Automatic coding and selection of causes of death: an adaptation of Iris software for using in Brazil

Leading causes of excess mortality in Mexico during the COVID-19 pandemic 2020-2021: A death certificates study in a middle-income country

Background

The death toll after SARS-CoV-2 emergence includes deaths directly or indirectly associated with COVID-19. Mexico reported 325,415 excess deaths, 34.4% of them not directly related to COVID-19 in 2020. In this work, we aimed to analyse temporal changes in the distribution of the leading causes of mortality produced by COVID-19 pandemic in Mexico to understand excess mortality not directly related to the virus infection.

Methods

We did a longitudinal retrospective study of the leading causes of mortality and their variation with respect to cause-specific expected deaths in Mexico from January 2020 through December 2021 using death certificate information. We fitted a Poisson regression model to predict cause-specific mortality during the pandemic period, based on the 2015–2019 registered mortality. We estimated excess deaths as a weekly difference between expected and observed deaths and added up for the entire period. We expressed all-cause and cause-specific excess mortality as a percentage change with respect to predicted deaths by our model.

Findings

COVID-19 was the leading cause of death in 2020–2021 (439,582 deaths). All-cause total excess mortality was 600,590 deaths (38⋅2% [95% CI: 36·0 to 40·4] over expected). The largest increases in cause-specific mortality, occurred in diabetes (36·8% over expected), respiratory infections (33·3%), ischaemic heart diseases (32·5%) and hypertensive diseases (25·0%). The cause-groups that experienced significant decreases with respect to the expected pre-pandemic mortality were infectious and parasitic diseases (-20·8%), skin diseases (-17·5%), non-traffic related accidents (-16·7%) and malignant neoplasm (-5·3%).

Interpretation

Mortality from COVID-19 became the first cause of death in 2020–2021, the increase in other causes of death may be explained by changes in the health service utilization patterns caused by hospital conversion or fear of the population using them. Cause-misclassification cannot be ruled out.

Funding

This study was funded by Conacyt.

Does the Application of International Classification of Disease Codes for the Cause of Death on Death Certificates Reduce Garbage Codes?

INTRODUCTION

This study aimed to determine if applying International Classification of Diseases (ICD) disease codes directly as the cause of death (COD) on death certificates (DCs) instead of writing or typing the COD could reduce the use of garbage codes.

METHODS

Beginning in April of 2016, a documentation process change was made, retiring the process of hand-writing or typing the COD onto DCs to directly applying ICD disease codes that were registered during the patient's course of treatment. The DCs issued at the emergency department (ED) 1 year before (Pre-code group) and after (Code group) applying ICD disease codes directly on DCs was instituted were retrospectively analyzed. The occurrence of garbage codes along with other major and minor errors was compared between the two groups. The investigation and judgment of errors were performed by four emergency physicians.

RESULTS

The overall garbage code occurrence in the Code group (25%) was significantly lower than that in the Pre-code group (49%). Fewer garbage codes were used in the Code group with an average of .5 in the Pre-code group and .3 in the Code group. No significant difference was identified in major error occurrences except for in the garbage codes. Minor errors were more common in the Pre-code group than in the Code group.

CONCLUSION

The overall use of garbage codes on DCs could be reduced by changing the process by which physicians complete DCs, that is, the application of documenting ICD disease codes directly as the COD on DCs.

Construction of a semi-automatic ICD-10 coding system

Background

The International Classification of Diseases, 10th Revision (ICD-10) has been widely used to describe the diagnosis information of patients. Automatic ICD-10 coding is important because manually assigning codes is expensive, time consuming and error prone. Although numerous approaches have been developed to explore automatic coding, few of them have been applied in practice. Our aim is to construct a practical, automatic ICD-10 coding machine to improve coding efficiency and quality in daily work.

Methods

In this study, we propose the use of regular expressions (regexps) to establish a correspondence between diagnosis codes and diagnosis descriptions in outpatient settings and at admission and discharge. The description models of the regexps were embedded in our upgraded coding system, which queries a diagnosis description and assigns a unique diagnosis code. Like most studies, the precision (P), recall (R), F-measure (F) and overall accuracy (A) were used to evaluate the system performance. Our study had two stages. The datasets were obtained from the diagnosis information on the homepage of the discharge medical record. The testing sets were from October 1, 2017 to April 30, 2018 and from July 1, 2018 to January 31, 2019.

Results

The values of P were 89.27 and 88.38% in the first testing phase and the second testing phase, respectively, which demonstrate high precision. The automatic ICD-10 coding system completed more than 160,000 codes in 16 months, which reduced the workload of the coders. In addition, a comparison between the amount of time needed for manual coding and automatic coding indicated the effectiveness of the system-the time needed for automatic coding takes nearly 100 times less than manual coding.

Conclusions

Our automatic coding system is well suited for the coding task. Further studies are warranted to perfect the description models of the regexps and to develop synthetic approaches to improve system performance.

Avaliação de um aplicativo para smartphone para aprimoramento da certificação médica da causa da morte

RESUMO Introdução: O aplicativo para smartphone AtestaDO foi desenvolvido para auxiliar o médico na certificação da causa de morte. Neste estudo se propõe avaliar a aceitabilidade desse aplicativo. Métodos: Médicos foram convidados para participar de reuniões sobre certificação correta da causa da morte e avaliar o aplicativo em três etapas dessas reuniões, realizadas em Natal e em dois grandes hospitais de Belo Horizonte. Resultados: Em Natal, 82% dos 38 médicos participantes tinham mais de 20 anos de graduação, e em Belo Horizonte, mais de 67% dos 58 médicos tinham menos de 5 anos de graduação. As seções “Interface do aplicativo”, “Como atestar as causas de morte”, “Prática com exercícios” e “Outras informações para o médico” foram bem avaliadas por mais de 50% dos médicos de Belo Horizonte. Em Natal, todas as seções foram bem avaliadas por pelo menos 80% dos médicos. Mais de 70% dos participantes de Natal e da segunda etapa de Belo Horizonte usariam o aplicativo para preencher a causa de morte. A probabilidade de usar o AtestaDO para dar aulas sobre preenchimento da Declaração de Óbito foi de 83% para médicos de Natal, mas inferior a 60% em Belo Horizonte. Nas três etapas, a maioria dos médicos recomendaria o uso do aplicativo para outros colegas. Conclusão: A avaliação do aplicativo AtestaDO mostrou boa aceitabilidade. Espera-se que o uso dessa ferramenta permita alcançar melhorias na certificação médica da causa do óbito.