When I was a newborn — a preemie struggling to survive in a hospital nursery’s incubator — an article deep inside The Washington Post saved me from becoming blind.

The article — on Page A22 — discussed research showing that too much oxygen in an incubator could cause babies to lose their sight. When my worried parents phoned the hospital, they were told doctors had also seen the piece and promptly adjusted the incubator’s air mixture. What none of them knew was that the sight in my right eye had already been destroyed by what is now called retinopathy of prematurity, or ROP. Fortunately, the vision in my left eye remained intact, saving me from a lifetime in the dark.

That was way back in 1953. Yet just a few months ago, a federal judge dismissed a lawsuit involving premature babies enrolled in a study of what incubator oxygen level was best. The infants’ parents said they hadn’t been fully informed of all the risks to their infants. I was stunned. In 2015, how can the oxygen level in incubators still be endangering babies?

The answer involves twists and turns in the tale of a disease blamed at one time for an epidemic of blindness. It also illuminates larger questions about medical research and how in medicine what seems right and sure at one time can be wrong, very wrong, later.

In the 1940s, the loss of 30,000 infants a year made infant mortality a leading cause of death. One out of five preterm babies died, a 1947 Post article reported. When doctors noticed that healthy babies had a more regular breathing pattern than struggling preemies did, they would often enrich the incubator air with more oxygen. At around this time, a physician in Boston identified the first case of retrolental fibroplasia, or RLF, an abnormal buildup of blood vessels that can scar the retina and irreversibly destroy eyesight.

Like thousands of other premature babies born in the United States, Michael Millenson’s vision was damaged by excess oxygen in his incubator. (Courtesy of Michael Millenson)
An epidemic amid the boom

By the time the post-war baby boom was in full swing, public health officials were warning of a “blindness epidemic” among preterm infants. (An RLF victim in Saginaw, Mich., would go on to achieve fame as Stevie Wonder.) Still, when a Washington ophthalmologist named Arnall Patz asked the National Institutes of Health in 1950 to fund a clinical trial examining whether too much oxygen might be the culprit, he received a three-sentence dismissal. It curtly began: “There is nothing scientific about added oxygen as a possible cause of RLF.”

Patz and two colleagues proceeded with their study anyway, funded by an anonymous local donor. The results were startling. Of 28 small babies getting a high concentration of oxygen, seven ended up with RLF, the men wrote in a 1952 American Journal of Ophthalmology article. In the lower-dose group of 37 newborns, not a single one was harmed.

In September 1953, I was born seven weeks premature at George Washington University Hospital, just a few miles from where Patz had done his study. On Oct. 23, 1953, a Post article on Patz’s research informed readers of “circumstantial, but impressive evidence” about “the blindness which strikes the infants after several weeks of incubator life.” As my parents instantly understood, I was that kind of high-risk infant.

Pediatrician William Silverman, who chronicled the history of RLF in a book and in journal articles, estimated that by 1953 about 7,000 U.S. babies had been blinded by oxygen therapy and what he called “well-meaning guesswork.” Countless others of us retained some vision due only to the sheer luck of one eye’s being more resistant to oxygen toxicity.

A September 1954 presentation of the results of a larger trial dispelled all doubts about oxygen’s role in blinding babies. Moreover, reducing oxygen did not seem to jeopardize the infants’ survival. The end of the disease seemed at hand.

Unfortunately, eliminating this disease wasn’t quite so simple.The earliest ROP researchers hadn’t investigated oxygen’s impact, and their successors hadn’t looked closely enough at infant mortality. In the 1960s, researchers began discovering that reducing oxygen could also be dangerous. A 1973 article in the Lancet concluded that for every case of blindness prevented by oxygen restriction, 16 infants died.

‘Depths of ignorance’

Decades passed with no firmer guidance about how to balance the preterm baby’s survival against the risks of blindness. Along the way, the clinical criteria changed and the disease name was changed from RLF to ROP. Still, seeming breakthroughs in ROP research — Fluorescent lights caused blindness! Vitamin E prevented it! — turned into blind alleys.

In a 2002 editorial, “50 Years of Uncertainty,” the journal Pediatrics bemoaned “the depth of our ignorance” concerning “how much oxygen these infants really need.”

In early 2005 came another promise of certainty. A study was launched to discover both the optimum incubator levels of oxygen for newborns and the best way to deliver it. Led by the University of Alabama at Birmingham, it was known as the Surfactant, Positive Pressure and Oxygenation Randomized Trial, or SUPPORT.

At the time, the standard of care for small preemies was 85 to 95 percent oxygen saturation. The SUPPORT trial randomly placed 1,300 newborns into two groups. One received 85 to 89 percent oxygen, the other 91 to 95 percent oxygen.

“Balancing the benefits of supplemental oxygen against the risks . . . has been a concern of doctors and parents for decades,” said an official at one of the NIH institutes funding the work.

In 2010, researchers reported in the New England Journal of Medicine that significantly more babies in the lower-oxygen group died, but among those who survived, fewer developed ROP than did their higher-oxygen counterparts. One researcher not involved in the study later told the New York Times that her hospital had immediately enriched its incubator oxygen flow.

A consent problem

However, any good feelings about the SUPPORT project abruptly stopped in early 2013. The Department of Health and Human Services’ Office of Human Research Protections sent a letter to the researchers saying their consent forms hadn’t made adequately clear the possible risks to babies and asking that the forms be changed. Shortly after the letter was made public, however, five parents filed a class-action suit accusing the researchers of inadequately disclosing the danger of brain injury, blindness or death.

Bioethicists and editorial writers came down on both sides of the controversy. In August, a federal judge in Alabama dismissed the suit, citing a lack of evidence that infants in the trial had suffered actual harm or were at greater risk than those not in the trial.

Although the court case is over, the problem of how best to treat tiny preemies persists. Today’s newborns are born earlier and emerge smaller and more vulnerable than ever: Those in the SUPPORT study weighed, on average, less than two pounds. What has become clear is that in ROP, as with so many other diseases, no simple answer explains everything.

“It’s not just an oxygen story,” says Graham Quinn, a pediatric ophthalmologist at Children’s Hospital of Pennsylvania, who has spent a career investigating ROP.

Since the 1980s, there has been enormous progress in early diagnosis and treatment of ROP. Special instruments let doctors examine a newborn’s eyes, and cryotherapy and lasers can prune blood vessels before they do irreparable harm. Meanwhile, investigation continues into genetic factors that may attenuate ROP’s toll.

Despite these advances, researchers estimate that 20,000 babies around the world go blind or become severely impaired from ROP annually. In the United States, 500 to 600 babies go blind every year, Quinn estimates. In the early 1950s, by contrast, 100 blind babies were being sent into the school system of just one state (New Jersey) every year.

Neonatologists, Quinn says, “walk a tightrope” between protecting brain function and preserving vision. Because much of ROP can now be treated, he adds, “I’d want to preserve brain function.”

Me, too. By dint of good fortune or good genes, my left eye is normal. Nor did I suffer brain damage or any of the other disabilities that freighted the lives of so many others. Most of all, I survived. Lowering the oxygen level saved my sight without sacrificing my life.

Of course, I wish I had good vision in both eyes. But age and a more careful reading of the ROP story have given me the wisdom to be deeply grateful for what I have been given.

Millenson is a health-care consultant and writer in Highland Park, Ill.