Off-label drug use in human immunodeficiency virus disease.

We wished to determine the extent to which drugs used to treat HIV disease and its clinical manifestations are prescribed for conditions other than those listed on the U.S. Food and Drug Administration's approved drug label, how such "off-label" use varies by patient characteristics and type of HIV-related medical condition, and the extent to which physicians alter the way they treat HIV-related conditions because of reimbursement problems associated with off-label drug use. We surveyed 1,530 primary care providers for people with HIV disease between February and May 1993. A three-part survey instrument was used to obtain data on the drugs prescribed for the last three patients with HIV disease treated by the provider, the preferred choice of therapy for 32 specific HIV-related conditions, and the extent to which providers faced reimbursement problems regarding the use of drugs for off-label indications. Three drug compendia were used as cited sources of off-label drug uses. In all, 387 (32%) evaluable surveys were returned, yielding data on 1,148 patients. The majority (81%) of patients received at least one drug off-label, and almost half (40%) of all reported drug therapy was off-label. Most off-label drug use was for treatment and prevention of HIV-related opportunistic infections, which frequently represented the community standard of practice (e.g., trimethoprim/sulfamethoxazole for prevention of Pneumocystis carinii pneumonia), or the de facto standard of practice when no licensed therapies were available (e.g., drugs for treatment of Mycobacterium avium complex, MAC). More than 75% of off-label usage was cited in at least one of the three authoritative medical compendia. The use of drugs for off-label indications in HIV care is common and frequently represents community standards of care. Reliance on drug compendia for support of off-label drug use accounts for the majority of such uses, although many legitimate off-label uses may not be included because of compendia publication lag. The prevalence of off-label drug use in routine clinical practice and the development of newer and more costly drugs for treatment of HIV and its medical complications argues for the articulation of an explicit national reimbursement policy for off-label uses of prescription drugs so that medically appropriate therapies will be available to those with insurance in a rational, consistent way.

Prevalence of resistance in patients receiving ganciclovir for serious cytomegalovirus infection.

Seventy-two AIDS patients treated with ganciclovir for cytomegalovirus (CMV) disease were prospectively monitored for the development of drug-resistant virus. No resistant strains were found in 31 patients before therapy or among seven culture-positive patients treated for less than or equal to 3 months. Of 13 culture-positive patients treated for greater than or equal to 3 months, 5 excreted virus resistant (ED50, greater than 12 microM, or ED90, greater than 30 microM) to ganciclovir. Thus, 38% of patients and receiving ganciclovir for greater than 3 months and excreting virus or, overall, 7.6% of the patients were excreting CMV resistant to the drug.

Use of the enzyme-linked immunosorbent assay (ELISA) for detection of Entamoeba histolytica antigen in faecal samples.

A promising new ELISA method developed by Root et al. to diagnose amoebiasis by the detection of Entamoeba histolytica coproantigen was tested in Mexico City and was reported to be more than twice as sensitive as microscopy. In this study the same procedure was compared to microscopy for paired specimens from 107 patients in San Francisco. Microscopy was more than twice as sensitive as ELISA: of 12 specimens positive by microscopy only 5 were also positive by ELISA. The ELISA did not detect a single specimen not detected by microscopy. Over-all, there was a 93% agreement between the two tests: 5% were positive and 89% negative by both tests. Possible reasons for the differences in the San Francisco and Mexican City findings are discussed.