An Evidence Based Position Presented By A Former Death Certificate Clerk
“Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark attributed to Disraeli would often apply with justice and force: ‘There are three kinds of lies: lies, damned lies, and statistics.'” – Mark Twain
Synopsis: Although revered as the guiding star for science, clinical practice and legislation aimed to save lives, cause-of-death reporting does not meet any basic criteria of objective fact. Across continents, from 40 years ago to present day, death certificates, which provide the basis for our beliefs as to why we die, have been found to be erroneous in their causal conclusions 20-60% of the time according to the peer-reviewed literature. The daily process of obtaining cause-of-death information, which I was an eyewitness to, is not a process of careful investigation, but rather a rushed and apathetic bureaucratic tumbling machine that incentivizes compliance over recording the complexity of truth.
In this piece I offer a personal account, a logical argument and the scientific evidence for the claim that mortality statistics derived from cause-of-death reporting on death certificates are an unstable material upon which to build actionable scientific or societal beliefs about risk. Then I provide an in-depth examination of the very particular situation of COVID death reporting manipulation that happened beginning in March of 2020, infused politicized bias into an already defunct system. Lastly, you will find a call to action, with steps that we, the individuals affected by the inaccurate data capture, can take to hold the regulatory bodies responsible for this to account, as well as volunteer and support opportunities to help those who need to get erroneous death certificates officially amended.
Being a former death certificate clerk, and having spent nearly 7 years in the funeral home industry ushering thousands of death certificates from digital creation to final registration, I am appalled that death certificate data is codified for use as our national mortality statistics. Aside from some basic demographic tracking of age, place and gender of the deceased, using death certificates for anything beyond closing bank accounts is a disservice to society.
With the rare exception of a medical certifier that has independently chosen to be conscientious and thorough in their certificate completion practices, or the special circumstances of car accidents, overdose, suicides and homicide deaths that lend themselves to robust investigation and reporting protocols, the average natural cause of death reporting on death certificates and the mortality statistics extrapolated from them are not the product of careful investigation, are known to have a 20-60% inaccuracy rate according to the peer-reviewed literature, and are, by definition, variable medical opinions, not facts.
It’s an extremely uncomfortable truth when you look around us at a world enslaved by the daily COVID mortality tallies being reported from every outlet. It’s especially disconcerting if you’ve assumed mortality statistics were somehow exempt from the Twain-ism about statistics being lesser in value to both lies and damn lies. But both the nature and the nurture of cause-of-death data capture flies in the face of any reliability in mortality statistics as structurally sound pillars of objective fact.
However, unlike the entrenched modern-day mores that demand unquestioning homage to those with special knowledge, I will not ask that you believe me simply because of my professional experience. I am here to offer you three considerations to help you develop your own understanding of cause-of-death data capture so as to create an independence in your own pursuit of truth regarding this underlying societal assumption about the infallibility of mortality data. Perhaps you will find, as I have, that mortality statistics tabulated from death certificates have no business steering public health recommendations or medical decisions, and using them as a metric for scientific research or public policy is about as prudent as building a skyscraper on a sand box.
The first harsh reality we need to come to terms with is that even though causes of death provided on death certificates are treated like gavel-dropping legal facts, especially with their prima facie status in a court of law, there’s not actually much scientific investigation happening behind the scenes as to what has caused a death.
The best way to describe the culture I witnessed being the middle-woman in the death recording process for nearly 5,000 death certificates, was not a culture of careful, unbiased scientific investigation but rather a demoralizing, bureaucratic game of hot potato.
The funeral home directors want the record registered ASAP so the family they are serving won’t have their burial or cremation services delayed and the next-of-kin can get their certified copies so as to start settling affairs (close bank accounts, access life insurance, etc.).
The doctor’s office, hospice or hospital decedent affairs staff wanted me (the mortuary representative) to stop calling them with urgent messages about the upcoming burial or cremation service and the need for doctor’s expedient cooperation in the multi-step process for record approval and attestation.
The doctor wants the request for causes of death off his/her desk and doesn’t want to deal with multiple rejections from either the mortuary or the vital records registrars if he/she put causes or contributory factors that don’t fit the narrow allowances under the “natural” manner of death umbrella.
The coroner/medical examiner office doesn’t want to take cases that they don’t absolutely have to, when they are understaffed and already up to their ears in car accident deaths, drug overdoses, suicides and homicide death investigations.
The local vital records registrars don’t want to approve a cause of death that will get flagged by their bosses at the state registrar office after the record has been sent for final registration, causing a whole mess of paperwork to fix the problem.
This bureaucratic tumbling machine results in bland, simple, broad brushstroke causes of death that are an easy ‘pass’ in the electronic system becoming the gold standard in death recording. Any time-intensive investigation is avoided at all costs. The system isn’t built to allow for investigation anyway. In fact, in the state where I worked, doctors are supposed to provide causes of death within 15 hours of the death occurring, and all the multi-step information gathering and verification process between the family, doctor, coroner and state registrar is supposed to be finalized within 7 days after the death.
Towards this end, I was regularly advised by the local registrar’s office to coach the doctors in submitting causes that passed the registrar’s easy filters for natural manners of death, despite the physician’s uncertainty.
The doctor doesn’t know why the person died? Just ask the doctor if the patient was on any medications (insinuating that the cause for a medication prescription, such as hypertension, diabetes, Alzheimer’s, etc. is an easy pass for the cause of death).
Oh, the doctor hasn’t physically seen the patient in over six months? They can still sign the death certificate; just ask them if a refill prescription was sent to the pharmacy for their patient in the past six months, then they are still the “attending” physician.
A 60 year old patient died unexpectedly at home? No autopsy needed, it’ll just be a coroner sign-out case.
A sign-out case, at least here in Los Angeles County, means that the local coroner/medical examiner just needs to stop by the mortuary and take a couple of pictures of the outside of the body to make sure there’s no evidence of physical trauma. Then, the last doctor to order a prescription refill can sign the death certificate with their best guess as to why the patient died, or if the doctor won’t cooperate, the coroner/medical examiner will just slap a catch-all diagnosis like “atherosclerotic heart disease” on the death certificate and call it good.
Everyone involved in death recording gets used to (read:demoralized by) the system, especially for those who died in hospice care or in long-term care facilities. Their causes of death will typically default to the primary diagnosis for which they were put in the nursing home or on hospice in the first place.
Some of the facilities I worked with had a cause-of-death worksheet sent to me minutes after the death occurred because the worksheet had been pre-filled out and was waiting in the patient’s file weeks or months before the person actually died.
For very few deceased, some scientific-ish investigation does occur, although that has dramatically trended down since the 1940s. Postmortem autopsy investigation has dramatically dropped from 20-50% postmortem autopsy rate as late as the 1970s to only 4-8% in our current postmortem protocols.
Because of a shortage in those who specialize in this type of investigation, combined with the requirement that a medical examiner/coroner must be involved in the death recording process for any unnatural or iatrogenic factors impacting the death, you probably shouldn’t expect your loved one’s doctor to be including any medical complications after medication or a medical intervention (such as vaccination) as a cause of death on the death certificate.
In fact, even if your doctor is bold enough to concede that your loved one’s health deteriorated significantly after a medical intervention, the death certificate process would then have to come to a screeching halt.
That’s an unnatural cause of death. Now the case gets bumped to the medical examiner/coroner. But even then, 30% of doctors have reported being instructed by the coroner to put an inaccurate cause of death on purpose so that the medical examiner/coroner office won’t need to take the case. And the metaphorical potato game continues.
However, if the case is accepted by the medical examiner/coroner office, things start getting really messy for the family and the funeral home. The medical examiner/coroner office can be likened to the DMV for death recording. The grieving family is now extremely likely to experience delays in what date the funeral or cremation services can be arranged. When I was a mortuary employee I personally saw situations where the doctor sent causes that required coroner involvement but the services had already been scheduled, and traveling family and friends had already flown in from across the country for the burial. The service schedule needed to be completely rearranged sometimes by up to two weeks out to allow for autopsy and death certificate completion before we could get the permit to bury (or cremate).
On top of that inconvenience, there’s hundreds of dollars in fees from the coroner investigation and post-autopsy body reconstruction services the mortuary must perform if the family had a viewing service in their wishes. Even after the burial, the traffic jam imposed on settling affairs and having closure can last up to a year while the coroner takes the time to determine the manner and cause of death.
What’s the understanding to take away from this behind-the-scenes look at death recording? A thorough picture of what impacted the health of your loved one is de-incentivized in a bureaucratic system, and the carefully investigated truth that ought to guide science research, public policy and medical decision-making for future generations becomes no more reliable than pulling a lever on a slot machine.
But what many don’t realize, and the second of my three offered considerations on this matter, is that the causes of death listed on a death certificate were never designed to be the immovable pillars of science, medicine or law in the first place. As laid out by the CDC, both the physician handbook and medical examiner/coroner’s handbook state that causes of death are a medical opinion, and that these opinions can change from provider to provider.
Let me tell you, they sure did change from provider to provider. When I worked as a death certificate clerk, I occasionally would send death certificate worksheets to multiple doctors involved in a patient’s care if we had a rush to bury or cremate. In these situations we needed to cast a wider net to find a rapidly responding doctor to accomplish the record before final disposition. Many times each physician would send me back a different cause of death. Same patient. Different opinions. Different causes of death.
In general, if someone died in a hospital, the hospitalist would put the acute condition they treated the patient for while leaving out pre-existing chronic conditions. The primary care or hospice physician would put a chronic condition like heart disease, diabetes or hypertension that they prescribed regular meds for, with very little information about the past few weeks or days of health decline. And a specialist would put the specific condition they were managing as the cause of death, such as stage 4 kidney disease and any disease-specific complications that, in their opinion, could explain the demise.
Occasionally there was some consensus on the causes of death between the worksheets sent back from different providers, but thoroughness of the contributory factors or the logical sequence of conditions that led to the decline was almost always lacking or inconsistent in the majority of worksheets received.
These data capture “captains,” who are in charge of supplying us with some of the most valuable data, exercise very little care or consistency in how they fill out these records. Yet their output is blindly guiding scientific assumptions, research funding, public health policy and clinical risk estimation for generations to come.
And I don’t think we can quite blame them. Physicians have received little-to-no education on the importance of death certification and most are unaware that this data is simply repackaged and regurgitated back to them in the news media, scientific literature or public health policy. In medical schools there is not much more than a couple of hours of discussion on death certificate completion, and sometimes the education is as basic as watching this 20 minute slideshow and being quizzed with a handful of questions. Doctors have no thorough or standardized training, and at time of a patient’s death they are not taking enough time to review each patient’s complete medical record and clinical course carefully before completing the causes-of-death worksheet. And even the few who are more thoughtful in the information they provide can still have a varying opinion on what qualifies to be reported as a cause.
Does this culture of data capture really support the weight of science, medicine and public health policy with any confidence? As my third and final consideration for you, let’s take a look at what the peer-reviewed literature shows us as to how this bureaucratic data tumbler spits out.
Here’s an international study of COPD patients, where 42% of clinical trial patients whose death certificates were analyzed by an independent committee did not have COPD listed anywhere on their death certificate. These were patients enrolled in a clinical trial for COPD therapy.
Then, in Norway, 17.6% of investigated death certificates required amendments to change the underlying cause of death.
A study out of Pakistan shows 62% of death certificates have errors that significantly changed the death certificate interpretation.
A Missouri DHSS 2009-2012 study found 45.8% of the underlying cause of death reporting inaccurate.
A blinded study based on reviewing medical records vs. death certificates in Vermont showed 60% as needing a change in the underlying cause of death.
Another Vermont study with a similar methodology found that 34% of hospital death certificates were wrong in the cause or manner of death.
This meta-analysis comparing clinical diagnoses against autopsy findings states: “At least a third of death certificates are likely to be incorrect and 50% of autopsies produce findings unsuspected before death.”
And how about 25% of adults dying within 30 days of being hospitalized with a Clostridium difficile infection in the UK? According to this study, if you were to die soon after being hospitalized for a C. diff infection, there’s only a 17% chance C. diff will be listed as the underlying cause of your death, and only a 31% chance it will be mentioned on your death certificate at all.
And did you know that even though tuberculosis is believed to be the leading infectious disease killer cited by global authorities to be taking 1.5 million lives every year, this South Africa’s study found 63% of decedents who were autopsied after receiving a tuberculosis diagnosis on their death certificate didn’t even test positive for TB by smear or culture. Whichever disease or situation that is killing the people falsely diagnosed with TB is not getting the research funding it deserves.
And the death certificates for infants bring this truth home about the lack of accuracy in causes of death even more:
This study found 48% of infant deaths in Mexico were not reported accurately compared to the patient’s medical chart. And 71% of those inaccurate death certificates had failed to mention an infectious, parasitic, or respiratory disease as either contributory or underlying factor.
This Ohio study of infant death certificates found 56.5% of death certificates were discordant with autopsy findings.
So across the board, reported causes of death are wrong 20-60% of the time. With the exception of a couple of cancer types, studies done on every continent have found an incompetence in death certificate data recording that is so shocking, it’s a wonder it hasn’t taken up enough headlines to actually effect change.
But there was a change made this past year. Not a data capture reform for all the erroneous death diagnoses, and not even a data capture reform to improve reporting for ALL the infections that significantly impact our health before death. The CDC’s National Vital Statistics System (NVSS) rolled out the data capture red carpet for one – and only one – disease-causing pathogen: SARS-CoV-2.
On March 24th, 2020, only 11 days after the first lockdown started, and well before widespread testing was available, the NVSS gave hand-holding guidance to the medical certifiers, local registrars and mortality statistics coders on precisely how they ought to spotlight COVID-19 as the underlying cause of death on death certificates. They boldly declared that COVID should be the underlying cause on a death certificate “more often than not” even without laboratory confirmation of infection. What’s crazier still, is that when they created this COVID alert in March and followed up by releasing this COVID death recording guidance a few days later, we couldn’t have possibly had enough country-specific statistics to justify such a drastic departure in coding COVID deaths compared to how other infectious disease fatalities are ascertained.
So the NVSS actually dictated a belief to the community of death certificate medical certifiers and vital records registrars (who are our cause-of-death approval “gate keepers”), before having any reasonable disease surveillance infrastructure established to support their claim of probability of undiagnosed COVID being the cause of death, thus greatly amplifying the perception of COVID mortality. This may have even been against Federal law on data collection changes, as this peer-reviewed research paper suggests, stating “Federal agencies that make changes to how they collect, publish, and analyze data without alerting the Federal Register and OMB [Office of Management and Budget] as a result, are in violation of federal law.”
Furthermore, their COVID-19 death certifying guidance, changed the death recording rule book completely when it declared: “…reporting “COVID–19” due to “chronic obstructive pulmonary disease” in Part I would be an illogical sequence as COPD cannot cause an infection, although it may increase susceptibility to or exacerbate an infection. In this instance, COVID–19 would be reported in Part I as the UCOD [underlying cause of death] and the COPD in Part II [as the contributory factor].”
The UCOD on a death certificate is what’s reported and tallied in our national mortality statistics as the reason that the death occurred. It is found on the last line of Part 1 on a death certificate. What needs to be provided for a death certificate is a logical sequence of conditions that explain why the death has occurred, not a logical sequence as to why an infection has occurred. So relegating an important chronic condition that logically explains why someone has died of an infection that most people survive is a drastic departure from previous cause-of-death guidance.
Here are four examples given to medical certifiers in the CDC training module and the CDC handbook on proper death certification of cases with infection-related deaths in patients with pre-existing conditions. (UCOD is shown in bold and the infection that has immediately led to death is italicized.) :
From slide 43 of the CDC training module on Improving Cause of Death Reporting:
Cause of Death Reporting Assessment – Answer 3 of 5
The correct sequence of conditions in Question #3 is:
a. Enterobacter aerogenes sepsis
b. Bilateral lower lobe pneumonia due to Enterobacter aerogenes
c. Chronic respiratory failure requiring mechanical ventilation
d. Quadriplegia due to C4 spinal cord injury
From the CDC handbook on death certification:
a. Pseudomonas aeruginosa sepsis
b. Pseudomonas aeruginosa urinary tract infection
c. In-dwelling bladder catheter
d. Left hemiparesis
e. Old cerebrovascular accident
a. Pneumocystis carinii pneumonia
b. Acquired immunodeficiency syndrome
c. HIV infection
a. Escherichia coli meningitis
b. Cystic fibrosis
In all these examples it is the pre-existing condition that made the patient susceptible to death from an infection (i.e., quadriplegia, stroke (cerebrovascular accident), HIV or cystic fibrosis) that is advised by regulatory bodies to be reported as the underlying cause of death (UCOD) which is then subsequently tallied in our mortality statistics as the reason for the death.
But the new COVID-19 guidance advises the exact opposite: medical certifiers are now to report the infection as the UCOD and tally it in our mortality statistics, while simultaneously demoting the previously revered underlying chronic condition (e.g., COPD) into a section of the death certificate that doesn’t impact mortality statistics and holds less sway in science, medicine, public health and law.
Here’s an example from the Hawaii Vital Records website showing how the COVID death certificate is supposed to look:
a. Acute respiratory distress syndrome
As you can see, reporting death in this way will naturally highlight the short term COVID illness resulting in death, instead of reporting the chronic illness like we have done in the past. This is another way how COVID mortality is being artificially amplified over any other infectious cause of death.
Finally, yet another biased standard of boosting COVID mortality specific to this year’s very odd death tallying was PCR testing for SARS-CoV-2 carriage performed after death, including on those whose cause of death was suicide or car accidents and obviously not COVID-related at all Testing for pathogen carriage after accidental death would have never been performed in the past. Similarly, any at-home deaths that used to be chalked up to “atherosclerotic heart disease” without any investigation were now presumed COVID deaths. And nursing home clusters of deaths in the elderly – which, by the way, I used to regularly witness multiple times a year in my capacity as a death recording clerk from 2013-2019 – were now opportunities to swab the dead to contribute to the COVID death toll in 2020, even without evidence of symptoms in the deceased.
As I mentioned previously, deaths that occurred in nursing homes and under hospice care almost always were attributed to the chronic condition that explained their decline in health – regardless of what final infection they suffered from… until now.
This year has provided an undue cause-of-death spotlighting for one pathogen, bolstered by a biased infrastructure of mortality statistics tabulation that has greatly skewed the scientific process of data capture needed to steer medicine, public policy and public perceptions rationally. Without consistent guidance from accurately reported cause-of-death information, science and medicine cannot apply their resources and recommendations wisely to save the highest number of lives. Our rights and freedoms are being lost because public policy and perceptions are being built on a foundation of risk estimation that is so erroneous that it crumbles under even the slightest academic examination. It’s time to have better conversations and create real solutions to the data capture crisis misleading our world. This year has shown us just how horrifically misled we can be by a set of fallacious assumptions.
Families look at the death certificate information of their deceased loved ones to steer their own medical decision-making when it comes to forming their beliefs about genealogical susceptibility to disease and perceptions of risk. Scientific, medical and legislative bodies are influenced by apparent conclusions drawn from the death certificate data and affect the well-being of nations around the globe. Cause of death reporting changes the world on a micro- and macro-scale for better or worse; thus, accuracy matters.
To this end, I’m personally stepping out of my comfort zone, and into the world of grassroots social impact. Many others are concerned about the issue of accuracy in death certification and we are starting a nonprofit to help families, funeral homes and medical certifiers amend death certificates so as to provide an accurate reporting of underlying and contributory health factors that played a role in a patient’s demise.