From mono to multi-causality: Towards a comprehensive perspective on understanding death.

Cause-of-death statistics are an age-old source of information for health policy and medical research. In these statistics, the presentation of data is based on the idea of an underlying cause of death, i.e. one ("the") cause of death per deceased. This idea reflects an 18th Century causal thinking and is less and less applicable to contemporary patterns of dying in high income countries with an aging population suffering from chronic diseases and multi- or comorbidity at the end of life. Therefore, today's clinical reality calls for an innovation of cause-of-death statistics. For this, I will consider contemporary philosophical ideas on causality and their application to death. I will argue multi-causality is a more comprehensive way to understand death than mono-causality, implying a change of perspective with regard to current cause-of-death statistics.

Risk of death due to COVID-19 among current and former smokers in the Netherlands: a population-based quasi-cohort study.

BACKGROUND: Research on smoking as a risk factor for death due to COVID-19 remains inconclusive, with different studies demonstrating either an increased or decreased risk of COVID-19 death among smokers. To investigate this controversy, this study uses data from the Netherlands to assess the relationship between smoking and death due to COVID-19. METHODS: In this population-based quasi-cohort study, we linked pseudonymized individual data on smoking status from the 2016 and 2020 'Health Monitor Adults and Elderly' in the Netherlands (n = 914 494) to data from the cause-of-death registry (n = 2962). Death due to COVID-19 in 2020 or 2021 was taken as the main outcome. Poisson regression modelling was used to calculate relative risks (RRs) and 95% CIs of death due to COVID-19 for current and former smokers compared with never smokers while adjusting for relevant confounders (age, sex, educational level, body mass index and perceived health). RESULTS: Former smokers had a higher risk of death due to COVID-19 compared with never smokers across unadjusted (RR, 2.22; 95% CI, 2.04-2.42), age-sex-adjusted (RR, 1.38; 95% CI, 1.22-1.55) and fully adjusted (RR, 1.30; 95% CI, 1.16-1.45) models. Current smokers had a slightly higher risk of death due to COVID-19 compared with never smokers after adjusting for age and sex (RR, 1.21; 95% CI, 1.00-1.48) and after full adjustment (RR, 1.08; 95% CI, 0.90-1.29), although the results were statistically non-significant. CONCLUSIONS: People with a history of smoking appear to have a higher risk of death due to COVID-19. Further research is needed to investigate which underlying mechanisms may explain this.

The Mortality of the Spanish Flu and COVID-19 in The Netherlands: a Historical Comparison.

The COVID-19 pandemic has been called the deadliest disease event in history. In this study, we compared the cause specific mortality of the Spanish flu (1918-1920) with the cause specific mortality of COVID-19 (2020-2022) in the Netherlands. During the period of exposure, around 50,000 people died from COVID-19 and 32,000 people from the Spanish flu. In absolute numbers, COVID-19 seems to be deadlier than Spanish flu. However, the crude mortality rates of COVID-19 and Spanish flu were respectively 287 and 486 per 100,000 inhabitants. Compared by an age standardized mortality, there would have been 28 COVID-19 and 194 Spanish flu related deaths in 1918-1920, or there would have been 214 Spanish flu and 98 COVID-19 related deaths in 2020-2022 per 100,000 inhabitants per year. Thus, taking the population differences into account, the Spanish flu would have been deadlier than COVID-19.

Estimating the lifetime risk of dementia using nationwide individually linked cause-of-death and health register data.

BACKGROUND: Previous estimates of the lifetime risk of dementia are restricted to older age groups and may suffer from selection bias. In this study, we estimated the lifetime risk of dementia starting at birth using nationwide integral linked health register data. METHODS: We studied all deaths in The Netherlands in 2017 (n = 147 866). Dementia was assessed using the cause-of-death registration, individually linked with registers covering long-term care, specialized mental care, dispensed medicines, hospital discharges and claims, and primary care. The proportion of deaths with dementia was calculated for the total population and according to age at death and sex. RESULTS: According to all data sources combined, 24.0% of the population dies in the presence of dementia. This proportion is higher for females (29.4%) than for males (18.3%). Using multiple causes of death only, the proportion with dementia is 17.9%. Sequential addition of long-term care and hospital discharge data increased the estimate by 4.0 and 1.5%-points, respectively. Further addition of dispensed medicines, hospital claims and specialized mental care data added another 0.6%-points. Among persons who die at age ≤65-70 years, the proportion with dementia is ≤6.2%. After age 70, the proportion rises sharply, with a peak of 43.9% for females and 33.1% for males at age 90-95 years. CONCLUSIONS: Around one-fourth of the Dutch population is diagnosed with dementia at some point in life and dies in the presence of dementia. It is a major challenge to arrange optimal care for this group.

The Role of Dementia as Cause of Death: Certifiers' Opinions versus Automated Coding.

BACKGROUND: Dementia is a major cause of death in many countries today. The way in which countries code causes of death determines the occurrence of dementia in statistics. The change over from manual to automated coding is accompanied by a 7-19% increase in the occurrence of dementia as the underlying cause of death. Because of this sudden change, researchers, physicians, policy makers, and press question the validity of the outcome of automated coding. Therefore, the role of dementia as a cause of death was investigated. METHODS: A questionnaire was sent to a random sample of 700 certifiers who mentioned "dementia" on a death certificate in the second half of 2017. They were asked questions about the role of dementia as a cause of death. For each certificate, the opinion of the certifier was compared with the outcome of automated coding. RESULTS: A response of 65% (n = 446) was obtained. The automated coding system selected dementia as the underlying cause of death 9.5% points (95% CI: 5.8-14.4%) more often than the certifier would do. This finding in the sample corresponded to an overestimation of dementia in the cause-of-death statistics with 22.7% (95% CI: 18-28%). Main reason for this overestimation was the selection of dementia as the underlying cause of death by the automated coding system, while it was noted as the contributory cause of death on part 2 of the death certificate by the certifier. CONCLUSION: For international comparisons of data on dementia as a cause of death, the outcome of automated coding can be used as the system adheres to international (ICD-10) guidelines and reduces coding variations in and between countries. However, for interpreting the local (national) impact of dementia as a cause of death, the opinion of the certifier should be taken into account.

Prevalence of diabetes mellitus at the end of life: An investigation using individually linked cause-of-death and medical register data.

AIMS: Although diabetes mellitus at the end of life is associated with complex care, its end-of-life prevalence is uncertain. Our aim is to estimate diabetes prevalence in the end-of-life population, to evaluate which medical register has the largest added value to cause-of-death data in detecting diabetes cases, and to assess the extent to which reporting of diabetes as a cause of death is associated with disease severity. METHODS: Our study population consisted of deaths in the Netherlands (2015-2016) included in Nivel Primary Care Database (Nivel-PCD; N = 18,162). The proportion of deaths with diabetes (Type 1 or 2) within the last two years of life was calculated using individually linked cause-of-death, general practice, medication, and hospital discharge data. Severity status of diabetes was defined with dispensed medicines. RESULTS: According to all data sources combined, 28.7% of the study population had diabetes at the end of life. The estimated end-of-life prevalence of diabetes was 7.7% using multiple cause-of-death data only. Addition of general practice data increased this estimate the most (19.7%-points). Of the cases added by primary care data, 76.3% had a severe or intermediate status. CONCLUSIONS: More than one fourth of the Dutch end-of-life population has diabetes. Cause-of-death data are insufficient to monitor this prevalence, even of severe cases of diabetes, but could be enriched particularly with general practice data.

The implementation of an automated coding system for cause-of-death statistics.

Background: The production of cause-of-death statistics requires the coding and selection of an underlying cause of death from death certificates. Nowadays, this is done manually in many countries around the world. However, automated coding systems have been available since the 1970s and more and more countries are switching from manual to automated coding. The introduction of an automated coding system is expected to change the coding process and its outcomes in a fundamental way. Therefore, we studied the implementation of such a system, called IRIS, in the Netherlands.Methods: We adapted the system to our situation step-by-step and measured the number of death certificates processed without any manual intervention. Medical coders analyzed and qualified death certificates that could not be processed by the system. We also performed a bridge (double) coding study on a set of death certificates so that we could compare the underlying cause of death assigned by IRIS with the underlying cause of death assigned by a medical coder to the same death certificate.Results: IRIS could handle 68.5% of the death certificates without any manual intervention. Of the (31.5%) rejected death certificates, the main reasons for rejection were the following: editing and/or coding problems (76%), not able to select an underlying cause of death (12%), ambiguous causal connection between codes (5%), a combination of the reasons mentioned (7%). In the bridge coding study, 78% of the death certificates coded by IRIS without any manual intervention showed exactly the same underlying cause of death (ICD-10, four digits) as death certificates coded manually.Conclusions: An automated coding system for causes of death reduces the workload for medical coders considerably. When editing or coding problems can be solved the system is expected to handle up to 85% of the death certificates without manual intervention. The performance of the system is strongly dependent on the quality of death certificates. A change from manual to automated coding brings about changes in the frequency of occurrence of major causes of death. Users of death statistics should be aware of these changes when studying trends in time or regional variations of causes of death.

A Novel Time Series Approach to Bridge Coding Changes with a Consistent Solution Across Causes of Death.

Revisions of the International Classification of Diseases (ICD) can lead to biases in cause-specific mortality levels and trends. We propose a novel time series approach to bridge ICD coding changes which provides a consistent solution across causes of death. Using a state space model with interventions, we performed time series analysis to cause-proportional mortality for ICD9 and ICD10 in the Netherlands (1979-2010), Canada (1979-2007) and Italy (1990-2007) on chapter level. A constraint was used to keep the sum of cause-specific interventions zero. Comparability ratios (CRs) were estimated and compared to existing bridge coding CRs for Italy and Canada. A significant ICD9 to ICD10 transition occurred among 13 cause of death groups in Italy, 7 in Canada and 3 in the Netherlands. Without the constraint, all-cause mortality after the classification change would be overestimated by 0.4 % (NL), 0.03 % (Canada) and 0.2 % (Italy). The time series CRs were in the same direction as the bridge coding CRs but deviated more from 1. A smooth corrected trend over the ICD-transition resulted from applying the time series approach. Comparing the time series CRs for Italy (2003), Canada (1999) and the Netherlands (1995) revealed interesting commonalities and differences. We demonstrated the importance of adding the constraint, the validity of our methodology and its advantages above earlier methods. Applying the method to more specific causes of death and integrating medical content to a larger extent is advocated.

The reliability of cause-of-death coding in The Netherlands.

Cause-of-death statistics are a major source of information for epidemiological research or policy decisions. Information on the reliability of these statistics is important for interpreting trends in time or differences between populations. Variations in coding the underlying cause of death could hinder the attribution of observed differences to determinants of health. Therefore we studied the reliability of cause-of-death statistics in The Netherlands. We performed a double coding study. Death certificates from the month of May 2005 were coded again in 2007. Each death certificate was coded manually by four coders. Reliability was measured by calculating agreement between coders (intercoder agreement) and by calculating the consistency of each individual coder in time (intracoder agreement). Our analysis covered an amount of 10,833 death certificates. The intercoder agreement of four coders on the underlying cause of death was 78%. In 2.2% of the cases coders agreed on a change of the code assigned in 2005. The (mean) intracoder agreement of four coders was 89%. Agreement was associated with the specificity of the ICD-10 code (chapter, three digits, four digits), the age of the deceased, the number of coders and the number of diseases reported on the death certificate. The reliability of cause-of-death statistics turned out to be high (>90%) for major causes of death such as cancers and acute myocardial infarction. For chronic diseases, such as diabetes and renal insufficiency, reliability was low (<70%). The reliability of cause-of-death statistics varies by ICD-10 code/chapter. A statistical office should provide coders with (additional) rules for coding diseases with a low reliability and evaluate these rules regularly. Users of cause-of-death statistics should exercise caution when interpreting causes of death with a low reliability. Studies of reliability should take into account the number of coders involved and the number of codes on a death certificate.