THE DEATH of 18-year-old Jesse Gelsinger during a gene therapy experiment in September was shocking to those with high hopes for gene-based medical treatment. But at least they heard about it. In at least six other cases, according to a troubling story by Post reporters Deborah Nelson and Rick Weiss, researchers supervising clinical trials of new gene therapies failed to inform the National Institutes of Health when patients died during the trials. In three more cases -- one death and two serious illnesses -- researchers made the required reports but only after asking the NIH for confidentiality.

Does this mean gene treatments are uniquely dangerous? Probably not, but that's what these trials are there to find out. Gene treatments typically involve injecting modified genes into a patient in the hopes that they will propagate and replace defective ones. The companies doing the research argue that they need only report deaths when they are caused by the trial -- a connection that wasn't clear in any of the cases -- and that that determination is one they can make themselves.

Regulators at the NIH dispute that interpretation and plan to make more explicit the rules that govern such research. The Food and Drug Administration, which also has jurisdiction, backed up the NIH's view Friday with a stern letter reminding researchers to report all deaths or "serious adverse events." The notion that companies need only report deaths or other side effects that they know are caused by the treatment is circular. How else are researchers to track the cause of negative symptoms except by comparing them with adverse effects from other studies?

Gene therapy trials fall under both FDA and NIH regulations because of federal funding for genome research and because the nature of the work is thought to call for heightened public scrutiny. But while the FDA, which traditionally supervises drug trials by private companies, keeps those companies' reports on the trials confidential, the NIH process is more public. Until the biotechnology boom, most medical research was done by federally funded scientists accustomed to sharing information as a matter of course -- and without personal or company fortunes riding on the outcome.

The migration of biotech research into a still fluid and speculative private sector has strained that model. The potential for conflict of interest is evident when companies stake their future on the outcome of clinical trials. (The researcher who asked the NIH for confidentiality strongly denied any connection between that request and an upcoming public stock offering, now postponed.) The privatization of cutting-edge gene research has some benefits, notably a huge influx of capital and a boom in start-up companies (particularly in the Maryland suburbs). But that benefit needs to be counterbalanced with redoubled vigilance. Nobody knows very much in practice about this form of treatment, and meddling in the genetic code induces a deserved squeamishness. It is hard to imagine a field where openness to public scrutiny is more urgently required.