 |
|
Problems With Proper Completion and Accuracy of the Cause-of-Death Statement
Ann E. Smith Sehdev, MD;
Grover M. Hutchins, MD
Arch Intern Med. 2001;161:277-284.
ABSTRACT
| |
Background Mortality statistics are largely based on death certificates, so it
is important that the data on the death certificate is accurate. At our institution,
clinicians complete cause-of-death statements (CODs) prior to autopsy. Since
May 1995, separate CODs have been included in autopsy face sheets.
Methods Clinical and autopsy-based CODs filled out separately on 494 cases between
June 1995 and February 1997 were compared for proper reporting and accuracy
using the published guidelines and definitions of immediate, intermediate,
and underlying causes of death put forth by the College of American Pathologists
and the National Center for Health Statistics.
Results Of the 494 death certificates, 204 (41%) contained improperly completed
CODs. Of these, 49 (24%) contained major discrepancies between clinicians'
and pathologists' CODs. Of the 494 death certificates, 290 (59%) had properly
completed CODs. Of the 290 properly completed CODs, 141 (49%) contained disagreements:
73 (52%) on underlying CODs; 44 (31%) on immediate CODs; and 47 (33%) on other
significant conditions (part II).
Conclusions The reliability and accuracy of CODs remain a significant problem. Despite
its limitations, the autopsy remains the best standard against which to judge
premortem diagnoses. The CODs of the death certificate may be improved if
death certificates are completed in conjunction with the postmortem examination
and amended when the autopsy findings show a discrepancy.
INTRODUCTION
THE INFORMATION from death certificates, specifically the cause-of-death
statements (CODs), is the basis for our national mortality database. The National
Center for Health Statistics (NCHS) uses this database to help with surveillance
of disease and proper allocation of funds for public health programs and research,
and to help prioritize governmental decisions and actions in regard to health
care. Because health statistics, national mortality and morbidity statistics,
and data on disease prevalence in society are largely derived from death certificates,
it is important to ensure proper completion and accuracy of the cause-of-death
section of the death certificate.1
At The Johns Hopkins Medical Institutions, Baltimore, Md, it is practice
of the clinician to complete the cause-of-death section of the death certificate
at the time of death of the patient even if an autopsy is to be performed.
In 1991, during the Second Workshop on Improved Cause-of-Death Statistics,
the NCHS and the National Committee on Vital and Health Statistics (NCVHS)
recommended adding CODs to autopsy reports.2
Since May 1995, a cause-of-death section has been included in the autopsy
face sheet of all postmortem examinations at The Johns Hopkins Medical Institutions.
As a result, residents and faculty in the pathology department have been instructed
on the importance of the proper completion and accuracy of the cause-of-death
section of the death certificate.
MATERIALS AND METHODS
Completed CODs from 494 autopsies performed at The Johns Hopkins Medical
Institutions between June 1995 and February 1997 were selected for study.
All clinical and autopsy-based CODs were first evaluated by the first author
(A.E.S.) for consistency of reporting and adherence to instructions for proper
completion. The CODs were subsequently evaluated by one of us (G.M.H.) for
validation and confirmation of the accuracy of the findings. Proper completion
of the CODs was determined using the published guidelines and definitions
of immediate, intermediate, and underlying causes of death put forth by the
College of American Pathologists and the NCHS3
(Table 1).
|
|
|
|
Table 1. Guidelines and Definitions* for Proper Completion of the Cause-of-Death
Statement as Put Forth by the CAP and the NCHS
|
|
|
The "properly" completed clinical and autopsy-based CODs were subsequently
compared for accuracy of diagnoses using the postmortem examination in conjunction
with clinical information as the standard for comparison. The overall disagreements
between clinical and autopsy-based CODs were subclassified into the following
categories: disagreement on immediate cause of death, disagreement on underlying
cause of death, and/or disagreement on other significant conditions. The categories
of disagreement were further subclassified into whether the disagreement involved
the same organ system or a different organ system.
The "improperly" completed CODs were further analyzed for the presence
of major discrepancies between the clinicians' and the pathologists' CODs.
These major discrepancies included either (1) a major finding listed in the
autopsy as immediate or underlying cause of death that was not listed in the
clinician's cause of death, or (2) a disease or manifestation of disease listed
in the clinician's cause of death that was not validated by postmortem examination.
RESULTS
Overall, 204 (41%) of the CODs were improperly completed (Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, and Figure 6). Of these, the CODs improperly
completed by clinicians (n = 191) significantly outnumbered those improperly
completed by pathologists (n = 27) (Table
2). The most common mistakes made by clinicians included (1) using
"mechanisms" as the immediate cause of death in 64 cases (34%); (2) not qualifying
nonspecific processes in 90 cases (47%); (3) listing the underlying and immediate
causes of death out of order in 35 (18%); and (4) placing underlying or immediate
causes of death in part II (other significant conditions contributing to death
but not resulting in the underlying cause of death given in part I) in 39
(20%). Other mistakes included using abbreviations, listing other significant
conditions (part II) in part I, and listing incidental findings in part II.
|
|
|
|
Figure 1. Mechanisms of death should not
be used in cause-of-death statements. What is the underlying cause of death?
|
|
|
|
|
|
|
Figure 2. Abbreviations should not be used
anywhere in cause-of-death statements. What else is wrong with this cause-of-death
statement?
|
|
|
|
|
|
|
Figure 3. Citing a nonspecific process as
the underlying cause of death without qualification, ie, unspecified etiology.
|
|
|
|
|
|
|
Figure 4. Nonsequential listing of conditions.
What is the underlying cause of death? The immediate cause of death?
|
|
|
|
|
|
|
Figure 5. The underlying cause of death
(disease or condition that initiated the morbid train of events leading to
death) is listed in Part II.
|
|
|
|
|
|
|
Figure 6. Hypertension is another significant
condition that likely contributed to but did not cause death. What is the
underlying cause of death?
|
|
|
|
|
|
|
Table 2. Comparison of Proper Completion of 494 Separate Clinical and
Autopsy-Based Cause-of-Death Statements*
|
|
|
Of the 494 CODs reviewed, 290 (59%) were properly completed according
to the guidelines of the College of American Pathologists and the standards
set forth by the NCHS. Of those 290 CODs, 141 (49%) contained disagreements
between the clinical and postautopsy versions with many CODs containing more
than 1 disagreement. The most common discrepancy involved the underlying cause
of death, with 73 CODs (52%) containing disagreements, 35 (48%) of which were
assigned to the same organ system and 38 (52%) of which were assigned to a
different organ system. Of the 44 CODs (31%) that contained discrepancies
regarding the immediate cause of death, 18 (41%) were assigned to the same
organ system, and 26 (59%) were assigned to a different organ system. In 47
CODs (33%) there was disagreement on other significant conditions when listed
by both clinicians and pathologists. Of the CODs improperly completed, 49
(24%) contained major discrepancies, including either a major finding listed
in the pathologists' immediate or underlying COD that was not listed by the
clinician, or an entity listed by the clinician that was not validated by
postmortem examination (eg, pulmonary embolus) (Table 3).
|
|
|
|
Table 3. Comparison of the Accuracy of 494 Clinical Cause-of-Death
Statements With Those Based on Clinical and Autopsy-Derived Information*
|
|
|
COMMENT
Based on the results of this study, CODs are not reliable or accurate
sources of information on which to base national mortality statistics
(Figure 7,
Figure 8,
Figure 9, and
Figure 10). Several studies from around
the world have addressed this topic and reached similar conclusions.4, 5, 6, 7, 8, 9, 10, 11, 12, 13
In each of these studies, it is emphasized that information gathered from
death certificates plays a key role in determining disease prevalence in society
and ultimately has a significant effect on decision-making processes regarding
the distribution of resources in the fields of medicine and health.
|
|
|
|
Figure 7. Disagreement in the underlying
cause of death. Disagreement is confined to the same organ system.
|
|
|
|
|
|
|
Figure 8. Disagreements on the immediate
and underlying causes of death. Disagreements involve different organ systems.
|
|
|
|
|
|
|
Figure 9. Significant finding listed in
autopsy face sheet that was not listed in clinician's cause of death.
|
|
|
|
|
|
|
Figure 10. Finding listed in clinician's
cause of death statement that was not validated by autopsy.
|
|
|
In a study done in London, Ontario, Jordan and Bass7
reviewed 426 death certificates to determine if they met the criteria for
proper completion. In addition, the authors looked at which clinical department
each death occurred in, whether staff physicians or residents completed the
death certificate, and whether a coroner was involved and an autopsy was performed.
Of the 426 certificates reviewed, 45% were filled out correctly and 23% contained
only minor errors (inappropriate information and absence of time intervals).
Of the 32% of death certificates with major errors, the most frequent error
was incorrect sequencing (22%), which resulted from recording information
out of order or in an illogical fashion when read vertically down the death
certificate. The second most common major error was recording 2 causes of
death in part I (17%). The most serious type of error, according to this study,
was listing a mechanism as a cause of death without an explanation. This error
occurred in 10% of death certificates. In 1993, Hanzlick6
reviewed 56 death certificates completed over a 10-day period: 35 (63%) of
the certificates showed either an omission or underlying COD that was nonspecific
or in need of further explanation. In 32 (91%) of the incorrectly completed
certificates, the immediate cause of death was cited in terms of a mechanism.
In 1981, Cameron and McGoogan4, 5
performed a prospective study of 1152 autopsies and compared the certified
clinical diagnoses with autopsy findings to assess the diagnostic accuracy
(or inaccuracy) of death certification. In their study, the main clinical
diagnosis was confirmed in 61% of cases. The clinical diagnoses, which were
not confirmed at autopsy, were either disproved (27% of cases) or determined
to be subsidiary to the cause of death (12% of cases). In a study by Kircher
et al,9 272 autopsy reports were reviewed of
3884 decedents in Connecticut in 1980. In this study the researchers compared
the International Classification of Diseases, Ninth Revision (ICD-9) disease categories of the underlying
cause of death listed by the clinicians with that listed by the nosologically
coded autopsy. Kircher et al9 reported that
a major disagreement (underlying cause of death assigned to different major ICD-9 categories) occurred in 29% of cases. In 1991, Nielsen
et al13 published a study comparing accuracy
of death certificates on all autopsies performed in 1976 and 1986 at the University
of Iceland, Reykjavík. These authors reported a 50% overall disagreement
between the death certificate and autopsy diagnosis with a disagreement on
COD in 25% of total cases. Of interest, the authors found that the overall
accuracy of premortem diagnoses remained unchanged between the years 1976
and 1986, during which time nonobstetric ultrasound and computed tomography
were reportedly introduced into practice in Iceland.
It has been suggested that clinicians may not be aware of the importance
of the COD in the generation of health statistics. In Hanzlick's attempt6 to improve accuracy of death certificates, he sent
form letters to 32 physicians who had improperly completed CODs. The letter
included a listing of omissions and the suggestion that the physician consider
amending the certificate. Instructions for amending were provided, and a contact
number was included. After 30 days, however, only 1 physician had amended
a COD. This suggested that clinicians might not consider amendment to be necessary
or a priority.
Recently there have also been several publications designed to help
educate physicians regarding the proper completion of the death certificate
and the standard definitions of immediate, intermediate, and underlying cause
of death.14, 15 In 1987, Kircher
and Anderson16 published a special communication
in JAMA to help provide medical students, house staff, and physicians with
information regarding the proper completion of the death certificate and the
standard definitions of immediate, intermediate, and underlying cause of death.
In October 1989, a national conference sponsored by NCHS and NCVHS made several
recommendations to improve the accuracy of death certificates.17
Their suggestions included increasing training of house staff and medical
students, developing quality improvement programs, revising the format of
the death certificate, and encouraging amendment of death certificates when
indicated. Currently, the medical students at our institution receive instruction
in death certificate completion during their second-year curriculum as well
as during a transitional course given in their last year of training, prior
to entering residency. In addition, for the past 5 years, our institution
has required that all new members of the house staff receive specific instruction
regarding death certification. Third, it has been the practice at our institution
that admissions personnel review each completed death certificate to find
and address inaccuracies in completion. This latter policy has been in place
for approximately the past 10 years. Despite these implementations and attempts
at clinician education, the proper completion of the CODs remains a significant
problem at our institution, suggesting that further guidance is needed on
death certification. In 1993, Hanzlick published a letter emphasizing this
fact.18 In his letter, Hanzlick writes that
many problems in CODs may result from inconsistency in wording, variations
in certification style, and confusion regarding published examples of recommendations
and guidelines. In an effort to educate and provide consistency, Hanzlick
and the Autopsy Committee of the College of American Pathologists3 published The Medical Cause of Death
Manual in 1994.
From our present study's comparison of clinical and autopsy-based CODs,
it seems that education of the pathology faculty and house staff does help
to improve the proper completion of the CODs. At our institution it is part
of the initial training of any house officer in anatomic pathology to receive
instruction regarding the importance of the proper completion and accuracy
of CODs through the use of a prosector's manual, the COD manual,3
and various didactic methods. In addition, the house officer works closely
with a more senior resident in the formulation of the COD. Finally, all completed
autopsy-based CODs after release by a faculty member are re-reviewed by the
director of the autopsy service for quality assurance. This study suggests
not only that education of clinical house staff can help to improve the proper
completion of CODs but also that pathologists may play an important educational
role in promoting standardization and accuracy of death certification.
Based on the findings of this study as well as other studies, the autopsy
remains a highly valuable educational and diagnostic tool that plays an invaluable
role in the final step in clinical investigation.19
Despite its limitations, the autopsy, in conjunction with clinical information,
remains the best standard by which to judge premortem diagnoses.4, 9, 13, 20
It has been suggested that, for quality improvement, a team approach between
the certifying physician and the physician trained and experienced in death
certification may help reduce the errors in death certificate completion.8 Ultimately, the reliability and accuracy of CODs may
be improved if death certificates were completed in conjunction with the postmortem
examination and amended when the autopsy findings show a discrepancy.21
AUTHOR INFORMATION
Accepted for publication July 11, 2000.
Presented in part at the annual meeting of the US and Canadian Academy
of Pathologists, Boston, Mass, March 1998.
From the Department of Pathology, The Johns Hopkins Medical Institutions,
Baltimore, Md.
Corresponding author: Ann E. Smith Sehdev, MD, Carnegie 400, Department
of Pathology, The Johns Hopkins Hospital, 600 N Wolfe St, Baltimore, MD 21287-6901.
REFERENCES
| |
1. Maudsley G, Williams EMI. Inaccuracy in death certification: where are we now? J Public Health Med. 1996;18:59-66.
FREE FULL TEXT
2. National Committee on Vital and Health Statistics. Report of the second workshop on improving cause-of-death statistics. Presented at: National Center for Health Statistics; April 21-23,
1991; Virginia Beach, Va.
3. Hanzlick R, ed. The Medical Cause of Death Manual: Instructions For
Writing Cause of Death Statements For Deaths Due to Natural Causes. Northfield, Ill: College of American Pathologists; 1994.
4. Cameron HM, McGoogan E. A prospective study of 1152 hospital autopsies, I: inaccuracies in
death certification. J Pathol. 1981;133:273-283.
FULL TEXT
|
ISI
| PUBMED
5. Cameron HM, McGoogan E. A prospective study of 1152 hospital autopsies, II: analysis of inaccuracies
in clinical diagnoses and their significance. J Pathol. 1981;133:285-300.
FULL TEXT
|
ISI
| PUBMED
6. Hanzlick R. Improving accuracy of death certificates [letter]. JAMA. 1993;269:2850.
FREE FULL TEXT
7. Jordan JM, Bass MJ. Errors in death certificate completion in a teaching hospital. Clin Invest Med. 1993;16:249-255.
ISI
| PUBMED
8. Kaplan J, Hanzlick R. Improving the accuracy of death certificates [letter]. JAMA. 1993;270:1426.
9. Kircher T, Nelson J, Burdo H. The autopsy as a measure of accuracy of the death certificate. N Engl J Med. 1985;313:1263-1269.
ABSTRACT
10. Maclaine GDH, Macarthur EB, Heathcote CR. A comparison of death certificates and autopsies in the Australian
Capital Territory. Med J Aust. 1992;156:462-468.
ISI
| PUBMED
11. McKelvie PA. Medical certification of causes of death in an Australian metropolitan
hospital: comparison with autopsy findings and a critical review. Med J Aust. 1993;158:816-821.
ISI
| PUBMED
12. Moussa MAA, Shafie MZ, Khogali MM, et al. Reliability of death certificate diagnoses. J Clin Epidemiol. 1990;43:1285-1295.
FULL TEXT
|
ISI
| PUBMED
13. Nielsen GP, Bjornsson J, Jonasson JG. The accuracy of death certificates: implications for health statistics. Virchows Arch A Pathol Anat Histopathol. 1991;419:143-146.
FULL TEXT
|
ISI
| PUBMED
14. Hanzlick R. Protocol for writing cause-of-death statements for deaths due to natural
causes. Arch Intern Med. 1996;156:25-26.
FREE FULL TEXT
15. Hanzlick R. Principle for including or excluding "mechanisms" of death when writing
cause-of-death statements. Arch Pathol Lab Med. 1997;121:377-380.
ISI
| PUBMED
16. Kircher T, Anderson RE. Cause of death: proper completion of the death certificate. JAMA. 1987;258:349-352.
FREE FULL TEXT
17. Barber JB. Improving accuracy of death certificates. J Natl Med Assoc. 1992;84:1007-1008.
PUBMED
18. Hanzlick R. Death certificates: the need for further guidance. Am J Forensic Med Pathol. 1993;14:249-252.
ISI
| PUBMED
19. Feinstein AR. Epidemiologic and clinical challenges in reviving the necropsy. Arch Pathol Lab Med. 1996;120:749-752.
ISI
| PUBMED
20. AMA Council on Scientific Affairs. Autopsy: a comprehensive review of current issues. Arch Pathol Lab Med. 1996;120:721-726.
21. Smith AE, Hutchins GM. Case of the month: making amends. Arch Intern Med. 1998;158:1739-1740.
FREE FULL TEXT
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati  Twitter
What's this?
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
|
Why Is the Educational Gradient of Mortality Steeper for Men?
Montez et al.
J Gerontol B Psychol Sci Soc Sci 2009;64B:625-634.
ABSTRACT
| FULL TEXT
Grade and Stage at Presentation Do Not Predict Mortality in Patients With Bladder Cancer Who Survive Their Disease
Messing et al.
JCO 2009;27:2443-2449.
ABSTRACT
| FULL TEXT
Causes of Death by Level of Dependency at 6 Months After Ischemic Stroke in 3 Large Cohorts
Slot et al.
Stroke 2009;40:1585-1589.
ABSTRACT
| FULL TEXT
Early and Late Mortality After Diagnosis of Wilms Tumor
Cotton et al.
JCO 2009;27:1304-1309.
ABSTRACT
| FULL TEXT
Cause-Specific Late Mortality Among 5-Year Survivors of Childhood Cancer: The Childhood Cancer Survivor Study
Mertens et al.
JNCI J Natl Cancer Inst 2008;100:1368-1379.
ABSTRACT
| FULL TEXT
Life expectancy and death from cardiomyopathy amongst carriers of Duchenne and Becker muscular dystrophy in Scotland
Holloway et al.
Heart 2008;94:633-636.
ABSTRACT
| FULL TEXT
Androgen Deprivation Therapy for Localized Prostate Cancer and the Risk of Cardiovascular Mortality
Tsai et al.
JNCI J Natl Cancer Inst 2007;99:1516-1524.
ABSTRACT
| FULL TEXT
Sex Differences in Lung Cancer Survival: Do Tumors Behave Differently in Elderly Women?
Wisnivesky and Halm
JCO 2007;25:1705-1712.
ABSTRACT
| FULL TEXT
Outcomes of Treatment vs Observation of Localized Prostate Cancer in Elderly Men--Reply
Wong et al.
JAMA 2007;297:1652-1653.
FULL TEXT
Defining Cause of Death in Stroke Patients: The Brain Attack Surveillance in Corpus Christi Project
Brown et al.
Am J Epidemiol 2007;165:591-596.
ABSTRACT
| FULL TEXT
Mortality in rheumatoid arthritis. Increased in the early course of disease, in ischaemic heart disease and in pulmonary fibrosis
Young et al.
Rheumatology (Oxford) 2007;46:350-357.
ABSTRACT
| FULL TEXT
Survival Associated With Treatment vs Observation of Localized Prostate Cancer in Elderly Men
Wong et al.
JAMA 2006;296:2683-2693.
ABSTRACT
| FULL TEXT
Geographical clustering of mortality from systemic lupus erythematosus in the United States: contributions of poverty, Hispanic ethnicity and solar radiation
Walsh and Gilchrist
Lupus 2006;15:662-670.
ABSTRACT
Declining death rates from hyperglycemic crisis among adults with diabetes, u.s., 1985-2002.
Wang et al.
Diabetes Care 2006;29:2018-2022.
ABSTRACT
| FULL TEXT
Heart Disease and Dementia: A Population-based Study
Bursi et al.
Am J Epidemiol 2006;163:135-141.
ABSTRACT
| FULL TEXT
Radiation Therapy for the Treatment of Unresected Stage I-II Non-small Cell Lung Cancer
Wisnivesky et al.
Chest 2005;128:1461-1467.
ABSTRACT
| FULL TEXT
Ethnic Disparities in the Treatment of Stage I Non-Small Cell Lung Cancer
Wisnivesky et al.
Am. J. Respir. Crit. Care Med. 2005;171:1158-1163.
ABSTRACT
| FULL TEXT
Postservice Mortality in Vietnam Veterans: 30-Year Follow-up
Boehmer et al.
Arch Intern Med 2004;164:1908-1916.
ABSTRACT
| FULL TEXT
The Effect of Tumor Size on Curability of Stage I Non-small Cell Lung Cancers
Wisnivesky et al.
Chest 2004;126:761-765.
ABSTRACT
| FULL TEXT
Medium-Term Survival after Hospitalization with Community-Acquired Pneumonia
Waterer et al.
Am. J. Respir. Crit. Care Med. 2004;169:910-914.
ABSTRACT
| FULL TEXT
A Prospective Study of Sleep Duration and Coronary Heart Disease in Women
Ayas et al.
Arch Intern Med 2003;163:205-209.
ABSTRACT
| FULL TEXT
Hepatitis C and Progression of HIV Disease
Sulkowski et al.
JAMA 2002;288:199-206.
ABSTRACT
| FULL TEXT
Determining Cause of Death in Prostate Cancer: Are Death Certificates Valid?
Penson et al.
JNCI J Natl Cancer Inst 2001;93:1822-1823.
FULL TEXT
|
