Reliability of cause of death due to violence from information systems in Belo Horizonte, Southern Brazil

OBJECTIVE: To determine reliability of data encoding for death due to violence; to assess the agreement between cause of death in the Mortality Information System and cause selected from the Forensic Medicine Institute database; and to assess the impact of adding information of non-specified injuries and undetermined death events subsequently obtained from Forensic Medicine Institute in the mortality statistics due to violence. METHODS: A random sample of 411 death certificates due to violence was obtained in Belo Horizonte, Southeastern Brazil, between 1998 and 2000. Based on data from death certificates and Forensic Medicine Institute database, causes of death were coded and the agreement between this information and that from Mortality Information System was assessed. Also, in all certificates including "non-specified injury" and "undetermined death events," the impact of adding information from Forensic Medicine Institute was assessed in the classification of cause of death. RESULTS: Coding agreement was significant (Kappa=0.782; 95% CI: 0.744; 0.819) and of the underlying cause was moderate to significant (Kappa=0.602; 95% CI: 0.563; 0.641). There were 12.9% and 5.7% misclassification of suicides and murders, respectively, for those causes classified as "non-specified injury" and "undetermined death events," which were overall reduced to 47.3% and 59.8% respectively. CONCLUSIONS: There is a need for further improving the process of underlying cause coding and selection. Also medical examiners need to provide more complete death certificates and medical and police information provided with bodies for Forensic Medicine Institute autopsy should be more complete, especially in those cases of road traffic injuries and falls.     

Reasons for the Increasing Hispanic Infant Mortality Rate: Florida, 2004–2007

Assess whether the 55% increase in Florida's Hispanic infant mortality rate (HIMR) during 2004-2007 was real or artifactual. Using linked data from Florida resident live births and infant deaths for 2004-2007, we calculated traditional (infant Hispanic ethnicity from death certificates and maternal Hispanic ethnicity from birth certificates) and nontraditional (infant and maternal Hispanic ethnicity from birth certificate maternal ethnicity) HIMRs. We assessed trends in HIMRs (per 1,000 live births) using Chi-square statistics. We tested agreement in Hispanic ethnicity after implementation of a revised 2005 death certificate by using kappa statistics and used logistic regression to test the associations of infant mortality risk factors. Hispanic was defined as being of Mexican, Puerto Rican, Cuban, Central/South American, or other/unknown Hispanic origin. During 2004-2007 traditional HIMR increased 55%, from 4.0 to 6.2 (Chi-square, P < 0.001) and nontraditional HIMR increased 20%, from 4.5 to 5.4 (Chi-square, P = 0.03). During 2004-2005, agreement in Hispanic ethnicity did not change with use of the revised certificate (kappa = 0.70 in 2004; kappa = 0.76 in 2005). Birth weight was the most significant risk factor for trends in Hispanic infant mortality (OR = 1.33, 95% CI = 1.10-1.61). Differences in Hispanic reporting on revised death certificates likely accounted for the majority of traditional HIMR increase, indicating a primarily artifactual increase. Reasons for the 20% increase in nontraditional HIMR during 2004-2007 should be further explored through other individual and community factors. Use of nontraditional HIMRs, which use a consistent source of Hispanic classification, should be considered.     

Inadequacies of death certification in Beirut: who is responsible?

OBJECTIVE: To assess the completeness of data on death certificates over the past 25 years in Beirut, Lebanon, and to examine factors associated with the absence of certifiers' signatures and the non-reporting of the underlying cause of death. METHODS: A systematic 20% sample comprising 2607 death certificates covering the 1974, 1984, 1994, 1997 and 1998 registration periods was retrospectively reviewed for certification practices and missing data. FINDINGS: The information on the death certificates was almost complete in respect of all demographic characteristics of the deceased persons except for occupation and month of birth. Data relating to these variables were missing on approximately 95% and 78% of the certificates, respectively. Around half of the certificates did not carry a certifier's signature. Of those bearing such a signature, 21.6% lacked documentation of the underlying cause of death. The certifier's signature was more likely to be absent on: certificates corresponding to the younger and older age groups than on those of persons aged 15-44 years; those of females than on those of males; those of persons who had been living remotely from the registration governorate than on those of other deceased persons; and those for which there had been delays in registration exceeding six months than on certificates for which registration had been quicker. For certificates that carried the certifier's signature there was no evidence that any of the demographic characteristics of the deceased person was associated with decreased likelihood of reporting an underlying cause of death. CONCLUSION: The responsibility for failure to report causes of death in Beirut lies with families who lack an incentive to call for a physician and with certifying physicians who do not carry out this duty. The deficiencies in death certification are rectifiable. However, any changes should be sensitive to the constraints of the organizational and legal infrastructure governing death registration practices and the medical educational systems in the country.     

The accuracy of occupation and industry data on death certificates

To assess the accuracy of occupation and industry data on death certificates, we compared the known occupation and industry of 3,789 individuals with information on their death certificates. All individuals were members of the National Institute for Occupational Safety and Health cohort studies and their occupation and industry were known from personnel or union records. We focused our analysis on 2,198 long-term workers who had 10 or more years of work experience (average, 18 years). Our principal finding was that for the long-term workers in our sample, the probability of the known occupation being listed on their death certificates was 64.7% (SE = 1.1) (for white workers this figure was 73.5% (SE = 1.2). Furthermore, for the long-term workers the probability of the plant or industry being named on their death certificates was 70.1% (SE = 1.2). Women and non-whites had lower matching rates. For women the probability of a matched listing was 45.1% for occupation and 60.4% for industry. For nonwhites, it was only 30.1% for occupation and 49.2% for industry. Our results for white long-term workers generally agree with the results of previous investigators who have used different methods.     

An assessment of occupation and industry data from death certificates and hospital medical records for population-based cancer surveillance

This study analyzed 30,194 incident cases and 4,301 death certificates for completeness of occupational reporting. Analysis of data accuracy was based upon a comparison of more than 2,000 death certificates with incident abstracts and 352 death certificates with interview data. Death certificates had a higher proportion with occupation (94.3%) and industry (93.4%) reported than did incident abstracts of hospital medical records (39.0% and 63.5%, respectively). Compared with occupational history data obtained by interview, 76.1% of the death certificates were exact matches for usual occupation and industry.     

A computer system for coding occupation

BACKGROUND: Occupation information is widely used in epidemiologic studies and is collected on most death certificates and many birth certificates in the United States. Coding the massive amount of occupation information collected has been a challenge. METHODS: A simple word-matching computer program to code occupation entries from vital records was developed. The accuracy of the program was evaluated by comparing its output to codes assigned by human coders. RESULTS: In routine use in the Washington State Department of Health (DOH), the computer system codes 96-97% of the occupation entries on birth and death records. It assigned the correct code on 89% (95% confidence interval (87%, 91%)) of the records it coded. CONCLUSIONS: The occupation-coding program is both efficient and accurate and can simplify the process of coding occupation entries from vital records. The system is adaptable and can be modified to use occupation classifications other than the one used by DOH.     

Using death certificates and medical examiner records for adolescent occupational fatality surveillance and research: a case study

Death certificates and medical examiner records have been useful yet imperfect data sources for work-related fatality research and surveillance among adult workers. It is unclear whether this holds for work-related fatalities among adolescent workers who suffer unique detection challenges in part because they are not often thought of as workers. This study investigated the utility of using these data sources for surveillance and research pertaining to adolescent work-related fatalities. Using the state of North Carolina as a case study, we analyzed data from the death certificates and medical examiner records of all work-related fatalities data among 11- to 17-year-olds between 1990-2008 (N = 31). We compared data sources on case identification, of completeness, and consistency information. Variables examined included those on the injury (e.g., means), occurrence (e.g., place), demographics, and employment (e.g., occupation). Medical examiner records (90%) were more likely than death certificates (71%) to identify adolescent work-related fatalities. Data completeness was generally high yet varied between sources. The most marked difference being that in medical examiner records, type of business/industry and occupation were complete in 72 and 67% of cases, respectively, while on the death certificates these fields were complete in 90 and 97% of cases, respectively. Taking the two sources together, each field was complete in upward of 94% of cases. Although completeness was high, data were not always of good quality and sometimes conflicted across sources. In many cases, the decedent's occupation was misclassified as "student" and their employer as "school" on the death certificate. Even though each source has its weaknesses, medical examiner records and death certificates, especially when used together, can be useful for conducting surveillance and research on adolescent work-related fatalities. However, extra care is needed by data recorders to ensure that occupation and employer are properly coded when dealing with adolescent worker deaths.     

Occupation and industry on death certificates of long-term chemical workers: concordance with work history records

This study evaluated the concordance between occupation and industry listed on death certificates with actual work history information for a group (n = 5,882) of long-term (10 years or more) workers at a chemical company. Match rates were calculated as the percent of death certificate occupation and company entries that were confirmed by work history data using 3-digit 1980 U.S. Census Bureau group codes. The concordance rate for industry differed by employment status at death: employed, 94.9%; inactive, 30.8%; and retired, 91.1%. Concordance on occupation was analyzed for employed (n = 467) and retired (n = 932) subjects who had computerized work histories (randomly done prior to the study) and who had matched on the company on the death certificate. Concordance ranged from 0 to 50% for the first job, to 50 to 70% for the last job, longest job, and longest job in the last 10 years of company employment. The most consistent predictor of concordance was job duration. Misclassification was reviewed by occupational category. Results from this and other investigations lead to the inevitable conclusion that usual occupation data from death certificates are grossly inadequate for studies of occupational risks.     

Accuracy of race coding on American Indian death certificates,Montana 1996-1998

OBJECTIVES: The purpose of this study was to describe the consistency of coding of American Indians on Montana death certificates and to identify the characteristics of American Indians in Montana associated with consistent classification on death certificates. METHODS: The Billings Area Indian Health Service (IHS) patient registration file was linked with Montana Department of Health and Human Services death certificate files for 1996-1998. RESULTS: A total of 769 Montana residents who had died in 1996-1998 were matched to the IHS registration file. Of these decedents, 696 (91%) were consistently classified as American Indian on the death certificate. Seventy-two (99%) of the 73 decedents not classified as Indian were classified as white. American Indians living in counties on or near the seven Montana reservations were more likely to be consistently classified than Indians living in other counties (95% vs. 70%); those with less than 12 years of education (93% vs. 88%) were more likely to be consistently classified than those with 12 or more years of education. Decedents whose cause of death was suicide were less likely than those with other causes of death to be consistently classified (72% vs. 95%). In contrast, a higher percentage of those with an alcohol-related cause of death than of those with other causes of death were consistently classified, although this difference was not statistically significant. CONCLUSIONS: The mortality rates for Montana American Indians are underestimated overall, and are differentially under- and overestimated for selected conditions.     

Agreement on cause of death between proxies, death certificates, and clinician adjudicators in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) study

Death certificates may lack accuracy and misclassify the cause of death. The validity of proxy-reported cause of death is not well established. The authors examined death records on 336 participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study, a national cohort study of 30,239 community-dwelling US adults (2003-2010). Trained experts used study data, medical records, death certificates, and proxy reports to adjudicate causes of death. The authors computed agreement on cause of death from the death certificate, proxy, and adjudication, as well as sensitivity and specificity for certain diseases. Adjudicated cause of death had a higher rate of agreement with proxy reports (73%; Cohen's kappa (κ) statistic = 0.69) than with death certificates (61%; κ = 0.54). The agreement between proxy reports and adjudicators was better than agreement with death certificates for all disease-specific causes of death. Using the adjudicator assessments as the "gold standard," for disease-specific causes of death, proxy reports had similar or higher specificity and higher sensitivity (sensitivity = 50%-89%) than death certificates (sensitivity = 31%-81%). Proxy reports may be more concordant with adjudicated causes of death than with the causes of death listed on death certificates. In many settings, proxy reports may represent a better strategy for determining cause of death than reliance on death certificates.     

Risk factors associated with the classification of unspecified and/or unexplained causes of death in an occupational cohort.

In a matched case-control study of an occupational cohort in East Tennessee, data from the death certificates of 608 cases and controls were abstracted and analyzed in order to investigate possible risk factors associated with the certification on the death certificate of an ill-defined cause of death. There was a very strong association between the ill-defined classification and the certifier, especially if the certifier was a medical examiner (OR = 10.4, 95% CI: 6.0, 18.0).     

Comparing Swedish hospital discharge records with death certificates: implications for mortality statistics

BACKGROUND: The quality of mortality statistics is of crucial importance to epidemiological research. Traditional editing techniques used by statistical offices capture only obvious errors in death certification. In this study we match Swedish hospital discharge data to death certificates and discuss the implications for mortality statistics. METHODS: Swedish death certificates for 1995 were linked to the national hospital discharge register. The resulting database comprised 69 818 individuals (75% of all deaths), 39 872 (43%) of whom died in hospital. The diagnostic statements were compared at Basic Tabulation List level. RESULTS: The last main diagnosis and the underlying cause of death agreed in 46% of cases. Agreement decreased rapidly after discharge. For hospital deaths, the main diagnosis was reported on 83% of the certificates, but only on 46% of certificates for non-hospital deaths. Malignant neoplasms and other dramatic conditions showed the best agreement and were often reported as underlying causes. Conditions that might follow from some other disease were often reported as contributory causes, while symptomatic and some chronic conditions were often omitted. In 13% of cases, an ill-defined main condition was replaced by a more specific cause of death. CONCLUSIONS: There is no apparent reason to question the death certificate if the main diagnosis and underlying cause agree, or if the main diagnosis is a probable complication of the stated underlying cause. However, cases in which the main diagnosis cannot be considered a complication of the reported underlying cause should be investigated, and assessments made of the feasibility and cost-effectiveness of routinely linking hospital records to death certificates.     

A comparison of occupational data from death certificates and interviews

A comparison was made of the occupational data reported on the death certificates of 586 men with their employment history obtained by interviews. Agreement was assessed for 19 occupational and 14 industrial categories of usual employment, with the highest levels of concordance (greater than or equal to 80%) found for agricultural, medical, and public administration activities. Between the two sources of information, there was overall agreement of 56% for usual occupation and 51% for usual industry of employment. Concordance was highest among the 68 self-respondents (usual occupation 66%; usual industry 53%). Among the 518 surrogates, spousal agreement was highest (58% for occupation and 51% for industry). For other surrogate types, agreement was 49% for both industry and occupation. Agreement varied by duration of employment and by level of education, with concordance tending to increase as length of employment and educational attainment rose. These relationships remained when examined by respondent type. Evaluation of agreement levels by age and other study subject characteristics showed little effect on concordance. Review of verbatim data from the interviews and death certificates revealed that most disagreements could be attributed to coding problems caused by vague or misleading information on the death certificates, although some disconcordance was due to uncodable and missing information in the interview history. Based on results from this and prior studies, the value of occupational data derived from death certificates in epidemiologic studies may be limited, although the addition of explicit instructions on the death certificate itself may aid in providing more useful and complete information for usual employment.     

Berufsbezogene sozio-ökonomische Gruppen für die Schweiz: sozialwissenschaftliche Grundlagen und Untersuchungen zur Validität

A system of 15 socio-economic groups (SOEG) based on occupation and position within occupation is proposed and discussed as a basis for statistical analyses of mortality data. In order to check the validity of this classification, a random sample of 3058 male death certificates is compared with the information obtained from the corresponding 1980 census records. The comparison reveals that most of the SOEGs have satisfactory validity. The same is true for a model emulating the widely used English system of Social Classes (SC). Estimates of the differences in mortality between SOEGs or SCs can be improved significantly applying correction factors derived from the sample of linked death certificates and census records.     

Reliability and accuracy of reported causes of death from cancer. I. Reliability of all cancer reported in the State of Rio de Janeiro, Brazil

Mortality records are often used in epidemiological studies, particularly in cancer studies. This paper aims to evaluate reliability and accuracy of cancer mortality data in Rio de Janeiro, Brazil. A systematic random sample of 394 death certificates was obtained from a total of 12615 cancer deaths. This sample was recoded by an independent codifier. A kappa coefficient of 0.89 (95% C.I. 0.86-0.92) was obtained to the third digit, which increases to 0.95 (95% C.I. 0.94-0.96) when restricted to the mortality list used in international publications. The positive predictive value was 95.7% for this sample. These results reveal a high standard reliability of cancer mortality records in the State of Rio de Janeiro making them suitable for use in epidemiological research.     

Suicide and Potential Occupational Exposure to Pesticides,Colorado 1990–1999

Abstract

A number of occupational studies have reported high rates of suicide among selected occupations, including farmers. Limited work has focused on occupational exposures that may increase the risk of suicide. The purpose of this study is to describe suicide among individuals potentially exposed to pesticides through their occupation. Data from Colorado death certificate files for the period 1990–1999 were obtained. Eligible records were those individuals who were Colorado residents at the time of death who had an occupation listed on their death certificates. Cases had suicide listed as the primary cause of death on the death certificates. The comparison group included Colorado residents who died from any cause during the same period other than cancer, mental disorders and injuries. A total of 4,991 suicide deaths were included and a total of 107,692 other deaths served as the comparison group. Occupations considered pesticide exposed included: veterinarians; pest control occupations; farmers and farm workers; farm managers and supervisors; marine life cultivators; nursery workers; groundskeepers and gardeners; animal caretakers; graders, sorters and inspectors of agricultural products; and forestry workers, supervisors and loggers. All other occupational categories were coded as unexposed. Logistic regression was used to compare the groups, separately for males and females. After controlling for age, race, Hispanic ethnicity, years of education, and marital status, males who were in pesticide exposed occupations had higher odds of suicide (odds ratio 1.14; 95% confidence interval 0.97, 1.34) and females in pesticide exposed occupations also had higher odds of suicide (odds ratio 1.98; 95% confidence interval 1.01, 3.88).     

Why race is differentially classified on U.S. birth and infant death certificates: an examination of two hypotheses

Among U.S. infants who die within a year of birth, classification of race on birth and death certificates may differ. I investigate two hypotheses: (1) The race of infants of different-race parents is more likely to be differentially classified at birth and death than the race of infants of same-race parents. (2) States with a greater proportion of infant deaths of a given race are less likely to differentially classify infants of that race on birth and death certificates than states with a smaller proportion of infant deaths of that race. Using the Linked Birth/Infant Death data tape for 1983-1985, I assessed the first hypothesis by comparing rates of differential classification for infants with different-race parents and same-race parents. To assess the second hypothesis, I examined the correlations between the proportion of infant deaths of each race in each state and the proportion of infants of that race consistently classified. Differential racial classification on birth and death certificates was more than 31 times as likely with different-race than with same-race parents. The second hypothesis was confirmed for white, black, American Indian, and Japanese infants. As the U.S. population becomes more heterogeneous, attention to these methodologic issues becomes increasingly critical for the measurement and redress of differential racial health status.     

Risk for commercial fishing deaths in Canadian Atlantic provinces.

The risk of mortality related to occupation was determined for commercial fishermen in the Canadian Atlantic coast provinces of Nova Scotia, New Brunswick, and Prince Edward Island. The subjects were a cohort of 31,415 fishermen licensed by the Canadian Department of Fisheries during 1975-83. Mortality and cause of death were obtained from the Canada Mortality Data Base and the Marine Casualty Investigation Unit (MCI), and were confirmed by examination of death certificates. Eighty four deaths likely to be related to fishing were recorded over 183,378 person-years of exposure for an annual mortality of 45.8 (95% confidence interval (CI) 36.0-55.6) per 100,000 fishermen. The rate of potential years of life lost up to age 75 was 1583 per 100,000 person-years of exposure. Inclusion of 14 additional deaths, which were possibly related to occupation, would increase these rates further. Bias in this study is likely to underestimate the risks. It is concluded that fishing is one of the most hazardous occupations in terms of mortality related to work.     

Risk for commercial fishing deaths in Canadian Atlantic provinces

The risk of mortality related to occupation was determined for commercial fishermen in the Canadian Atlantic coast provinces of Nova Scotia, New Brunswick, and Prince Edward Island. The subjects were a cohort of 31,415 fishermen licensed by the Canadian Department of Fisheries during 1975-83. Mortality and cause of death were obtained from the Canada Mortality Data Base and the Marine Casualty Investigation Unit (MCI), and were confirmed by examination of death certificates. Eighty four deaths likely to be related to fishing were recorded over 183,378 person-years of exposure for an annual mortality of 45.8 (95% confidence interval (CI) 36.0-55.6) per 100,000 fishermen. The rate of potential years of life lost up to age 75 was 1583 per 100,000 person-years of exposure. Inclusion of 14 additional deaths, which were possibly related to occupation, would increase these rates further. Bias in this study is likely to underestimate the risks. It is concluded that fishing is one of the most hazardous occupations in terms of mortality related to work.     

Traffic accidents in a city in Southern Brazil: an evaluation of coverage and quality of data]

The aim of this study was to analyze police coverage and the validity of data on emergency and hospitalization records as well as on death certificates for traffic casualties in Londrina, Paraná State, Brazil. Victims (3,643) of road accidents during the first semester of 1996 were investigated and followed up after 180 days to confirm whether death was due to the accident. Police data recorded only 32.5% of the casualties, the coverage being higher for car occupants (71.6%) and lower for cyclists (8.1%) and pedestrians (24.8%). Agreement was low between original information and that derived from investigation of death certificates (Kappa coefficient 0.10; 95% CI: 0.02-0.17), fair for hospitalization records (Kappa coefficient 0.33; 95% CI: 0.27-0.40), and substantial for emergency records (Kappa coefficient 0.63; 95% CI: 0.61-0.65). Results suggest that police data underestimate the number of traffic casualties and that it is necessary to improve the validity of medical records.