Most of the occupational HIV infections documented in the United States have been associated with parenteral injuries inflicted by hollow-bore needles used in veins or arteries, but other sharp instruments have also been involved in transmission. Six instances of HIV infection have occurred after either exposure of breaks in skin to HIV-contaminated fluids or exposure of mucous membranes to HIV-contaminated materials. Each of these six instances involved a large-volume exposure, an extended duration of exposure, or both. In one case of mucous membrane exposure in Europe, the inoculum was smaller.⁶ An additional case was associated with several exposures over an extended time period.⁷ To date, contamination of intact skin with blood or other infectious material, close personal contact with infected patients, and contact with contaminated environmental surfaces or fomites have not been linked to occupational HIV transmission. Aerosolization of blood occurs during dental, pathologic, laboratory, and surgical procedures, and conventional surgical masks do not prevent inhalation of aerosols. Nonetheless, to date, no data indicate that aerosol exposure is a route of HIV transmission in any setting.

Exposures to blood from HIV-infected patients account for all but 5 of the 58 documented occupational infections in the United States. Of the 5, one infection resulted from exposure to bloody pleural fluid and three involved exposures to concentrated preparations of HIV in scientific laboratories; for the fifth, the source material was not reported.

Definition of Occupational HIV Exposure

The instances of documented occupational HIV transmission have been helpful in developing a definition of what constitutes an exposure associated with HIV transmission risk. As noted in the previous section, exposure routes implicated in occupational HIV transmission include (1) percutaneous injury (e.g., needle puncture or cut caused by a needle or other sharp object); (2) mucous membrane contamination; and (3) contamination of nonintact skin (e.g., skin that is chapped, abraded, or afflicted with dermatitis).^8,9 Whereas HIV-infected blood contamination of intact skin has not been implicated in occupational infection, exposures of intact skin to contaminated blood for extended periods (several minutes or longer) or exposures involving extensive areas of skin should be considered potentially infective. This potential exists because unrecognized areas of inadequate skin integrity could serve as portals of entry for the virus.

Sources of HIV that may pose a risk for transmission through these routes include blood; visibly bloody fluids; tissues; and other body fluids, including semen, vaginal secretions, and cerebrospinal, synovial, pleural, peritoneal, pericardial, and amniotic fluids.^8,9 In addition, any direct cutaneous or mucosal contact (i.e., without barrier protection) to concentrated HIV in a scientific or research laboratory or production facility should be considered an exposure.

Although nonoccupational HIV transmissions have been attributed to contact with blood-contaminated saliva, these incidents were not analogous to the contact with saliva that occurs during dental or medical care.^9,10 In the absence of visible blood in saliva, exposure to saliva from an HIV-infected person is not thought to pose a risk for HIV transmission. Exposure to products that are not visibly bloody (tears, sweat, urine, or feces) from infected patients does not constitute an HIV exposure. Whereas human breast milk has been implicated in perinatal HIV transmission, this route of transmission is not analogous to occupational exposure, and contact with breast milk from a patient infected with HIV does not constitute an occupational exposure.⁸

Worldwide, more than 20 prospective studies have helped to quantify the transmission risk associated with discrete occupational HIV exposures.^2,11 In each of these studies, health care workers who sustained occupational HIV exposures were tested for HIV antibody at baseline (i.e., at the time of exposure) and periodically thereafter, at regular follow-up intervals, to detect new infections.

Pooled data from these studies suggest that the average risk for HIV transmission associated with percutaneous exposures to blood-contaminated sharp objects that have been used on HIV-infected individuals is 0.32% (21 infections associated with 6498 exposures; 95% confidence interval of 0.18% to 0.46%).² The estimated risk for mucocutaneous transmission is 0.03% (1 infection associated with 2885 HIV exposures involving mucous membranes or nonintact skin); however, this estimate may be biased because the single transmission event was actually reported before prospective data were collected from the involved institution.¹² The risk for infection associated with intact skin exposure to HIV is too low to be detected in these studies.¹³

Since the onset of the AIDS epidemic in the early 1980s, only four instances of HIV transmission from infected health care workers to one or more patients have been reported.^38–43,44 Of these instances of transmission, one occurred in the United States in 1990^40–43; two were reported from France^38,44,45; and the fourth was reported from Spain.³⁹

The episode in the United States involved six patients whose HIV infections were linked epidemiologically and through DNA sequencing to a dentist who had AIDS. Although the investigation indicated that HIV transmission occurred in the dentist’s office and probably represented transmission from dentist to patient, rather than from patient to patient, precise mechanisms of transmission were never determined. Although the dentist was a patient in his own practice, no deficiency of infection control that would readily explain HIV transmission to the six patients could be identified. The dentist did not report occupational injuries that could have created opportunities for cross-contamination, nor was it proved that the infections were intentionally transmitted. The high rate of transmission in this situation remains unexplained.

The second episode of iatrogenic HIV transmission involved an orthopedic surgeon in France whose HIV transmission to one patient was confirmed through DNA sequence analysis of viral isolates obtained from the surgeon and the patient.^38,45 The surgeon in this case probably became infected from an occupational injury sustained during a surgical procedure in 1983. The surgeon was not aware of his infection until AIDS was diagnosed in 1994. Investigators initiated a retrospective investigation of the 3004 patients who had undergone at least one invasive procedure performed by the infected surgeon since 1983. Investigators were able to contact 2458 of these 3004 patients and were able to assess the HIV infection status of 983 of these 2458 patients. One patient, who had a negative result of an HIV antibody test before undergoing the first of three procedures performed by the index surgeon, was found to be infected with HIV when she underwent preoperative testing before a third procedure. Although the precise mechanism of transmission is unknown, the duration of the initial procedure (10 hours) and a presumably high viral titer in the surgeon were hypothesized as contributing factors to the transmission event. No breaches in recommended infection control practices were identified. In the third instance, also in France, the virus was thought to have been transmitted from an infected nurse to a patient (according to the phylogenetic analysis of the isolates from the patient and the nurse), although no route of transmission could be identified.⁴⁴ In a subsequent retrospective study of 7580 patients of the infected nurse, investigators were able to notify 5308 of the patients concerning the potential for exposure.⁴⁶ No additional HIV infections were detected in the 2293 (of the 5308) who were tested. The nurse who was identified as the source of the patient’s infection was coinfected with HCV and was found to have both a high HIV viral burden and advanced HCV-induced hepatic disease, including clotting abnormalities.

In the fourth case of iatrogenic transmission of HIV in Spain (about which only limited information has been published),³⁹ a woman was infected with HIV by her gynecologist, presumably during the conduct of a cesarean delivery. Spanish officials conducted a retrospective study in which 250 of 275 of the gynecologist’s patients were tested; no additional infections were identified.³⁹

Nosocomial HIV Transmission from Infected Patients to Other Patients

Episodes of nosocomial HIV transmission from one patient to another have most frequently involved breaches in infection control practices and disinfection procedures. Re-use or improper sterilization of blood-contaminated injection needles or syringes has been linked to HIV transmission to hospitalized children in Russia,⁵⁴ Libya,^55,56 Romania,^54,57–59 several countries in Africa,^55,60–63 and India⁶⁴ and probably in other developing countries. Medical errors in three institutions (two in the United States and one in the Netherlands) resulted in inadvertent exposures of patients to HIV, as a result of injection of blood from HIV-infected patients during nuclear medicine procedures.⁶⁵ Contamination of multidose vials frequently has been incriminated as a vehicle for transmission of HIV and other bloodborne pathogens in several instances in both industrialized and developing countries,^55,66–69 and one modeling study has demonstrated the clear potential for such transmission.⁷⁰

Five patients in Australia who underwent minor outpatient surgical procedures necessitating local anesthesia that were performed on the same day by an HIV-negative surgeon were subsequently found to be HIV positive.⁷¹ The first infected patient had known risk factors for HIV and was the probable source of infection for the other four patients. The exact mode of patient-to-patient transmission in this practice has not been elucidated. No cases of patient-to-patient transmission of HIV have been reported from hemodialysis centers in the United States. Conversely, HIV transmission to at least nine patients in a hemodialysis center in Colombia has been reported and was attributed to inadequate disinfection and re-use of contaminated access needles; similar cases have been reported from Argentina and Egypt.^72–76

Injury Prevention during Routine Patient Care

Needle punctures are the most frequent cause of occupational HIV infection, and priority for their prevention is highest. All health care workers, including those who actually perform or assist with procedures, those employed as housekeepers and laundry workers, and other nonclinicians are at risk for injury and infection. For this reason, prevention efforts must incorporate strategies that prevent injuries while the needle is being used for its intended purpose, as well as strategies that decrease the risk for injuries after use or disposal of the device.

One component of injury prevention that may be overlooked is avoidance of unnecessary needle use. Phlebotomy is a common indication for using needles and is also the procedure most commonly associated with occupational HIV infection. To reduce opportunities for injury, health care workers should avoid “routine” blood drawing that does not contribute to patient care; use better planning to minimize the number of phlebotomies necessary to obtain the necessary blood tests; and use needleless vascular access ports for blood withdrawal and injection of medication. Similarly, avoiding unnecessary placement of intravenous catheters when alternative routes are available for administering therapy will decrease the opportunity for needle-related injuries.

Implementation of needleless or protected-needle infusion systems can reduce the frequency of needle-related injuries.^89–92 The effect of this intervention on disease transmission is less certain because most needles used for intravenous infusion are not contaminated with blood. Not every institution that has implemented one of these systems has found it to be effective in preventing disease transmission or cost-effective.⁹³ Needles used for heparin flushes and needles in contact with ports close to the site of intravenous line insertion are more likely to be contaminated with blood and hence are more hazardous. In reality, determining whether an infusion needle is or has been contaminated is often difficult, if not impossible. Preventing injuries associated with intravenous infusions is therefore an important component of risk management, even though such injuries may be substantially less likely to transmit infection. Improving worker safety must not increase the risk for infection or complications among patients; results of several studies have suggested an increased risk for device-associated bacte­remia in association with the use of certain needleless intravenous systems.^94–96

Safer needle devices that have been engineered to retract, cover, or blunt the needle are now in widespread use. Some of these devices have safety features that are activated while the needle is being used for its intended purpose. Other safety features are activated after withdrawal from the patient. The most effective devices are passive (i.e., do not require the user to activate the safety feature), do not require extensive training, and are cost-effective. A multicenter study conducted and reported by the CDC demonstrated that implementing safer needle devices for phlebotomy procedures is an effective strategy for preventing percutaneous injuries.^97–99 Improved product design and lower cost may lead to even more effective programs for protecting workers during procedures that require the use of needles.

All needles and other sharp instruments, with or without safety features, should be discarded in puncture-resistant containers. Such containers should be located as close as possible to the point of use in emergency departments, in operating rooms, and in other patient care areas. Proper disposal also prevents injuries caused by needles that have been carelessly discarded. Needle disposal programs can significantly reduce the incidence of injury. Finally, some investigators have suggested that employees who have certain personality profiles associated with risk-taking may have increased risks for occupational exposures to blood.¹⁰⁰

Injury Prevention during Invasive Surgical, Obstetric, Dental, and Radiologic Procedures

Preventing intraoperative and intraprocedural injuries that confer a risk for blood exposure is an important priority for preventing HIV transmission among health care providers and their patients. For operative procedures, data from observational studies indicate that the risk for provider injury is highest during procedures lasting longer than to 3 hours, when intraoperative blood loss exceeds 250 to 300 mL, and during certain categories of major procedures (e.g., intra-abdominal gynecologic procedures, vaginal hysterectomies, major vascular procedures, and orthopedic procedures).^{51,101,102–109}

Prevention priorities in the operating room are based on the same principles used in other health care settings.^110–116 The least invasive surgical approach that will achieve the desired patient outcome is preferable. For example, fiberoptic techniques usually pose a lower risk for injury and blood exposure than do more invasive surgical approaches. Similarly, when patient safety allows, alternatives to needles and other sharp implements (e.g., adhesive tape, staples, and tissue glue rather than sutures; electrocautery rather than scalpels) should be used.

Suture needles are the most frequent cause of injuries in operating and delivery rooms. Curved suture needles with blunted tips are now available and appear to be an acceptable replacement for standard curved suture needles for suturing many types of tissue.^98,117–121 Use of these needles is effective in preventing intraoperative injuries. In one multicenter study, 1.9 injuries per 1000 curved suture needles used were observed during gynecologic surgery, but no injuries were associated with the use of blunted suture needles.⁹⁸ The estimated odds of sustaining an injury with a curved suture needle were reduced by 87% when 50% of the suture needles used during a procedure were blunted. Use of blunted suture needles is also associated with a lower incidence of glove perforation. Overall, surgeons involved in these studies were accepting of the blunted needle and no adverse outcomes among patients were noted.

Another approach advocated to reduce risk for percutaneous exposures during the conduct of invasive procedures is the so-called no-touch technique. Aspects of this technique include using instruments, rather than hands, for retracting and exploring tissue; avoiding the simultaneous presence of the hands of two or more operators in the procedural field; avoiding hand-to-hand passage of sharp instruments by using a “neutral zone” (e.g., emesis basin, Mayo stand, or magnetic pad); and announcing the transfer of sharp instruments from person to person.

Gloves provide an important barrier between potentially infectious materials and health care providers. Sterile surgical gloves prevent microbial contamination of patient wounds and sterile instruments and also protect surgical personnel from cutaneous blood contact. Surgical gloves do not provide a barrier to sharp object penetration, but they may reduce the volume of blood transferred to the skin and hence decrease the risk for infection by a bloodborne pathogen.

Unfortunately, glove perforation is extremely common, especially during major surgical procedures of long duration. Breakdown in glove integrity can cause contamination of exposed tissue and blood contamination of the provider’s hands; and if the provider sustains an injury that results in bleeding (needle puncture) or tissue trauma (e.g., suture-induced “shear injury”), the patient may be exposed to the provider’s blood or interstitial fluids. Double gloving is one strategy that may attenuate these problems. Without exception, in all studies of double gloving, the prevalence of inner glove perforation was significantly lower than that of the outer glove.^{102,122–130} In addition, double gloving reduces the frequency of visible blood contamination of providers’ hands.

Overall, the thumb, index, and middle fingers of the nondominant hand are the most common glove perforation sites.^51,131,132 Reinforcement of these areas is one approach to prevent perforation.^{110,133–136} The use of gloves that increase the thickness of the barrier between a patient and the provider creates concern about manual dexterity and tactile sensitivity.^134,136,137 Nonetheless, in a study that measured two-point discrimination and the ability to tie surgical knots, double gloving did not affect performance. Some measures of tactile sensitivity are reduced, but not the ability to discriminate between suture pairs. In a subjective assessment, surgeons reported that double gloving did impair comfort, sensitivity, and dexterity, but acceptance was better if the inner glove was larger than the outer glove.¹³⁷

The benefits of double gloving, glove reinforcement, and new glove materials in preventing disease transmission have not been proved. Nevertheless, double gloving decreases inner glove perforation and reduces blood contamination of operators’ hands. Most authorities now recommend routine double gloving during invasive surgical and obstetric procedures.^52,131

As emphasized previously, preventing intraoperative injuries to surgical care providers is the most important strategy for preventing the transmission of HIV and other bloodborne pathogens to patients. In two studies of intraoperative provider injuries, 11.4% to 29% of the sharp objects that injured a provider subsequently recontacted the patient.^50,51 These exposures are preventable by immediately replacing the contaminated suture needle or other sharp object before re-use. Recontacts can also occur when the provider is injured by bone spicules or materials permanently embedded in the patient’s body.^50,51 Exposure to these types of sources might be prevented by the use of reinforced gloves,^138,139 liners, or other devices or materials to protect the provider’s hands.^{110,134–136,138,140} Gloves constructed of monofilament polymers or other materials resistant to tears have become available for use when manipulation of bone fragments or of suture wires is needed, but, as noted previously, their use is not universal because of the associated decrease in tactile sensation.

The frequency of blood exposure among dental personnel has declined since the 1990s. Surveys conducted at annual meetings of the American Dental Association revealed that the mean number of injuries involving blood or body fluid contact reported by dentists decreased from 12.0 to 2.2 per year between 1986 and 1993.¹⁴¹ This impressive decline may be the result of the widespread implementation of universal precautions in dental practices, safer instrumentation, and educational programs for dental professionals and patients. Nonetheless, a significant number of exposures continue to occur in the dental health care setting.¹⁴²

Specific practices designed to prevent injuries include use of the one-handed “scoop” technique and mechanical devices for recapping needles used to administer local anesthetic; restricting the use of fingers during suturing and administration of anesthetic; controlling the placement of sharp instruments (e.g., scalers and laboratory knives); and improvements in the ergonomic design of dental operatories.¹⁴³ Safer devices such as self-sheathing anesthetic needles, dental units designed to shield burs in handpieces, and plastic finger guards might also contribute to safer dental care.

Today, most injuries to dental personnel actually occur outside the patient’s mouth, involve very small amounts of blood, and are unlikely to pose a risk to patients.^{141,143–145} In a 7-month observational study of dentists and of oral and maxillofacial surgical residents in two New York City teaching hospitals, injuries were observed during 0.1% of dental procedures, and 86% of these injuries occurred outside the patient’s mouth.¹⁴⁴ Only one needle puncture was observed during 16,000 anesthetic injections.

Low exposure rates have been observed during outpatient oral surgical procedures as well. However, oral procedures performed in the operating room are associated with injuries caused by surgical wires during fracture reduction.^123,146,147 The use of small plates instead of wires during the surgical treatment of some mandibular fractures, as well as reinforced gloves, may help prevent some of these injuries.

Interventional radiologists, stimulated by increasing awareness of bloodborne pathogen risks, have developed similar approaches to risk aversion in the interventional radiology suite,^148–150 a venue in which particular attention must be paid to the risk for splashes, spattering, and mucous membrane exposures.¹⁵¹

Exposure Site Management

Wounds and skin sites that have been in contact with blood or body fluids should be washed with soap and water.^8,162 Exposed mucous membranes should be flushed with tap water. Eyes should be flushed with sterile water or a commercial eye irrigant when available or else with clean tap water. Antiseptics can be used to flush the wound, but they are not known to reduce the incidence of infection, and decontamination should not be delayed until they are obtained.

Counseling and Triage

The emotional impact of a known or suspected HIV exposure is usually significant, especially in the first hours to days after the episode.^163,164 During this time, it is helpful to have access to supportive counseling by experienced clinicians who are familiar with the special medical and psychological needs of exposed persons. The clinician must function as an effective translator. Objective information about exposure risk and the pros and cons of chemoprophylaxis must be communicated to an individual who is usually preoccupied with very subjective emotions.^152,165 Although trying to talk an exposed worker out of “irrational” fear when objective data indicate that the risk is low may seem intuitively tempting, such reassurance is rarely successful. To the worker, any exposure risk may feel like 100%, and no amount of epidemiologic data is likely to change this impression in the short run. The most important initial messages to communicate are probably empathy (e.g., “I can see how frightening this is for you”), validation (e.g., “Most people in your situation feel the way you do now”), and reassurance (e.g., “This is difficult, but I’ll help you get through it”). Because the exposed individual is very likely to be preoccupied, the counselor should be patient and prepared to answer the same questions repeatedly.

Health care workers who are too upset or confused to make decisions about chemoprophylaxis can sometimes be helped by suggesting that treatment be started immediately, with the option to stop it later (e.g., “Start treatment now, and then tomorrow we can decide whether continuing is your best option”). Buying some time in this manner alleviates the additional pressure to make an immediate decision about initiating the full 4-week course of treatment; empowers workers to be able to change their minds about treatment when they are able to evaluate the risks and benefits more objectively; and (on the basis of animal data) provides the best opportunity for therapeutic efficacy.^152,165,166

Several often overlooked points should be included in counseling any health care worker who is facing a decision about postexposure chemoprophylaxis, among them: (1) most persons exposed to HIV do not become infected, even if no treatment is administered; (2) treatment can be stopped at any time; (3) data about the efficacy and safety of chemoprophylactic regimens are incomplete; (4) to date, zidovudine is the only drug for which there are any data suggestive of efficacy in preventing HIV transmission in humans; and (5) despite the logic associated with administering two or three agents for chemoprophylaxis, no data prove that combination treatment is more effective than single-drug therapy for HIV prevention.¹⁵³

Health care workers who sustain exposure to HIV should be counseled to avoid transmission to others during the follow-up period, especially during the first 6 to 12 weeks after exposure, when seroconversion is most likely to occur.^8–10 Recommended practices to avoid transmission include sexual abstinence or the use of condoms to prevent sexual transmission, as well as the avoidance of blood and organ donation. If the exposed person is breast-feeding, discontinuation of breast-feeding should be considered, especially for high-risk exposures. Modifying an exposed health care worker’s responsibilities for patient care to prevent transmission to patients is not necessary.

Counselors should also provide reassurance, review information about the degree of risk present, and inform the worker about procedures to protect the confidentiality of the exposure medical records. As noted earlier, the most important message to communicate to most workers is that occupational HIV transmission is very unlikely; 99.7% of exposures do not result in HIV infection, even if chemoprophylaxis is not administered. Continued reassurance from a supportive clinician, coupled with practical advice about measures to prevent future exposure, enables the worker to cope successfully with the exposure and its aftermath, although some exposed workers have major difficulty adjusting.¹⁶⁷ Counselors should also be alert to the concerns of sexual partners, co-workers, family, and friends of the exposed worker. Referral for ongoing supportive therapy during the follow-up interval is helpful for the minority of exposed persons who experience difficulty in adjusting to the stress inherent in waiting the 6 months for testing to be complete. For such individuals, testing by using one of the newer fourth-generation platforms, which detect P24 antigen and HIV antibodies, may be beneficial, if available. Finally, adherence to chemoprophylaxis regimens may be enhanced if skilled counselors provide advice to drug recipients. In the San Francisco Post Exposure Prevention Project, the frequency of side effects was similar to those in health care worker studies; however, nearly 80% of participants completed 4 weeks of therapy. The authors ascribed this success largely to the intensive, skilled counseling that enrollees received.^168,169

Exposure History

When an exposure is reported, the first priority is to evaluate the risk for infection and the need for immediate wound care and prophylactic treatment. After these issues have been addressed and the exposed worker is calm enough to engage in a more detailed discussion, the interviewer can elicit additional details about the exposure. Information should be recorded about when, where, and how the exposure occurred, the type of device involved, the presence or absence of safety features (and, if present, their state of activation), and when in the course of handling the device the exposure occurred (e.g., during use, after use, during disposal). Each institution should pool these data and should periodically evaluate them systematically to identify common circumstances or flawed processes that may be modified to reduce these risks and to increase patients’ and workers’ safety.

If exposure to HIV (or hepatitis B or C virus) (see Chapters 305 and 306) has occurred, the risk for transmission for any or each of the pathogens to which the individual is exposed should be assessed. If the source patient is known to be infected with HIV, the interviewer should determine the stage of illness, recent results of viremia testing (if available), and recent antiretroviral treatment history.^8,9,152,165 If the source patient’s HIV status is not known, information relevant to the probability of infection (e.g., presence or absence of risk behaviors), as well as clinical and epidemiologic clues that suggest undiagnosed HIV infection, should be recorded and considered when recommendations for follow-up management are made.

For needle punctures or similar percutaneous injuries, information about the source material and exposure characteristics known to be associated with increased risk for HIV transmission (deep injury, visibly bloody device, device used in an artery or vein) should be carefully collected and recorded.^8,9,22 In addition, the practitioner should inquire about factors likely to increase the transmission risk (e.g., injection of a volume of blood, exposure to hollow-bore needle, exposure to large-gauge needle).

For mucosal exposure, the body fluid or material involved and the exposed site, volume of material, and duration of contact before decontamination should be recorded. In addition to these data, for reported skin contacts, the condition of the skin at the site of contact should be evaluated to detect lesions that could provide a portal of entry and influence the risk for infection. For intact skin exposures without an obvious portal of entry, infection is so unlikely that further evaluation and treatment are not necessary unless the contact is prolonged or involves a large area of intact skin. Even then, the risk for HIV infection is extremely small.¹³

Human bites rarely transmit HIV (see later discussion).^170–173 The person who inflicted the bite may sustain a mucosal blood exposure to HIV, but only if the skin was penetrated, the bite wound bled, or both. The person who is bitten is usually not at risk for HIV infection unless blood or visibly bloody saliva was in direct contact with the bite wound. Penetrating bite wounds do pose a risk for bacterial wound infection, and appropriate wound care and antibacterial prophylaxis should be provided, when indicated (see Chapter 320).