CAN PATIENT confidentiality and medical research coexist?

The answer could lie in Health and Human Services Secretary Tommy G. Thompson's proposed changes to the rigid privacy rules announced by Bill Clinton as he was leaving office.

Privacy zealots had convinced Mr. Clinton that highly restrictive rules were needed to prevent abuses of patients' medical records. But the draft regulations went too far. They could prove difficult to implement, seriously retard life-saving medical research and impede a patient's access to care.

Here's one example:

For the past 20 years, I've conducted research on preventing or diminishing epileptic seizures in children. One of the treatments my team has used is a surgical approach, removing the diseased half of a child's brain.

This sounds horrendous, but most of these children resume a normal life, with normal intelligence, even attending college. But roughly 10 percent continue to have some seizures.

Recently, my colleagues and I decided to do a "look-back study," reviewing records to see if we could find factors that interfered with total success. We needed to analyze brain scans done after the operation to see if we had missed some epileptic tissue. Perhaps we could learn to do the operation better.

All of the scans involved children who were my patients. Their parents had signed informed-consent documents allowing the children to have the operation and the results to be used in a research report of the outcomes. I didn't need to examine the children again. I just wanted to look at their records and their brain scans stored in my institution's records archive.

Nonetheless, I had to apply to the Institutional Review Board (IRB) for permission, since this was a new research study. I expected this would be routine. But much to my surprise, the IRB said I must contact the parents of every child and obtain signed permission just so I could look at these old records.

Why? The IRB was anticipating implementation next year of the original Health Insurance Portability and Accountability Act (HIPAA).

The privacy standards set out in the act were so restrictive I could not use my own patients' records for a follow-up study to advance good care of children.

This situation requires more common sense and less extremism.

Clinical research studies are carefully designed to answer specific questions while ensuring a favorable risk-to-benefit ratio for the patient. Informed consent from patients or their surrogates is mandatory.

Yet sometimes as research progresses, we stumble on unexpected questions that can only be answered by looking at the records once again. Under strict interpretation of the new health privacy regulations, we'd be barred from conducting such a routine follow-up study. We'd have to track down every single parent -- if we could find them -- and get written consent.

It's true that the revolution in information technology has opened the door to potential invasions of privacy. Pharmacists have sold the names of patients obtaining certain drugs to firms marketing other medications for similar illnesses. There are clear, and not unreasonable, concerns that insurers and health care providers might have access to prior medical records and use them to discriminate against groups of patients.

But quick access to a patient's electronic medical records could mean proper treatment during an accident or might alert an emergency room physician to potential drug interactions or drug allergies.

Privacy and progress in health care can coexist.

The Bush administration has taken steps to correct the worst provisions of HIPAA. One of the common-sense changes proposed by Mr. Thompson recognizes that important research sometimes requires reviewing and analyzing patients' records. The proposed new rule simplifies obtaining a waiver of consent for records research. Yet the administration is under attack by privacy purists who want no change in the rules.

Health information should not be used for commercial or discriminatory purposes. But over-zealous rules, rigidly designed to protect privacy at all cost, will only harm us all. Let reason -- not fear -- prevail.

John M. Freeman, M.D., is Lederer Professor of Pediatrics and Neurology and director of the Pediatric Epilepsy Center at the Johns Hopkins University School of Medicine.