Physician-to-Patient HIV-1 Transmission

abstract & commentary

Source: Lot F, et al. Ann Intern Med 1999;130:1-6.

Reasonable evidence of possible transmission of hiv from medical personnel to patients has been limited to one well-publicized cluster of cases involving a dentist in Florida. Following the recognition of HIV infection in a 53-year-old orthopedic surgeon practicing in a Parisian suburb who presented with HIV encephalopathy in 1994, Lot and colleagues from France began attempts to contact 3004 former patients. It was believed that the surgeon had likely been infected while performing surgery 12 years earlier, and his CD4+ cell count on presentation was 46/mm3. A total of 983 patients responded to letters requesting information regarding their HIV status or were referred for HIV testing.

Attempts to contact nonresponders were made by obtaining current addresses through the hospital registry or the French National Health Insurance Registry. In addition, with the agreement of the French Committee on Information and Privacy, all untested patients were "matched" to the National AIDS case registry.

Only one (0.1%) person with HIV infection was identified—a 67-year-old woman who had undergone a difficult 10-hour hip replacement in 1992 whose retroviral infection was recognized during a preoperative screen in 1994. Phylogenetic analysis revealed that the env sequences of the HIV from the patient and surgeon were similar and probably belonged to a unique and as-yet unidentified HIV-1 subtype.

On in-depth questioning, the surgeon reported frequent cutaneous blood exposures and almost weekly percutaneous injuries. His orthopedic surgical colleagues reported a similar frequency of percutaneous injuries, especially while suturing in a blind cavity or while placing metal wires or pins. Few, if any, of these exposures were reported. This remarkably high frequency of high-risk exposures was common to orthopedic surgeons and was not shared by other surgical subspecialties.

Comment by Carol A. Kemper, MD

Despite having performed thousands of surgical procedures, many of them high risk, during a period of time when he was probably viremic, only 0.1% of this surgeon’s patients may have acquired HIV. It’s reassuring to note that no other cases of HIV as the result of transmission from health care provider to patient have been documented to date. Nevertheless, these data suggest that physicians in specialties with frequent blood exposures should be aware of their HIV status and should take extra precautions to prevent transmission whenever possible. Improved reporting of percutaneous blood exposures should be encouraged.

A possibly more disturbing aspect of this story was the use of the French National Health Insurance system to track patients, especially those who had changed address or phone numbers or those who did not respond to the first direct mailing. Furthermore, the French National Committee on Information and Privacy gave Lot et al permission to examine the mandated National AIDS case registry in an attempt to match AIDS cases with the surgeon’s list of patients who had not voluntarily reported for testing.

Not a single additional patient was identified through these measures. The ethical and scientific merits for the use of these patient databases in this instance should be carefully examined as we, in the United States, contemplate similar (supposedly, confidential) systems. One could argue that patients would want to know their potential risk and that the availability of successful therapeutic intervention provides a rationale for early case identification. However, based on the lack of documentation of any similar cases of HIV-1 transmission, despite thousands of surgeries and invasive procedures having been performed since the beginning of the HIV epidemic, the justification for the invasiveness of this investigation does not seem warranted. On the other hand, when is the sacrifice of civil liberties justified? How many lives is it worth? And who decides? I ask myself if I would have seen this article in a different light if, for example, 30 patients had been discovered, none of whom knew they were HIV infected.