Pratt & Whitney Aircraft got its message out today, burying evidence of higher cancer rates at a local factory.
“No cancer link found at P&W,” blared the front page of Friday’s New Haven Register.
“Study: Pratt & Whitney Workers Got Brain Cancer At Same Rate As Overall Population,” trumpeted the Hartford Courant. “Study Shows No Cluster At North Haven Plant.”
The headlines echoed Pratt & Whitney’s official statement about the most recent results of a massive brain cancer study, released June 3. “We are reassured,” said the statement from the the giant, East Hartford-based jet engine maker, “that the study does not show an increased rate of brain cancer among our Connecticut employees.”
Reassuring, yes. But not entirely true.
In fact, Phase 2 of the $12 million, eight-years-and-counting study found slightly elevated rates of a deadly brain tumor among people who worked at Pratt’s North Haven plant—where a rash of death ignited concerns a decade ago. Thousands of workers made jet engines in that plant for decades before it closed in 2002.
The gap gets bigger when you compare P&W employees who worked only in North Haven with those who were never assigned there. The biggest increase was among salaried employees in North Haven.
Those two comparisons didn’t make the headlines.
Most, but not all, of the increases were too small to be “statistically significant”—meaning scientists think they likely occurred by chance. But all are important enough to investigate further, say the scientists who are conducting the study.
It was Pratt’s North Haven plant where workers and their families first raised the alarm about a possible brain cancer cluster a decade ago. Just six months apart in 1999, two machinists, John Shea and John Greco—friends who worked next to each other—were diagnosed with a rare and fatal brain cancer called glioblastoma.Their wives, soon to become their widows, started compiling information about other P&W employees with brain cancer.
By 2001, the widows and some union activists had involved the state Department of Public Health. With attention mounting, Pratt & Whitney agreed to hire Gary Marsh, a biostatistician at at the University of Pittsburgh, to investigate.
Marsh and his team examined data on more than 212,000 people who worked at eight of Pratt’s Connecticut facilities between 1976 and 2002. In 2008 they announced their Phase 1 findings: 489 of those employees had brain cancer; glioblastoma, the most aggressive form, accounted for 275 of the cases.
Those numbers are big, but so is Pratt & Whitney, which used to be Connecticut’s largest private employer. The brain cancer rates did not exceed the state’s overall average.
In Phase 2, Marsh and his team took a closer look at types of brain cancer and subgroups of P&W employees. Again, the overall rate of glioblastoma caused no concern. But some of the finer-grained analysis yielded possible problems:
* North Haven employees had 8 percent more glioblastoma than the Connecticut average.
* P&W employees who worked only in North Haven had 40 percent more glioblastoma than those who never worked at that plant.
* The rate among North Haven’s salaried employees was double the state average.
Which Numbers Count?
Pratt workers had less glioblastoma than the general Connecticut population. The company on Thursday and the newspapers on Friday emphasized the comparisons between Pratt workers and the general population.
But that might not be the most relevant comparison. Epidemiologists warn of a “healthy worker effect.” As a group, people who work—in any occupation—are healthier than the general population, which includes people who are too sick to work. That can make it misleading to compare disease rates.
According to Marsh, only the twofold increase among salaried employees is big enough to be considered statistically significant. A two-fold increase doesn’t necessarily mean the cancers were work-related; it just raises suspicion.
“The increases we’re seeing are very small,” Marsh said Thursday night at a public meeting, where he presented the Phase 2 findings. (They’ll be published will be published online in the journal Neuroepidemiology on June 7.) “But you don’t pooh-pooh them.”
Marsh’s co-investigator, Nurtan Esmen of the University of Illinois at Chicago, wasn’t so quick to dismiss the 40 percent glioblastoma gap between Pratt’s “North Haven only” and “never North Haven” workers.
“I don’t think we’re ready to make a judgment,” he said, explaining that an overall cancer rate that’s normal or near-normal can mask significant increases among small subgroups.
The study’s third phase will dig deeper into those subgroups. Esmen’s team is painstakingly reconstructing work conditions—who made what parts, using what processes and what toxic chemicals and metals. Phase 3, which is supposed to be finished early next year, will match that reconstruction with data about individual brain cancer victims, looking for patterns and possible causes.
One of those possible causes, Esmen said, is the “blue haze” of the now-closed North Haven plant. Created by metal-working fluids, coolants that (usually) kept the high-speed grinding machines from catching fire, the haze always hung in the air, those who worked there say.
The cancer victims’ families are hoping that the study’s third and final phase will penetrate that haze and—at long last—provide some answers.
An overwhelming proportion of our elementary school children in the New Haven school district (and many in other surrounding towns) are overweight, come from families with dire food insecurity, and/or have early onset diabetes, meaning they will likely lose 20-30 years of life compared to an average American... and we are investing exactly how much money worrying about a few (most likely statistically insignificant) cases of brain cancer?