Original article
Factors associated with errors in death certificate completion: A national study in Taiwan

https://doi.org/10.1016/S0895-4356(00)00299-7Get rights and content

Abstract

To identify characteristics of certifying physicians and the deceased that are associated with errors in death certificate completion in Taiwan, we retrospectively reviewed 4123 systematically sampled death certificates issued in 1994. Multivariate analyses were used to assess the associations of various characteristics of the certifying physicians and the deceased with four types of error. Of the 4123 death certificates reviewed, 2525 (61%) were completed correctly. In 289 (7%), only the mechanism(s) of death was given (Major Error 1); in 146 (4%), multiple causal sequences were given in part I (Major Error 2); in 800 (19%), a single causal sequence was given but was not specific enough (Minor Error 1); and in 363 (9%), a single causal sequence was given but the order was incorrect (Minor Error 2). Multiple logistic regression analyses revealed that the probability of error in death certification increased as the age of the deceased increased, the age of the certifier decreased, and the level of the hospital decreased. These findings suggest that training in death certificate completion should focus on younger certifiers and those working at lower level teaching hospitals and nonteaching hospitals. Given the high rate of Minor Error 1, physicians should be reminded to state information as specifically as possible to render cause-of-death statistics more informative.

Introduction

Cause-of-death statistics derived from death certificates are an important tool for health planning and epidemiologic studies. To ensure valid comparisons within and among countries, the World Health Organization (WHO) has recommended a standard cause-of-death diagnosis form to be used on death certificates [1]. Failure to follow the requirements of the standard form would be expected to adversely affect the accuracy and usefulness of cause-of-death statistics.

Many instruction handbooks and articles are available to teach physicians how to complete death certificates correctly 2, 3, 4, 5, 6, 7, 8. Nevertheless, previous studies showed that the error rates in death certification completion are still very high, ranging from 25% to 78% in hospital-based studies 9, 10, 11, 12, 13, 14, 15, and from 16% to 56% in population-based studies 16, 17, 18. The reasons behind these high error rates, however, remain obscure.

Only five of the above studies investigated factors affecting the likelihood of errors 12, 14, 16, 17, 18. Most of these studies assessed only one factor (i.e., the type of doctor) and found that general practitioners made fewer errors than hospital doctors 12, 17, 18. Two studies evaluated more than one factor, but did not use a multivariate analysis technique 14, 16. One of these, a hospital-based study in Taiwan revealed that the age of the deceased and the seniority and specialty of the certifying physician were significantly related to the error rate [14]. In a population-based study in Western Australia, the error rate did not vary significantly between city and county areas, or between teaching hospitals and other locations [16].

Without knowing the factors that affect error rates in death certificate completion, it is hard to design a relevant intervention program to improve the quality of death certification. Thus, we performed a national population-based study to determine the frequencies of various types of error in death certificate completion in Taiwan and to identify characteristics of the certifying physicians and the deceased that are associated with these errors.

This study was part of a project evaluating the quality of cause-of-death statistics in Taiwan. In the first phase of the project, we examined the accuracy of coding and selection of the underlying cause of death by coders and their effects on national mortality, as reported earlier [19]. The present study used the same data set as our previous investigation, but focused on the manner in which the specificity and the sequence of morbid events is reported on the medical certificate by certifying physicians.

Section snippets

Methods

All original death certificates in Taiwan are centralized to the Provincial department of Health, Office of Statistics. In 1994, 112,238 death certificates were issued. Systematic sampling (one in every 20) of these yielded 5621 death certificates. After excluding 1498 in which the underlying cause-of-death was not natural or the name and license number of the certifying physician could not be identified, 4123 death certificates were left for analysis.

Two authors (T.H.L. and C.K.L.)

Frequencies of errors

Of the 4123 death certificates reviewed in this study, 2525 (61%) were completed correctly. In 289 (7%), only the mechanism(s) of death was given (Major Error 1); in 146 (4%), multiple causal sequences were given (Major Error 2); in 800 (19%), a single causal sequence was given but was not specific enough (Minor Error 1); and in 363 (9%), a single causal sequence was given but the order was incorrect (Minor Error 2).

Factors associated with errors

The rates of the various types of error according to the characteristics of the

Discussion

In this national population-based study, one third of the death certificates issued in 1994 were completed incorrectly. The most frequent type of error was Minor Error 1 (19%). Important factors associated with the error rates were the age of the deceased, the age of the certifier, and the level of the hospital.

Acknowledgements

This study was supported by the Department of Health, Republic of China (DOH85-TD-042). The authors wish to thank Show-Ru Shau for assisting with the data management and Jeff Radcliff for English revision of this manuscript.

References (36)

  • T Weeramanthri et al.

    Death certification in Western Australia—classification of major errors in certificate completion

    Aust J Public Health

    (1992)
  • M.C White et al.

    A study of multiple causes of death in California1955 and 1980

    J Clin Epidemiol

    (1989)
  • Manual of the international statistical classification of diseases, injuries, and causes of death. Vol 1

    (1977)
  • Medical certification of cause of deathinstructions for physicians on use of International Form of Medical Certificate of Cause of Death

    (1979)
  • Physicians' handbook on medical certification of death

    (1987)
  • R Hanzlick

    The medical cause of death manualinstructions for writing cause of death statements for deaths due to natural causes

    (1994)
  • T Kircher et al.

    Cause of deathproper completion of the death certificate

    JAMA

    (1987)
  • E.C Klatt et al.

    Death certificationpurposes, procedures, and pitfalls

    West J Med

    (1989)
  • R Hanzlick

    Protocol for writing cause-of-death statements for death due to natural causes

    Arch Intern Med

    (1996)
  • B.P Megrane et al.

    Certification of death by family physicians

    Am Fam Physician

    (1997)
  • S Leadbeatter

    Semantics of death certification

    J R Coll Phys London

    (1986)
  • R.E Zumwalt et al.

    Incorrect death certification—an invitation to obfuscation

    Postgrad Med

    (1987)
  • R Fernando

    Medical certification of cause of death in the general hospital, Colombo

    Ceylon Med J

    (1990)
  • J.M Jordan et al.

    Errors in death certificate completion in a teaching hospital

    Clin Invest Med

    (1993)
  • D.N Slater

    Certifying the cause of deathan audit of wording inaccuracies

    J Clin Pathol

    (1993)
  • T.H Lu et al.

    Analysis of formative errors and validity of cause-of-death diagnosis in a teaching hospital

    Chin J Public Health (Taipei)

    (1996)
  • K.A Myers et al.

    Improving the accuracy of death certification

    Can Med Assoc J

    (1998)
  • A Armour et al.

    Nosological inaccuracies in death certification in Northern Ireland

    Ulster Med J

    (1997)
  • Cited by (86)

    • Common errors in reporting cause-of-death statement on death certificates: A systematic review and meta-analysis

      2021, Journal of Forensic and Legal Medicine
      Citation Excerpt :

      Therefore, the results are not based on the same criteria and the findings of different studies are not completely comparable. Although there are variable ways to stratify DC completion errors, in general, errors are categorized into two main types: minor and major errors.9,11,16–18,20,21,27–31 The latter refers to errors that have a serious impact on the selection and classification of the underlying cause-of-death (UCOD) and the former refers to errors that have little effect on classifying and selecting the underlying cause, although the absence of such errors can help in more precise determination and classification of the underlying etiology 11, 17, 26, 32, 33.

    View all citing articles on Scopus
    View full text