and 1981 and found that the duration of employment for laboratory workers, surgical staff members, and medical staff members was associated with increased risk for having HBV markers. In this latter study, the highest gradient of risk in these occupations occurred during the first 5 years of employment. Another large multiinstitutional study of nearly 5700 hospital employees conducted by Hadler and colleagues¹⁴ controlled for nonoccupational risk factors and confirmed the earlier findings of Dienstag and Ryan that occupational blood exposure, but not patient contact, was associated with risk for prior HBV infection. Hadler and coworkers also found that the frequency of needle accidents during daily work was directly related to HBV seroprevalence. The occupational group with the highest HBV infection rate was clinical laboratory and blood bank technicians, who routinely handled large numbers of blood specimens. In general, these and similar studies in the pre-HBV vaccine era may be summarized by noting that HCPs who have occupational exposure to blood had a prevalence of HBV markers several times both that of HCPs who did not have blood exposure as well as that of the general population. This prevalence of HBV infection increased with increasing years of occupational exposure. HBV infection was related to the degree and frequency of blood exposure and not to the degree of patient contact. West reviewed studies evaluating the risk for HBV infection in HCP and found the risk to be ˜4 times elevated when compared to the risk for infection in the at-large adult population.¹⁵ In West’s review, physicians and dentists were found to be 5-10 times more likely to experience hepatitis B infection and surgeons, dialysis personnel, personnel providing care for developmentally disabled individuals, and clinical laboratorians to be at 10-fold or higher risks for HBV infection.¹⁵

The risk of occupational exposure to HBV depends on several other factors besides occupation and frequency of occupational exposures. The prevalence of HBV infection in the patient population also influences the risk for occupational exposure. Because HBV prevalence is generally higher in urban settings, providers in urban hospitals have been found to be at higher risk for HBV infection¹⁰ than are providers in rural hospitals.¹⁶ Renal dialysis patients (see Chapter 35, Dialysis-Associated Infections) who require frequent blood transfusions and have suppressed immune responses have long been known to be at high risk, both for acquiring HBV infection and for developing chronic HBV infections. For this reason, staff caring for dialysis patients are at increased risk for occupational HBV infection.^15,17,18 HCPs in hospitals serving large numbers of other patient population groups at risk for HBV infection, such as intravenous drug users, homosexual men, prison inmates, the developmentally disabled, or immigrants from highly endemic areas, are also at higher risk for occupational exposure and infection with HBV.¹⁹ Patients who are asymptomatic HBV carriers are the primary reservoir for HBV infection in the healthcare setting. Broad-scale testing to identify infected patients is neither practical nor cost-effective. In one study, testing patients who reported a history of hepatitis would have detected fewer than 20% of HBV-infected patients.²⁰

The infectivity of the source material also influences the risk of acquiring HBV infection. Although hepatitis B surface antigen (HBsAg) has been detected in nearly all body fluids, blood is considered the most infectious and is probably responsible for most occupationally acquired infections. The infectivity of blood is generally correlated with the presence of increased circulating viral burdens, HBV DNA polymerase activity, or hepatitis B e antigen (HBeAg) in the blood. The risk for HBV infection after a percutaneous (“needlestick”) exposure to blood from an HBV-infected individual has been estimated to range from 19% to 37% if the donor blood is HBeAg-positive.^21,22 In the dental setting, saliva, particularly bloody saliva, is also considered to represent a substantial infectious risk.

The type of exposure to blood or other potentially infectious materials also influences the risk of acquiring infection. Percutaneous exposures, such as needlesticks or injuries with contaminated sharp instruments, are associated with the highest risks for occupational infection. Very small inocula of HBsAg-positive blood may produce infection, since the blood of acute or chronic HBV carriers may contain as many as 10¹³ virus particles of HBV per milliliter of blood.²³ Infectivity studies in chimpanzees have demonstrated that serum positive for HBeAg is infectious in dilutions up to 10⁸.²⁴ Despite the fact that percutaneous exposures are the most efficient route of infection, the CDC estimates that fewer than 20% of HBV-infected HCP recall an injury/exposure of this type.²⁵ Thus, other so-called “inapparent parenteral” exposures account for a substantial fraction of occupational HBV infections. Preexisting cuts, dermatitis, other skin lesions, or mucous membranes may provide portals of entry for HBV infection. Blood-contaminated inanimate objects or environmental surfaces also have been implicated in occupational transmission in certain settings. In one study, sustaining paper cuts while handling laboratory computer cards in a hospital clinical laboratory was associated with an outbreak of HBV infection.²⁶ Before strict infection prevention measures were implemented in hemodialysis centers, environmental contamination with blood that subsequently resulted in contaminated provider’s hands was hypothesized to facilitate HBV transmission.^27,28 Contamination of mucous membranes of the eye or mouth, which may occur with accidental splashes or pipetting accidents, also may result in HBV transmission.²⁹

In the past 25 years, seroprevalence studies in HCP have documented the importance of hepatitis B vaccine in preventing infections. Thomas and associates studied 943 HCPs in an inner-city hospital. Their multivariate analysis identified only one risk factor—absence of HBV vaccination—to be independently associated with HBV infection in this population of HCP.³⁰ Similarly, Panlilio and colleagues studied 770 surgeons for markers of HBV infection and found two risk factors—not receiving hepatitis B vaccine
and practicing surgery for at least 10 years—for HBV infection.³¹ Another study in 114 operating room personnel in Pakistan also documented that nonvaccinated providers were more likely to be infected with HBV.³¹ Supplementing these seroprevalence studies, Lanphear and coworkers investigated the incidence of clinical HBV infection in HCP and found a dramatic decrease associated with increased immunity due to vaccination.³²

Hepatitis D virus (HDV), formerly called the delta agent, is a defective virus that needs HBV as a helper virus. Thus, HDV may infect HCPs either as a coinfection with HBV (ie, a simultaneous exposure) or as a superinfection when HCPs already have HBV infection. The extent of HDV infection in HCP has not been determined, because HDV antibody testing is not routinely performed.³³ Even if HDV antibody screening was routine, the prevalence would be difficult to determine, because infection may elicit only a transient and low-titered response.³⁴ Nevertheless, anecdotal evidence suggests that occupational HDV transmission occurred in a hemodialysis nurse,³⁵ and evidence documents transmission to a surgeon following a deep needlestick injury.³⁶

Because of its dependence on HBV, the epidemiology and mode of transmission of HDV are similar to those of hepatitis B. Worldwide, ˜5% of HBsAg carriers are infected with HDV.³⁷ However, not all HBV-infected individuals have the same risk for HDV infection, because geographic and risk group distribution vary substantially. Patient populations that include HBV-infected persons from HDV-endemic areas, such as southern Italy, the Amazon basin, the Middle East, and certain Pacific islands, are more likely to be coinfected with HDV and, therefore, present a greater risk to HCP. Among risk groups for HBV infection, HBV-infected hemophiliacs, intravenous drug abusers, and hemodialysis patients are more likely to be coinfected with HDV than are homosexual men. A major benefit of the efficacy of the HBV vaccine has been a significant decrease in HDV infections in high prevalence areas.³⁸

Our current understanding of the role of hepatitis C virus (HCV) in occupationally acquired infections is less clear than for HBV and is complicated by the evolving understanding of the pathogenesis and immunopathogenesis of exposure and infection with this flavivirus. Since the parenteral mode of transmission of HCV has been clearly established as a primary route of infection for transfusion recipients and intravenous substance users, by analogy to HBV, occupational transmission of HCV in the healthcare setting—including transmission from patients to staff, from patient to patient, and from infected providers to their patients—is likely linked to apparent and inapparent parenteral exposure to blood. To date, exposure to blood remains the primary vehicle for occupationally acquired HCV infection as is evidenced by the overwhelming majority of the cases of occupational infection that have been described in the literature.^{39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55} HCV also has been transmitted by a punch.⁵⁶ HCV RNA has been detected in saliva,^57,58,59 and two cases suggest that transmission of HCV followed human bites.^60,61 Abe and colleagues also provided experimental documentation of HCV transmission by saliva.⁶² When present in saliva, HCV titers are lower than in blood. The potential infectivity of saliva may have important implications for patient to provider transmission, primarily in the dental healthcare setting. HCV RNA also has been detected in a variety of other body fluids from infected patients, including menstrual fluid,⁶³ semen,^59,64,65 urine,⁵⁹ spinal fluid,⁶⁶ and ascites.⁵⁹ The relevance of these latter body substances to the transmission of HCV is unclear. One recent study demonstrated transmission of HCV as a result of a nurse providing care for a patient with severe epistaxis and concluded that transmission occurred as a result of exposure of the nurse’s nonintact skin to the patient’s blood.⁶⁷ In summary, blood is the body substance that presents the most risk for HCV transmission in the healthcare setting. Despite the fact that transmission of HCV resulting from exposures to body fluids other than blood has not yet been documented, presumably because viral titers in these fluids are substantially lower than in blood, other body substances may present measurable risks for occupational infection, particularly if the healthcare provider is exposed by the parenteral route and/or receives a large inoculum.

Parenteral exposures represent the primary mode of occupationally acquired infection, as is evidenced by the overwhelming majority of the cases of occupational infection that have been described in the literature.^{43,44,45,46,47,48,49,50,51,52,53,54,55} However, two cases of HCV infection have been documented following mucosal exposures to blood,^68,69 and one case has been associated with exposure of “nonintact skin” to blood.⁶⁷ Extensive HCV environmental contamination of instruments and surfaces in hemodialysis^{70,71,72,73,74} and dental surgery settings⁷⁵ can occur, and such HCV environmental contamination has been suggested to play a role in transmission of HCV.⁷⁶ However, to our knowledge, transmission of a specific HCV strain through environmental surface contamination has not, yet, been specifically documented, although transmission via semicritical devices has been reported.⁷⁷ Transmission resulting from environmental contamination should be precluded if proper sterilization and disinfection procedures are practiced and if current standards of infection prevention, particularly hand hygiene, are followed.

Numerous cases of nosocomial transmission from patient to patient (often as a result of the use of cross-contamination from an index case, eg, in hemodialysis, from multidose vials for sequential patients, reuse of spring-loaded finger-stick devices, and contamination of endoscopes and other devices for invasive procedures) have been reported in the literature. In the past decade, the literature documents a disturbing increase in the detection of such cases.^{78,79,80,81,82,83,84,85,86,87,88}

Recognizing the epidemiological similarities between HCV and HBV, several investigators attempted to assess the risk of occupational infection by testing HCP for the serological prevalence of HCV antibodies, when serologic tests for HCV became available. Interpretation of these studies must take into account both the limitations of the serological assays⁸⁹ and the inadequacy of assessing only the humoral immune response as a measure of exposure and HCV infection.⁹⁰ Many of the published studies employed the first-generation anti-HCV test that detects an antibody directed against a nonstructural HCV protein, anti-c100-3, and that has low sensitivity and specificity for diagnosing HCV infection when compared with second- and third-generation tests. Even later generation anti-HCV antibody tests still may not detect 100% of infected persons, and tests designed to detect
circulating HCV RNA may be necessary to identify some infected individuals. In addition, the anti-HCV tests have a high rate of false positivity in populations with a low prevalence of infection, and supplemental tests for specificity are necessary. The recombinant immunoblot assay (RIBA) or another supplemental HCV neutralization assay should be used to verify repeatedly reactive enzyme immunoassays. Even HCV RNA detection assays are problematic. These tests are subject to false-positive and false-negative results following improper collection, handling, or storage of test samples, and their interpretation is not conclusive: a single negative test may not indicate lack of infection but may be due to fluctuating RNA levels,⁹¹ and a single positive test should be repeated to exclude the high likelihood of contamination and a falsely positive assay. In summary, the evolving diagnostic technology has complicated comparisons of HCV seroprevalence and incidence among the various published studies.

In addition to the substantial variations in study design, the differences in HCP populations studied, and the differences in the technologies used for detection, other considerations further complicate comparing and interpreting these studies. HCV seroprevalence varies geographically, so similar occupational groups from different locations cannot be compared directly, and local comparison groups are needed for determining if particular healthcare provider groups are at increased risk. Because blood donor seroprevalence data are readily available, blood donors were often used for comparison in these prevalence studies. However, blood donors are not a good comparison group because they are preselected to avoid a history of hepatitis as well as a history of risk factors for blood-borne infections.⁹² Most of these studies were not designed to investigate risk factors for HCV seroprevalence or had too few HCV-seropositive subjects to do so. Those studies that did identify risk factors for HCV infection found associations with increasing age,^93,94 years in healthcare occupations,^95,96,97 a history of blood transfusions,^97,98 and a history of prior needlestick injuries.^98,99 In aggregate, given the limitations of the study designs, testing methodology, and selection bias, studies suggest that healthcare provider’s risk of HCV infection is only minimally higher than that of volunteer blood donors and appears to be ˜10-fold lower than the occupational/nosocomial risks posed by HBV in the healthcare setting.

Table 38-1 summarizes the results of HCV incidence studies conducted in various populations of HCP who had sustained occupational exposures to HCV.^{47,51,94,97,100,101,102,103,104,105,106,108,110,113,114,115,116,117,118,119,120,121,122,124,125,126,127,128,129,130,131,132} Although most of the studies employed anti-HCV antibody testing as the primary detection system for HCV infection, nine of the studies used polymerase chain reaction (PCR) technology to attempt to detect HCV RNA as a marker for infection among individuals who had sustained parenteral exposures to blood from patients known to harbor HCV infection.^{51,103,104,113,115,116,117,119,121}

Several factors contribute to the wide variance in the transmission rates (0%-22.2%) observed in these studies, among them are different study designs and testing methods, widely differing sample sizes, variable populations of workers followed, different types of exposures, different infectivity of source patients, and potential geographical variability. Recognizing these limitations and acknowledging that the studies are not directly comparable, the pooled infection rate following percutaneous exposures was 0.83%. The risk for infection following other types of exposures has been less intensively studied, but, to date, no infections have been identified in the longitudinal studies following either mucous membrane or other less commonly occurring exposures. Monitoring for infection by measuring HCV RNA may be a more reliable marker for HCV viremia and infectivity^{105,133,134,135}; but even when PCR monitoring is combined with antibody testing, the risk for infection may still be underestimated because neither of these technologies will identify individuals who mount only a brisk cellular response and quickly clear the infection.⁹⁰ Noting all of these limitations, if one pools the data from the nine studies that used RNA PCR testing, the calculated transmission rate for percutaneous injuries is somewhat higher (3.6%) than is found in the studies assessing incidence by anti-HCV antibody tests alone.

At least four cohort studies of hospital providers initially negative for anti-HCV have attempted to measure the incidence of HCV infection. In the first study, samples collected from 960 dental staff during 1979-1981 were retrospectively tested for anti-HCV, and 2 were found to seroconvert, for an incidence of 0.15 per 100 person-years of follow-up.¹³⁶ In the second study, in a cohort of hospital staff in Cincinnati followed from 1980 to 1989, 6 cases of occupationally acquired non-A, non-B hepatitis occurred, for an incidence of 21 cases per 100 000 healthcare providers per year.¹⁰⁶ Four of the 6 cases were confirmed to be HCV infection. This incidence was ˜3 times higher than that of non-HCP. The third study followed 765 hospital providers in Italy who were screened for HCV in 1986 and retested in 1992.¹³⁷ One provider became infected, for an annual incidence of HCV infection of 0.02%. The fourth cohort study, conducted in San Francisco, observed a single seroconversion between 1984 and 1992 and found an incidence density rate of 0.08 per 100 person-years.¹³⁸ For perspective, this study also measured an incidence density rate of 3.05 per 100 person-years for HBV among nonvaccinated susceptible providers and 0.055 for HIV. A population-based surveillance system for acute viral hepatitis in Italy found that in 1991 HCP were 2.95 times as likely to acquire acute hepatitis C compared to the general population, and in 1994 they were 1.72 times as likely.¹³⁹

The findings of a low seroprevalence of HCV infection among HCPs and the moderate risk of documented transmission by needlestick injury suggest that the occupational risk of HCV infection exists and is intermediate between the 0.3% per percutaneous exposure risk for occupational HIV exposure¹⁴⁰ and the 19%-37% risk for parenteral exposure to an “e” antigen-positive, HBV-infected source.^21,22 The most probable reason for the lower risk is that titers of HCV circulating in blood are relatively low (probably 2-3 log₁₀ lower than HBV titers, as noted above),^24,139 so that transmission by small inocula, such as needlesticks or other injuries in the occupational setting, is less efficient than is the case for HBV. However, because most HCV infections are persistent, the prevalence of HCV infection in some patient populations actually may be higher than for hepatitis B,¹⁴¹ providing a larger pool of potential sources for occupational infection. Because of the wide variability in HCV prevalence by geographic region and patient populations, occupational risk will necessarily vary by these conditions.

As noted above, recent studies of the immunopathogenesis of HCV infection suggest that none of the techniques that have been applied in the longitudinal studies of risk for occupational HCV infection may provide a true denominator of HCP sustaining occupational HCV infections. Anecdotal case reports document HCV antigen circulation in individuals who never made anti-HCV antibody, despite the development of productive HCV infection.¹⁴² Additionally, some investigators have suggested that both antibody tests and tests for circulating HCV nucleic acid underestimate the true denominator of exposures, further suggesting that the most sensitive measure of past exposure may well be an assessment of specific cellular immunity directed against HCV.⁹⁰ As noted above, none of the longitudinal studies of HCP measured cellular immune responses.

HIV-1 is the only retrovirus that has been associated with serious occupational morbidity and mortality. Several cases of simian immunodeficiency virus (SIV) seroconversion have been reported,^143,144 but this virus has not yet been shown to cause disease in humans, and the SIV-seropositive laboratory workers remain well. Because several other human retroviruses have routes of transmission similar to those of HIV-1 (eg, HIV-2, human T cell lymphotropic virus I, and HTLV II), occupational transmission of these viruses may someday be detected, although no reports of occupational infection with these other agents have been published. Nonetheless, risks of transmission associated with other retroviruses are likely to be extremely low, and current guidelines for prevention of transmission of HIV-1 are thought to be adequate to prevent transmission of all blood-borne viruses, including other retroviruses.

Occupational injuries and exposures to blood and body fluids continue to be commonplace in virtually every healthcare setting. HCPs who sustain these injuries often react immediately with anxiety, fear, and concern over their risk for acquiring HIV. Framing the issue of HIV transmission risk is quite complex. Nonetheless, several decades of dealing with HIV infection in the healthcare workplace have led to a fairly extensive database characterizing these occupational risks.

Healthcare provider’s perceptions of risk were initially affected by the news media and publicity regarding cases of occupational infection. The sensationalism that traditionally accompanied HIV-related issues in the media artificially inflated perceptions of occupational risk. We frequently find that both the lay public and, particularly, HCP believe that large numbers of occupational HIV infections have been documented. Depending on the definition of “occupational infection” chosen for the analysis, one can arrive at quite disparate assessments of the number of occupational HIV infections documented in the United States.¹⁴⁵ The number of cases of occupational HIV infections in HCP has clearly decreased dramatically over the past decade.

Reports of Occupational Infections A wide variety of sources have provided information about HIV infection in HCP.¹⁴⁵ Several general types of case reports have appeared in the literature, ranging from healthcare providers in whom HIV seroconversions have been documented following an occupational exposure to HCPs who are found to be seropositive but in whom the seropositivity cannot be linked to a discrete injury or exposure.

Documented seroconversions are generally defined as cases in which a healthcare provider sustains an injury with a device contaminated with blood from an HIV-seropositive or indeterminate source; the healthcare provider is documented to be HIV-seronegative at the time of the exposure, and then the healthcare provider develops serologic evidence of HIV infection within the ensuing 6 months. More recent studies have included genomic studies documenting the close relationship of the source and recipient isolates of HIV. Such documented seroconversions are the source of the most detailed and reliable epidemiologic information about occupational infections and are, in fact, the standard against which other types of information about occupational HIV infection can be measured. Through December 2013,⁵⁸ cases of occupational seroconversions had been documented either in the medical literature or in individual case reports to the CDC that meet the criteria established for this category of occupational infection.^146,147 Of the 58 infected HCPs, 49 had percutaneous injuries or cuts, 5 had mucocutaneous exposures, 2 had both percutaneous and mucous membrane exposures, and 2 had unknown routes of exposure. Since 2001, and through 2018, only two confirmed cases (first a laboratory technician sustaining a needle puncture while working with a live HIV culture and the second an anonymous healthcare provider who sustained a needlestick exposure) have been reported¹⁴⁷ (Kuhar, D. CDC, personal communication).

In addition to these documented seroconversions, 150 additional cases of HIV infection have been categorized by the CDC as “possible” occupational infections.¹⁴⁷ This “possible occupational infection” category exhibits different demographics from the set of individuals who have documented occupational infections and likely includes individuals who have confounding community-based risk for infection.¹⁴⁵ Since the overwhelming majority of these cases have been reported as anecdotes, these data provide only limited insight into the magnitude of risk for occupational infection (ie, based on these data, one can state only that HCPs are at risk for occupational HIV infection). Some conclusions can be drawn, however, from the cases of documented seroconversions regarding the epidemiology of occupational infection. For example, by examining cases of documented seroconversion for circumstances of occupational exposure, one can gain substantial insight into the types of exposures likely to result in transmission of HIV. Even these relatively small databases provide evidence that the risk associated with mucocutaneous exposures appears to be lower than the risk associated with percutaneous injuries.

Data Describing the Magnitude of Risk of HIV Transmission in the Healthcare Setting Longitudinal cohort studies of HCP involved in the day-to-day care of HIV-infected patients and in the handling and processing of specimens from such patients provide the best available data regarding the magnitude of risk for transmission in the healthcare setting. A number of prospective studies have followed HCPs who have sustained documented exposure to blood or blood-containing body fluids from HIV-infected patients. In all of these studies, healthcare providers undergo baseline and follow-up HIV serologic testing (at a minimum) any time a healthcare provider sustains a percutaneous exposure to blood from an HIV-infected patient. The average risk of HIV infection following percutaneous exposure to HIV-infected blood has remained at ˜0.3% (95% confidence interval = 0.2%-0.5%) for a number of years.

Similarly, other prospective studies examined the risk associated with mucous membrane exposures to blood or body fluids from HIV-infected patients. Although mucous membrane exposures that resulted in HIV transmission have been reported anecdotally,^148,149 no seroconversions have occurred following the mucous membrane exposures that were prospectively collected from enrollees in these longitudinal studies.

Factors That Might Influence the Risk for Transmission Although these data are reasonably specific, and confidence intervals around the calculated risks of transmission are narrow, we still lack sufficient information to predict which injuries will result in transmission of infection. Many of the percutaneous injuries that have been associated with documented seroconversions have been quite deep or extensive or have involved injection of a volume of blood into the healthcare provider, whereas other percutaneous injuries associated with transmission have been relatively minor. Mucous membrane or nonintact skin exposures that resulted in transmission have almost uniformly been quite extensive (eg, the contact with blood has been for a prolonged period (>15 minutes) or has involved large areas of skin surface). Occasionally, injuries that one might intuitively think would have a higher than average risk for infection have not resulted in infection. For example, in the preantiretroviral era, a healthcare provider at the Clinical Center of the National Institutes of Health sustained a severe injury with a bone marrow aspiration needle that had been used on a patient with end-stage HIV disease; the needle actually penetrated through the palm and was visible from the dorsum of the provider’s hand. This exposure did not transmit HIV infection.

The epidemiologic factors contributing to the risk for occupational infection have been explored using the case-control method.¹⁵⁰ Thirty-three cases of occupational HIV seroconversion following percutaneous exposures to HIV-infected blood and 665 controls who did not seroconvert were studied by Cardo and her colleagues at the CDC.¹⁵⁰ Multivariate logistic regression identified several risk factors associated with HIV transmission after percutaneous exposure: deep injury (odds ratio [OR] 15, 95% CI 6.0-41), visible blood on device (OR 6.2, 95% CI 2.2-21), procedure involving needle in artery or vein (OR 4.3, 95% CI 1.7-12), terminal illness in source patient (OR 5.6, 95% CI 2.0-16), and postexposure use of zidovudine (OR 0.19, 95% CI .06-.52). Increased risk was associated with factors that are indirect measures of the inoculum size (ie, the quantity of blood transferred in the exposure) or higher viral burden (ie, source patient in the terminal stage of AIDS). Thus, although the average risk of HIV transmission following a percutaneous exposure is 0.3%, the risk of transmission following exposures involving large quantities of blood or high viral titers may be substantially higher than the average risk. Corroborating evidence for the factors identified by the case-control study was supplied by a laboratory study that demonstrated that more blood is transferred by deeper injuries and hollow-bore needles.¹⁵¹ Mast and his colleagues also determined that glove use reduced the transferred blood volume by nearly 50% in their laboratory model.¹⁵²

Despite our inability to predict with precision which exposures will result in transmission of HIV infection, the documented seroconversions have provided us with specific information about which body fluids have resulted in transmission. Of the 58 documented seroconversions, 49 exposures were to HIV-infected blood, 1 to visibly bloody pleural fluid, 4 to an unspecified fluid, and 4 to a concentrated viral preparation in a laboratory.¹⁴⁷ Thus, blood appears to be the major clinical risk associated with transmission. One case report documented transmission of HIV to a laboratory technician from Germany who sustained an accidental splash of serum from an infected patient to his eye.¹⁵³ Transmission in this case may have been facilitated by failure to wash the eye and by concomitant conjunctivitis related to a contact lens present in his eye at the time of exposure. Blood, visibly bloody body fluid, and now serum clearly remain the primary risk for occupational transmission of HIV in the healthcare setting.¹⁴⁸

The type (and, likely, size) of needle or sharp object involved in the injury also appears to affect the risk of transmission. To date, to our knowledge, no cases of occupational infection have been definitively linked to an exposure resulting from a solid (ie, suture) needle. Transmission has been associated with several types of hollow-bore needles (including injection needles and intravenous catheters) and other sharp objects (including contaminated broken glass, scalpels, and an orthopedic pin¹⁴⁸).

Finally, certain source patient variables, and, perhaps, even several factors relating to the recipient healthcare provider’s status, likely affect transmission. Source patients with terminal HIV disease were found to be associated with higher risks of HIV transmission in the case-control study discussed previously.¹⁵⁰ Although data regarding specific measurement of HIV viral burden were not available to the CDC researchers, the increased risk of HIV transmission from source patients who are in the late stage of HIV infection likely is a surrogate marker for the source patient’s circulating viral burden. Some also have postulated that the recipient healthcare provider’s histocompatibility with the source patient (ie, human leukocyte antigen (HLA) type, etc.) or any concurrent viral illnesses such as Epstein-Barr virus, cytomegalovirus infection, or infection with human herpesvirus-6 that results in increased CD4 expression, or the presence of chronic inflammation at or around the skin entry site, might also influence the risk of transmission. Despite this educated speculation, the numbers of cases of documented seroconversions with these data available are too few to permit adequate characterization of these risks.

Comparison of the Risk of Transmission of Various Blood-Borne Pathogens When assessing the risk of acquiring occupational HIV infection, healthcare workers must be able to place that risk into the broader context of risks associated with other blood-borne pathogens such as hepatitis B and hepatitis C. Hepatitis B has long been recognized as a significant cause of healthcare provider morbidity and mortality; HCP risks associated with hepatitis C have been documented and partially characterized in the past several decades.^140,154

The CDC estimated in 1987 that 12 000 cases of hepatitis B infection per year occurred among HCPs in the United States and that 500 HCPs were hospitalized each year because of the complications of occupationally acquired hepatitis B.¹⁹ Additionally, prior to the full-scale implementation of hepatitis B immunization, ˜200 HCPs died each year from occupational hepatitis B or its complications.⁶ More recently, Mahoney and colleagues found that the calculated number of hepatitis B virus infections among HCPs declined from 17 000 in 1983 to 400 in 1995.¹⁵⁵ This dramatic decline was associated with the implementation of universal precautions policies, with licensure of recombinant-DNA hepatitis B vaccines, and with the implementation of
the Occupational Safety and Health Administration (OSHA) Bloodborne Pathogens Standard.^25,156

The risk associated with hepatitis C appears to be lower than the risk associated with hepatitis B: HCPs with frequent blood contact account for 1%-2% of reported cases of hepatitis C infection,¹⁵⁷ and seroprevalence studies indicate that healthcare provider’s risk of hepatitis C infection is only slightly higher than that of volunteer blood donors. Several prospective studies have measured the risk of transmission after percutaneous exposure to average 1.9%⁹¹ with a range from ˜0% (in six studies, summarized in reference¹⁴⁰) to 22%,¹¹⁶ depending on the size of the population studied and the assays used to test source patients and employees, among other important variables. Lower rates of transmission have been associated with the use of the (much less sensitive) first-generation hepatitis C serologic test and with an interesting geographic distribution. In addition, a few newer studies or relatively large cohorts of exposed HCPs suggest an even smaller risk (Table 38-1).

The diagnosis of hepatitis infections in HCPs who have sustained occupational exposures is no different from diagnosis in a patient presenting with a hepatitis syndrome. One diagnosis-related issue that is worthy of some emphasis (especially when the source patient is known) is that of determining the hepatitis infection status of the source patient. When the source patient is identifiable and the hepatitis infection status is not known, documenting the source patient’s infection status will facilitate both risk assessment and the healthcare provider’s postexposure management and follow-up and, in the event that the source patient is found to be infected with the same virus, will likely solidify the healthcare provider’s compensation claim. As is done with postexposure testing of source patients for HIV infection, we feel strongly that such testing should be done with the informed consent of the source patient. State laws vary regarding the need for informed consent for testing. In occupational exposure settings, some states permit testing of available serum without consent. Hospitals and infection prevention committees should construct (and follow) policies that are consonant with their state and local laws.

Similarly, the clinical manifestations of the viral hepatitis infections arising as a result of occupational exposures in HCP are not distinct from those in other adults. An exception may exist for occupationally acquired HCV. One follow-up study suggested that, when hepatitis C develops following occupational exposure, the disease tends to be mild and transient¹⁰³ in contrast to post-transfusionacquired hepatitis C, which tends to become persistent and chronic. Another study of community-acquired HCV infection, however, found that the frequency of development of chronic hepatitis is similar regardless of how the HCV infection is initially contracted but that severe chronic disease in the form of chronic active hepatitis is more common following transfusion-acquired infection (perhaps because of an inoculum effect).¹⁵⁸ Further studies are needed to confirm these preliminary observations.

The clinical and laboratory manifestations of HIV infection are generally no different for HCPs who acquire occupational infections than they are for persons infected through other routes. Findings that may be useful in establishing the diagnosis of HIV infection of HCP are discussed in detail in the following paragraphs.

HIV-specific antibodies usually appear from 6 weeks to 4 months following exposure. An analysis of 51 seroconversions in HCP determined that the estimated median interval from exposure to seroconversion was 46 days, with a mean interval of 65 days.¹⁵⁹ Serodiagnosis consists of screening enzyme-linked immunosorbent assays (ELISAs) followed by a diagnostic Western blot when the ELISA is positive. On evaluation using the Western blot technique, antibodies to the group-specific antigen/core (GAG) proteins (ie, p18, p24, and/or p55) may be the first to appear, but antibodies to the envelope (ENV) (eg, gp120, gp160, gp41) and polymerase (POL) gene products (eg, p31) develop thereafter, confirming the serodiagnosis of HIV infection. Rapid HIV antibody testing with high sensitivity and specificity (99.6% and 100%, respectively) and 20-minute turnaround time is now widely available. Rapid testing may facilitate source patient testing and decrease the length of time HCP take postexposure prophylaxis pending the source patient HIV test result. Delayed seroconversion has been suggested following sexual exposures,^160,161 and the relatively low-inoculum exposures sustained by HCP could result in latent HIV infection and delayed seroconversion. Postexposure prophylaxis does not appear to prolong time to development of HIV antibodies.¹⁶² Ninety-five percent of healthcare providers seroconverting after occupational HIV exposure have occurred within 6 months of the exposure¹⁶² when routine testing has been performed. According to the CDC, two cases of delayed seroconversion occurring in HCPs have been reported.¹⁶² These healthcare providers had both tested seronegative for HIV at least 6 months following exposure but were seropositive within 12 months after the exposure. One of these delayed seroconversions was associated with concomitant exposure to hepatitis C virus, and this individual developed coinfection with hepatitis C that was rapidly fatal.⁵³ CDC models indicate that the upper 95th percentile of the distribution of time between exposure and seroconversion is 190 days and that 5% of HCP are estimated to seroconvert in <6 months following exposure.¹⁶³ Acute retroviral syndrome^164,165,166 associated with primary HIV infection has been a relatively common finding among HCPs in whom documented occupational HIV infection has occurred. This syndrome usually occurs 4-6 weeks after the occupational exposure. The CDC reported that 81% of HCP experienced a syndrome compatible with primary HIV infection with a median of 25 days after exposure.¹⁵⁹ This clinical syndrome has been described as resembling acute infectious mononucleosis: fever, rash, malaise, myalgias/arthralgias, headaches, night sweats, pharyngitis, and lymphadenopathy have been documented.^164,165,166 Laboratory abnormalities have also been described, including reduced total lymphocyte count, elevated sedimentation rate, and elevated transaminase and alkaline phosphatase levels.

Core (ie, p24) antigenemia may be detected coincident with the onset of symptoms and usually resolves within several weeks to months, as antibodies to p24 are produced and become detectable in the peripheral circulation.¹⁶⁷ One can also detect the presence of virus, either by culture or by PCR, in cerebrospinal fluid, peripheral blood mononuclear cells (PBMCs), and plasma before the development of an antibody response in persons who have sustained nonoccupational exposures.^{165,168,169,170} Plasma HIV RNA levels are highest immediately after acquisition and then rapidly decrease.¹⁷¹ These direct virus assays (including HIV p24 antigen EIA, PCR for HIV RNA, and the branched-chain DNA assay) consistently detect infections 1-2 weeks earlier than the most sensitive antibodies, but they still do not become positive until weeks or months postexposure, and they may revert to negative following antibody seroconversion.¹⁶³ Interestingly, no association between plasma HIV RNA levels at the time of seroconversion and subsequent rate of CD4+ cell loss or AIDS progression has been detected.¹⁷¹ The use of PCR to detect circulating viral RNA will likely supplant the use of the p24 antigen test, although the p24 assay turnaround time is much shorter than for the PCR assay in some centers.

Although direct virus assays have been used as ancillary tests in the diagnosis of occupational HIV infection, these tests should not routinely be used to detect infection in exposed HCP.¹⁶³ These tests may be helpful in defined adjunctive circumstances, such as when the ELISA is positive, but the Western blot is indeterminate, or when symptoms are consistent with the acute retroviral syndrome, but serologic testing remains negative for more than several weeks. A negative direct virus assay should never be the basis for excluding infection. Although ultrasensitive direct virus assays are available that detect 20-40 copies/mL of HIV RNA, the risk for false-positive results increases accordingly.

Symptoms consistent with the acute retroviral syndrome signal that HIV antibodies will appear, usually within 1-10 weeks¹⁶⁵ if infection has indeed occurred. HCPs who sustain occupational exposures should be educated about the symptoms of the seroconversion illness and should be instructed to seek urgent attention in the employee health clinic if these symptoms appear. In most occupational seroconversions, HIV seropositivity has not been documented as part of the routine serologic follow-up but has been detected after the healthcare provider seeks medical attention for an illness consistent with seroconversion. Nonetheless, the CDC recommends HIV antibody testing at 6 weeks, 3 months, and 6 months following the occupational exposure.¹⁷² If a newer fourth-generation combination HIV p24 antigen-HIV antibody test is used, then HIV testing may be concluded at 4 months after exposure.¹⁷² The NIH Clinical Center Occupational Medical Service also elects to check HIV antibody status at 12 months following exposure, although this is not routinely recommended by the CDC because of the rarity of delayed seroconversion events.¹⁷³ Because of the anecdotal experience with delayed HIV seroconversion occurring following concomitant exposures to HIV and HCV, most authorities would recommend extending follow-up to 12 months following simultaneous exposures to hepatitis C and HIV.

In the past three decades, nosocomial transmission of blood-borne viruses from one patient to another has most frequently been linked to inappropriate injection practices,^{174,175,176,177} breaches in infection control practices, and inadequate disinfection procedures. Reuse or improper sterilization of contaminated injection needles or syringes has resulted in transmission of blood-borne viruses to hospitalized children in Europe, the Middle East, Africa, and Asia.^{178,179,180,181,182,183} Blood-borne infections have also been transmitted to patients in both developed and developing countries through the use of contaminated multidose vials.¹⁸⁴ In addition, improper reuse and improper disinfection or sterilization of specific devices, such as spring-loaded finger-stick devices,¹⁸⁵ glucometers,¹⁸⁶ and podiatric tools,¹⁸⁷ have also been linked to patient-to-patient spread of blood-borne pathogens. Such instances of patient-to-patient spread are often detected when clusters of infection are identified. These substandard practices—inappropriate injection practices, breaches in accepted infection control procedures, and inadequate disinfection and/or sterilization procedures—remain far too common, even in developed countries.

Adherence to the following set of principles should minimize the risk for patient-to-patient spread of blood-borne viruses.¹⁸⁸ In general, reuse of needles and syringes should be strongly discouraged. In addition, medications should never be administered from the same syringe to more than one patient, even if the practitioner changes the needle in between patients. A used needle and syringe should never be used to puncture a medication vial, and medications packaged for single use should never be administered to more than one patient. Medications prepared in multidose vials should, whenever possible, be used for only one patient. Single containers of intravenous solution should not be used as a source of diluent or supply for more than one patient. All staff involved in the preparation and administration of injected medications should maintain meticulous adherence to proper infection control practices.¹⁸⁸

Prevention efforts in the healthcare setting have focused aggressively on occupational HBV infection. Results of these efforts are encouraging, but opportunities exist for further improvement. The CDC estimates that the incidence of HBV in HCP declined from 17 000 per year in 1983 to ˜400 annual infections in 1995.¹⁵⁵ This decline is generally attributed to the introduction of the hepatitis B vaccine in 1982, the institution of universal blood and body fluid precautions (universal precautions) in 1987, and the issuance of OSHA’s blood-borne pathogens standard in 1991.^{32,155,190,191,192}

Because the source patients for most occupationally acquired HBV infections are never identified, all patients should be assumed to be infectious. This concept is the cornerstone of universal/standard precautions and was originally developed in 1987 to address concerns about the transmission of HIV. All JCP who have potential occupational exposure to blood or other potentially infectious materials must receive training in the various aspects of universal/standard precautions: administrative and engineering controls, appropriate work practices, and use of protective barrier equipment to minimize occupational exposures. Such training is required both by the OSHA final rule on blood-borne pathogens²⁵ and by federal law. Engineering controls include the provision of hand washing facilities and equipment designed to minimize percutaneous injuries (eg, impervious needle disposal units and selfblunting, shielded, or needleless devices). Work practices include appropriate hand washing, safe handling of needles and other sharp devices, and avoiding risky behaviors such as oral pipetting, recapping needles, and improper handling or disposal of needles and other sharp instruments. Employees must also know how and when to use appropriate protective barrier equipment, such as gloves, gowns, masks, and eye protection, to prevent occupational exposure to blood or other infectious substances. Healthcare provider training should also include safe disposal of infectious wastes, housekeeping practices to prevent environmental contamination, and first aid procedures and injury reporting procedures to follow in the event of an occupational exposure. A summary of specific methods to reduce exposure to blood and other body fluids in the higher-risk operating room setting has also been published.¹⁹³

Management of Employees Sustaining Occupational Exposures The CDC recommends,^173,194,195 and the OSHA final blood-borne standard requires, that postexposure prophylaxis be provided to employees experiencing adverse exposures to hepatitis B. Healthcare institutions should have established protocols for providing immediate appropriate first aid for injuries and exposures, mechanisms for reporting provider injuries/exposures, and protocols to manage these exposures.¹⁹⁶ Table 38-2 summarizes the management of HCP following exposures to blood or other potentially infectious materials as practiced at the Clinical Center of the National Institutes of Health (NIH).¹⁹⁸ Management of exposures includes assessing the type, source, and circumstances of the exposure incident; evaluating the source (donor) patient for clinical, epidemiological, and laboratory evidence of hepatitis; and evaluating the hepatitis B vaccination history and hepatitis infection/immunity status of the exposed healthcare worker. Prophylactic treatment must be provided to susceptible healthcare providers as soon as possible following accidental occupational percutaneous or mucosal exposures to HBsAg-positive blood. A regimen combining hepatitis B immune globulin (HBIG) and hepatitis B vaccine will provide both short- and long-term protection and is the treatment of choice. At the Clinical Center, we often already know not only the vaccination status and HBV immunity status of the exposed healthcare provider but also the hepatitis status of many of the patients participating in research protocols. In many hospitals, this information will not be readily available, and employee treatment may have to be initiated pending laboratory test results. The most current CDC recommendations¹⁹⁵ for prophylaxis should be followed. As soon as possible following the exposure, the vaccination status and immunity status of the exposed provider should be reviewed. If the exposed provider has not been vaccinated or has not completed vaccination, the vaccine series should be started, and a single dose of HBIG (0.06 mL/kg) should be given as soon as possible, preferably within 24 hours of exposure. The vaccine should be administered in the deltoid at a separate site and can be given simultaneously with HBIG or within 7 days of exposure. If the exposed provider has already been vaccinated for hepatitis B and is known to have detectable antibody (anti-HBs ≥10 mIU/mL), no treatment is indicated. If the exposed provider has already been vaccinated and is known to be a nonresponder (anti-HBs <10 mIU/mL), either administration of a single dose of HBIG and initiation of revaccination is indicated or a dose of HBIG should be given as soon as possible followed by a second dose 1 month later. If the exposed provider has been vaccinated but the provider’s anti-HBs response status is unknown, the provider should be tested for antibody; if adequate, no treatment is indicated; if <10 mIU/mL, a single dose of HBIG and, in our institution, we administer a vaccine booster dose, though the CDC does not currently recommend a booster dose.¹⁹⁵ CDC recommendations^{173,194,195,199} should be consulted for prophylaxis of HCP when the source is HBsAg-negative, not tested for HBsAg, or the status is unknown.