Coding Of Causes Research Articles

Factors associated with the quality of death certification in Brazilian municipalities: A data-driven non-linear model.

Studies evaluating the local quality of death certification in Brazil focused on completeness of death reporting or inappropriate coding of causes of death, with few investigating missing data. We aimed to use missing and unexpected values in core topics to assess the quality of death certification in Brazilian municipalities, to evaluate its correlation with the percentage of garbage codes, and to employ a data-driven approach with non-linear models to investigate the association of the socioeconomic and health infrastructure context with quality of death statistics among municipalities. This retrospective study used data from the Mortality Information System (2010-2017), and municipal data regarding healthcare infrastructure, socioeconomic characteristics, and death rates. Quality of death certification was assessed by missing or unexpected values in the following core topics: dates of occurrence, registration, and birth, place of occurrence, certifier, sex, and marital status. Models were fit to classify municipalities according to the quality of death certification (poor quality defined as death records with missing or unexpected values in core topics ≥ 80%). Municipalities with poor quality of death certification (43.9%) presented larger populations, lower death rates, lower socioeconomic index, healthcare infrastructure with fewer beds and physicians, and higher proportion of public healthcare facilities. The correlation coefficients between quality of death certification assessed by missing or unexpected values and the proportion of garbage codes were weak (0.11-0.49), but stronger for municipalities with lower socioeconomic scores. The model that best fitted the data was the random forest classifier (ROC AUC = 0.76; precision-recall AUC = 0.78). This innovative way of assessing the quality of death certification could help quality improvement initiatives to include the correctness of essential fields, in addition to garbage coding or completeness of records, especially in municipalities with lower socioeconomic status where garbage coding and the correctness of core topics appear to be related issues.

Formulation of forensic and pathoanatomical diagnoses, selection, and coding of causes of death according to ICD-10, edited by Professor V.A. Klevno and Professor O.V. Zairat’yants

The article analyzed the book Rules for the formulation of forensic and pathoanatomical diagnoses, selection, and coding of causes of death according to ICD-10, edited by Doctor of Medical Sciences, Professor Vladimir Aleksandrovich Klevno and Doctor of Medical Sciences, Professor Oleg Vadimovich Zayratyants.
 The review aimed to provide an objective assessment of the information presented in a scientific publication and determine their value for science and practice.
 The authors note that this publication is essentially a national guide for doctors in the Russian Federation on the formulation of forensic and pathoanatomical diagnoses, choice and coding of causes of death in accordance with the requirements of the 10th revision of the International Classification of Diseases (ICD-10), and issuance of a medical death certificate to improve statistics on population mortality. The authors noted that the manual contains unified rules for the formulation of forensic and pathoanatomical diagnoses, selection and coding of causes of death, registration of the medical death certificate in accordance with the provision of the Federal Law of November 21, 2011, No. 323-FZ On the Fundamentals of Protecting the Health of Citizens in the Russian Federation, requirements of ICD-10, and orders and guidelines of the Ministry of Health of Russia, and clinical and guidelines of the Russian Society of Pathologists and the Association of Forensic Experts.
 With absolute certainty, the book will retain its practical significance for a long time. It complies with all regulatory requirements, significantly simplifies the work of a doctor, and may well become a reference book for a practicing specialist.

Modeling of Causes and Consequences of Human Error in Mining Processes Design: A Qualitative Study

Given the significant role of mining in sustainable development and its intrinsic characteristics, the hazards and potential consequences are a great concern for the industry. A design error is one of the main reasons behind accidents and environmental disasters. This study aims to identify and categorize effective factors influencing design errors and their health, safety, and environmental consequences. The study was carried out based on the theme analysis of 12 Iranian surface miners’ opinions from 14 October to 25 December 2021. The data were collected using semi-structured interviews. The data analysis procedure was conducted based on the Strauss Model using MAXQDA2022. In the open coding section, 120 and 146 primary codes were identified regarding causes and consequences, respectively. As for the codes for causes, 26 main categories and five subcategory codes were identified, including organizational, personal, environmental, occupational, and external factors. As for the identified codes for consequences, 11 subcategories and three main categories were identified, including safety, health, and environmental effects. The findings of the study revealed that among causes, the external factor (p = 0.3703) had the weakest, and the personal factor (p = 0.003) had the strongest correlations with human error in design. In line with the opinion of the expert participants, design error had significant relationships with safety (p = 0.002), environmental (p = 0.01), and health effects (p = 0.034). The cause-consequence model introduced in this study can help many organizations, particularly surface mines, to provide a good basis for achieving sustainable safety, health management, and sustainable development.

Transformer-based models for ICD-10 coding of death certificates with Portuguese text

Natural Language Processing (NLP) can offer important tools for unlocking relevant information from clinical narratives. Although Transformer-based models can achieve remarkable results in several different NLP tasks, these models have been less used in clinical NLP, and particularly in low resource languages, of which Portuguese is one example. It is still not entirely clear whether pre-trained Transformer models are useful for clinical tasks, without further architecture engineering or particular training strategies. In this work, we propose a BERT model to assign ICD-10 codes for causes of death, by analyzing free-text descriptions in death certificates, together with the associated autopsy reports and clinical bulletins, from the Portuguese Ministry of Health. We used a novel pre-training procedure that incorporates in-domain knowledge, and also a fine-tuning method to address the class imbalance issue. Experimental results show that, in this particular clinical task that requires the processing of relatively short documents, Transformer-based models can achieve very strong results, significantly outperforming tailored approaches based on recurrent neural networks.

High Error Rates in Coding Causes of Death in Adults with Congenital Heart Disease

Mortality statistics calculated from death certificates are vital for national and global health systems planning, informing strategic priorities and health care funding. As congenital heart disease (CHD) is often complex and associated with comorbidities, and coding of cause of death is often done by junior doctors and/or those unfamiliar with the decedent, we hypothesised that errors might be common in coding causes of death in adults with CHD.

Excess mortality in England during the COVID-19 pandemic

IntroductionExcess mortality monitoring is not dependent on completeness or consistency of coding of causes of death and captures deaths caused directly and indirectly by COVID-19. It takes account of existing inequalities in the baseline rates and shows the additional impact of COVID-19. We report weekly on excess mortality by age, gender, geographical area, deprivation, and ethnicity.MethodsA quasi-Poisson model was fitted to deaths data for 2015-2019 to estimate expected mortality in population sub-groups each week during the COVID-19 pandemic. This was compared with observed mortality reported each week. Information on deaths registered was obtained from death certificates. Ethnicity was obtained from hospital records linked to deaths data.ResultsEarly in the pandemic, excess deaths exceeded official reported COVID-19 deaths every week. Throughout the pandemic there were stark inequalities in excess death rates. Based on data from 21 March 2020 to 9 April 2021, people living in the most deprived areas in England experienced more excess (1.21 times; 95% CI 1.21-1.22) compared with people living in the least deprived areas (1.17; 1.16-1.18). Black (1.56; 1.53-1.58) and Asian groups (1.55; 1.53-1.56) experienced more excess than the White group (1.16; 1.16-1.17). There was wide variation between English regions.ConclusionsReporting overall excess mortality early in the pandemic is important to inform policy makers and public about the true scale of the death toll when coding is inconsistent and developing. This approach has highlighted the large increases in already established inequalities in mortality. It provides rapid surveillance of developing disparities for national and local decision makers. It has been used to inform policy around reducing disparities and to revise messaging to encourage hospital attendance for non- COVID-19 health problems.Key messages Excess deaths from all causes is a robust measure of overall impact, unaffected by coding consistency.COVID-19 has dramatically worsened established inequalities between ethnic groups and deprivation groups.

Cause of death for patients with breast cancer: discordance between death certificates and medical files, and impact on survival estimates

BackgroundRegistration and coding of cause of death is prone to error since determining the exact underlying condition leading directly to death is challenging. In this study, causes of death from the death certificates were compared to patients’ medical files interpreted by experts at University Hospitals Leuven (UHL), to assess concordance between sources and its impact on cancer survival assessment.MethodsBreast cancer patients treated at UHL (2009–2014) (follow-up until December 31st 2016) were included in this study. Cause of death was obtained from death certificates and expert-reviewed medical files at UHL. Agreement was calculated using Cohen’s kappa coefficient. Cause-specific survival (CSS) was calculated using the Kaplan-Meier method and the relative survival probability (RS) using the Ederer II and Pohar Perme method.ResultsA total of 2862 patients, of whom 354 died, were included. We found an agreement of 84.7% (kappa-value of 0.69 (95% C.I.: 0.62–0.77)) between death certificates and medical files. Death certificates had 10.7% false positive and 4.5% false negative rates. However, five-year CSS and RS measures were comparable for both sources.ConclusionFor breast cancer patients included in our study, fair agreement of cause of death was seen between death certificates and medical files with similar CSS and RS estimations.

Abstract PS7-63: Cause of death discordance between death certificates and medical files: Impact on cancer survival assessment in a Belgian case study

Abstract Introduction Accurate information on cause of death is essential for correct breast cancer-specific mortality assessment. However, registration and coding of cause of death is prone to error since determining the exact underlying condition related to the death is challenging. In this study, an expert review of medical files was done to determine the principal cause of death for breast cancer patients of a Belgian tertiary hospital. The retrieved cause of death was compared to death certificate information to assess concordance between both sources. Secondly, the impact of discordant reporting on cause-specific survival (CSS) and other net survival approaches were examined. Methods Breast cancer patients diagnosed and treated at University Hospitals Leuven (UHL) between 2009 and 2014 with follow-up until December 31st, 2016, were included in the study. Information on cause of death was obtained from death certificates (following ICD-10 rules) and medical files. The latter were reviewed by a board of experts at UHL. Agreement was calculated using Cohen’s kappa coefficient, and reasons for discordant reporting were assessed. CSS was calculated based on cause of death information from both sources using the Kaplan-Meier method. These survival estimates were compared to the relative survival probability (RS) using the Ederer II and Pohar Perme method. Results A total of 2,862 patients were included, of whom 354 died after a median follow-up of 54.6 months. We found overall substantial agreement (kappa-value of 0.69 (95% C.I.: 0.62-0.77)) between cause of death reported by death certificates and medical files (Table 1). In 84.8% of cases, there was concordance between both methods. When comparing to medical files, misattribution of breast cancer-specific death in death certificates (4.5% of cases) was linked to the presence of comorbidities (43.7%), metastases (37.5%), or unspecified causes (18.8%). Five-year CSS based on medical files (93.1% (95% C.I.: 91.9-94.1)) was only slightly higher compared to CSS based on death certificates (92.3% (95% C.I.: 91.2-93.4)). RS measures using Ederer II and Pohar Perme were comparable to CSS measures. Conclusions Overall, substantial agreement of cause of death was seen between death certificates and medical files. Attribution of cause death to comorbidities was the most common reason for discordant reporting of breast cancer-specific death. Five-year breast cancer-specific survival was slightly higher based on cause of death information from medical files, compared to death certificates. Periodic reviews and implementation of ICD-10 guidelines for classification of cause of death could improve accuracy in cause of death annotation. Table 1: Discordance for the principal cause of death between medical files and death certificatesmedical filesother causesbreast cancerdeath certificatesother causes136 (38.4%)16 (4.5%)152 (42.9%)breast cancer38 (10.7%)164 (46.3%)202 (57.1%)174 (49.2%)180 (50.9%)354 (100%) Citation Format: Hava Izci, Tim Tambuyzer, Jessica Vandeven, Jérôme Xicluna, Hans Wildiers, Kevin Punie, Nynke Willers, Eva Oldenburger, Els Van Nieuwenhuysen, Patrick Berteloot, Ann Smeets, Ines Nevelsteen, Liesbet Van Eycken, Harlinde De Schutter, Patrick Neven, Geert Silversmit, Freija Verdoodt. Cause of death discordance between death certificates and medical files: Impact on cancer survival assessment in a Belgian case study [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS7-63.

Improving the Quality of Clinical Coding through the Training of Health Records and Information Officers in Selected Hospitals, Nairobi City County, Kenya

Improving the Quality of Clinical Coding through the Training of Health Records and Information Officers in Selected Hospitals, Nairobi City County, Kenya

Analysis of Coding Quality of Death Certificates and Autopsy Rates in Chemnitz: 2010-2013

Postmortem examination as the last service of physicians has been increasingly criticized recently. The aim of this project was to study the influence of place of death, type of postmortem physician, age and sex of the deceased person on the completeness and quality of ICD-10 coding of death certificates issued by physicians in Chemnitz, Germany. We reviewed 15 612 death certificates and 353 autopsy reports of the health office of Chemnitz for the years 2010-2013. We extracted age and sex of the deceased, causes of death as based on the primary postmortem examination and after autopsy if applicable, type of death (natural or unnatural cause of death), and complications as discovered at the second postmortem examination before cremation. In addition, we documented the place of death and type of physician carrying out the postmortem examination. Furthermore, we corrected the inappropriate or missing ICD-10 codes followed by agreement analysis. The analysis was in agreement with the rules of the German Institute of Medical Documentation and Information. The autopsy rate was 2.3%. The underlying condition was not documented in 34.1% of the cases. The agreement between quality-controlled ICD-10 coding of the cause of death and the original ICD-10 coding (three-digit codes) was high (96.5%). High age at death, death not occurring in hospitals, and postmortem examination not carried out in hospitals were associated with a higher probability of ICD-10 miscoding. The observed agreement of 3-digit ICD-10 coded underlying condition after primary postmortem examination and after autopsy was 26.3%. The underlying cause of death contained garbage codes in 24.2%, that is, codes that are inappropriate for cause of death statistics. Physicians have to be sensitized to the need for appropriate documentation of causes of death. Autopsy very frequently resulted in a change of the assessment of the underlying cause of death and therefore remains an important quality control measure for certification on the cause of death. ICD-10 coding of causes of death is not a primary task of physicians and physicians should not be required by law to carry out this task.

MOSCOW REGION ANALYZES ACHIEVEMENTS OF TARGET INDICATORS OF MORTAL REDUCTION IN THE FRAMEWORK OF IMPLEMENTATION OF MAY DECREES OF THE PRESIDENT OF RUSSIA

The article focuses on the problems of mortality rate and the ways of its reducing in the Moscow Region as well as correct coding of causes of death according to ICD-10.

Causes of renal allograft failure in the UK: trends in UK Renal Registry and National Health Service Blood and Transplant data from 2000 to 2013.

Improvement in long-term renal allograft survival is impeded by incomplete or erroneous coding of causes of allograft loss. This study reports 13-year trends in causes of graft failure across the UK. National Health Service Blood and Transplant (NHSBT) and UK Renal Registry data were linked to describe UK kidney patients transplanted in 2000-13. NHSBT graft failure categories were used, with 'other' recoded when free text was available. Adjusted analyses examined the influence of age, ethnicity and donor type on causes of graft failure. In 22730 recipients, 5389 (23.7%) grafts failed within a median follow-up of 5years. The two most frequent causes were death with a functioning graft (40.8%) and alloimmune pathology (25.0%). Graft survival was higher in recipients who were younger (mean 47.3 versus 50.7 years), received a pre-emptive transplant (20.2% versus 10.4%), spent less time on dialysis (median 1.6 versus 2.4 years) and received a living donor transplant (36.3% versus 22.2%), with no differences by sex, ethnicity or human leucocyte antigen mismatch. Allograft failure within 2years of transplantation fell from 12.5% (2000-4) to 9.8% (2009-13). Surgical- and alloimmune-related failures decreased over time while death with a functioning graft became more common. Age, ethnicity and donor type were factors in recurrent primary disease and alloimmune pathology. Since 2000 there have been reductions in surgical and alloimmune graft failures in the UK. However, graft failure codes need to be revised if they are to remain useful and effective in epidemiological and quality improvement trials.

Causes of burn based on hospitals reports in mainland France in 2014

Introduction Burns are defined as an injury to the skin or other organic tissue primarily caused by thermal or other acute trauma. Every year in France, burns are responsible for several hundreds of deaths and about 11,500 hospitalizations. Furthermore, burns often result in heavy physical, psychological or social sequelae. They are one of the most expensive traumatic injuries due to costly wound treatment and very long care (hospitalization and rehabilitation). Prevention of burns is therefore a necessity. In order to prevent burns, knowledge about their causes is essential. There is a lack of large-scale results about causes of burns in France. Therefore, the aim of this work it to analyze causes of burns in mainland France in 2014, using data from the hospital nationwide database - “Programme de medicalisation des systemes d’information (PMSI)”. Coding of causes is expected to present high-quality because three French agencies “Agence technique de l’information sur l’hospitalisation” - (ATIH), “Societe francaise de brulologie” (SFB) [formerly “Societe francaise d’etude et de traitement des brulures” (SFETB)] and “Sante publique France” [formerly “Institut de veille sanitaire” (InVS)] implemented in 2011 a thesaurus of burns causes in order to facilitate and harmonize the coding in the PMSI and make it mandatory in 2013. Methods Analyses were performed on first hospital stays in a burn center, in 2014 in mainland France, registered in the PMSI database with a principal diagnosis of burns, coded from T20 to T32 in the International Classification of Diseases, 10th revision (ICD10), and at least one associated diagnosis mentioning a cause of burn. Data collected for patients whatever the age included: demographics, circumstances (accident, assault, suicide attempt), etiology (scald, contact, fire and flames, radiation, electrical, chemical). Results In 2014 in mainland France, 2712 patients were hospitalized in a burn center and presented information regarding the cause of burn, 61.5% were men, mean/median age of 31/28 years. Burns occurred mostly in accidental circumstances (93.7% of patients) and are more rarely due to suicide attempt (4.9%) or assault (1.4%). Half of accidental burns were scald (50.8%), 27.2% fire and flame burns, 8% contact burns, 4.5% chemical burns, other etiologies are scarcer. Causes of burns varied according to age and sex. Scald was the first cause of burns for children (83.8% for boys and 88.2% for girls among 0–4 years) and for women whatever the age. Fire and flames burns were the first cause of burns for men after 15 (42.6% among 15–59 years, 56% among 60 years and older). Conclusions For the first time in France, this study provides results about causes of burns at a large-scale. In spite of numerous prevention measures, burns and in particular scald, are very frequent in France. Further research aiming at improving knowledge about causes of burns could be useful to better target prevention measures. Moreover, a precise evaluation of prevention measures already implemented could be useful.

Deep neural models for ICD-10 coding of death certificates and autopsy reports in free-text

We address the assignment of ICD-10 codes for causes of death by analyzing free-text descriptions in death certificates, together with the associated autopsy reports and clinical bulletins, from the Portuguese Ministry of Health. We leverage a deep neural network that combines word embeddings, recurrent units, and neural attention, for the generation of intermediate representations of the textual contents. The neural network also explores the hierarchical nature of the input data, by building representations from the sequences of words within individual fields, which are then combined according to the sequences of fields that compose the inputs. Moreover, we explore innovative mechanisms for initializing the weights of the final nodes of the network, leveraging co-occurrences between classes together with the hierarchical structure of ICD-10. Experimental results attest to the contribution of the different neural network components. Our best model achieves accuracy scores over 89%, 81%, and 76%, respectively for ICD-10 chapters, blocks, and full-codes. Through examples, we also show that our method can produce interpretable results, useful for public health surveillance.

Przyczyny zgonów dzieci i młodzieży w wieku 1-19 lat w Polsce w świetle międzynarodowych statystyk od 2000 roku.

BackgroundAnalyses of children and young people mortality continue to be an important component of health monitoring of this population. Such analyses provide the basis to assess the overall trends, the structure of the causes of death over longer periods, and the differences between Poland and other countries.PurposeThe purpose of the current study is to present the current status and the direction of changes since 2000 with regard to the level and underlying causes of mortality in children and adolescents aged 1-19 years in Poland on the background of statistics for leading European countries.Material and methodsInteractive databases available online: the National Demographic Database provided by the Central Statistical Office and the International WHO-MDB Database were used. Poland, constantly belonging to Eur-B category, was compared with the combined group of 27 leading countries, classified as a very low total mortality group (Eur-A) according to WHO. Linear trends of overall and cause-specific mortality in 2000–2013 were estimated. The causes of death have been presented according to the main classes of the 10th Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10). External and other causes were adopted as the two principal categories.ResultsIn 2015, 1471 deaths of persons aged 1-19 were recorded in Poland (19.9 per 100 000, 25.4 and 14.2 for boys and girls, respectively). Changes in children and adolescents mortality by age have a nonlinear nature (U-shaped), and the lowest level is recorded at the age of 5-9 years.According to 2014 data, 50.2% of deaths of children and adolescents aged 1-19 years occurred due to external causes, including non-intentional and intentional ones. This percentage increased from 18.4% in the 1-4 age group to 68.6% at the age of 15-19 years. Apart from external causes, the dominating causes of death are malignant neoplasms, congenital defects, or nervous system and respiratory system diseases. The ranking of those causes of death changes in successive age groups and over time. When age is considered, a higher proportion of congenital defects and respiratory system diseases was found in mortality younger children and a higher proportion of circulatory system diseases and undefined cases in mortality of adolescents. When trends were studied, a continuing elimination of infectious diseases was observed together with growing impact of rare diseases in all age groups.The excess mortality of Polish population at age 1-19 by comparison to Eur-A countries increased from 21% in 2000 to 56% in 2013, mainly due to unfavourable trends in adolescents. The rate of decline in the mortality of young children (1-4 years) was greater than in Eur-A countries, both in case of external and other causes. In the age group 5-14 years the higher rate of change was sustained only with regard to external causes. Among adolescents and young adults, the distance between Poland and Eur-A countries increased during the studied period. The shape of trend in the 15-24 age group was unfavourable for Poland, mainly with respect to external causes. This observation could be in part explained by increasing suicide trend in Poland since 2008, coexisting with rather constant level in Eur-A countries.ConclusionsThe mortality rate among the population aged 1-19 years in Poland is systematically decreasing, but it still exceeds the average level recorded in leading European countries, particularly in relation to adolescents. When assessing the ability to reduce mortality in Poland to the level of Eur-A countries, attention must be paid to the causes considered as avoidable. Further studies ought to focus on the trends and international comparisons only foreshadowed in this study with regard to individual diagnoses, discussing possible preventive measures. Introduction of an ICD-11 classification will enable more accurate coding of causes of death, including a more precise analysis of the burden of rare diseases, which are an increasing challenge to public health in the population at the developmental age.

Medical death certification by forensic physicians in the Netherlands: Validity and interdoctorvariation

The aim of the study was to assess interdoctorvariation and validity in death certification by forensic physicians using 19 written scenarios. The scenarios described typical cases from forensic-medical practice. Physicians were asked to determine the manner of death (natural/unnatural) and to provide an ICD-10 code for the cause of death. In contrast to most studies on this topic, the measure of agreement among physicians was chance-corrected and a standard was used to assess the correctness of the assigned cause and manner of death. Forty-seven physicians participated in the survey. The study demonstrated that forensic physicians varied widely in their conclusions. With respect to manner of death, adequate agreement (defined as kappa>0.70) was achieved in six scenarios (32% of all scenarios). Concerning the underlying cause of death, adequate agreement was reached in three cases (16% of all scenarios). Furthermore, predictors for the correctness of manner and cause of death were studied using logistic regression. Years of experience as a forensic physician significantly predicted the correctness of cause of death (p < 0.05). Other predictors remained insignificant. With regard to manner of death, none of the studied predictors proved to be significant. To conclude, there appears to be a lack of consistency among forensic physicians regarding death certification. The ICD-10 coding of causes of death applied by forensic physicians is questionable. Less experienced physicians need supervision by more experienced colleagues when making judgments concerning the cause of death. Altogether, there is an urgent need to work out consensus-based guidelines for forensic physicians on how to certify deaths.

The accuracy of the causes of death and the estimated trend: the case of cervix uteri

reduction in cervical cancer mortality is the ultimate goal of the screening. Quality of death certificate reports has been improved over time, but they are still inaccurate, making it difficult to assess time trends in mortality. to evaluate the accuracy of the topographic coding of causes of death and to estimate the mortality time trend for cervical cancer through the method of incidence-based mortality (IBM) using cancer registry (CR) data. from the mortality registry (MR), we extracted data on deaths for cervix uteri cancer, corpus uteri cancer, and uterus cancer not otherwise specified (NOS) referred to residents in Reggio Emilia (Emilia-Romagna Region, Northern Italy) from 1997 to 2013. Deaths were checked with the CR to verify the topographical site of the primary tumour. Furthermore, by using CR data, we constructed a cohort of incident cervical cancer cases diagnosed between 1997 and 2009 with a 5-year follow-up. We calculated cause-specific IBM (excluding ovary) and IBM for all cause, crude and standardized, and annual percentage change (APC). out of 369 deaths for uterine cancer, 269 were reported in the RT: 32 for cervix uteri cancer, 76 for corpus uteri cancer, 161 for uterus cancer NOS. 28 of the 32 persons who died for cervical cancer were incidents for cervix uteri cancer. 63 of the 76 who died for corpus uteri cancer were incidents for corpus uteri cancer. Of the 161 who died of uterus cancer NOS, 80 were incidents for corpus uteri cancer, 45 for cervix uteri cancer, 28 for uterus cancer NOS, 5 for vagina cancer, and 3 for cancer of other non-specified organs. Applying these proportions of misclassification, we can estimate that the real number of cervical cancer deaths is 2.4 folds the number of cases reported in the MR as cervical cancer. IBM for all causes decreased significantly over the years (APC: -9.5; 95%CI -17.1;-1.1); cause-specific IBM decreases, but not significantly (APC: -5.1; 95%IC -16.1;+7.3). There is no improvement in survival (r2=0.02; p=0.6), while the incidence shows a decrease (APC: -6.6;95%CI -10.0;-3.0). mortality for cervical cancer is still underestimated by deaths certificates: for each reported case, there are other 1.4 cases that are reported with other less specific causes.

National Mortality Register

One additional benefit of the post mortem external examination is to collect valid information about cause of death and to set up a national mortality register. To date, the only German federal state to have established a mortality index for the purpose of epidemiological studies is the city of Bremen (1). Up until now, national epidemiological mortality follow-ups have only been feasible by requesting pseudonymized death certificates from the local public health departments. Because in some German federal states death certificates only need to be kept for a limited period of time (10 years), significant data losses, especially in the case of historical cohorts, may be assumed when follow-ups are based on the individual collection of death certificates (2). To provide reliable follow-up information, not only the death certificate needs to be correctly completed, but the underlying cause of death must also be coded with the correct ICD. Coding is done by coders in the state statistical offices who work according to official coding rules (3). Most coders do not have any specific medical training or education. They learn the coding rules in seminars and regular refresher courses. Notwithstanding, even among experienced coders, differences in coding behaviour arise because of the complexity of the applied rules and the erratic quality of death certificates. In particular, when death certificate entries are ambiguous, coders have to try to interpret the underlying cause of death. This was confirmed in our experience of mortality studies in which coders carried out the reference coding of the cause of death from death certificates. In addition to the specific medical qualification for carrying out post mortem external examinations as proposed in the current political debate, an automated multicausal coding of causes of death, comparable to the existing IRIS program (4), should be required. This would ensure valid, reproducible coding of causes of death. With the concurrent establishment of a national mortality register, an intensification of epidemiological research using mortality data should be expected in Germany, which should also reveal implausibilities in the determination of causes of death. This should also contribute to improved data quality in the cause of death statistics.

Narrative verdicts and their impact on mortality statistics in England and Wales

Annually, there are around 30,000 coroner's inquests held in England and Wales that conclude with a verdict. 'Short form' verdicts such as accident or misadventure; natural causes; suicide; and homicide make up the majority of all verdict conclusions. 'Narrative' verdicts can be used by a coroner or jury, instead of a short form verdict, to express their conclusions as to the cause of death following an inquest. Since 2001 narrative verdicts have been more widely used, with over 3,000 narrative verdicts returned in 2009. In some cases, it can be difficult to code the underlying cause of death from the information provided in the narrative. For some time, the Office for National Statistics (ONS) and other organisations have been concerned about the impact of narrative verdicts on the quality of the statistics on cause of death. Our research investigated the impact of narrative verdicts on trends for deaths attributed to injury and poisoning in England and Wales. The research considered narrative verdicts received by ONS between 2001 and 2009. All available information provided by the coroner from the narrative verdict, together with the underlying cause of death, was used in the analysis. All causes of death where a narrative verdict was returned were investigated. More in-depth analysis of accidental deaths was undertaken, as classification of these deaths by intent is more difficult when the information from the coroner is imprecise. A sensitivity analysis of suicide rates (intentional self-harm and event of undetermined intent) was carried out. This involved using two different scenarios of reclassifying selected proportions of accidental hanging and poisoning deaths, where a narrative verdict was returned, as intentional self-harm. An exercise to measure the consistency of coding cause of death from narrative verdicts was also undertaken. The increasing proportion of narrative verdicts involving injury and poisoning has not significantly affected published mortality rates for suicide (intentional self-harm and injury or poisoning of undetermined intent). However, if the rise in narrative verdicts continues at the same rate, the accurate reporting of injury and poisoning deaths, including suicides, is likely to be affected. The exercise to establish the consistency of coding the cause of death by ONS cause coders showed that the current coding rules were being applied uniformly. The increase in the use of narrative verdicts by coroners has not had a statistically significant impact on published suicide rates in England and Wales and so no revision to these rates is needed. A review of current coding practices and the handling of narrative verdicts will be undertaken by ONS with particular reference to deaths from intentional self-harm. A recommendation has been made to coroners to consider ways of recording narrative verdicts to allow more accurate coding of cause of death. This will ensure that mortality statistics are maintained to the highest standards.List of Tables, 83.

Revisiting the mortality of France and Italy with the multiple-cause-of-death approach

Abstract In this paper we revisit mortality profiles of France and Italy in 2003 using multiple-cause-of-death approach. The method leads to a substantial upward reassessment of role played by certain conditions - e.g. diseases of blood and diseases of skin - in overall mortality. Regarding associations of causes, we distinguish three patterns of pairwise joint occurrence of causes that are common to both countries. The numerous similarities that emerge from comparison of two countries are a positive signal of reliability of multiple-cause-of-death data. (ProQuest: ... denotes formula omitted.) 1. Introduction Although monitoring of mortality trends is mostly based on underlying cause of death - that is the disease or which initiated train of morbid events leading directly to death, or circumstances of accident or violence which produced fatal injury (WHO 1949), a growing number of studies use entire set of causes listed on death certificate, hereafter referred to as multiple causes of death (MCOD), to portray mortality profile of a country (Chamblee and Evans 1982; Manton and Stallard 1982; Manton 1986; Manton and Myers 1987; White, Selvin, and Merrill 1989; Mackenbach et al. 1995; Stallard 2002; Desesquelles and Mesle 2004; Redelings, Sorvillo, and Simon 2006; Redeling, Wise, and Sorvillo 2007; Frova et al. 2009) or to re-evaluate contribution of a specific cause (Wing and Manton 1981; Israel, Rosenberg, and Curtin 1986; Nizard and Munoz-Perez 1993; Coste and Jougla 1994; Mannino et al. 1998; Wise and Sorvillo 2005; Fuhrman et al. 2006; Romon et al. 2008). As first advocates of MCOD approach (Janssen 1940; Dorn and Moriyama 1964) have stressed, underlying-cause-of-death approach misdirects attention away from conditions that tend to be reported as contributory causes. As a consequence their contribution to overall mortality is underestimated. The MCOD approach should be especially useful for describing medical circumstances surrounding death of older people. At old ages, death is indeed often final stage of a long morbid process involving several conditions. Clearly MCOD approach is a potentially valuable tool for demographers, who are necessarily concerned with portraying mortality profiles of aging populations as accurately as possible. Notwithstanding these developments, studies that use multiple causes of death in order to compare a set of countries are still very scarce. This situation may be partly due to quite widely shared scepticism about quality, and consequently comparability, of multiple-cause-of-death data. To our mind, apart from certification issues, recent technological innovations in coding systems represent a major advance towards improved quality of cause-of-death data. In a growing number of countries recording of information on death certificates and coding of causes have been automated (optical recognition and automatic coding system). Human intervention is limited to problematic cases that cannot be processed automatically and, at least in theory, World Health Organization coding rules can be applied systematically and uniformly, irrespective of coding agent or country. Another consequence of these automated systems is that number of coded causes is no longer limited. Researchers often have access to all medical information reported on certificate instead of a limited selection of causes. In this paper we use MCOD approach in order to revisit mortality profiles of France and Italy in 2003. Our goals are both substantive and methodological. First we re-evaluate cause-specific mortality profiles of two countries, taking into account all conditions reported on death certificates. We then measure pairwise joint occurrence of any underlying and any contributory cause (i. …