Between 1998 and 2004, some 17 percent of HIV-positive women observed in a U.S. survey—the Women’s Interagency HIV Study (WIHS)—were taking HIV med regimens deemed ineffective or dangerous by federal treatment guidelines. Some of the combos included meds from only one or two HIV-drug classes; others combined meds known to interact badly. An author of the Journal of AIDS study report, Jennifer Cocohoba, PharmD, of the University of California San Francisco, calls the regimens “contraindicated” (they shouldn’t be used) and the findings “shocking,” though she warned that the 217-woman survey hinged on interviews, not prescription records.

Because the federal guidelines for HIV treatment are periodically adjusted to reflect new research, each woman’s first combo was rated based on guidelines for the year she began meds. Doctors weren’t interviewed about why they’d prescribed the combos, but one possible explanation emerged: The 37 women on contraindicated regimens had lower starting viral loads than the other women, suggesting that their doctors might have thought only partial regimens were required. That proved wrong: The women in the bad-combo group had slower, lower CD4 gains and higher viral loads than the others—even after switching to a better combo. (An additional 64 women were on regimens that were not recommended, but not deemed dangerous, by the guidelines.)

Fewer women took contraindicated drugs after 2001, says Cocohoba, “but it’s a reminder that women seek HIV specialists less often than men.” Whatever the reason for those bad regimens, the study results serve notice on all positive people—and their doctors—to triple-check that combo.