This article compares April’s federal guidance for recording the cause of death that could be due to the COVID-19 virus to assessments of the cause of death among patients infected with a “superbug.” (Article by: Lawrence F Muscarella PhD).
Highlights — This article:
- calls into question the soundness of the CDC’s April 2020 guidance for certifying deaths due to COVID-19
- differentiates between people dying “with” COVID-19 and dying “from” COVID
- suggests the CDC’s April 2020 guidance be revised and account for this differentiation
- questions certifying someone as dying from COVID-19 without having first necessarily confirmed the person was infected with COVID-19 (i.e., using a reliable, validated laboratory or rapid diagnostic test) — a practice that could result in over-counting the actual number of deaths caused by COVID-19
- recommends validation of death certificate data associated with the COVID-19 disease to ensure the data’s accuracy and quality, and to prevent the over-counting (or under-counting too) of actual COVID-19 deaths
- notes that an accurate accounting of deaths due to infection, particularly trends in these data over time, is necessary to identify and implement effective strategies and mitigations that prevent additional infections and deaths
- discusses and compares assessments of the cause of death associated with superbugs, COVID-19 and influenza
- argues for a better understanding of “superbug” infections as the cause of patient deaths
- argues for adoption of policies that reduce, if not eliminate, subjectivity by requiring testing to confirm infection with a respiratory virus
- recognizes that assessments of the cause of death of a patient infected with a superbug, the COVID-19 virus, or another pathogen can be complicated and more challenging in the presence of pre-existing underlying health conditions
August 20, 2020 (updated Sept. 11, 2020) — The Centers for Disease Control and Prevention (CDC) in April published guidance for recording the deaths of suspected or confirmed cases of SARS-CoV-2 infection. This guidance may be as notable for its advice as for its definition and classification of a COVID-19 case.
While advising that “testing for COVID–19 should be conducted whenever possible,” the CDC’s guidance does not require, as a prerequisite, such testing be conducted to confirm infection.
Specifically, the document advises that “it is acceptable to report COVID-19 on a death certificate,” without confirmatory testing, “if the circumstances are compelling within a reasonable degree of certainty.” This advice can introduce subjectivity, however, as one physician’s assessment of what is a reasonably certain case may not be in line with another’s.
The CDC classifies COVID-19 cases as suspect, probable or confirmed based on clinical criteria, laboratory evidence and epidemiologic linkage. Of these three types, only a confirmed case “meets confirmatory laboratory evidence” (i.e., detection of SARS-CoV-2 RNA in a clinical or autopsy specimen using a molecular amplification test).
To help explain these types of cases, this CDC’s April guidance includes an appendix with three practical examples of death certifications, or cause-of-death statements. The SARS-CoV-2 virus — also known as the severe acute respiratory syndrome coronavirus 2 virus — causes COVID-19 disease.
One of the appendix’s examples details the death of an 86-year-old woman with an unconfirmed case of COVID-19 disease. Testing to verify infection was not performed, although the coroner concluded that “given the patient’s symptoms and exposure to an infected individual,” the likely underlying cause of death was COVID–19.
Cases that do not include testing to confirm infection, however (i.e., suspect and probable cases), are not confirmed cases, which lays open the possibility (if not the likelihood, at least in some cases) of a certifier, despite adhering strictly to the CDC’s April guidance, recording COVID-19 on the death certificate of a decedent not infected with the virus — a practice that could result in over-counting the actual number of deaths caused by COVID-19.
“Probable” or “presumed” infection
The CDC’s April guidance further advises that if a case cannot be definitely diagnosed but the disease is “suspected or likely (e.g., the circumstances are compelling within a reasonable degree of certainty)” — as in the case of the 86-year old woman — COVID–19 may be reported on the death certificate as “probable” or “presumed.”
The guidance again adds a similar caveat that, in such instances, “certifiers should use their best clinical judgement in determining if a COVID–19 infection was likely.”
This April advice is consistent with a presentation the CDC posted on-line the same month discussing this topic. During the webinar, the CDC stated it encourages use of the words “probable” and “likely” when testing for the virus is not available.
The CDC verified during this presentation, however, that it is not concerned that writing “probable” or “presumed” on a death certificate could result in the “overcounting the number of deaths.” (Other CDC guidance too can permit listing a condition as “probable” even if it has not been definitively diagnosed.)
Clarifying whether the CDC might “count” a probable (or unconfirmed) case as a true, definitive COVID-19 infection, a representative for the Agency stated during this webinar that, “If you’re pretty sure that it was a COVID-19 infection and you’ve written probable COVID-19, I don’t think that there’s necessarily a need to amend the death certificate. And, of course, you may do that if you like.” All but eliminating any possible confusion, the representative added that “in a case of probable COVID-19 we will count it as COVID-19.”
The potential for misclassifying cases notwithstanding, a death certificate’s accurate recording of the cause of death that could be related to an infection — whether caused by COVID-19, a multidrug-resistant organism, or another infectious agent — or, alternatively, to a preexisting condition or comorbidity, such as diabetes, is integral to sound estimates of the pathogen’s mortality rate and public health threat, to the appropriate allocation of limited resources, and to the identification and implementation of effective strategies and mitigations that prevent additional infections.
For instance, an increase or spike in a pathogen’s mortality rate could compel health officials to adopt more robust measures to improve public safety, while a decline in this rate might cause the easing or removal of imposed precautions.
Expert “On-Call” Consulting Services: LFM-Healthcare Solutions, LLC provides expert guidance for healthcare facilities (hospitals), device manufacturers and others, specializing in medical device designs and both the causes and prevention of healthcare-associated infections linked to medical equipment.
Multidrug-resistant organisms — or “superbugs” — are another public health scourge. For perspective, while the COVID-19 death rate has been estimated to be less than 1%, the U.S. mortality rate for some superbug infections, such as “CRE” and Candida auris, is significantly higher, and can be as high as 50%.
A peer-reviewed report in 2018 re-estimated the number of deaths in the U.S. in 2010 due to a superbug infection to be as high as 162,000, which is not starkly different from the number of deaths in the U.S. attributed to the COVID-19 disease reported to date.
CRE are members of the Enterobacteriaceae family of bacteria that are resistant to “last resort” carbapenem antibiotics. Examples of CRE include NDM-1-producing E. coli and carbapenemase-producing Klebsiella pneumonia.
In today’s clinical settings, COVID-19 and serious bacterial infections can become intertwined. A study published in Lancet reporting on the deaths of 54 hospitalized patients with laboratory-confirmed COVID-19 found that half (27) experienced a secondary infection, such as ventilator-associated pneumonia. Fifty three of these patients (98%) received antibiotics and 12 (22%) received antivirals.
Underscoring the dangers and epidemiology of superbugs, U.S. federal health officials reported earlier this year that “antibiotic resistance is one of the biggest public health challenges of our time.”
Last year, referring to antimicrobial resistance, a United Kingdom public health representative warned about “a danger to humanity” that is “as big a danger as climate change or warfare.” Also last year, the CDC stated in a report discussing certain superbugs as “urgent threats” that the agency “has taken aggressive action,” in coordination with other stakeholders including state and local health departments, “to rapidly detect, prevent spread, and innovate against antibiotic resistance.”
In May, a former director of a CDC division warned that superbugs could be another “hidden danger” of COVID-19. A September news report quoted the biosecurity research director at Australia’s national science agency to say “Covid is not anywhere near the potential impact” of antimicrobial resistance.
And, last year the World Health Organization concluded that if no action is taken, “drug-resistant diseases could cause 10 million deaths each year by 2050.”
OBJECTIVES and METHODS
This article reviews the CDC’s April guidance for recording the cause of deaths related to the COVID-19, and compares its policies and criteria to the standards and practices state and U.S. health officials have used to assess the cause of death of patients infected with a superbug.
Several publications were reviewed in addition to the CDC’s COVID-19 guidance, including several superbug outbreak investigations, other federal guidelines, state guidance documents, news reports, and other relevant sources.
Federal and state guidance for reporting cases and deaths that could be due to the influenza virus, which causes the flu in humans, was also reviewed for some perspective.
FINDINGS and DISCUSSION
The standards the CDC’s guidance uses for the certification of deaths that could be due to COVID-19 appear to differ, in some respects, from the practices that health officials have used to assess the potential contribution of a superbug infection to a patient’s death.
For instance, establishing a relatively low watermark for causality, the CDC advises that COVID-19 be specified on the death certificate if it “played a role in the death.” Health officials in Illinois investigating a superbug outbreak in 2013 linked to reusable medical equipment, however, like other reviewed cases, appear to have applied a different, if higher, standard of proof.
Instead, these health officials reported that the patient deaths in Illinois “did not appear” to be related to the superbug infection, that officials “could not say whether the bacteria had contributed to any deaths,” and that “no one had died as a direct result of infection or exposure” — this, despite the documented high mortality rate associated with this outbreak’s superbug.
Similarly lacking a definitive assessment in a different case involving four deaths linked to the deadly C. auris superbug, health officials apparently told a news source that “it’s not clear if the deaths were due to the (fungal) infection or the patients’ underlying health conditions.”
And, discussing a superbug outbreak in Washington, between 2012 and 2013, that was also linked to reusable medical equipment, the investigators observed that “[b]ecause affected patients all had severe underlying disease, the role of the outbreak strain in patient deaths cannot be determined.”
That a comorbidity can apparently hinder, interfere with or even obscure a clinical assessment and determination of the role a superbug played in the death of infected patients is salient, particularly in the context of today’s pandemic when certifying and recording the COVID-19 disease as a cause of death has become common in patients with comorbidities.
For example, the aforementioned March Lancet study reported that two thirds of the COVID-19 patients who died were diagnosed with a comorbidity, 26 of them (48%) with hypertension.
The same month, the CDC reported that approximately 90% of hospitalized patients identified through a COVID-19 surveillance network “had one or more underlying conditions.” And in updated August findings, the CDC reported that 94% of the individuals in the U.S. who died of COVID-19 had at least one pre-existing condition, such as obesity or heart disease.
In a second fungal case, three chronically ill patients in New York with pan-resistant C. auris infection expired. Appearing again to apply a different burden for causality than its April COVID-19 guidance, the CDC reported that for two of these three patients “the role of C. auris in their deaths is unclear.” The third infected patient died, the CDC reported, “from underlying medical conditions” — not apparently, either directly or indirectly, from the pan-resistant superbug.
Notwithstanding its assessment that the potentially life-threatening C. auris infection was not to blame for any of these three deaths, the CDC has reported that 30-60% of individuals infected with C. auris die.
In short, none of the superbug investigations that were reviewed for this article found that the life-threatening infection (which in some cases could be associated with a mortality rate of as high as 50%) caused, or directly contributed to, patient death.
Whether the assessments of the cause of death by officials, including the CDC, in these three investigated superbug cases are anomalistic, or instead reflective of a common standard health officials may apply to superbug-related deaths, is unclear. (This article acknowledges that bacterial and fungal superbugs, like the infections they may cause, differ in many important regards from the COVID-19. disease.)
(Article continues below)
The following set of three peer-reviewed articles by Dr. Muscarella (and in one case, a co-author) may be the most comprehensive and historical collection published to date discussing associations between deadly superbugs and reusable medical equipment:
(1) Muscarella LF. Use of ethylene-oxide gas sterilisation to terminate multidrug-resistant bacterial outbreaks linked to duodenoscopes. BMJ Open Gastroenterol 2019 Aug 5;6(1):e000282. || (2) Mehta AC, Muscarella LF. Related “Superbug” Infections. Chest 2019 Aug 14;pii:S0012-3692(19)31457-6. || (3) Muscarella LF. Risk of transmission of carbapenem-resistant Enterobacteriaceae and related “superbugs” during gastrointestinal endoscopy. World J Gastrointest Endosc 2014 Oct 16;6(10):457-74.
(Article continues from above)
It is not only assessments by health officials of the cause of death for superbug-infected patients that appear to employ a different standard or measure than the CDC’s April guidance for recording deaths that could be due to COVID-19.
In fact, some of the criteria the CDC and state health departments use for defining cases and reporting deaths associated with the influenza viruses appear, too, to differ substantively, in terms of the testing requirements to confirm infection, from the CDC’s April COVID-19 guidance.
In the context of pediatric mortality, the CDC defines an influenza-associated death (for surveillance purposes) as “a death resulting from a clinically compatible illness that was confirmed to be influenza by an appropriate laboratory or rapid diagnostic test.”
This CDC definition elaborates that there “should be no period of complete recovery between the illness and death,” and that influenza-associated deaths “in all persons aged <18 years should be reported.” (While flu-associated deaths in children are nationally notifiable, states are not required to report to the CDC adult deaths related to influenza infection.)
Re-emphasizing the importance of testing to confirm infection, the CDC clarifies that a death (in this context of pediatric mortality) should not be reported if, among other exclusionary criteria, there is “no laboratory confirmation of influenza virus infection.” This policy reduces, if not eliminates, the potential for subjectivity and for false-positive assessments of infection.
Nevertheless, unlike the CDC’s case classification for influenza-associated pediatric mortality, which features only confirmed cases of infection, the CDC’s case classification for COVID-19 includes, additionally, suspect and probable cases, which in contrast (and as stated previously) lack confirmatory laboratory evidence.
Influenza infection can range from asymptomatic illness to viral pneumonia and death. The seasonal “flu” is caused by influenza A and B viruses, and like COVID-19, it is a contagious respiratory illness.
Also addressing a confirmatory testing requirement in the context of the influenza, the CDC advises in an on-line document that infection “cannot be diagnosed accurately based on signs and symptoms alone. Laboratory testing is necessary to confirm the diagnosis.”
Adding an important detail, the CDC maintains in this document that “individual cases of influenza cannot be distinguished from other respiratory virus infections based on clinical information alone.”
Affirming the importance of confirmatory testing in the context of deaths possibly related to influenza, states in the U.S., similar to the CDC’s pediatric mortality guidance, may require laboratory testing to confirm infection.
For instance, the Washington State Department of Health advises that “[o]nly lab-confirmed fatal cases are reportable.” According to this state’s reporting policies (which do not appear to distinguish between pediatric and adult cases), the purpose of reporting and surveillance of (laboratory confirmed) influenza-associated deaths is, in part, to determine mortality rates and to “detect emerging threats such as avian and other novel influenza strains.”
Similarly, the North Carolina Department of Health and Human Services advises that, for both adult and pediatric cases, all influenza-associated deaths “are reportable to the Local Health Department. An influenza-associated death is defined for surveillance purposes as a death resulting from a clinically compatible illness that was confirmed to be influenza (any strain) by an appropriate laboratory or rapid diagnostic test.”
This state health department emphasizes that, among other exclusion criteria, a “death should not be reported” if there is “no laboratory or rapid test confirmation of influenza virus infection.“
Unlike flu deaths in adults, however, COVID-19 is a nationally notifiable disease. And in contrast to state and federal definitions of influenza-associated deaths, which include only cases confirmed by testing, the CDC’s case classification for COVID-19 additionally includes suspect and probable cases and deaths.
As a consequence and in summary, the CDC’s guidance permits recording and certifying COVID-19 as a cause of death without the requirement that testing be conducted to confirm infection (provided “the circumstances are compelling within a reasonable degree of certainty”).
Expert Scientific Advisor: LFM-Healthcare Solutions, LLC provides medical expertise and guidance for healthcare facilities (hospitals), device manufacturers and legal representatives of the public, specializing in medical device designs and both the causes and prevention of healthcare-associated infections linked to reusable medical equipment.
Validation can be a critical tool for confirming the quality of healthcare safety data. According to the CDC, validation is “double-checking, or confirming” the accuracy and completeness of reportable infection data. Central line-associated bloodstream infections are reportable data that have been validated for years to assess accuracy and the potential for inadvertent reporting bias.
Infection data can be validated using different methodologies (e.g., “external validation”). In 2012 the CDC published guidance for the validation of central line-associated bloodstream infections in intensive care units.
A 2017 inspector general report states that “accurate data are fundamental to the integrity of the (Centers for Medicare and Medicaid Services’) quality-based payment programs,” adding that the reporting of imprecise data could affect payment accuracy.
Validation of death certificate data associated with the COVID-19 disease is also important to ensure the data’s quality and integrity. Death certificates state a certifier’s medical opinion and therefore can have mistakes — for example, listing a pathogen like COVID-19 as the cause of death of a decedent who was neither tested for nor infected with the pathogen.
Such validation is a type of “data auditing” that might reasonably involve reviewing a statistically reliable and randomly selected sample of patient medical records (or autopsy findings) and comparing the findings with the associated death certificate documentation and data to evaluate their consistency with one another. Other data validation model types could also be used.
Update (March 31, 2021): On March 31, 2021, the CDC published a report — “Death Certificate–Based ICD-10 Diagnosis Codes for COVID-19 Mortality Surveillance — United States, January–December 2020” — that assessed documentation of diagnoses co-occurring with an ICD-10 code for COVID-19, which is U07.1, on U.S. death certificates from 2020 that had been reported to CDC as of February 22, 2021. (“ICD-10” refers to the International Classification of Diseases, Tenth Revision.)
The report concludes that the CDC’s analysis “support the accuracy of COVID-19 mortality surveillance in the United States using official death certificates. High-quality documentation of co-occurring diagnoses on the death certificate is essential for a comprehensive and authoritative public record.”
Whether the data do indeed support this conclusion is a fair point to raise, because CDC acknowledges in the report that, as a limitation of its analysis, it “was unable to compare death certificate data with decedent’s medical records or autopsy reports for end-of-life events and co-occurring diagnoses.”
Lacking this specific type of validation, conclusions about the accuracy of COVID-19 mortality surveillance in the U.S. using official death certificates could be questioned.
Highlighting the importance of validation using medical record data (or autopsy reports), the CDC wrote in this March 31st report that: “Medical record review is needed to confirm findings from this study and elucidate more information for decedents with only COVID-19 listed on their death certificate or those that could not be plausibly categorized as attributable to COVID-19 based on death certificate data alone.”
CONCLUSIONS and SUMMARY
This review’s findings suggest that the standards and measures the CDC’s April 2020 guidance use for recording deaths that could be due to COVID-19 differ in some salient regards from the practices and paradigms that U.S. health officials have adopted to assess the contribution of a superbug infection to patient death.
For instance, the CDC’s guidance establishes a standard of inclusivity that permits the recording of COVID-19 on a descendant’s death certificate if it “played a role in the death,” is “assumed to have caused or contributed to death,” or is “presumed” to be at fault.
In contrast, the policies and procedures that health officials have adopted to assess the cause of death in a superbug-infected patient have employed (at least in the several cases reviewed for this article) a narrower standard or burden for causality, with officials reporting at times that mortality “was not necessarily attributable” to the superbug infection, the superbug infection’s role in death was “unclear,” or the infected patient’s death was due to underlying medical conditions, not the superbug.
Despite the relatively high mortality rates of superbugs, which can be as high as or exceed 50%, none of the outbreak cases reviewed for this article “presumed” the multidrug-resistant organism to be a cause of patient death or concluded that the superbug infection caused or directly contributed to patient death.
That a comorbidity could potentially hinder or interfere with clinical assessments of the role played by a pathogen, at least apparently a superbug, in an infected patient’s death is salient, particularly in today’s context of certifying and recording COVID-19 as a cause of death in patients with comorbidities.
Moreover, lacking confirmatory testing, it therefore becomes plausible that, in some cases, adherence to the CDC’s April guidance could inadvertently introduce unwanted subjectivity and result in recording COVID-19 on the death certificate of a decedent not infected with the virus (since this guidance does not, under all circumstances and as a prerequisite, require testing to confirm infection).
Also suggesting different rules for different pathogens, public health definitions and requirements in the U.S. for reporting an influenza-associated death require confirmation of influenza virus infection using an appropriate laboratory or rapid diagnostic test. As discussed in this article, however, the CDC’s April guidance, like the case definition for COVID-19, does not require confirmatory testing to record COVID-19 as a cause of death.
If the criteria and policies for reporting and recording the cause of death of a patient infected with other types of pathogens such as a superbug were to mimic more closely the guidance the CDC applies today to COVID-19 — for example, not requiring confirmatory testing to ensure a superbug caused an infection, or specifying the superbug-related infection on the death certificate as long as it “played a role” in death (but did not necessarily cause the death) — the impact on healthcare and infection prevention could be significant, possibly with the number of recorded deaths due to a superbug increasing significantly. (In 2016, a California state senator introduced a bill that would require clinicians to record antibiotic-resistant infections on death certificates if the superbug “played a role in the death,” the Los Angeles Times reported at that time.)
Certainly, accurate accounting of the number of deaths due to a pathogen, particularly trends in these data over time, is necessary to the targeted allocation of limited recourses and more so to identify and implement effective strategies and mitigations that prevent additional infections. The accuracy of infection and death data, whether due to a virus or superbug, can be evaluated by validating these data (e.g., reviewing a statistically reliable and randomly selected sample of patient medical records (or autopsy findings) for accuracy and completeness.
In closing, this article discusses and compares assessments of the cause of death associated with superbugs, COVID-19 and influenza (understanding, however, that there are important differences between bacteria and viruses). Enhanced efforts to standardize the methods used to count infections and to improve surveillance for antibiotic resistance and to advance public awareness about superbug infections, particularly during today’s COVID-19 pandemic, are recommended to understand with greater accuracy the impact, mortality and costs of primary and secondary superbug-related infections in patients with and without underlying medical conditions. The adoption of policies that reduce, if not eliminate, subjectivity by requiring testing to confirm infection with a respiratory virus is also encouraged.
LIMITATIONS, COMMENTS: This article has several limitations. First, the several superbug outbreak investigations that this article reviewed could have conceivably introduced inadvertent publication bias into its methodology, findings and conclusions due to the lack of randomization, although this is unlikely and efforts were taken to avoid such bias.
Second, it is recognized and appreciated that: (a) assessments of the cause of death of a patient infected with a superbug, the COVID-19 virus, or another pathogen can be complicated and more challenging in the presence of pre-existing underlying health conditions; and (b) the circumstances surrounding the discussed superbug cases may be significantly different in important respects than the discussed cases discussing COVID-19 and influenza, which might, in some conceivable way, justify the application of different standards based on the pathogen type. The characteristics of bacteria and viruses can be markedly different. Nevertheless, a diagnosis of infection would ideally be performed by a physician specializing in infectious diseases, and the number of deaths attributed to an infection be based on the patient’s medical records, not solely the death certificate.
Third, an association between an infection and the use of medical equipment does not indicate the equipment caused the infection. Fourth, the rationale for the CDC’s April guidance not mandating, as a prerequisite, confirmatory testing to record COVID-19 infection as the cause of death was not determined. Fifth, the justification for this federal COVID-19 guidance to appear to differ in certain respects from the practices and standards health officials have used in the past to assess the role that a superbug infection may have played in patient death also was not investigated.
Sixth, the percentage of recorded deaths attributed to COVID-19 that may have been due instead to a comorbidity (and/or were associated with someone not infected with the virus) could not be evaluated or approximated.
Last, also not studied were the clinical and diagnostic similarities and differences between COVID-19 and the influenza viruses (the latter of which are not coronaviruses), or the impact that implementation of data validation might have on published mortality rates of either COVID-19 or superbug-related infections.
Article by: Lawrence F Muscarella, PhD. Posted August 2020. Copyright (2020). LFM Healthcare Solutions, LLC. All rights reserved. Dr. Muscarella is the president of LFM Healthcare Solutions, LLC, an independent quality improvement company. Click here for more details about his quality improvement healthcare services.
Case Reviews, Expert “On-Call” Consulting Services: LFM-Healthcare Solutions, LLC provides medical expertise and guidance for healthcare facilities (hospitals), device manufacturers and legal representatives of the public, specializing in the causes and prevention of healthcare-associated infections linked to medical equipment.