By R Carter
How does the CDC collect data on prescription opiate overdose deaths? With the CDC’s hyper focus on prescription opiates as the leading cause of opiate abuse and overdose deaths, it’s a good question. After all, the CDC is a science based organization so wouldn’t there be data to back up its claims? And it only follows they use scientific methods in collecting, evaluating and reporting, right? In the next series of posts I will take a critical look at how the CDC collects, compiles and interprets data when reporting on prescription opiate overdoses.
|Death Certificate Data|
The CDC’s data on opiate overdoses comes primarily from death certificates prepared by physicians or coroner reports prepared by medical examiners. Each state in the US collects these reports as part of keeping vital statistics. The CDC in turn purchases this data from the States and makes it publicly accessible through a database known as CDC Wonder. From this database the CDC performs research which is the basis for policy recommendations to federal and state agencies as well as a source for the private sector.
As you review the specific steps on how data is collected, coded, classified and interpreted, I believe you will see what I am seeing. The CDC’s claim about prescription opiates is at best a wild guess, built on assumptions, not science, from a system which in reality is incapable of collecting specific prescription opiate overdose data. And at best, can only make poorly qualified assumptions on the role prescription opiates play in opiate overdose deaths. I will identify key elements in this process which show how the data is hijacked from its original purpose and used to build an illegitimate case for prescription opiate overdose deaths.
|Death Certificate Layout|
Although states can offer their own variations, most comply with a standardized form called “U.S. STANDARD CERTIFICATE OF DEATH 2003” issued by the CDC. As most would agree, the most important information on a death certificate is the cause and manner of death. A death certificate is the only legal proof that someone has died. The State uses it to stop social security payments, pensions and other benefits. Families use it to settle their affairs and insurance companies use it for determining and awarding death benefits. The CDC uses it to compile statistics on the leading causes of death, organized by age, gender, ethnicity, education, time and place. So one could say, the death certificate is likely one of the most important documents in our society.
For a full review of how the standard death certificate is organized, revised and implemented for use by all fifty states, follow this link to CDC -> National Center for Health Statistic (NCHS) -> National Vital Statistics System and review the section titled “Specifications for Collecting and Editing the U.S. Standard Certificates and Reports”.
The World Health Organization has prepared an online tutorials for physicians and medical examiners, on how to code a death certificate with the cause of death. The CDC publishes a handbook and guide as well. The reader can follow these links to see the two areas I will focus on in this series, Box 32 Part 1 and Part 2.
In Part I, the certifier must describe the immediate cause of death — the conditions and sequence of events that led up to the death. In Part II, they must also list the underlying causes of death over time. This can be something that happened in the hours before a person died, or be a condition revealed in the medical history from several years ago. And because there’s often confusion and ambiguity, certifiers can use words like “probable” or “presumed” to qualify their decisions but they must always fill out the underlying causes. There are two sections on each death certificate where the underlying cause of death is identified,
|Who fills out the form and signs it?|
According to the National Association of Medical Examiners 20 percent of all death certificates are filled out and signed by a coroner or a medical examiner. Autopsy rates have plummeted in hospitals resulting in death investigators performing the majority of the nation’s autopsies. Yet autopsy’s remain a vital barometer for revealing causes of death which might otherwise have been missed. In 2007 the CDC reported 201,000 autopsies performed, accounting for just 8 percent of all deaths. So who records the remaining 92 percent of all deaths? This requirement falls to the primary care provider or attending physician at the time of death, a key factor which has significant implications, leaving room for mistakes, misrepresentations and falsification on the cause of death. Many health professionals commonly complain that if they see “cardiac arrest” written as the cause of death, it’s often a catchall, meaning the physician couldn’t determine the cause of death or maybe, just didn’t think the cause was worth mentioning.
It should also be noted that along with 80% of death certificates coming from a PCP or an attending physician, there is also no requirement to perform any lab tests to support the cause of death. So it’s completely conceivable that when a cause of death is listed as opiate poisoning, the physician coding the form has no means of making a distinction between illegal opiates verses prescribed opiates. This fact immediately raises serious concerns about the reliability of 92% of the data collected by States and the CDC.
Typically, deaths have to be recorded with local health departments within 72 hours of the death, and to the state within five to seven days. This constraint further limits the chances an attending physician will draw blood and send it off to a lab for testing to identify specific substances which were the cause of death. And a Medical Examiner only gets involved in the event of a death associated with a crime or some unnatural or mysterious cause.
|How Does the CDC get the Data?|
Once information is recorded by a clerk at the State Vital Statistics office, it is purchased by the National Center for Health Statistics (NCHS) — the division of the CDC responsible for compiling mortality data in the CDC Wonder database. The information is then used by government agencies as well as the private sector, to direct funding and future policy.
The NCHS has been collecting mortality data since 1979, but in 2007 Dr. Robert Anderson, head of the division, said states are often slow to deliver. And sometimes the agency lacked the funds to pay for all the information a state could provide. Budget cuts in 1995 forced the CDC to stop collecting data on the number of autopsies being performed. Although the process began again 2003, it left an eight-year gap in vital records. This fact alone automatically disqualifies any reporting on opiate overdose deaths up through 2003. On the flip side, it could also explain why CDC was blind to the growing opiate death rate until somewhere around 2010 as well as why the CDC reaction now, seems so overboard.
|Theft of Prescription Opiates|
Another area which confuses the subject of prescription opiate overdose deaths is the theft of prescription opiates which make their way to the black market. Several large scale thefts have been reported over the years, some from storage facilities such as Ely Lilly, others from distribution warehouses for leading retail pharmacies, such as Walgreens. These thefts cloud efforts to make an accurate assessment on the amount of prescription opiates which are diverted from patients who have received a prescription from a physician. So, when the CDC makes claims on the number of overdose deaths attributable to prescription opiates, exactly how are they accounting for these large scale thefts? More importantly, are black market opiates in any way connected to patients who are legally prescribed opiates and should they factor into setting caps for prescribers? How the CDC come up with numbers to justify capping prescribed opiates for legitimate medical conditions seems a bit of a reach when you consider how opiate overdose deaths are accounted for and considering the large scale theft of prescription medications. Without an accurate reporting system on both deaths and diversion, its really anyone’s guess.
|What are the facts?|
- Only 8% of death certificates are backed by an autopsy, 20% from a Medical Examiner.
- ME’s are not required to perform lab analysis on blood or tissues samples to empirically document toxicological causes of death.
- Assuming toxicological tests are performed, no records exists on whether those test are immunoassay, which cannot specifically identify a prescribed opiate, or gas chromatography.
- 80% of all death certificates are filled out by a family medicine, primary care provider, or attending physician.
- More than 92% of the time, cause of death is only as good as the information provided to the physician certifying the death or the time and effort invested by the certifier.
- Death certificates must be sent to State Vital Statistics within 72 hours of death, leaving no time for lab tests to determine cause of death.
- Data in CDC Wonder is only as good as the data purchased by the CDC. What records we do have indicate the CDC samples less than 50% of all states in the U.S.
- Of the known States where data is purchased by the CDC, those states have a statistically higher percentage of opiate overdose deaths as reported by individual states. Raising further questions on the accuracy of data reported for the entire U.S.
- The lack of accurate reporting on diversion of prescribed opiates.
Other References: Frontline – PBS