Christof Koch's article on “How to Make a Consciousness Meter” discusses a technique purporting to deduce whether severely brain-injured patients are conscious by sending magnetic pulses to the brain while measuring its electrical activity and then creating a mathematical measure of the response called the perturbational complexity index (PCI).

This article is problematic on several levels. Koch interprets the PCI data in one study as showing that some unresponsive brain-damaged patients have something akin to consciousness because they have a PCI value above a threshold that the researchers had established for conscious subjects. But that range includes the state of REM sleep, which he defines as “Unresponsive, Conscious” in the “Zapping and Zipping” box. If consciousness is “the quality or state of being aware,” as per Merriam-Webster, then to equate any neural activity with consciousness would require evidence that the activity was a response to a sensory input. Only if data emerge demonstrating that these patients have a prognosis that is less dismal than for other persistently unresponsive patients should the PCI be used as a possibly hopeful sign for physicians and families.

Further, Koch blatantly misstates the nature of Terri Schiavo's death: he calls it “medically induced,” when, in fact, Schiavo was kept alive for years with medical intervention and died when it was stopped and her body was allowed to follow the natural course of severe brain damage. Her death was delayed by medical treatment, not caused by it.

Sutter Independent Physicians, Sacramento

Koch's article focuses on the cutoff for consciousness, but what about the spread of the PCI? Might the value slide as one is ravaged by Alzheimer's disease? Does it correlate with age or IQ in healthy individuals? Baby steps come first, but I am excited at the prospect that the PCI might elucidate the connection between consciousness and the physical structure of the brain.

Regents Professor of Physics, Washington State University


“Health by the Numbers?” by Claudia Wallis [The Science of Health], argues that individuals should have personalized goals for levels of, say, glucose and cholesterol rather than one-size-fits-all targets, which certainly makes sense to me.

The article says that for decades doctors have told patients who are prediabetic or who have diabetes to aim for a blood level of hemoglobin A1C below 7 percent. Actually, while many doctors do try to control A1C below that level, not all of them follow that criterion. Rather if the A1C is found to be above the prediabetes threshold of 5.7 percent, they start prescribing medication, often metformin. I know friends who have been put on metformin with levels of around 6.1 percent.

I wonder how many patients are being overtreated by adherence to the 5.7 percent criterion. I do understand the need to be proactive in preventing diabetes, and there may be other medical factors that might call for tight control. As pointed out in the article, however, there are possible negative consequences of overtreatment (comprehensively studied in a 2016 paper by René Rodríguez-Gutiérrez and Victor M. Montori, both at the Mayo Clinic).


Wallis references a published risk-estimation tool for heart attacks and stroke that factors in cholesterol levels. My online searches have produced other references to the tool but no links. Is it available to the public?

ROBERT GRANE via e-mail

WALLIS REPLIES: Regarding Holtzman's letter: Although doctors will sometimes prescribe metformin for patients with prediabetic A1C levels (between 5.7 and 6.4), a report by Eva Tseng of Johns Hopkins University and her colleagues published last year in Diabetes Care indicates that fewer than 1 percent of people with prediabetes take metformin. Typically the first-line treatment for such patients is exercise, weight loss and modification of diet. That said, metformin is believed to be safe and effective for both prediabetes and type 2 diabetes and does not cause hypoglycemia or other risks associated with some diabetes drugs.

In answer to Grane: You can find a risk estimator from the American College of Cardiology online at!/calculate/estimate


Thanks to Carl L. Hart for his insightful article “People Are Not Dying Because of Opioids” [Forum], which suggests reasoned public health approaches to address addiction and overdose that won't harm patients in need.

My husband and I are physicians. We are also disabled because of painful conditions that are expected to worsen until death. Like so many others, thanks to misinformation about the “opioid epidemic” and misguided efforts to do something about it, we have lost access to care more than once. Patients are dying from complications of untreated pain and committing suicide. Where you live determines whether you will be overprescribed or underprescribed pain medicine, but many physicians are now refusing to prescribe at all. Hart's article was a breath of fresh air.

NAME WITHHELD via e-mail


“The Zoomable Universe,” an excerpt of a book by Caleb Scharf and Ron Miller, invites simultaneous consideration of the age and size of the knowable universe. It estimates the diameter of the universe to be 93 billion light-years. If the speed of light is a limiting factor and the universe is approximately 13.8 billion years old, how can the radius of the universe be more than 13.8 billion light-years and thus the diameter be greater than approximately 27.6 billion light-years?

Sarasota, Fla.

THE EDITORS REPLY: Current estimates place the radius of the observable universe as just more than 45 billion light-years, yielding a diameter slightly in excess of 90 billion light-years. The reason this value is larger than 27.6 billion is because of the universe's expansion over the course of its 13.8-billion-year existence. While it's true that nothing can travel faster than light through space, the expansion of space itself can exceed this cosmic speed limit.


“The Radical Groundwater Storage Test,” by Erica Gies, proposes several methods of recharging underground aquifers in California, including the controlled flooding of farm fields adjacent to rivers and streams. Such flooding may help refill the aquifers, but it can be troublesome. When a government official persuaded my family to flood our ranch in Santa Barbara County sometime in the 1950s, it resulted in severe erosion, and we never did it again. A more efficient method is to pump the water down wells drilled into the aquifer that needs to be recharged. These could even be the same wells that are used to pump water out for use during the growing season. And injecting directly into the aquifer mitigates the risk of transporting pesticides and fertilizers and may result in a more rapid fill rate.

JOSEPH A. RUSSELL via e-mail