Has CDPHE Really “Monitored” People Near Rocky Flats?

Cancer-registry analysis is not the same as long-term health monitoring of a potentially exposed population

For decades, people who lived near Rocky Flats have asked whether releases from the former nuclear weapons plant may have affected their health.

In communications about Rocky Flats, the Colorado Department of Public Health and Environment has said:

“CDPHE has monitored cancer incidence rates near Rocky Flats since the 1980’s and has not found elevated rates of cancer among the residents living near the area.”

At first glance, that statement may sound reassuring. It may also leave readers with the impression that Colorado has been monitoring the health of people who lived near Rocky Flats since the 1980s. That is not what the cancer-registry work has done.

CDPHE has used data from the Colorado Central Cancer Registry to conduct analyses of cancer incidence among geographic populations around Rocky Flats. Cancer registries are important public-health tools, and those analyses provide useful information. But analyzing cancer-registry data is fundamentally different from identifying a potentially exposed population and monitoring or conducting surveillance of that population's health over time.

That distinction should be clear whenever CDPHE describes what it has learned about cancer and Rocky Flats.

What does a cancer registry do?

The Colorado Central Cancer Registry collects information about cancers diagnosed among Colorado residents. Among other purposes, registry data can be used to examine whether particular cancers are occurring more or less frequently in one geographic population than another. For Rocky Flats, CDPHE has periodically analyzed cancer-registry data for geographic areas surrounding the former plant. That tells us something important: What cancers were diagnosed among people residing within the areas being studied during the periods included in those analyses?

But it does not necessarily tell us: What happened to the health of the people who lived near Rocky Flats during the years when they may have been exposed? Those are different research questions.

What would health monitoring or surveillance look like?

If the goal were to monitor the long-term health of a population potentially exposed to contaminants from Rocky Flats, the starting point would be the people and their exposure histories, not simply their address when cancer was diagnosed. Such an effort could identify people who lived, attended school or worked in potentially affected communities during relevant periods of Rocky Flats operations and then examine their health outcomes over time.

Depending on the design, that could include information such as residential history, years lived near Rocky Flats, age during the period of potential exposure, movement away from the area, latency between potential exposure and disease, possible exposure pathways and relevant health outcomes.

That is much closer to what an ordinary reader might understand when hearing that the health of people near Rocky Flats has been “monitored since the 1980s.” It is not what a cancer registry alone does.

The difference becomes especially important when people move

Cancer can develop many years—or decades—after an environmental exposure. People also move. Consider someone who grew up in Arvada during the 1960s and 1970s, moved away in 1990 and was diagnosed with cancer somewhere else years later. A geographic analysis based on where that person lived when diagnosed may not associate that cancer with the person's earlier residence near Rocky Flats. Now consider the reverse. Someone could move into the study area years after Rocky Flats ceased plutonium production and subsequently be diagnosed with cancer. Depending upon the study period and methodology, that person's cancer could be associated with the geographic population surrounding Rocky Flats.

Those two individuals have very different histories with respect to possible Rocky Flats exposure.

Cancer-registry geography does not, by itself, reconstruct those histories. That limitation becomes particularly important when trying to understand the long-term health of a mobile population following a potential environmental exposure.

Monitoring the area is not the same as monitoring the people

This is the distinction Rocky Flats Downwinders believes needs to be made much more clearly. CDPHE has analyzed cancer incidence in geographic areas around Rocky Flats.

That should not be confused with saying that Colorado has monitored or conducted long-term health surveillance of Rocky Flats Downwinders since the 1980s. The first statement describes cancer-registry work. The second suggests that an identified population with a history of potential Rocky Flats exposure has been followed over time. Those are not equivalent.

Why the wording matters

Words such as “monitored” and “surveillance” carry meaning. When the public hears that cancer rates near Rocky Flats have been monitored since the 1980s, people may reasonably assume that researchers have been following the health of people who actually lived near Rocky Flats during the plant's operating years. That impression matters because it can make the evidence sound more comprehensive than it is.

Rocky Flats Downwinders is not arguing that the cancer registry has no value. Nor are we arguing that cancer-registry data should be disregarded. We are asking that the work be described accurately.

A clearer statement would be:

CDPHE has periodically analyzed Colorado Central Cancer Registry data to compare cancer incidence among geographic populations around Rocky Flats with other populations. These analyses have important limitations for evaluating the long-term health of people based on their historical residence or potential exposure during Rocky Flats operations.

That description tells the public both what was done and what it was not designed to do.

What can the cancer-registry studies tell us?

They can help determine whether certain cancers occurred at statistically elevated rates among the populations and geographic areas included in a particular analysis during a particular period. That is useful information. But a finding that an analysis did not detect elevated cancer incidence does not automatically establish that there was no increased risk among people historically exposed to contaminants from Rocky Flats. That would require answering a different question about a differently defined population.

The distinction is particularly important when interpreting phrases such as “no elevated cancer rates near Rocky Flats.” As we explained in our earlier article, What Does “No Elevated Cancer Rates” Near Rocky Flats Actually Mean?, the meaning of that statement depends heavily on who was studied, where they lived, when cancers were diagnosed and what the underlying data can actually measure.

The unanswered question

The question Rocky Flats Downwinders continues to ask is straightforward: Has Colorado identified people who lived, worked or attended school in potentially affected communities during Rocky Flats' operating years and systematically monitored their health over time—including after they moved away?

The cancer-registry analyses we have reviewed are not the same thing. That does not tell us what a properly designed long-term health study would find. It tells us that the study has not been done simply because cancer-registry analyses have been conducted. Rocky Flats operated for decades. Many of the people who lived around it during those years have since moved elsewhere. Diseases such as cancer can have long latency periods. If Colorado wants to understand the long-term health of that population, geography at the time of diagnosis cannot answer every question.

We have to follow the people—not simply the map.