David K. Henderson
Context
Blood borne pathogens (BBP) present clearly identifiable occupational hazards to healthcare workers (HCWs). Transmission of so-called serum hepatitis to HCWs was detected as early at the 1940s [1]. The past seven decades have provided a clear explication of the various risks associated with handling and processing blood in the healthcare setting. The purpose of this chapter is to identify the BBPs that have been clearly characterized as associated with occupational risks for infection, to describe the factors associated with occupational risk for infection, to discuss other aspects of the epidemiology of occupational infections with BBPs, and to address both primary and secondary prevention/control strategies that have been developed to mitigate the risks for occupational infections with these pathogens. The chapter specifically addresses risks associated with managing patients who are infected with hepatitis B, hepatitis C, hepatitis D, or the human immunodeficiency virus (HIV). The chapter does not discuss issues related to several other potential blood borne agents, including cytomegalovirus, West Nile virus, bovine spongioform encephalopathy/prion disease, hepatitis A virus, the agent called hepatitis French (origin) virus; the blood borne “GB” agents that seem to infrequently cause disease in humans; or the hepatitis G virus. This chapter focuses on the etiology of the major occupationally acquired BBP infections in HCWs, the epidemiology of these viruses in the healthcare setting, and the specific prevention and control strategies that may be relevant to these occupational risks.
Etiology/Pathogens
The agents and selected epidemiological characteristics considered in this chapter as major occupational risks are listed in Table 42-1. Whereas the risks vary substantially by pathogen, all of these agents present risk to HCWs in association with occupational exposures to blood and blood-containing body fluids. Occupational infection with each of these agents is influenced by a host of factors including the relative infectivity of each of the individual pathogens, the occupations and work responsibilities of the individual HCWs under consideration, the prevalences of infection of each of these pathogens in the populations of patients being served, and the individual HCW's attention to detail of and adherence to accepted work-practice standards and accepted infection prevention strategies among numerous others.
An issue that remains perplexing with respect to occupational infections with BBPs is the fact that many HCWs are not aware of precisely what constitutes an occupational exposure. For purposes of this chapter, occupational exposures are considered transcutaneous injuries with blood- (or other
P.702
blood-containing body fluid-) contaminated devices, mucous membrane contamination with blood (or other blood-containing body fluid); and blood-(or other blood-containing body fluid-) contamination of non-intact skin (e.g., chapped, abraded, or integrity compromised by dermatitis) [2].
|
TABLE 42-1 |
|||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||
Hepatitis B Virus (HBV)
As the primary agent responsible for serum hepatitis of the 1930s, 1940s, and 1950s, HBV has long presented the most significant risk for occupational infection to HCWs. Before the development and use of the hepatitis B vaccine, HBV was characterized as the single largest occupational risk for HCWs whose jobs entailed exposure to blood [3].
Several studies have demonstrated definitively that exposure to blood is the single most important risk factor for occupational infection with HBV [4,5]. In an elegant seroepidemiologic survey assessing occupational risk for HBV infection among HCWs, Denes et al. [6] found that occupational risk for infections increased with practice in an urban location, with the number of years in practice, and with careers in either surgery or pathology. In one of the most definitive studies of its type ever published, Dienstag and Ryan [7] correlated the presence of HBV serological markers in HCWs with (1) frequency of direct contact with blood in the individual's healthcare occupation, (2) years in a healthcare occupation (also identified as a risk for infection among providers who have frequent exposures to blood by Snydman et al.) [8], and (3) practitioner age. Interestingly, several factors that intuitively might have been suggested as likely to be associated with occupational risks for BBP infection including extent of the HCW's contact with patients, years of medical education, documented history of needlestick exposures, past history of receipt of a
P.703
transfusion, and a prior history of having received a gamma-globulin injection, were found not to be associated with serological markers for HBV infection. Among the healthcare occupations studied, the highest seroprevalences of HBV markers were found among HCWs in several occupations that are known to have high risks for occupational exposures to blood, including surgeons and surgical house officers, laboratory medicine (clinical pathology) staff, emergency room nurses, and transfusion medicine staff.
|
TABLE 42-2 |
|||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||
Subsequently, Hadler et al. conducted a similarly designed study that also controlled for nonoccupational risk factors [9], again underscoring the findings of Dienstag and Ryan's study and demonstrating that occupational blood exposure, not patient contact, was associated with risk for serological markers of HBV infection among HCWs. In a retrospective review, West found the risk for HBV infection among HCWs to be four times the risk for the general adult population [10]. He found that surgeons, dialysis personnel, personnel providing care for developmentally disabled individuals, and clinical laboratorians were at 10-fold higher risk for infection and that physicians and dentists were between 5- and 10-fold more likely to have serological markers of prior HBV infection than the general population [10].
In addition to occupational exposure to blood, several additional factors influence the risk for HBV infection in
P.704
HCWs, among them the prevalence of HBV infection in the population being served, practice in urban settings (because the prevalence of infection is higher there than in rural settings), practice involving dialysis patients, and HCWs who provide care for other populations of patients known to be at increased risk for HBV infection (e.g., injecting drug users, men who have sex with men, prison inmates, the developmentally disabled, and/or immigrants from highly endemic areas) [11].
|
TABLE 42-3 |
|||||||||||||||||||||||
|
|||||||||||||||||||||||
The source-patient's viral burden (i.e., presumably due to an inoculum effect) also influences the risk for transmission. Thus, patients who are hepatitis B “e” antigen positive (who generally have substantially higher circulating viral burdens than those who are “e” antigen negative) present higher levels of risk. HBV infectivity also correlates directly with the levels of hepatitis B virus DNA in the circulation.
The characteristics of the exposure itself influence the risk for acquiring occupational infection. For example, parenteral exposures are associated with increased risks for occupational infection. Conversely, because of the extraordinary levels of viremia in hepatitis B “e” antigen positive patients, even what might be considered trivial inocula of hepatitis B “e” antigen-positive blood may produce infection. Patients who are hepatitis B “e” antigen-positive chronic carriers of HBV may harbor as many as 1013 virus particles of HBV per milliliter of blood [3]. Because of these remarkable levels of viremia, miniscule amounts of blood contaminating inanimate objects or environmental surfaces actually may present significant occupational risks for infection. Whereas parenteral exposures account for the large majority of occupational infections with HBV, several episodes document that contaminating mucous membranes also may produce HBV infection [12].
P.705
One of the most significant developments in the past 50 years in terms of mitigating risks for occupational infection with BBPs was the development of the HBV vaccine. Studies conducted since the introduction of the HBV vaccine demonstrate its substantial efficacy in preventing occupational infections. For example, a seroprevalence study conducted by Thomas et al. at Johns Hopkins Hospital in Baltimore identified “absence of hepatitis B vaccination” as the only factor independently associated with risk for HBV infection in a large cohort of HCWs [13]. Similarly, Panlilio et al. evaluated a cohort of surgeons for prior HBV infection and found only two factors associated with risk for infection: (1) not having been immunized with the HBV vaccine and (2) having practiced surgery for at least the previous 10 years [14].
Hepatitis C Virus (HCV)
HCV continues to present occupational risks for infection to HCWs for a variety of reasons. The population of patients chronically infected with HCV, particularly among injecting drug users, continues to expand aggressively. Whereas the epidemiology and pathogenesis of HBV infection are understood with a great deal of clarity, our current understanding of the pathogenesis and immunopathogenesis of HCV infection remains far less clear, especially with respect to the early events in the course of the infection. In addition, despite the fact that the pathogen responsible for this disease was identified in 1989 [15], we still have no vaccine for it and have no proof that any intervention is efficacious in preventing infection following occupational exposure to the virus.
Because HCV is known to be a major cause of posttransfusion hepatitis, the thought that HCWs would be at occupational risk for HCV transmission makes implicit sense. Numerous anecdotal case reports of occupational infection have been reported in the literature (summarized in [16]). Whereas, similar to HBV, parenteral inoculation (e.g., needlestick exposure) presents the highest level of risk for occupational HCV infection, inapparent parenteral transmission (including mucous membrane exposures) likely accounts for many of the remaining episodes. To date, all instances of occupational HCV infection have been associated with exposures to blood despite the fact that HCV has been isolated (albeit generally in much lower concentrations) from a variety of other body fluids. With respect to circumstances of occupational exposure, the most frequent type of exposure resulting in HCV infection in the healthcare setting to date has been a needlestick with a hollow-bore needle.
Several serological prevalence studies evaluating HCWs for the presence of antibody directed against HCV have been published. Although these studies have substantial limitations, they demonstrate that HCWs' risk for acquiring occupational (HCVI) infection is only minimally higher than that of volunteer blood donors and approximately 10-fold lower than the comparable occupational risks associated with exposure to HBV in the healthcare setting. Most of these studies were designed as simple seroprevalence surveys, not to investigate risk factors for HCV infection in their respective cohorts. The few studies that were designed to detect risk factors for HCV infection found increasing age [17,18], years of employment in healthcare occupations [17,19,20], a history of blood transfusions [17,21], and a history of prior needlestick injuries [21,22] to be associated with risk for HCV infection (as detected by assays for circulating antibody directed against HCV).
Study and technological limitations cloud the issue of the risk for transmission of HCV associated with single parenteral exposures. The primary technological limitation is the fact that a variety of different tests have been used in these studies to detect prior infection. Some of the published studies used only the first-generation antibody test (that was neither highly sensitive nor specific). Others have used subsequent iterations of the antibody tests that have substantially improved sensitivities and specificities. Some have used direct detection of the HCV genomic material by polymerase chain reaction to detect infection. These studies provide highly disparate estimates of the risk for HCV infection following discrete occupational exposures.
Studies conducted during the past decade have suggested that the detection strategies for HCV infection in these seroprevalence and longitudinal cohort studies may have been relatively insensitive with some studies suggesting that both antibody tests and tests for circulating HCV nucleic acid underestimate the risk. In these studies, the investigators suggest that the most sensitive test for detecting prior infection/exposure to HCV is the measurement of specific cellular immunity directed against this flavivirus [23].
Hepatitis D Virus (HDV)
HDV by itself presents no occupational risk to HCWs. The HDV is an “incomplete” virus that requires co-infection with HBV to produce infection. In addition, although approximately 5% of all HBV carriers are co-infected with HDV, substantial demographic, risk-group, and geographic variations exist. For example, HDV infection is particularly endemic in the Middle East, parts of the Amazon River basin, a few of the Pacific Islands, and southern Italy. Injecting drug users and hemodialysis patients are more likely to be co-infected with HDV than are other groups known to be at risk for HBV infection (e.g., men who have sex with men).
Exposure to HDV represents a risk to those HCWs who are already chronically infected with HBV and to uninfected HCWs who experience an exposure to blood from someone who is chronically infected with both viruses. Occupational HDV infection has been infrequently detected to date, in part because of the requirement for simultaneous infection
P.706
with HBV, and in part because tests to detect HDV infection are rarely conducted [24,25].
Human Immunodeficiency Virus (HIV)
The introduction of a new BBP—HIV—into the healthcare workplace in the 1980s was associated with significant fear and anxiety on the part of HCWs. Despite the fact that the risk for occupational infection with other BBPs (e.g., HBV) had been common knowledge since the late 1940s, the epidemic of HIV infection in the United States and its association with almost monumental societal fear and anxiety fueled HCWs' concerns. In the years since the widespread introduction of HIV infection into society, we have learned that exposure to blood from an HIV-infected patient is associated with measurable occupational risks for infection, that such occupational infections occur infrequently, and that sensible procedural interventions can reduce the risk for exposure (and therefore infection) with this blood borne retrovirus. We also have learned that postexposure interventions may further reduce the risk for occupational infections.
In the past 25 years, there have been only 57 documented HIV infections in U.S. HCWs, and the majority of those infections occurred in the first 15 years of the epidemic [26]. These definitive episodes are instances in which an HCW sustained an occupational exposure to blood from someone known to be HIV-infected, the HCW had a baseline sample drawn to demonstrate that she or he was not infected at the time of the exposure, was followed by serological evaluation over time, and, in temporal association with the exposure, the HCW developed serological and, in some instances clinical, evidence of HIV infection. In addition to these definitive infections, the U.S. Public Health Service (USPHS) has identified nearly 150 other episodes that could be categorized as possible or probable occupational HIV infections among U.S. HCWs. These latter individuals did not have a baseline serological sample drawn at the time of the occupational exposure to demonstrate absence of infection at the time of the exposure. Despite the fact that these individuals all denied having nonoccupational risks for HIV infection, when one compares the demographics of this population with those of the definitive episodes as described, and the substantial differences in these two populations become apparent, strongly suggesting that confounding community-based risks are present in the possible/probable population [27].
Pathogenesis
Transmission of each of the major BBPs identified here has been most closely associated with transcutaneous injuries with sharp objects in the healthcare setting. The preponderance of these exposures have been needlestick injuries with hollow-bore needles; however, a variety of other blood-contaminated sharp objects have been implicated in the transmission of one or more of these infectious diseases. Mucosal exposures to blood also transmit infection. For example, six or seven of the definitive occupational HIV infections described here were associated with mucosal exposures to blood from patients known to harbor HIV infection. Whereas other body fluids may present some risk for occupational infection, the primary risk for all of the pathogens discussed is associated with exposures to blood from infected individuals.
Magnitude of Risk for Occupational Infection Associated with Occupational Exposure
For all of these significant BBPs, the risk for infection associated with a single discrete exposure to blood from a patient known to be infected with one of these viruses depends on a number of variables, including (but not limited to) circulating viral burden in the source patient, inoculum size, significance of the exposure, and the route of exposure, among others. For example, the risk for transmission of HBV following parenteral (e.g., needlestick) occupational exposure ranges from 6–37% per exposure, depending on a variety of factors including the type of exposure, the inoculum size and type, and the source-patient's circulating viral burden and/or hepatitis “e” antigen status [28].
For HCV infection, considering all of the codicils about the different methods that have been used to detect infection (discussed previously), when taken together, these studies suggest that the risk for HCV transmission following a parenteral exposure to blood from a patient known to be infected with HCV is between 1–3% per exposure [16]. Thus, for HCV, the 1–3% risk for transmission per exposure places the occupational risk for transmission of HCV between the infection risk for occupational HIV exposures discussed later and the risk for HBV exposures discussed earlier.
For HIV exposures, >20 longitudinal studies have been conducted attempting to measure the risk for transmission following an individual occupational exposure [29,30]. When data from all of these studies are combined, the risk for transmission of HIV associated with a single percutaneous exposure to blood from a patient known to be HIV-infected is 0.32%, or roughly 1 infection for every 325 percutaneous occupational exposures to blood from an HIV-infected patient [29]. Many of these same studies also attempted to assess the risk for infection associated with mucous membrane exposures to blood from patients known to be HIV infected. Pooling data from these studies, the risk for occupational infection associated with a single mucosal exposure to blood from an HIV-infected patient
P.707
is estimated to be 0.03%; however, this approximation may be an overestimate because the single infection in this series was actually reported as an anecdotal case-report in the literature [31] before prospective data purportedly had begun to be collected for the longitudinal study. Thus, this episode would have clearly occurred before the prospective study of risk began [30].
For HDV, no prospective studies have been able to measure the risk associated with either a single exposure to blood from a patient harboring HDV or for a single exposure to blood from a patient co-infected with HBV and HDV.
Several other factors likely influence the risk for transmission of these viruses associated with individual exposures. Clearly, inoculum is a major determinant, and the viral inoculum is related both to the volume of material involved in the exposure and the source patient's circulating viral burden. As might be anticipated, studies of needlestick exposures demonstrate that the volume of the exposure increases with the size of the needle causing the injury and the depth of penetration of the needle. In several studies, hollow needles have been shown to be associated with higher inocula of blood than are comparably sized suture needles [32,33].
The amount of virus present in the source material may vary by several logs, depending on the stage of the source patient's illness and the effect of antiviral or immunomodulator. For most, if not all, of these infections, viral burden is likely to be the single best predictor of infection risk.
In 1997, the Centers for Disease Control and Prevention (CDC) published the results from a retrospective case-control study of percutaneous exposures to HIV among HCWs to attempt to define factors associated with transmission risks [34]. The investigators identified four factors to be associated with an increased risk for occupational infection with HIV following percutaneous exposures: deep rather than superficial injuries; injuries with sharp devices that were visibly bloody when compared with devices on which no blood was visible; injuries with sharp devices that had been used in arteries or veins as compared with those that had not been in vascular channels; and injuries associated with source-patients who had preterminal AIDS (defined as source-patients who expired within 2 months of the time of the exposure) than when the sources had earlier stages of infection. Each of these factors is likely a surrogate marker for viral inoculum.
The specific characteristics of the pathogens to which an HCW is exposed also may influence the risk for infection. For example, some strains of HIV may be more aggressive than others (e.g., some strains may produce infection by inducing syncytia more efficiently than others, and some clearly are able to attach to macrophages more efficiently than others). Patients with late stage HIV or HCV infection harbor numerous quasispecies of these viruses, also likely increasing the transmission risk.
A final factor that likely influences the risk for occupational infection relates to host factors in the exposed HCW. Variation in an individual HCW's immunological responses also likely affects the probability of HIV transmission. Three possible outcomes have been postulated to result from occupational exposures to HIV: infection (usually with detectable antibody responses directed against the invading pathogen); no infection with absent immunologic responses; or so-called transient infection that is characterized by measurable and persistent T-cell responses (i.e., directed against HIV peptides and envelope antigens or HCV envelope proteins and absence of protracted or systemic infection and antibody response directed against the infecting virus). With respect to HIV exposures, several populations of exposed but uninfected individuals have been studied to gain these insights, including the steady sexual partners of infected individuals, children born to mothers who are infected with HIV, prostitutes [35,36], and occupationally exposed HCWs [35,36,37,38,39,40,41,42,43,44]. Neither the efficacy nor the precise role for this cellular immune response in the overall defense against initial HIV infection is well understood.
Primary Prevention—Preventing Occupational Exposures
If one examines possible strategies for preventing occupational infections with BBPs, by far the most efficient approach is to implement strategies designed to prevent occupational exposures to blood. Despite the fact that this prevention is, perhaps, counter culture in medicine, the strategy affords the HCW the most cost-effective, most efficient strategy for reducing the risks for occupational infection with all BBPs. In 1987, CDC published its universal precautions guidelines [45]. The recommendations were designed to reduce the risk for exposure to blood and, therefore, to reduce the risk for BBP transmission. Effective use of these precautions will unquestionably reduce cutaneous, mucocutaneous, and percutaneous exposures to blood. Thus, effective implementation of universal/standard precautions will decrease risks for occupational infection with all BBPs.
When one assesses the specific detailed recommendations in these guidelines, the specific components—the effective use of hand-hygiene strategies, the use of appropriate personal protective equipment (e.g., gloves), and the need for attention to the appropriate use and disposal of needles and other sharp objects—the reasons for the efficacy of these precautions becomes readily apparent. A number of additional approaches have been demonstrated as effective in reducing occupational exposures and injuries including the comprehensive education of staff about the attendant occupational risk associated with providing care for patients who have BBP infection, education of staff about the occupational risks that are present and highly
P.708
prevalent in the healthcare workplace, the need to modify procedures and work practices that are intrinsically associated with risks for occupational exposure, and the need to monitor staff for adherence to standard/universal precautions and other relevant infection control guidelines. Institutions also should develop strategies to be able to monitor the healthcare marketplace for technological advancements that can be implemented to replace existing approaches while simultaneously reducing occupational risks. All healthcare institutions should collect information prospectively about occupational exposures that occur in their institutions and should use these data to drive performance improvement activities to reduce the attendant risks.
Last, the appropriate use of vaccines (e.g., HBV vaccine) already plays a crucial role with respect to HBV in primary prevention of occupational infection with BBPs. When additional vaccines become available (e.g., HCV, HIV), such vaccines would play an increasingly important role in primary prevention.
Immediate Postexposure Management
One of the most important considerations for immediate postexposure management is first to determine that an occupational exposure presenting a risk for transmission of one of these BBPs has actually occurred. To make such a determination, the practitioner must thoroughly evaluate the exposure event, the potential for susceptibility of the exposed HCW (e.g., immunity to HBV, preexisting infection with HBV, HCV, or HIV), and the information available about the patient who was the source for the exposure. If the source-patient's BBP infection status is not known, the source-patient should be tested for all of these BBP infections, making certain that the testing is appropriately conducted within the constraints of relevant state and local laws. Currently, marketed rapid tests for HIV are highly reliable when negative. Positives must be followed up with standard immunoassay and confirmatory tests. In instances in which the source-patient is determined to be infected with one or more of these pathogens, obtaining as much additional information about the source-patient's infection(s) makes implicit sense. Determining the duration of the source-patient's infection, the current therapy for the infection(s), key immunologic and/or virologic parameters for each pathogen, such as viral burden(s), and a variety of other risk factors that relate to each of the pathogens, may help the practitioner understand the significance of the exposure. If information about the source-patient's viral isolates is available (i.e., phenotypic or genotypic information, information about prior resistance), this information should be considered as well. In instances in which such practices are practical, saving a sample of the source-patient's pathogen is entirely sensible. The practitioner responsible for management also should obtain as much information as possible about additional factors likely to increase the risk for transmission of BBP infections (e.g., if a volume of blood was injected; if the exposure was to a hollow-bore, rather than solid, needle; if the needle was of a large-, rather than small-gauge; if blood was visible on the device causing the injury; or if the device had been placed in one of the source-patient's arteries or veins) [34].
In instances in which source-patient testing is either not possible or readily available, immediately offer prophylaxis; if the HCW elects to take the prophylaxis, sort out the exposure data as quickly as possible. If the source-patient's infection status for these pathogens cannot be discerned, the practitioner should make his or her best epidemiological assessment about the likelihood of exposure and manage the HCW in accord with that assessment. Factors that may be considered in making such an epidemiological assessment include (but are certainly not limited to) the severity of the exposure; the precise circumstances of the exposure; the location where the exposure occurred and the likelihood that pathogens were present; the demographics of the source-patient; and the presence of other epidemiological factors known to be associated with risk for one or more of these infections. Such source-unknown exposures must, of necessity, be managed on a case-by-case basis.
Although determining whether or not an exposure has actually occurred seems straightforward, this determination in actuality is the Achilles-heel of postexposure management. Summary data from the National Clinicians' Post-Exposure Prophylaxis Hotline (PEPLINE) at the University of California at San Francisco have consistently suggested that postexposure prophylaxis often is prescribed and administered for instances in which the PEPLINE professionals felt that an exposure had not occurred [46]. Although these data appear to be improving gradually over time, the number of instances in which post exposure prophylaxis is prescribed for circumstances felt not to represent exposures remains far too high. One reason for the problem of overtreatment may be that the practitioners who ultimately end up managing these exposures (very often emergency room practitioners) often are unfamiliar with the exposure definitions and with the drugs administered for postexposure prophylaxis. Furthermore, because the practitioners often are colleagues of the exposed individuals, they may be more easily influenced by their putatively exposed colleague's anxiety. Institutions should develop systematic procedures and a multidisciplinary team approach to occupational exposures to ensure that these exposures are managed both consistently and with the highest possible quality. Occupational medicine, hospital epidemiology, hospital safety, and the infectious diseases/HIV team should be key members of this team. Qualified, knowledgeable staff should be available 24 hours a day, 7 days a week to assist with the management of these exposures. As part of this multidisciplinary approach, the team should collect information about occupational
P.709
exposures to blood in their institution to be able to assess their data for common circumstances of exposure or for intrinsic problems with patient-care processes that might be improved to mitigate risk.
Effective exposure management is a healthcare institution's responsibility. Staff need to know precisely what to do when an exposure occurs and precisely when to do it. Access to information about appropriate exposure management procedures must be readily available to all potentially exposed staff and also must be user friendly.
Healthcare institutions are required by law to provide systems for reporting exposures and ensuring rapid access to appropriate postexposure care [47]. Despite the development of elegant strategies to facilitate reporting and management of such exposures, many occupational exposures are never reported. Since the early 1980s, underreporting of these injuries has been identified as a significant problem, and it persists as a major problem well into the 21st century.
We recommend that wounds, punctures, or other skin areas that have had direct contact with blood or body fluids should be initially washed thoroughly with soap and water [2,48]. Some authorities have recommended that antiseptics be used to decontaminate the wound; however, no data actually provide scientific support this recommendation. Flushing and washing the wound should not be delayed until antiseptics can be obtained.
For mucous membrane exposures, we recommend flushing the exposure site aggressively with tap water; eyes should ideally be flushed with sterile water or a commercial eye irrigant; if neither is readily available, clean tap water will suffice.
An important aspect of early postexposure management is counseling. The emotional impact of an occupational exposure to a BBP should never be underemphasized. In addition to ensuring that the staff who are providing care to exposed HCWs are knowledgeable about the epidemiology, risks for transmission, treatment options, and known complications of treatment, institutions also should ensure that staff who sustain these exposures have access to skilled counseling. The clinician providing care for the exposed HCW must be able to provide the exposed staff member understandable, objective information about the risks for infection associated with the type of exposure that the employee has sustained and what is known about the risks and benefits of the various possible treatment options. Clinical staff should guard against minimizing or trivializing the risks and should work hard to express empathy and reassurance. These events are incredibly troubling to the exposed HCW, who may not be able to assimilate all of the information that the clinicians provide. Staff providing care for individuals sustaining these types of exposures should be ready and willing to answer the same questions repeatedly, both for the exposed HCW and her or his spouse or significant other. Regardless of the treatment course elected, the exposed HCW should be scheduled for a follow-up appointment 48 hours following the initial appointment to assess how she or he is doing and to answer any outstanding questions.
HCWs who are too upset or confused to make a decision about chemoprophylaxis can sometimes be helped by suggesting that treatment be started immediately with the option to stop it later (i.e., “Because some evidence suggests that the timing of the first dose influences the success of treatment, I suggest that you start treatment now and then tomorrow, or even later, if need be, we can decide whether continuing is your best option”). This approach modulates the acute pressure the HCW may feel to make an immediate decision and empowers workers to be able to decide about their own treatments.
Counseling the HIV-exposed HCW should include a clear discussion of several important issues related to occupational exposures and their management: (1) More than 99% of individuals who sustain occupational exposures will not become infected, even if they elect not to take postexposure antiretroviral chemoprophylaxis (PEP); (2) although a great deal of indirect evidence suggesting efficacy for postexposure antiretroviral prophylaxis (discussed later) has been assembled over the years, no agent or combination of agents has been approved as safe and effective by the Food and Drug Administration (FDA) for use in this setting; (3) data about the efficacy and safety of the use of these potentially toxic agents in this setting are far from complete; and (4) exposed HCWs should be counseled to take precautions to prevent secondary transmission, especially during the first 3 months following exposure when seroconversion, including precautions to prevent sexual transmission (e.g., abstinence or condom use), and the avoidance of blood and organ donation and discontinuation of breast-feeding.
Exposed staff should be counseled about the magnitude of risk associated with occupational exposures, the institutional measures that have been put in place to protect the confidentiality of exposed HCWs' medical records, and the typical concerns of sexual partners, co-workers, family, and friends of the exposed worker. Finally, counseling staff should be prepared to answer questions for spouses, significant others, and family who have major concerns about associated risks.
Pathogen-Specific Postexposure Management and Follow-Up
Hepatitis B Virus
A large body of evidence demonstrates that PEP (both active and passive) is effective in preventing infection with HBV following occupational exposures. PEP should be provided to all susceptible HCWs who sustain occupational HBV exposures. As for each of the commonly encountered BBPs, healthcare institutions should establish protocols for providing appropriate management of HBV
P.710
exposures. As described, management of HBV exposures includes assessing the type, source, and circumstances of the exposure; evaluating the source-patient for clinical, epidemiologic, and laboratory evidence of hepatitis; and evaluating the HBV vaccination history and HBV infection/immunity status of the exposed HCW. Additionally, institutions should strive to provide prophylactic treatment as soon as possible following occupational HBV exposures. Current recommendations include the administration of both HBV immune globulin and the HBV vaccine series [11] (Table 42-2). Practitioners must be aware that certain BBPs travel together. Simply because a patient is admitted for treatment of complications of HBV infection does not mean that the other pathogens should be ignored. For this reason, even when the source known to harbor HBV, the source-patient should also be tested for HCV and HIV in addition to the current HBV infection status.
The issue of booster doses of HBV vaccine for HCWs who have occupational exposures and who are shown to have undetectable levels of antibody remains controversial. The USPHS does not currently recommend booster doses for HCWs who initially responded to the vaccine but whose antibody levels have declined to the undetectable level. Several studies have addressed the issue of the durability of vaccine-induced immunity. One study suggested that between 30–60% of vaccines had suboptimal levels of antibody 8 years following immunization [49]; however, several other studies have suggested that the vaccine response may be quite durable even 10 years after immunization [50,51,52]. In fact, productive HBV infection has been relatively rare in vaccine recipients regardless of antibody levels. Nonetheless, some institutions offer booster doses of vaccine to HCWs who have previously responded to HBV vaccination, whose antibody levels have become undetectable, and who remain in jobs associated with risk for blood exposure and HBV infection.
Hepatitis C Virus
For occupational exposures to HCV, the immediate PEP is identical to that described previously for all BBPs. As for known HBV exposures, practitioners must be aware that certain of these BBPs travel together. Thus, even when known to harbor HCV, the source-patient should also be tested for HBV and HIV, in addition to testing for his or her current HCV infection status. Practitioners ordering these tests must be cognizant of the local and state laws relevant to informed consent for these tests. Furthermore, practitioners should be mindful of the fact that the majority of the screening tests are designed to detect antibody directed against some of these pathogens and that these tests clearly do not detect all patients who have been infected previously. Specifically with respect to HCV, detecting antibody against HCV in a source-patient's serum is not an accurate indicator of the individual's infectivity.
The most recent CDC guidelines for managing occupational HCV exposure recommend (1) testing the source-patient for antibody against HCV at the time of exposure; (2) baseline testing of the HCW sustaining the exposure for both antibody against HCV and alanine aminotransferase levels; (3) repeat testing of the HCW (for HCV antibody and alanine aminotransferase levels) at six months following the exposure or at any time when symptoms suggest possible infection; (4) using confirmatory tests, including direct antigen detection, direct genome detection, and supplementary or confirmatory antibody tests (e.g., recombinant immunoblot assay [RIBA]) to investigate positive results of HCV antibody testing in more detail; (5) not providing immunoglobulin, antiviral agents, or immunomodulators as PEP; and (6) providing comprehensive information to the exposed HCW about the magnitude of occupational risks associated with the exposure, the risks for secondary transmission, and strategies known to be effective in preventing exposure to blood and/or transmission of HCV in occupational settings [53].
Some individuals have advocated more aggressive monitoring and interventional strategies [16,54]. One approach is to follow HCWs using periodic monitoring with PCR detection of HCV RNA at some defined interval (e.g., monthly, every 6 weeks, at 3 months following exposure) in addition to the antibody testing described earlier. If an exposed HCW is repeatedly detected as positive for HCV RNA by the PCR (because of frequent false positives in low prevalence settings, one should never rely on a single positive sample), the HCW can be referred to a specialist in the management of HCV for definitive therapy (see the following discussion).
Historically, some investigators advocated the use of immune serum globulin to attempt to decrease the risk for transmission of what has been subsequently identified to be HCV [55,56,57]. As additional details of the pathogenesis and immunology of HCV infection have been delineated, most experts now agree that postexposure immunoglobulin prophylaxis is of no value. Similarly, no information has provided scientific support for a role for interferon or other immunomodulators administered as true PEP (i.e., in the immediate postexposure period) for occupational HCV exposures. Finally, to date, no agents with specific antiviral activity against HCV have been marketed. Nonetheless, antiviral agents with specific, defined HCV targets (e.g., protease, helicase, and polymerase inhibitors) are in drug development and may ultimately play a role in immediate postexposure prophylaxis (e.g., analogous to antiretrovirals in the management of occupational exposures to HIV discussed later). Currently, in the absence of data purporting efficacy for any of these compounds administered as PEP, no recommendation can be made about their potential use.
Although a clear recommendation for the use of immediate PEP following occupational HCV exposures cannot be made, some postexposure management strategies do show promise. One PEP management strategy that has been
P.711
adopted by many centers throughout the United States centers on the early treatment of occupational HCV infection. This strategy incorporates the periodic monitoring of exposed HCWs by HCV-PCR combined with early therapy with immunomodulators once infection is confirmed. This strategy was first advocated by Schiff in 1992 [58]. Once occupational infection has been definitively documented, two separate approaches have been advocated: pre-emptive therapy and watchful waiting [16].
The preemptive therapy strategy involves the aggressive initiation of immunomodulator therapy (e.g., interferon, with or without ribavirin) as soon as the diagnosis of occupational HCV infection is established with certainty. The rationale for this strategy is at least in part based on studies documenting the successful treatment of patients with so-called acute HCV infection. In several studies, treatment of patients with “early” HCV infection has been far more successful (i.e., some with efficacies of therapies exceeding 90–95%) [59,60,61,62,63,64]. In perhaps the largest of these series published to date, Jaeckel et al. treated 44 patients who had acute HCV infection with interferon alpha-2b, initially daily (for 1 month) and then three times a week for 5 months. Of the 44 patients studied, HCV-RNA was undetectable by PCR, and alanine aminotransferase levels were entirely normal in 43/44 patients who were treated, both at the end of their treatments and six months following completion of therapy [62]. Although this study has limitations [16], particularly with respect to generalizing the results to the occupational exposure setting, the success rate is remarkable and substantially exceeds cure rates of even the best studies of the treatment of chronic HCV infection. The 98% “cure rate” observed in this study is extraordinary. In addition to the documented clinical successes of treating acute infection, one can assemble a strong intellectual argument for the preemptive therapy approach primarily because the clinician would be treating the patient when the HCV would have produced a limited number of quasispecies. In fact, a National Institutes of Health (NIH) Consensus Conference published in 2002 argued strongly for the aggressive treatment of individuals who have acute HCV infection [59].
The second approach to the management of occupational HCV infection employs the watchful waiting strategy. Following this strategy, once the definitive diagnosis of occupational HCV infection has been established (by repeatedly detecting HCV-RNA in the circulation of the exposed HCW), the clinician caring for the exposed/infected HCW observes the patient for a defined period of time (e.g., 2–3 months) to see whether the infected HCW spontaneously clears the infection. HCWs who spontaneously resolve their infections obviously would be spared both the potential toxicities and expense of immunomodulator therapy. The rationale for the watchful waiting strategy is discussed shortly.
Although the data describing the success of early interventions in the treatment of patients who have acute HCV infection are undeniably encouraging, there are no data demonstrating that either of these two approaches to managing occupational HCV infections is efficacious, let alone that one approach is advantageous over the other. One anecdotal case-report has documented resolution of HCV infection following early treatment (the HCW became PCR-positive, received immunomodulator treatment, became PCR-negative, and never produced anti-HCV antibody) [65].
Several factors favor the watchful waiting approach to the management of occupational HCV exposures. First, because we have no clear understanding of the early events in the pathogenesis of occupational HCV infection, we need to keep open minds about the factors that do (and might) influence transmission. For example, we do not know what fraction of HCWs who sustain an occupational HCV exposure and subsequently develop PCR evidence of HCV infection will develop a cellular immune response to the insult that allows the individual to clear the infection spontaneously without developing antibody against HCV. Larghi et al. reported that 50% of the individuals exposed to HCV in a point-source epidemic spontaneously resolved their infections without long-term infectious or serologic sequelae [66]. Clearly, clinical scientists need additional insight into the early events in the pathogenesis, immunopathogenesis of, and host immunological response to HCV infection to understand the best strategies for interventions including those involving immunomodulators. Additional support for the watchful waiting approach can be derived from the study by Seeff et al. that found that 20% of patients who acquired transfusion-associated HCV infection spontaneously cleared their infections [67]. This latter finding is perhaps even more striking because individuals who acquire HCV infection from contaminated transfusions presumably receive much higher inocula of virus (and presumably many different quasispecies) than do individuals who become infected from an occupational exposure. Extrapolating from these two studies, administering interferon to 100% of individuals detected by PCR as having circulating HCV-RNA would unnecessarily expose between 20–50% to toxicity with no benefit.
A variety of additional factors qualifies the interpretation of the available data relevant to the management of occupational HCV exposure management [16]. For example, the use of immunomodulating agents as postexposure prophylaxis is substantially different from the use of antiviral agents directed against specific viral targets (e.g., PEP with antiretrovirals for occupational HIV exposures). Finally, administering interferon before the exposed HCW's cellular immune response has matured (i.e., has had time to begin to develop a specific response to the invading pathogen) may be far less effective than waiting until the relevant cells have been activated and expanded (and, thus, the specific response can then be further stimulated by immunomodulators). Even in the absence of definitive scientific support
P.712
for these approaches, many U.S. healthcare institutions have adopted either the preemptive therapy or watchful waiting strategy for managing occupational HCV exposure [68]. PCR monitoring for viremia, monitoring hepatic function by alanine aminotransferase levels, closely monitoring developing antibody responses, and, finally, levering a decision for intervention from the clinical and chemical data obtained from the actions detailed here represent an entirely sentient approach to this complex management problem.
Human Immunodeficiency Virus
Specific PEP interventions for occupational exposures to the HIV were initially controversial. CDC first published guidelines that included “considerations” with respect to the administration of antiretrovirals for PEP for occupational exposures to HIV in 1990 [69]. Since this initial publication, CDC has published several sets of updated recommendations. The most recent guidelines were published by CDC in 2005 [2].
The rationale for administering postexposure antiretroviral chemoprophylaxis is derived from several disparate sources: (1) in vitro studies of the efficacy of antiretrovirals in preventing retroviral infection of susceptibles; (2) studies delineating the early sequence of events in the process of HIV infection of cells that provide a sense of biological plausibility for the efficacy of these agents administered as prophylaxis; (3) studies in relevant animal models of the safety and efficacy of antiretroviral prophylaxis for retroviral infection; (4) clinical studies demonstrating the efficacy of antiretrovirals in preventing vertical transmission; (5) epidemiologic data collected over the past 20 years describing occupational HIV infections in HCWs; and (6) substantial clinical experience over the past 20 years administering these agents in the postexposure setting. Despite the fact that we likely will never have definitive scientific evidence of the efficacy of these agents administered as postexposure prophylaxis, when all of these sources of indirectly relevant data are considered together, they provide substantial rationale for the use of these drugs in this setting. A more detailed discussion of the various aspects of the rationale for administering these agents as chemoprophylaxis for occupational HIV exposures appears.
Laboratory studies conducted in the late 1980s demonstrated that adding nucleoside analogs to the tissue culture milieu could prevent infection of tissue culture cells known to be highly susceptible to HIV infection [70]. These studies provided definitive evidence that reverse transcriptase inhibitors could actually prevent infection of susceptible cells.
Several studies conducted over the past decade have dramatically increased our understanding of the early events in the pathogenesis of HIV infection. This increased clarity has provided additional support for the biological plausibility of postexposure antiretroviral chemoprophylaxis. These studies suggest that dendritic cells in the mucosa and skin are initial targets for HIV infection and that these cells also play an important role in disseminating HIV to cells in the regional lymph nodes [71]. In animal models, pathogenic retroviruses remain localized with dendritic cells for approximately 24 hours following inoculation with cell-free virus [72]. After 24–48 hours, the cells migrate to regional lymph nodes, resulting in productive infection of the T-cells in these lymph nodes [72].
Thus, our current understanding suggests that HIV infection occurs in a sequence of events, initially involving dendritic cells near the exposure site that subsequently move to and transmit infection to susceptible T-cells in regional lymph nodes. Early antiretroviral intervention appears most likely to prevent infection by preventing infection of susceptible T-cells. Delaying the infection of susceptible T-cells also could allow time for the exposed individual to develop a cellular immune response against HIV (see).
Although several pieces of evidence support the concept that the cellular arm of immunity plays an important role in host defense against HIV, the precise role of the cellular immunity has not been completely elucidated. Among the evidence suggesting a significant role for cellular immunity in host defense against HIV are (1) studies of prostitutes and seronegative sexual partners of HIV-infected individuals that demonstrate HIV-specific cytotoxicity in the uninfected sexual partners, studies of HCWs who have sustained occupational exposure to HIV and did not become infected but did develop cellular cytotoxic responses directed against HIV [35,36,37,38,39,40,41,42,43,44], and two anecdotal reports of individuals who sustained exposures (one by receipt of contaminated blood products [73], the other an HCW who sustained a substantial occupational exposure [74] both of whom became PCR-positive for HIV, both of whom cleared their infections [each also received three antiretrovirals], and both of whom developed substantial cellular responses directed against HIV antigens). Neither developed antibody directed against HIV. Interestingly, in two animal PEP models, successful prophylaxis was associated with the presence of effective cellular immune response in both mice and macaques [75,76]. These studies provide reasonably convincing, albeit indirect, evidence for the hypothesis that antiretroviral chemoprophylaxis administered soon after an occupational exposure in concert with a specific cellular immune response directed against HIV envelope antigens may be effective in preventing or inhibiting systemic HIV infection.
Although the initial animal model studies published failed to show any benefit of antiretroviral chemoprophylaxis, subsequent studies in several different models have clearly demonstrated efficacy for PEP. Most of the very early experiments employed intravenous injection of very high inocula. In most of these studies, the inocula were far in excess of what might be anticipated to be associated with a typical occupational exposure.
P.713
In one of the most elegant studies published to date, Tsai et al. demonstrated true prophylactic efficacy of phosphonylmethoxypropyladenine (PMPA, Tenofovir®) in a macaque model [77]. In a subsequent study, these same investigators demonstrated that all animals that received postexposure treatment for 28 days remained uninfected; only half of the animals that were treated for 10 days and none of the animals that received only three days of treatment were protected [78]. Similarly, in the latter study, delay in initiating prophylaxis was detrimental in this model. All of the animals that were treated within 24 hours of intravenous simian immunodeficiency virus (SIV) infection remained uninfected, whereas only 50% of the animals that received treatment beginning 48 hours following infection and only 25% of the animals that received treatment beginning 72 hours after exposure were protected [78].
Antiretrovirals also have been shown to be effective in preventing maternal-fetal transmission of HIV during childbirth [79,80,81]. Several studies have demonstrated the efficacy of single or combinations of agents in preventing neonatal infection. Perhaps even more important, two studies demonstrated efficacy of antiretrovirals in preventing infection when only the child was treated after birth (i.e., PEP efficacy) [82,83]. Although neither of these studies was designed to test the PEP hypothesis, both studies provide compelling evidence that these agents can work to prevent vertical transmission of HIV even well after the exposure has occurred.
Both clinical studies and clinical experience since these agents initially began to be used for PEP in the late 1980s provide additional rational for the use of antiretrovirals as postexposure prophylaxis for occupational HIV exposures. CDC published a retrospective case-control study that was designed to identify factors associated with increased risk for occupational HIV infection. In this study, a number of factors relating to the exposure itself were found to be associated with increased risk; however, the study also found that PEP with zidovudine was associated with an 81% reduction in risk [34,84]. Admittedly, the case-control design is far from the optimal study design to try to demonstrate efficacy of an agent or combination of agents in preventing infection; however, this study provided additional evidence that these agents may be useful in this setting. Antiretroviral chemoprophylaxis for occupational HIV exposures has been in use in the United States since the late 1980s [85], and, although a variety of factors has contributed to the reduction in incidence of occupational HIV infections, at least coincident with the expanded use of antiretroviral chemoprophylaxis, the numbers such infections over the past decade have decreased substantially (Table 42-4). The two anecdotal case-reports of transient infection (in the blood product recipient [73] and in the HCW who sustained an occupational exposure [74]) described earlier also provide indirect anecdotal support for chemoprophylaxis efficacy.
|
TABLE 42-4 |
|||||||
|
Current USPHS Recommendations for HIV Postexposure Chemoprophylaxis
A lengthy list of factors influences the selection of antiretroviral drugs for PEP, among them (1) the severity of exposure and the estimated risk of HIV transmission associated with the specific type of exposure that has occurred (e.g., transfusion of a volume of contaminated blood would be associated with a much higher risk for transmission than a needlestick with a solid surgical needle); (2) the source-patient's experience with antiretroviral agents as therapy for her or his infection and the influence this experience has on the likelihood that the source-patient harbors resistant isolates; (3) the source-patient's adherence to his or her current treatment regimen and the influence that this factor has on the likelihood that drug-resistant isolates were present in the circulation at the time the occupational exposure occurred; (4) the known toxicities of the agents proposed for the prophylaxis regimen and the likelihood that the HCW will be able to adhere to the recommended course; (5) the extent to which clinicians have experience administering these agents to uninfected individuals and the known safety profiles of the agent(s); and (6) cost.
Agents from at least five different classes of drugs—nucleoside reverse transcriptase inhibitors, nucleotide reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitors, protease inhibitors, and fusion inhibitors—have been marketed for the treatment of HIV disease, and, thus, are available for use as PEP. In 2006,
P.714
we unquestionably have the most experience administering agents of the nucleoside analog class for PEP, although we have reasonable clinical experience with agents in three of the other classes of drug. Perhaps only because of its historical role, we have the most experience with the nucleoside analog, zidovudine. Since their marketing, we also have substantial experience using drugs of the protease inhibitor class in the PEP setting. Whereas these agents are extremely potent, they also are associated with a variety of toxicities and complex drug–drug interactions (see).
|
TABLE 42-5 |
|||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||
Recommendations for managing occupational HIV exposures are summarized in Table 42-5. Current recommended drug combinations and their alternatives, with the advantages and disadvantages associated with the various regimens are detailed in Table 42-6. Whereas combinations of antiretroviral drugs have been definitively proven to be more effective than single agents for treating established HIV infection, we do not have such data for the PEP setting. In fact, as noted, because of the way in which we have come to learn about the use of these agents in this setting, we may never have true efficacy data for either individual drugs or combinations.
A major point of controversy involves the issue of whether to offer the basic or expanded regimen, that is, should you offer a third agent—usually a protease inhibitor or a newer non-nucleoside reverse transcriptase inhibitor—as part of the regimen. Historically, CDC has recommended adding the third drug to the regimen when
P.715
P.716
P.717
P.718
the risk for transmission is known to be increased, for example, for exposures known to be associated with higher inocula of blood or virus (Table 42-6) [2].
|
TABLE 42-6 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Adverse Effects Associated with Postexposure Chemoprophylaxis for Occupational Exposures to HIV
Antiretroviral agents are not benign drugs. All of these agents have both known and substantial side effects. One curious finding has been that healthy individuals taking these agents seem to have more and more severe side effects than HIV-infected patients have when they take these agents for therapy. In particular, subjective side effects are remarkably common among HCWs taking PEP for occupational HIV exposures.
Untoward effects have been uniformly associated with each of these agents and each of the varied regimens that have been used for PEP [86,87]. Known or anticipated side effects represent one of several important considerations when selecting a chemoprophylaxis regimen. Toxicities reported with nucleoside analogues include bone marrow
P.719
suppression (including neutropenia and anemia), nausea, vomiting, diarrhea, abdominal pain, headache, neuropathies, aminotransferase elevations, myalgias, lassitude, malaise, and insomnia. Very severe toxicities including instances of severe pancreatitis, dermatitis, severe hepatic dysfunction, lactic acidosis, or seizures rarely have been reported.
Toxicities associated with the use of protease inhibitors in a chemoprophylaxis regimen include nausea, vomiting, diarrhea, abdominal pain, hyperglycemia, hyperlipidemia, hypercholesterolemia, galactorrhea [88]; hyperprolactinemia [88]; cholestasis [89]; and headache, jaundice, anorexia, altered taste, and/or paresthesias [86]. Less commonly reported side effects associated with protease inhibitor use in chemoprophylaxis include nephrolithiasis [87] and lipodystrophy [90]. Another important issue in the use of the protease inhibitors in chemoprophylaxis regimens is drug–drug interactions that occur extremely commonly. If protease inhibitors are prescribed as part of a chemoprophylaxis regimen, the responsible clinician should evaluate all other drugs currently being taken by the exposed HCW with an eye toward these interactions. For example, simultaneous administration of either rifampin or the nutritional supplement St. John's wort can reduce plasma levels of protease inhibitors well into the subtherapeutic range. Protease inhibitors can potentiate the effects of antihistamines, ergot alkaloids (increasing risk for ergot toxicity, vasospasm, and ischemia), benzodiazepines (increasing risks for central nervous system [CNS] depression), and statins (increasing risk for severe toxicities, such as rhabdomyolysis) and can induce cardiac arrhythmias when administered with diltiazem or cisapride. Another important interaction occurs with oral contraceptives. Protease inhibitors may accelerate their clearance, decreasing their efficacy. For this reason, if protease inhibitors are included in a chemoprophylaxis regimen, women taking the regimen should use alternative contraceptive measures.
Although non-nucleoside reverse transcriptase inhibitors have never been primary choices for PEP, some authorities have advocated their use. This author has generally avoided recommending these agents for a variety of reasons, including the fact that rash is a commonly occurring side effect that could easily be confused with the seroconversion illness. Some instances of dermatitis associated with the use of nevirapine and other agents in this class have been quite severe (e.g., two reported episodes of Stevens-Johnson syndrome) [91], but perhaps the major concern relates to two instances of severe hepatic dysfunction (one requiring hepatic transplantation) and 10 episodes of moderate hepatic toxicity in HCWs who took nevirapine as part of a chemoprophylaxis regimen [91,92,93]. Concern also has been expressed about the use of efavirenz in pregnancy because of studies suggesting potential for teratogenicity in animal models. Because of its method of metabolism, efavirenz has extensive drug–drug interactions similar to those of the protease inhibitors including the interactions described previously with antimicrobials, ergots, and benzodiazepines. Other toxicities associated with the use of non-nucleoside reverse transcriptase inhibitors include mild CNS dysfunction (e.g., somnolence, insomnia, difficulty concentrating, abnormal dreams, and dizziness). Nonetheless, because of its substantial potency, CDC has recommended in its most recently published guidelines that efavirenz be included in the list of agents to be considered as the third drug in an alternative expanded regimen [2].
A major lesson learned over the past 15 years is that many, if not most, of these side effects can be anticipated and managed symptomatically prospectively (e.g., acetaminophen for headache and myalgia; prochlorperazine for nausea; antimotility drugs for diarrhea).
Failures of Postexposure Chemoprophylaxis for Occupational HIV Exposures
PEP failures occur. Most of the failures that have been reported involve the use of zidovudine as a single agent (again, perhaps an historical artifact). Additionally, five instances of failure have been reported in association with the use of regimens involving >1 agent (two failures of a 2-drug, three failures of 3-drug, and one failure of 4-drug regimen) [2,94,95]. In the overwhelming majority of these episodes, the source-patient was highly experienced with antiretrovirals and very likely may have harbored resistant isolates. Furthermore, a variety of additional factors may have contributed to chemoprophylaxis failures including exposure to very high inocula; delayed initiation of PEP; failure to achieve adequate drug concentrations; inadequate treatment duration; and so on. Occasionally, what appears to be a chemoprophylaxis failure turns out to be something else. Two publications have detailed episodes that initially were thought to be a result of occupational exposure and infection but with more investigation were found to represent instances of community infection that were entirely unrelated to occupational exposures [96,97].
Unresolved Issues
Exposure to Blood from Source-Patients Known or Suspected to Harbor Resistant Isolates of HIV
The basic and expanded CDC drug regimen recommendations represent excellent choices for instances in which the source-patient is unlikely to harbor resistant viral isolates. Drug resistance is most likely among patients who do not consistently adhere to their treatment regimens. When drug resistance is suspected, the clinician providing care should get expert counsel from individuals knowledgeable about HIV therapy to tailor a regimen to which the
P.720
source-patient's isolates have not been exposed. Basically, the same principles used to select drugs for HIV-infected patients who are failing treatment should be used to craft the chemoprophylaxis regimen [98,99].
Although the responsible clinician should not wait to begin therapy until discussing the patient with an expert in HIV therapy given all of the complexities inherent in selecting antiretroviral drugs, consultation with an expert is highly recommended when legitimate concerns about the possibility of exposure to drug-resistant HIV have arisen. This author recommends beginning an expanded regimen and immediately seeking the counsel of a colleague experienced in tailoring regimens for HIV-infected patients who have resistant isolates. If such expertise is not immediately available, clinicians can call or e-mail the experts at the National Clinicians' Post-Exposure Hotline (PEPLINE) (1-888-HIV-4911 or www.pepline.ucsf.edu/pepline).
Managing Occupational Exposure to HIV in Pregnant Staff
The decision to offer a pregnant woman postexposure antiretroviral chemoprophylaxis should be based on the same considerations that apply to all HCWs who have sustained occupational HIV exposures. In counseling the exposed pregnant HCW, the counselor must consider the risks and benefits for the worker and her fetus. Specific issues that should be discussed include the magnitude of risk for HIV transmission to the mother and the fetus associated with the exposure she has sustained, what is known about the potential for teratogenicity and other toxicities associated with the agents being prescribed in the context of the stage of pregnancy, and what is known about the safety and side effects of the specific antiretroviral agents being administered during pregnancy. In general, the data on which to base such discussions are extremely limited. For example, the risk to the fetus of administering a course of postexposure antiretroviral chemoprophylaxis is essentially unknown. Additionally, virtually all marketed antiretroviral agents have the potential for carcinogenicity, teratogenicity, and/or mutagenicity, and a few have been demonstrated to be mutagenic in premarketing animal studies. Furthermore, safety and pharmacology data addressing the risk of administering antiretrovirals to HIV-uninfected pregnant women are extremely limited. Because of the complexity of administering these agents to healthy pregnant women, this is another setting in which the prescriber should seek the counsel of someone who has substantial expertise in using these drugs on a daily basis.
In the final analysis, the HCW herself must make the decision as to whether or not to proceed with postexposure antiretroviral treatment. The role of the clinician providing care has to be to deliver accurate, thorough, balanced, and unbiased counseling.
An almost embarrassing paucity of safety data addresses the risk of administering antiretrovirals to HIV-uninfected pregnant women; similarly, data about the pharmacology of antiretroviral drugs in this setting is extremely limited. Studies evaluating the efficacy of antiretrovirals in preventing vertical HIV provide useful but not directly comparable information about the use of these drugs in the postexposure setting. A large French study identified fetal neurological/mitochondrial toxicity associated with administration of nucleoside analogues in pregnancy. Two infant deaths and six additional instances of probable mitochondrial toxicity were identified in HIV-uninfected offspring of HIV-infected mothers in this large trial [100]. Both deaths were associated with mitochondrial toxicity that led to progressive neurological disease. Interestingly, no fetal deaths attributable to, or associated with, antiretroviral-induced mitochondrial toxicity have been identified among several large U.S. vertical transmission studies. The differences between the French and U.S. experiences remain unclear.
Concern also has been expressed that the didanosine/stavudine (ddI/d4t) regimen also is associated with increased risk in pregnancy. The FDA published a warning about the use of this regimen in the treatment of HIV-infected pregnant women, noting that several instances of severe pancreatitis and lactic acidosis have occurred in pregnant patients, and that some of these episodes were associated with maternal or fetal death (or both) [101]. There are no reports of complications of this severity associated with administration of this combination as PEP. Nonetheless, based on the experience with this regimen in HIV treatment, CDC has decided to recommend against its use for PEP management for pregnant HCWs sustaining occupational HIV exposures.
“Source Unknown” Exposures
Among the most complicated issues with respect to the administration of postexposure antiretroviral chemoprophylaxis are decisions about treatment when the source-patient and/or material is suspected but not known to contain HIV. Each such episode should be handled individually and should be based on a careful risk assessment, including a determination of (1) the probability of HIV infection in the source-patient, (2) the type of exposure and the associated risk of HIV transmission with such an exposure if HIV was, in fact, likely to have been present, and (3) the risks associated with treatment for the HCW. For many such exposures, the risk for HIV transmission is so small as to be considered entirely negligible. In such settings, the risks associated with administration of the antiretrovirals likely outweigh the risks for infection, and treatment should not be recommended. Only in instances in which the risk assessment suggests that the exposure risk outweighs the risks associated with chemoprophylaxis (always a subjective assessment) should treatment proceed,
P.721
keeping in mind that, if additional data become available suggesting that the risk is lower than initially perceived, the treatment can be discontinued.
Decisions to Administer Chemoprophylaxis When Reporting Has Been Delayed
Based on the data from relevant animal models described previously, treatment should be initiated as soon as possible after exposure. In a number of animal studies, efficacy is reduced when treatment is delayed for more than 24 hours [77], but the relevance of this observation to low-inoculum transcutaneous and transmucosal occupational HIV exposures is completely speculative. Occupational HIV exposures should be considered medical emergencies. Antiretroviral agents should be administered as soon as the rationale for them is apparent. Institutions should use this measure as an HCW safety performance measure, and this issue should become the focus of ongoing institutional performance improvement activities. When definitively indicated, PEP should be started as soon as practical (i.e., within hours rather than days). As noted, if because of the likelihood of resistance, the HCW is pregnant or if the practitioner encounters other complexities, consider initiating the basic or, perhaps more sensibly, the expanded regimen until consultation can be obtained. In instances in which the risk of transmission is high (e.g., an instance in which a surgeon sustains a scalpel cut when the source-patient is known to have a high viral burden), initiate treatment even after a long delay (e.g., even 1 to 2 weeks after the exposure).
Because no data definitively demonstrate the efficacy of PEP, the optimal duration of chemoprophylaxis can obviously not be known. Animal models have provided highly variable results. At the NIH, a four-week regimen is used.
Follow-Up for Occupational HIV Exposures
Individuals sustaining occupational exposures should undergo baseline testing at the time of the exposure to demonstrate that they have not been previously infected with the pathogens being considered as occupational risks. In addition to baseline HIV testing, serological testing for a documented occupational exposure is usually performed six weeks, three months, and six months after exposure [2].
Some exposure characteristics may be associated with increased risks for transmission (e.g., injection of a volume of contaminated blood, simultaneous exposure to HIV and HCV). In such instances, extending the testing period makes implicit sense.
Routine follow-up is crucial to effective postexposure management. All HCWs sustaining occupational exposures should be re-evaluated at 48 hours postexposure to get a clear reading on how they are managing their exposures. This author recommends that individuals be seen weekly if possible while on PEP to make certain that they are tolerating therapy and that they do not have unanswered questions or unresolved issues.
Most (i.e., >80%) instances of documented HCW seroconversions have been associated with the symptoms typical of the acute retroviral infection (fever, lymphadenopathy, pharyngitis, rash, headache, profound fatigue). For this reason, HCWs sustaining occupational exposures should be counseled to return for evaluation and HIV testing if such symptoms occur. Exposed HCWs should be counseled that these symptoms do not always indicate acute HIV infection. A variety of other circumstances (e.g., reactions to nevirapine or other antiretrovirals or other viral infections can produce virtually identical symptoms).
Providers evaluating HCWs who develop symptoms suggesting acute HIV infection must be aware that HIV antibody tests may be negative or indeterminate during the early phases of the illness. Direct tests for the viral genome, viral load tests (quantitative HIV RNA PCR), or viral cultures may be more valuable in making the initial diagnosis. Whereas these latter tests may be of value in differentiating the acute seroconversion illness from other diagnoses, these tests may produce more uncertainty than help in the routine management of occupational exposures and should not be used routinely in follow-up. Positive tests should be repeated to confirm the result.
As noted, HCWs who elect to take chemoprophylaxis after occupational HIV exposures should return 48 hours after the exposure and then, at a minimum of two weeks after the initial return visit. The clinician should consider seeing these individuals weekly while on treatment to assess them carefully for signs and symptoms of drug toxicity and to make certain that their symptoms are being managed appropriately. The visit should include a careful interim history, a focused physical examination, questioning about signs and symptoms of drug toxicity, inquiry as to whether the individual has unanswered questions about the exposure or the treatment, and the collection of specimens for laboratory tests relevant to the antiretrovirals being administered. As a general rule, all patients should have a complete blood count as well as renal and hepatic function tests. If protease inhibitors are included in the HCW's regimen, specimens should be drawn for random blood glucose and a lipid profile. HCWs electing to take PEP also should be counseled to return for re-evaluation if intractable side effects of therapy occur.
A major goal for the provider is to ensure that the exposed HCW completes the course of chemoprophylaxis. At each visit, information should be provided to the exposed HCW about potential drug interactions, emphasizing the drugs that should not be taken with the prophylactic drug regimen. In addition, the clinician should focus on the side effects being experienced by the exposed HCW and should address how to manage them. In addition to the earlier discussion about the symptoms associated with acute HIV infection, HCWs receiving postexposure
P.722
antiretroviral chemoprophylaxis also should be informed about symptoms suggestive of serious toxicity (e.g., back or abdominal pain, jaundice, pain on urination, or blood in the urine and symptoms of hyperglycemia such as increased thirst or frequent urination).
The most common reason for having HCWs spontaneously discontinue their chemoprophylaxis relates to the myriad side effects associated with these regimens. Common side effects, including nausea and diarrhea, often can be managed successfully with antiemetic and/or antimotility agents without modifying the chemoprophylaxis regimen. Prescribing these drugs prospectively (i.e., at the time the chemoprophylaxis regimen is initiated) often makes implicit sense. HCWs should be told what to expect in side effects and what to do to manage them. Dealing with these problems prospectively will likely increase adherence with the chemoprophylaxis regimens substantially. In instances in which the side effects are not easily managed by antimotility, analgesics, or antiemetics, modifying the dose interval, dose reduction of the prescribed antiretrovirals, or regimen modification may be necessary to make it possible for the HCW to complete the chemoprophylaxis regimen.
Provider-to-Patient Transmission of Blood Borne Pathogens
In general, the risks for provider-to-patient transmission of each of the important BBPs are substantially smaller than the corresponding risk for patient-to-provider transmission. Individuals chronically infected with any of the most prevalent blood borne viruses—HBV, HCV, or HIV—are extremely unlikely to transmit infection during routine patient contact. Nonetheless, instances in which each of these viruses were transmitted from an infected HCW to one or more of her or his patients are well documented in the literature. Because the risks for needlestick transmission of each of the major pathogens differs substantially, a reasonable hypothesis is that the risks for provider-to-patient transmission also varies substantially from pathogen to pathogen. This substantial variation argues for management of providers on a rational, pathogen-by-pathogen basis or, perhaps even more rationally, based on the provider's circulating viral burden.
Hepatitis B
Despite the fact that chronic HBV carriers may have remarkably high levels of circulating viremia, routine aspects of patient care really pose virtually no risk for provider-to-patient HBV transmission. Providers who routinely conduct what CDC has termed exposure-prone invasive procedures do present some risk to their patients for provider-to-patient spread of this BBP. The risk clearly is associated with the provider's HBV circulating viral burden. Thus, providers who have high circulating levels of HBV viral DNA or those who are hepatitis B “e” antigen positive are associated with the highest (albeit still extremely low) levels of risk for provider-to-patient transmission. In a review of articles published in the mid-1990s, CDC reported that 42 different HBV-infected HCWs (most of whom were hepatitis B “e” antigen positive) had been detected as having infected ≥1 patients (with >375 patients in these 42 individuals' practices having acquired infection from their providers) [102]. In many of these instances, investigation revealed that the practitioners involved used minimal or, in some instances, inadequate infection control procedures. Only two provider-to-patient HBV clusters have occurred in instances in which the provider was aware of her or his infection and was paying particular attention to infection control procedures that were intended to decrease the risk for transmission. In the first of these two instances, four patients acquired clinical HBV infection from an orthopedic surgeon [103]. In the second cluster, 19 of a thoracic surgery resident's intraoperative patients became infected [104]. Investigation of both of these clusters failed to identify a route of transmission or problem with technique.
The management of HCWs who are chronic HBV carriers is complex. Historically, with the exception of providers who were shown to transmit infection to their patients, no restrictions were put in place for providers who were HBV carriers. In part in response to a cluster of provider-to-patient transmission of HIV discussed subsequently in more detail, in 1991 CDC issued guidelines recommending that HCWs who perform so-called exposure-prone invasive procedures should personally be aware of their HBV infection statuses. These guidelines recommend that HCWs who learn that they are chronically HBV-infected and are HBV “e”-antigen positive should not perform exposure-prone procedures unless they have sought the counsel of an expert review panel and been advised under what circumstances (if any) they would be allowed to perform these procedures [105]. An HCW who is allowed to perform exposure prone procedures must first inform a prospective patient about the HCW's infection status [105]. The Society for Healthcare Epidemiology of America (SHEA) also issued a set of recommendations addressing the management of providers who are chronic carriers of HBV and other BBPs [106]. The guidelines recommend that providers who are HBV “e”-antigen positive take the additional precaution of routinely double gloving. Furthermore, the guidelines suggest that these providers not conduct procedures that have been epidemiologically associated with a risk for provider-to-patient transmission. SHEA further recommends that HBV-infected HCWs should not volunteer their infection status to patients except in circumstances in which the patient sustained an exposure to the provider's blood or other potentially infectious body fluid [106]. The U.S. Congress subsequently mandated that all states either implement the 1991 CDC guidelines or certify that state guidelines were equivalent to them.
The United Kingdom has taken a more conservative approach to the management of practitioners who are chronic
P.723
HBV carriers. Revised guidelines issued in 1993 require vaccination of all nonimmune HCWs who perform “exposure prone procedures” and postvaccination testing to document a protective response. The guidelines also clearly delineate practice restrictions for providers who are found to be HBV “e” antigen positive chronic carriers [107].
In 1997, several instances of provider-to-patient transmission of HBV from practitioners who were found to be HBV “e”-antigen negative but who were infected with strains of HBV that are “pre-core mutants” (i.e., strains that are genetically unable to express “e”-antigen but still capable of assembling infectious virions and producing high-viral-burden infection) [108]. Some authorities have proposed using the quantitative measure of circulating HBV-DNA as a basis for restricting the practices of infected providers who perform “exposure-prone procedures” [109,110]. In general, whereas most authorities seemed to have embraced this concept, substantial disagreement still exists about where to set the threshold for restriction; some have recommended that a viral burden as low as 100 copies/ml be used, and others have suggested 1,000, 10,000, or even 100,000 as perhaps more appropriate thresholds. Current USPHS guidelines do not address the issue of viral burden as a determinant for practice restrictions.
Hepatitis C
Provider-to-patient HCV transmission has been an extremely uncommon event except in some unusual circumstances. As noted for all of these BBPs, provider-to-patient HCV transmission is extremely unlikely in the setting of routine (i.e., non-invasive) patient care. The risk for provider-to-patient HCV transmission appears to be even smaller than the risk for HBV in this setting, presumably because most individuals chronically infected with HCV have circulating viral loads that are several factors of 10 lower than those of HBV “e”-antigen positive carriers. Despite the low risk for transmission, several instances of provider-to-patient HCV transmission have been reported in the past few years [111,112,113,114,115,116,117,118,119,120,121]. Although the precise mode of transmission for the overwhelming majority of these patients remains unknown, the circumstances surrounding several suggest that transmission was associated with percutaneous exposures. Interestingly, a number of the instances of provider-to-patient HCV transmission have been associated with HCW injecting drug use. The contribution of injection drug use is well documented in a few of these instances (cf. the epidemic of patients associated with an anesthesiologist in Spain who was addicted to opiates who was using some of patients' narcotics and then injecting the patients with the same syringe that he had used, in the process infecting >200 patients [111]). Detection of underlying injection drug use in this setting is difficult at best, so one cannot say for certain the extent to which this behavior may have influenced the other published reports.
Thus, provider-to-patient HCV transmission has been documented only rarely and in only one published episode in the United States. A second instance of HCV transmission from a U.S. provider to a patient has been reported in the lay press but has been described only obliquely in the medical literature. In this latter report, based on a personal communication, the authors report that a “look-back” study of the patients of a cardiac surgeon who according to lay press accounts had transmitted infection to three patients [122] found that the surgeon “likely transmitted HCV to as many as fourteen of the 937 patients who could be evaluated” [121].
In the absence of injection drug use in the HCW, provider-to-patient HCV transmission appears extremely uncommon in association with highly invasive procedures and associated with HCWs who have high circulating HCV viral burdens. To date, because of the paucity of data documenting the occurrence of provider-to-patient HCV transmission, the USPHS has not issued recommendations suggesting that the practices of HCV-infected HCWs be limited in any way. Conversely, public health authorities in the United Kingdom have recommended practice restrictions for HCV-infected providers, specifically noting that HCV-infected providers who have circulating HCV-RNA may not conduct “exposure-prone invasive procedures”; that trainees found to have circulating HCV-RNA should be restricted from starting training in “exposure-prone invasive procedures”; that HCV-infected providers who have circulating HCV-RNA who receive antiviral treatment and become HCV-RNA negative for a period of six months can be permitted to return to performing “exposure-prone invasive procedures” (but must be retested in six months to ensure that they remain HCV-RNA negative) [123]. A European conference convened to construct rational guidelines for infected providers could not reach consensus about restrictions for HCV-infected providers, ultimately concluding that, “on balance, it is not recommended that exposure-prone procedures be forbidden for hepatitis C-infected healthcare workers” [124]. Thus, the information accrued to date does not suggest a need for additional intervention. Ultimately, if transmission is detected with more regularity and an intervention becomes appropriate, creating restrictions based on the HCW's viral burden and transmission history may be appropriate. Deciding where to set the threshold for restriction clearly would be a challenge.
Human Immunodeficiency Virus
Four instances of HIV transmission from an infected HCW to ≥1 of her or his patients have been reported in the literature with a total of nine provider-to-patient infections detected in these four instances of transmission [125,126,127,128,129,130,131].
One of the four instances of transmission occurred in the United States [126,127,128], two occurred in France [125,129], and the fourth occurred in Spain [131]. The episodes in the
P.724
cluster of six infections that were detected in the United States in 1990 were linked to an HIV-infected dentist in Florida. Although these episodes were investigated thoroughly, the precise mechanism of transmission was never identified. The extraordinarily high transmission rate in the U.S. dentist's practice has never been explained. The two French episodes were reported in 1999 and 2000. The initial French episode involved HIV transmission from an HIV-infected orthopedic surgeon to one of his patients [125,130]. The surgeon was not aware of his infection until AIDS was diagnosed in 1994. The one iatrogenic infection that was detected was identified in a retrospective investigation of the surgeon's patients. As in the U.S. dental office episode, the precise mechanism of transmission could not be determined; however, the patient who had become infected had undergone an extremely lengthy procedure at a time that the surgeon very likely had a very high viral burden. The second French episode is more puzzling. In this instance, transmission was thought to have occurred from an infected nurse to a patient despite the fact that the nurse did not conduct any sort of invasive procedure. As with the other episodes, no route of transmission could be delineated. The nurse also was infected with HCV and at the time of the iatrogenic HIV transmission, had both advanced HIV disease and advanced HCV infection [129]. The fourth instance of transmission occurred in Spain and was reported in a news report in the medical literature in 2003 [131]. The details of this episode have never been described in the literature; however, the news report suggests that the transmission occurred during a caesarian section. None of the 250 other patients who had procedures performed by the obstetrician were found to be infected [131].
Given our 25-year experience with this disease, the paucity of episodes that have been detected underscores the fact that the risk HIV transmission from infected providers to patients is extremely low. Following the detection of the cluster in the Florida dentist's practice, the U.S. Public Health Service issued guidelines for providers infected with BBPs [105]. Those guidelines concluded that HCWs who are infected with HIV or HBV and are “e” antigen positive should not perform “exposure-prone procedures” unless they have sought the counsel of an expert review panel and have been advised under what circumstances, if any, they may continue to perform these procedures. The document also noted that the circumstances under which infected practitioners were permitted to continue performing exposure-prone procedures would include prospectively notifying patients of the practitioner's infection status before the procedure. After the guidelines were published, the U.S. Congress passed a statute (P.L. 102–141) mandating that all states adopt the CDC (or equivalent) guidelines. Subsequently, the CDC director at that time wrote a letter to all state health departments, emphasizing that the states, not the CDC, would certify the equivalency of the individual state's guidelines. He also concluded that, in his view, exposure-prone procedures would best be determined on a case-by-case basis, taking into consideration the specific procedure and the skill, technique, and possible impairment of the infected HCW. Many states created their own guidelines and certified them as equivalent. Thus, as a result, substantial variability exists in state guidelines. Whereas the U.S. guidelines are state based, the U.K. guidelines state that HIV-infected providers may not conduct exposure-prone invasive procedures [132].
Curiously, none of the U.S. guidelines take intos consideration that fact that the risk for provider-to-patient transmission of each of these BBPs almost certainly is associated with the provider's viral burden. In an era in which many HIV-infected patients have their circulating viral burden suppressed below detectable levels, in which a substantial fraction of patients who have chronic HCV infection can be either cured or have their infections substantially suppressed, and that includes promising new therapies for chronic HBV infection, this important factor should certainly be considered in the recommendations for managing infected providers. Whereas the United Kingdom, the European consortium, and the Netherlands all have included viral burden in assessing risks associated with HBV-infected providers, no one has attempted to do this with HCV or HIV. The 1991 USPHS guidelines clearly need updating. Until these guidelines are brought to currency, this author i recommends basing practice restrictions on evidence suggesting (1) that an infected HCW is impaired, (2) that the HCW does not adhere to accepted infection-control procedures, (3) that a documented risk for transmission has been established for the pathogen and procedure under consideration, or (4) that BBP transmission to a patient has occurred or is suspected to have occurred (modified from reference [106]).
References
P.725
P.726
P.727