By MaryJo Webster and Jessica Glenza

Digital First Media

Elected officials, law-enforcement officers and others proclaim there's a heroin "epidemic" sweeping the country, and it's taking hold in rural and suburban communities once considered unlikely places to find illicit drugs.

But nobody knows how many people have died.

Nobody knows how many have overdosed and survived.

Nobody even knows for certain where the problem is most severe.

The Centers for Disease Control and Prevention reported that 3,036 people died in 2010 from heroin overdoses, but due to problems with how death investigations are conducted and how those deaths are documented, the CDC estimates that its tally is at least 25 percent short, possibly more.

"I'm very scared for our nation in how fast this has grown and spread," said John Roberts, a retired Chicago police officer who created The HERO Foundation after his son died of a heroin overdose in 2010. "This is an epidemic. But it's not getting the attention that it needs because we don't know how bad it is."

Public-health researchers, called epidemiologists, say long-standing flaws in data about fatal and non-fatal overdoses compound the already difficult task of monitoring an illicit activity.

Having an accurate and timely picture of heroin abuse, and its effects, is crucial, researchers said, because heroin has much higher risks of overdose and infectious disease than other drugs, particularly compared to the prescription painkillers that many of today's heroin users started with.

"We need to know what they're dying from, because our ability to measure drug use is very poor," said Caleb Banta-Green, a research scientist in Seattle and former adviser to the White House Office of National Drug Control Policy. "Death is what you want to prevent, and that data can tell you something about how we keep people alive."

To make matters worse, it has taken three years for the CDC to make the 2011 death data available to the public and researchers -- a full year longer than usual -- at a time when local reports from throughout the country indicate alarming increases in heroin use and overdoses.

The 2011 data is expected to be released later this week, and researchers expect it will show a dramatic spike in the number of heroin-related deaths.

"The rapidly changing picture of substance abuse in the country now demands in the 21st century a more timely reporting of indicators," said Jim Hall, an epidemiologist with the Center for Applied Research on Substance Abuse and Health Disparities in Miami.

Hall and others said more complete, detailed and timely data would be helpful for informing the policy makers who decide how to fight drug abuse, and heroin in particular.

"Policy should rest upon good data. The better your data, the better policies you make in the first place and the better evaluations of the policies are," said Len Paulozzi, an epidemiologist with the CDC's National Center for Injury Prevention and Control. "If you have to wait four years to get data, you're well behind in terms of trying to arrive at an effective policy."

Deaths investigated differently in different states and counties

The CDC can say with confidence that more than 200,000 people died from drug overdoses between 2005 and 2010.

During that same time, its data shows that about 15,000 of those deaths involved heroin and about 86,000 involved other opioids, such as prescription painkillers.

But those numbers -- especially for heroin -- are most certainly an undercount.

The CDC thinks these figures are short because of a study by Margaret Warner, an epidemiologist at the CDC, and other researchers that found one-quarter of the death certificates for overdose victims between 2008 and 2010 did not specify the drug or drugs involved.

The study also found that the accounting varies by state. In Louisiana, 65 percent of the investigations of overdose deaths didn't specify what drug was involved; less than 1 percent failed to mention the drug in overdose deaths in Vermont, New Hampshire and West Virginia.

It's this discrepancy that makes it impossible for researchers to say there are more heroin overdose deaths in one part of the country than another. 

The differences stem from two things: an inconsistent death investigation system and the fact that it's difficult to detect heroin on toxicology tests.

Some medical examiner or coroner offices might not have the budget to run and interpret toxicology tests on every autopsy.

If the tests are not done or they are inconclusive, the death certificate might simply say the person died from a "narcotic" overdose, or, least definitive of all, "undetermined."

Even with a toxicology test, it's possible an investigator may not feel confident enough to say the person died from a heroin overdose. Heroin metabolizes in the body very quickly, and toxicology tests typically show evidence of morphine, a different opiate.

Falling behind

The bigger problem is that much of the data is slow to get to researchers, policymakers and others interested in assessing the problem at a national level. This has been exacerbated in recent years due to computer problems at the CDC and the shutdown of a surveillance system that measured non-fatal overdoses.

"(Timeliness) is a challenge, because we won't know for a couple years whether the efforts we're engaged in right now to fight prescription opioids are having an effect on overdoses," said Rafael Lemaitre, spokesman for the White House Office of National Drug Control Policy.

The death certificate data was delayed by more than a year due to problems with a new computer system at the CDC. Normally, these numbers are released two years after deaths occur, but the 2011 data isn't getting to researchers and the public until this month.

The CDC launched a new computer system and found flaws early last year, according to Bob Anderson, chief of the mortality statistics branch. The new computer system, Anderson said, will give his agency more ability to monitor health trends in near real-time.

"Within the next five years, things will be a lot different," Anderson said. "We'll be doing a lot more surveillance work using the mortality data ... We'll be able to say things about the previous year early in the new year."

Roberts, who lost his son to heroin, would especially like to see better accounting of non-fatal overdoses, and he's been asking Illinois lawmakers to require hospitals to report those numbers to the state.

With increasing use by paramedics and police officers of naloxone, an antidote to an opiate overdose, Roberts expects the number of non-fatal overdoses treated in emergency rooms would likely be "staggering."

"People are starting to embrace the idea that it's a health crisis," Roberts said. "How do we turn anecdotal evidence into real evidence? Let's do real-time reporting from a reliable source -- our hospitals.

"Let's just find out how bad it is," Roberts added. "And then try to solve it."