There’s a well-known problem with studies on humans. While we’re willing to let rats and even monkeys die in order to learn more about the mysteries of life and health, we aren’t willing to increase the risk of death in humans. Good studies even have to be double-blind, which means that even the person providing the new procedure or medicine doesn’t know what which group each patient is in: if they’re giving some people a new medicine and others a placebo, even the doctor doesn’t know which patients are getting medicine and which patients are getting the placebo.

Studies that start killing people will be halted before the study is complete. That’s human nature, and it’s good.

But human nature has its dark side, too, and that dark side tends to show up in government bureaucracies. That’s the tendency to hide people behind paperwork, and the tendency, when their programs run up against reality, to treat reality as defective. And when people start questioning decisions that deviate further and further from reality, to obfuscate and hide their true intentions.

That’s what struck me about the SUPPORT¹ study by the National Institute of Health and the Department of Health and Human Services:

Medical personnel routinely give supplemental oxygen to babies who are born with immature lungs. Too much oxygen can cause severe eye damage, including a blood vessel disease and blindness called retinopathy. Too little oxygen can lead to brain damage and death.

The NIH-funded experiment used the test babies in an attempt to find the sweet spot for preemies yet to be born: the lowest level of oxygen that would preserve vision, yet be sufficient to prevent brain damage and death.

To get the answer, researchers arbitrarily assigned infants to either a high-oxygen or low-oxygen group. Because, researchers say, all oxygen levels fell within the generally accepted range, they argue the babies received the same “standard of care” as babies not in the study. None of the consent forms mentioned a risk of death from the oxygen experiment.

The problem with the standard of care argument is that it was untrue. In real life, babies don’t get a single oxygen level throughout their time on life support. Their oxygen level is varied depending on their response to it. This is because too much oxygen can blind them, and too little can kill them, and every baby is different.

This study not only did not allow varying oxygen levels, it hid the restriction from doctors! The machines were altered to display changes when no changes were occurring. The medical staff thought they were working to save the babies, when in fact their efforts were blocked and the feedback faked so that they wouldn’t know.

It will thus not come as a surprise that the study found that high-oxygen babies ended up with more serious vision disorders, and low-oxygen babies were more likely to die.

Death before discharge occurred more frequently in the lower-oxygen-saturation group (in 19.9% of infants vs. 16.2%; relative risk, 1.27; 95% CI, 1.01 to 1.60; P=0.04), whereas severe retinopathy among survivors occurred less often in this group (8.6% vs. 17.9%; relative risk, 0.52; 95% CI, 0.37 to 0.73; P<0.001).

The study designers not only didn’t treat the babies as human, they didn’t treat reality as real. The results were worthless from a real-world perspective. That is, the study shows the tradeoff between two different static levels of oxygen. It does not show the tradeoff between static levels of oxygen and dynamically-managed levels of oxygen.

Someone, somewhere, had gotten funding to study oxygen levels. As the needs of medical research ran up against the realities of care, the study was modified. At some point, the study morphed into one that had no real-world effects except to put premature babies in danger of blindness and death. And the bureaucracy continued past that point.

At some point, the researchers realized that informing parents of the dangers of the experiment—increase their child’s risk of blindness or death—would mean parents would not agree to be part of it. The only way to get a reasonable sample was to lie to parents about the risks. And the bureaucracy continued past that point.

Government bureaucracies don’t exist to help; they exist to continue to exist. This experiment took place under the NIS/HHS without the authority that the “Affordable” Care Act gave to the government. Imagine what wonderful experiments are possible now.