Certifying Death: Health Department Seeks Less Cryptic Causes

Texas Medicine Logo(1)

Public Health Feature -- September 2000  

By Johanna Franke
Associate Editor

You begin another workday only to receive a call that a patient's life has ended. Her son notifies you that she passed away in her sleep at her home during the night. No one was there to witness her death.

You know the elderly woman suffered from several conditions, including diabetes and advanced Alzheimer's disease. But none of these sticks out as the immediate cause of death, a section you must complete on her death certificate.

Could you be more specific?  

Cases like these prompted Galveston internist Michael O'Malley, MD, to write a resolution for the Texas Medical Association Young Physician Section to present to the TMA House of Delegates at its 1999 interim session last November.

The resolution, which passed at TexMed 2000 in May, asked that TMA, through the Council on Public Health, request the Texas Department of Health (TDH) to permit the diagnosis of "unknown, but not suspicious" or "natural causes" as the primary cause of death on a death certificate. The physician then could be more speculative in the "Secondary To" or "As A Likely Consequence Of" lines on the certificate.

"This happens to a lot of physicians," Dr O'Malley said. "When they first get out of residency, they're sticklers about things -- they're very scientific and by-the-book. They don't want to sign a certificate saying a person died from congestive heart failure, or myocardial infarction, or stroke when they don't know. So they have a tendency to defer to the medical examiner a lot of cases the older docs just sign because they don't want to hold up a funeral."

But TDH doesn't even allow justices of the peace/coroners, medical examiners, and funeral directors to use "natural causes" as the primary cause of death, and private physicians are held to that same reporting standard. This is because "natural causes" doesn't mean much when collecting vital statistics for the state and the nation.

"When a patient shows no indication of cause of death, and the justice of the peace and the police have investigated and found nothing suspicious, we will accept 'natural causes,'" said Steve Elkins, field services program administrator for the TDH Bureau of Vital Statistics.

Those who perform autopsies and list "natural causes" as the immediate cause of death receive a letter from TDH asking for more information. TDH sends query letters on about 7% of all death certificates (nearly 10,000 in 1998), but not because of the overuse of "natural causes." More often, a physician receives a letter requesting more details like the type of cancer causing the patient's death, Mr Elkins says.

"If the physician writes back that he or she absolutely cannot narrow it down any more than 'natural causes,' then our coders will code it that way," Mr Elkins said. "But it just doesn't give us much information."

Definite death data  

The cause of death is the most important piece of public health information on the death certificate, Mr Elkins says.

Statisticians gather information from this section to assess the general health of the population, examine medical problems that may be found among specific groups of people, and indicate areas in which medical research could have the greatest impact on reducing mortality.

"Legislators look at these data to determine budgets and policy, and the health department allocates funding and programs based on those statistics," Mr Elkins added. "Information gathered on fetal deaths along the Texas-Mexico border helped TDH determine that pregnant women in those areas need to take folic acid to prevent neural tube defects and other problems in the womb."

But being more specific about cause of death doesn't mean a physician should slap down just anything other than "natural causes" or "unknown, but not suspicious," says Bryan emergency medicine specialist Robert Emmick, Jr, MD, chair of TMA's Council on Public Health.

"You can't use 'cardiopulmonary arrest.' That's not a reason of death," Dr Emmick said. "Everybody dies because their heart stops. Why did the heart stop?"

However, listing cardiac arrhythmia in the "Immediate Cause of Death" section of the death certificate, followed by myocardial infarction in the "Due To (Or As A Likely Consequence Of)" section, and hypertensive cardiovascular disease in the secondary "Due To (Or As A Likely Consequence Of)" section is acceptable, TDH officials say.

Dr O'Malley believes the cop-out of "heart attack" might have skewed mortality statistics and sent funding toward prevention of the wrong diseases.

"The No. 1 cause of death in the United States is coronary artery disease or myocardial infarction. And because of that, when people don't know why somebody dies, they put 'heart attack,'" he said. "And because they put 'heart attack,' it gets ranked as the No. 1 cause of death, and because of that, people put 'heart attack.' It's a self-perpetuating diagnosis."

For all we know, Dr O'Malley says, the No. 1 cause of death in the country and the state truly may be "unknown" or "natural causes." But if TDH allowed listing of these causes, physicians might get sloppy and put "unknown" on everything, he says.

"So Texas doesn't want to permit this, unilaterally, because it will mess up funding from the CDC [Centers for Disease Control and Prevention]," Dr O'Malley said. "But do we really want to secure federal funding with data we know are bad?"

Qualifying causes  

Texas health officials aren't asking anything extra of their private physicians, justices of the peace/coroners, medical examiners, and funeral directors. "The practice of completing a death certificate is no different in Texas than in the rest of the United States," Dr Emmick said.

During his training at the National Center for Health Statistics (NCHS), Mr Elkins learned that Texas has a unique mix of medically untrained justices of the peace/coroners and medical examiners, who are always physicians. The large metropolitan areas generally have medical examiners, while rural areas, which sometimes lack a physician of any kind, use justices of the peace to investigate deaths and order autopsies as needed. But one thing all the states have in common is the need for education on completing death certificates, Mr Elkins says.

The national cause-of-death determination and death registration procedure is based on World Health Organization (WHO) guidelines. On the federal level, the NCHS collects death certificate information in a format that enables state-to-state and country-to-country comparisons. The primary classification system for coding causes of death is based on the latest revision of the International Classification of Diseases (currently ICD-10), which follows WHO guidelines and recommendations.

Medical personnel rarely are trained specifically in writing accurate and appropriate cause-of-death statements, and few physicians complete death certificates often enough to be thoroughly familiar with the basic principles of writing them. But physicians are considered the experts on their patients and are ultimately responsible for the accuracy of entries into the national health database, Mr Elkins says.

"We get the best information from the family physician or the specialist who treated the deceased for the cause of death," Mr Elkins said. "A lot of times, we get death certificates completed by the ER physician, who knows nothing about the history of the patient. We're looking for family physicians, whether or not they were seeing these patients at the time of death. The ER physician needs to track down the family physician or specialist to complete the death certificate."

But even if a primary care physician or a specialist cannot be located, the attending physician is the expert, Mr Elkins says.

"No matter how uncertain they are about the cause of death, physicians are the people in the best position to give us guesses," he said. "They're not liable in any way, as it says on the certificate 'to the best of my knowledge' and 'in my opinion, death occurred in the cause and manner as stated.'"

Of course, knowingly falsifying a statement on a death certificate can earn a physician 2 to 10 years in prison and a fine of up to $10,000, according to the Texas Health and Safety Code.

Physicians have 5 days to complete the cause of death statement and forward the certificate to the funeral director or registrar. To improve accuracy and promptness in reporting deaths, most states allow death certifiers to use the qualifiers "presumed" and "possible" to describe the disease or injury that initiated the events resulting in death.

TDH accepts both the unqualified cause-of-death statement, which implies that it has been made with a high degree of certainty, as well as the qualified cause-of-death statement, for which the certifier felt it necessary to note a specific condition or a nonspecific process as being "presumed" or "probable." Physicians need not wait until the "secondary to" or "as a likely consequence of" sections of the death certificate to get more speculative, Mr Elkins says.

"The words 'presumed' and 'probable' are going to limit the meaning of the immediate cause of death," Mr Elkins said. "They're going to give us a general indication of the category -- heart disease or cancer, for example -- that may have led to death. 'Unknown' or 'natural causes' gives us no indication whatsoever."

So if a physician cannot cite a specific condition as the underlying cause of death, the cause must consist of a qualified specific condition (eg, probable myocardial infarction) or a qualified nonspecific process (eg, end-stage liver disease; congenital biliary atresia).

These general indications eventually will help statisticians produce data for the leading causes of death in Texas. In 1998, the top 10 causes of death in the Lone Star State were heart disease, malignant neoplasms, cerebrovascular diseases, accidents and adverse effects, chronic obstructive pulmonary diseases, diabetes mellitus, pneumonia and influenza, suicide, chronic liver disease and cirrhosis, and septicemia.

The next step  

The TMA resolution also asked TDH to revise the examples listed on the back sheet of the death certificate to include a sample of "probable" or "presumed." TDH will consider this addition in revising its death certificate based on a proposed US-standard death certificate, Mr Elkins says. The new death certificate is set to debut January 1, 2003.

In addition, TMA Council on Public Health members requested that TDH provide continuing medical education (CME) options, through the Internet and hard copy, for physicians and others on the accurate completion of death certificates.

TDH officials have created an online CME course available at www.tdh.state.tx.us/phpep. They also are working on a handbook on the medical certification of the cause of death, Mr Elkins says.

Dr Emmick and TMA public health staff continue to meet with Mr Elkins and other TDH officials to follow up on changes in the death certificate and make sure the state's immediate cause-of-death requirements ask no more of Texas physicians than the CDC's.

"These are CDC requirements, and the state's ability to receive funding from the federal level depends on complying with these requirements," Dr Emmick said. "So it's in our best interest to report deaths as accurately as we can."


Death certificate CME

Still unclear on how to complete a death certificate? Then visit www.tdh.state.tx.us/phpep.

The Texas Department of Health has created an online tutorial worth 3 hours of American Medical Association Physician's Recognition Award Category 1 continuing medical education (CME) credit. The tutorial also contains 1 hour of ethics CME.

Anyone can view the site for free, but those who complete the course and pay $30 online by credit card will receive CME credit along with a certificate for the CME ready for immediate printing.


Who should certify the cause of death?

  1. Are external causes involved in the death?  



Answer/ Decision  

Answer/ Decision  


1. Was trauma involved at any point in the sequence leading to the death?

If "No," proceed to the next question.

If "Yes," refer to a medical examiner or justice of the peace.

Burns, falls, coma, gunshot wounds, exsanguination, car accidents

2. Was the death a result of a person's action inflicted on another person?

If "No," proceed to the next question.

If "Yes," refer to a medical examiner or justice of the peace.

A passenger in a car wreck, a baby who dies from inadequate nutrition, a person shot by another person

3. Was the death a result of a person's action to him- or herself?

If "No," proceed to the next question.

If "Yes," refer to a medical examiner or justice of the peace.

Overdosing on drugs, driving drunk, playing Russian roulette

If the answer to all the above questions is "No," a physician should proceed to certify death.

  1. Do you know enough about the death to certify the cause?  




Answer/ decision  


1 . Do you know why this person died?

If "Yes," proceed to the next question.

If "No," refer to a medical examiner or justice of the peace.

Unknown cause of death, eg, apparent or suspected SIDS

2. Do you know how the sequence of events leading to the death began?

If "Yes," proceed to the next question.

If "No," refer to a medical examiner or justice of the peace.

Person was under your care for a long time, or you were treating the person for the disease or process that led to the death

3. Do you know if this death was expected?

If "Yes," proceed to the next question.

If "No," refer to a medical examiner or justice of the peace.

Person was under your care for a condition or disease that is usually fatal

Source: Texas Department of Health Public Health Professional Education Program  



The lost art of autopsy

Determining cause of death may be more difficult today because of the decreased number of autopsies that are performed.

As always, a physician's reluctance to ask family members' permission to perform an autopsy on a loved one has circumvented some autopsies. And the myths among the lay public -- such as sensational stories about people claiming their recently deceased relative's organs were taken and sold for profit -- also contribute to the lack of autopsies. But in recent years, increased technology and lack of reimbursement have curtailed the number of autopsies.

Hospital autopsy rates have fallen dramatically over the past 40 years, according to the College of American Pathologists (CAP). In the 1950s, autopsies were performed in approximately 40% to 50% of hospital deaths. In the late 1990s, average autopsy rates at non-teaching hospitals were thought to be below 10%. This means there are hospitals where an autopsy is an extremely rare occurrence or is not provided at all -- and there are pathologists who rarely practice their autopsy skills.

"While pathology residents in the 1950s performed around 2,000 autopsies during their 4-year training periods, by the late 1980s, most pathology residents performed 250 autopsies during a 4-year training period," said Austin pathologist Randy Eckert, MD. A pathology resident finishing training today probably has done between 75 and 100 autopsies, says Dr Eckert, chair of the CAP delegation to the American Medical Association.

"A lot of clinicians will tell you that the diagnostic capabilities we have currently -- CT scans, MRIs, and others -- have improved their ability to detect disease premortem, so autopsies are not needed," Dr Eckert said.

But an article in the October 1998 issue of The Journal of the American Medical Association titled "Autopsy Diagnoses of Malignant Neoplasms: How Often Are Clinical Diagnoses Incorrect?" begs to differ.

A 10-year retrospective study (1986-1995) of all autopsies performed at the Medical Center of Louisiana in New Orleans showed a 44% difference between clinical and autopsy diagnoses of malignant neoplasms and confirmed the importance of postmortem examination, the authors say.

Lack of reimbursement for autopsies also has led to fewer pathologists performing the procedures.

"Pathologists are unwilling to embark on such intensive studies for little, and in many cases, no reimbursement whatsoever from any third-party payers, hospitals, and patients' families," Dr Eckert said.

A thorough autopsy -- which can cost up to $2,500 -- takes about 2 hours at the table and another few hours to review a patient's entire clinical record, perform microscopic examinations, and compile a final report.

"A good 10 to 12 hours of physician time is necessary in some cases to do a good autopsy," Dr Eckert added.

Traditionally, autopsies have been reimbursed poorly by Medicare, Dr Eckert says. "When you submit an autopsy bill for a Medicare recipient who has expired, you get one of two responses: 'The cost has been covered in the DRG [diagnostic-related group] payment made to the hospital, so the hospital must pay for the autopsy,' or 'The patient is dead, therefore, he is no longer a Medicare beneficiary.'"

This topic popped up at the AMA annual meeting in June, as the Pennsylvania delegation asked AMA to "study the restoration of a Joint Commission on Accreditation of Healthcare Organizations requirement of a medically productive and realistic autopsy rate, premised on the provision of adequate designated reimbursement by health insurers for requesting and performing autopsy services."

Physicians aren't the only ones who would reap the rewards of higher autopsy rates, says Bryan emergency medicine specialist Robert Emmick, Jr, MD, chair of the Texas Medical Association Council on Public Health. "By completing death certificates correctly and obtaining autopsies when appropriate, we're creating more information that would better serve our patients."R 


Death by chocolate?

By Richard Viken, MD  

It has been said that death is the most definitive measure of ill health. Morbidity and mortality statistics derived from death certificates are the only continuously collected, population-based, disease-related information available in most parts of the world, including the United States. Therefore, in order to accumulate meaningful data, accurate and concise terminology should be used to describe, in a logical order, a sequence of events leading to death.

This sounds pretty dogmatic coming from a freewheeling primary care physician, who admits to being all over the map when trying to place a precise address on a death certificate. But I think that I have reoriented my approach after researching this subject quite extensively over the last several weeks.

When I was in training, only 25 years after introduction of the current death certificate, my mentors consistently reminded me that people die from only two things: cardiorespiratory arrest and cerebral anoxia. Now I find out that you can die from other things -- such as congestive heart failure, liver failure, selected arrhythmias, and exsanguination -- and that these are considered mechanisms, not causes, of death. That's why the death certificate has specific instructions (which I have always ignored) not to use these terms. My gosh! After all these years, my patients have been dying for the wrong reasons! I have repented, and on my latest death certificate, I wrote down "subarachnoid hemorrhage" rather than "cerebral anoxia" as the immediate cause of death. That will make someone happy, particularly if his or her research funding is linked to diagnostic coding frequency.

So why has filling out the death certificate correctly suddenly become a higher priority for me? Well, first of all, I happened to read the black box warning written sideways in the left margin of the certificate. It's easy to miss. It allows you the opportunity to contribute your time and money to the Texas Department of Corrections if you knowingly falsify any information.

Second, I do support the notion that the primary care physician should be the one most responsible for completing the death certificate, as well as for interacting with the family and/or the medical examiner. The issue that bothers many physicians is how to proceed in cases for which you truly don't know the cause, ie, the unwitnessed nursing home death. My advice: Make an educated guess, based upon your personal knowledge or chart knowledge of the patient's medical condition! I firmly believe that it's a cop-out to write "unknown" or "natural causes" as advocated by some.

Let's say that I am summoned to pronounce Old Mother Hubbard. I could miss an ice pick hole at the base of her skull; or I might not acknowledge the subtle findings of multiple keratoses, puffy eyelids, and alopecia, characteristic of slow arsenic poisoning. But the odds are pretty favorable that her death will be due to one or more of the 5 to 10 morbid conditions that nearly all old people have, eg, dementia, diabetes, ischemic heart disease, hypertension, chronic obstructive pulmonary disease, or stroke.

Finally, I am intrigued by the possibility that death certification may be one of the better applications of the current stampede toward computerized medicine. What could be simpler than entering a list of the patient's known disease states -- past, present, and presumptive -- and letting the computer work out "Immediate Cause," "Underlying Cause," and "Other Significant Conditions Contributing …"? Of course, the attending physician could tweak the information as appropriate before writing and signing the final form.

National and local efforts, including this issue of Texas Medicine , are under way to provide renewed education in death certificate completion. Software-based and online methods are also in production. And after writing this, I plan to prepare an interactive presentation to give to my family practice residents -- black box warning and all.

Suggested reading  

  • Huffman GB. Death certificates: why it matters how your patient died. Am Fam Physician . 1997;56:1287-1290.
  • Kircher T, Anderson RE. Cause of death: proper completion of the death certificate. JAMA . 1987;258:349-352.
  • Mackenbach JP, Kunst AE, Lautenbach H, Oei YB, et al. Competing causes of death: a certificate study. J Clin Epidemiol . 1997;50:1069-1077.
  • Magrane BP, Gilliland MGF, King DE. Certification of death by family physicians. Am Fam Physician . 1997;56:1433-1438.
  • Maudsley G, Williams EM. "Inaccuracy" in death certification -- where are we now? J Public Health Med . 1996;18:59-66.
  • Messite J, Stellman SD. Accuracy of death certificate completion. JAMA . 1996;275:794-796.

Richard Viken, MD, is a professor of family medicine at The University of Texas Health Center at Tyler and a member of the Texas Medicine Editorial Committee.  


September 2000 Texas Medicine Contents
Texas Medicine Back Issues