DALLAS -- Nine-week-old Kyle Gilbert had been inconsolable for hours when his weary parents decided late on the night of Aug. 6, 1989, to take him to the emergency room at Humana Hospital-Medical City Dallas.

The infant, otherwise normal and healthy, had suffered for several weeks with stomach problems that had been diagnosed as "irritable bowel syndrome," a disruption in the normal flow of food through the intestines. For most of the day on Aug. 6, "he was screaming," said his mother, Karen Gilbert. However, added his father, James, Kyle's condition was hardly critical. "I thought I'd be bringing him home (from the hospital) that night," he said.

The Gilberts would instead leave the hospital the next day with an empty infant carrier. Their son was dead.

For several weeks before Kyle's death, Alan Strickland, a Dallas pediatric gastroenterologist to whom the Gilberts had been referred, had tried orthodox treatments on the infant with little success. On July 23, he opted for a different approach -- one that would become the subject of a malpractice suit, as well as a complaint pending before the Texas State Board of Medical Examiners.

Attorneys on both sides agree that Strickland's treatment of Kyle on July 23 and again in August was unusual. He administered internally -- through an enema and a nose-to-stomach tube -- an iodine-containing antiseptic meant to be used topically.

But as Strickland's attorneys noted during a hearing before the medical examiners board last March, unusual or unapproved uses of drugs and other substances are common practice in pediatrics.

Fewer than half of the drugs and substances regularly used in pediatrics have been approved for use on children, according to Michael Shaffer, a spokesman for the Food and Drug Administration. Some doctors say the number is closer to one fourth.

Because of the enormous expense involved in bringing new drugs to market -- an average of 12 years at a cost of $231 million per drug -- they are not routinely tested or certified for children, Shaffer said.

"Virtually all medications are tested only on adults, so it becomes incumbent on the physician to extrapolate from the data when they are treating younger patients," Shaffer said. "If you follow the labeling, you have the FDA to back you up. If you don't, you're on your own."

Pediatricians often are in uncharted territory when they prescribe drugs for children, officials say.

In some cases, they must select drugs intended for adults and estimate the proper dosage for children -- sometimes with the assistance of studies or articles in medical journals. In other cases, drugs intended to be used in one manner on adults will be applied in a different way for children.

"The pediatrician is placed in a tremendously difficult position," said Ralph Kauffman of Detroit, chairman of the Committee on Drugs at the American Academy of Pediatrics. "If he is restricted to drugs approved by the FDA, children are deprived of important drugs. If he prescribes outside the labeling, he must be prepared to defend it."

"We still, after all these years, don't know for sure what the proper dosage is in children for something as generic as morphine," said Wayne Snodgrass, a member of the committee and a professor of pharmacology, toxicology and pediatrics at the University of Texas Medical Branch in Galveston. "The bottom line is we just do not have access to the same sort of data that exists for adults."

Jeff Trewhitt, a spokesman for the Pharmaceutical Manufacturers Association in Washington, said manufacturers agree that "more . . . labeling information could help pediatricians when they write their prescriptions." The cost of testing for children is "a factor, and no one is denying that," he said.

But a greater consideration, in Trewhitt's view, is "How does one overcome the clear ethical ramifications of widespread testing of possibly toxic experimental pharmaceuticals in infants and children? There are very serious concerns and challenges that need to be resolved."

Federal officials say they are unaware of increases in the number of children who die because of drug or substance misapplication.

Inspired partly by an article he had read by three doctors in Iran, Strickland sought to sterilize Kyle's colon with Betadine, a topical antiseptic that contains a small amount of iodine. Betadine is the dark yellow substance often smeared on the skin of people about to undergo surgery.

In testimony before the State Board of Medical Examiners, Strickland said he prepared a solution composed of one part Betadine and nine parts GoLYTELY, or Colyte, a saline solution that contains polyethylene glycol.

Kyle was given 10 cc of Betadine in the first administration of the substance, on July 23, and his condition improved slightly. When the infant's symptoms recurred, Strickland administered a second dose of the solution, which contained 25 cc of Betadine, on Aug. 6.

About 3 1/2 hours after the last of the solution was given to him through a nasogastric tube, Kyle was found dead in his crib at the hospital by his father. The Gilberts, who said they were not advised by Strickland about how he planned to use the Betadine solution, initially were told that their son had died of sudden infant death syndrome (SIDS).

Later, when an autopsy was completed by the Dallas County medical examiner's office, they would find that Kyle died from "electrolyte and fluid imbalance" probably caused by "iodine toxicity." Although attorneys for Strickland vigorously disputed the autopsy conclusions during the hearing before the medical examiners board, Charles Petty, the county's chief medical examiner, said his office has not been swayed from its assessment.

"Let me put it this way: We feel the death was directly related to the introduction of the iodine-containing material," said Petty.

Melvin D. Morgan, one of the Gilberts' attorneys and a medical doctor, said a toxicology report found an inorganic serum iodine level in the boy's blood that was 11,000 times the normal level.

Petty declined to comment on the iodine level, citing the lawsuit. But he did acknowledge that the level was "awfully high."

The Gilberts, who have a 3-year-old daughter, said they have yet to come to terms with their son's death. "If it was SIDS, I knew it was God's doing, and I could accept that," said Karen Gilbert, 32. "But to know a doctor took him . . . I'm still shocked."

Neither Strickland, 41, his attorney nor attorneys for Humana Hospital-Medical City Dallas would discuss the case, citing the complaint and malpractice suit.

The complaint against Strickland before the board of medical examiners was brought by the Dallas County medical examiner's office. "We simply considered the case significant enough to be brought to their attention," Petty said.

During the hearing last March in Austin, one of the witnesses called by Strickland's attorneys said children would be deprived of many new treatments if pediatricians followed FDA rules or drug manufacturers' guidelines.

"The drug companies have been very reluctant to put pediatric doses (in the Physicians' Desk Reference) and have always said in there, 'Don't use in children,' " said Janice Ophoven, medical director of the Department of Quality Management at St. Paul's Hospital in Minnesota. "There isn't a subspecialty in medicine that hasn't been . . . breaking that law for years. We would be depriving children of all kinds of access if we followed those rules."

In explaining how he came to use Betadine the way he did, Strickland said he had seen it administered in "multiple ways" for kidney and stomach problems and had seen "ear, nose and throat surgeons pouring it into the mouth to sterilize the mouth prior to operations in the mouth area."

"I was confident that it (Betadine) was something that could be used . . . to sterilize the colon," he testified at the board hearing.

Last May, a hearing examiner for the board recommended that all charges against Strickland be dropped. "Every single expert testified that they did not use Betadine in this fashion, nor did they know of other physicians who did," wrote hearing examiner Guy F. Gebbia. "This testimony is quite troubling, as is his (Strickland's) argument that physicians need to be 'innovative,' especially pediatricians, since there are few medications that are FDA-approved for children."

Those concerns notwithstanding, Strickland "administered the Betadine in a cautious, scientific manner," Gebbia concluded.

The board is scheduled to consider the recommendation late this month.