Skip to main content
Have a personal or library account? Click to login
Analysis of the Predictors of Mortality from Ischemic Heart Diseases in the Southern Region of Brazil: A Geographic Machine-Learning-Based Study Cover

Analysis of the Predictors of Mortality from Ischemic Heart Diseases in the Southern Region of Brazil: A Geographic Machine-Learning-Based Study

Global Heart
Volume 19 (2024): Issue 1
By: Amanda de Carvalho Dutra,  Lincoln Luis Silva,  Isadora Martins Borba,  Amanda Gubert Alves dos Santos,  Diogo Pinetti Marquezoni,  Matheus Henrique Arruda Beltrame,  Rogério do Lago Franco,  Ualid Saleh Hatoum,  Juliana Harumi Miyoshi,  Gustavo Cezar Wagner Leandro,  Marcos Rogério Bitencourt,  Oscar Kenji Nihei,  João Ricardo Nickenig Vissoci and  Luciano de Andrade  
Open Access
|Nov 2024
Figures & tables

Figures & Tables

Table 1

Data source.

VARIABLESPERIODCONSTANTSOURCE
Mortality from Ischemic Heart Diseases (IHD)2018–2022100.00027
Population2018–2022100.00032
Electrocardiograph rate2018–202210.00033
Cardiac catheterization rate2018–202210.00034
Access index to hemodynamics2018–2022E2SFCA33
Access index of pre-hospital mobile units2018–2022E2SFCA33
Cardiologists rate2018–20221.00033
Myocardial scintigraphy rate2018–202210.00034
Stress test rate2018–202210.00034
Stress echocardiogram rate2018–202210.00034
Shapefile from southern Brazil2022119135
Table 2

Number of deaths attributed to IHD in subjects aged 30 to 70 years in the Southern region of Brazil, 2018–2022.

VARIABLESSOUTH REGION N = 59,093
PARANÁ
N = 21,565
(36.5%)		SANTA CATARINA
N = 13,296
(22.5%)		RIO GRANDE DO SUL
24,232
(41.0%)		TOTAL
N = 59,093
(100%)
Gender
    Male14,171 (65.7%)9,034 (68.0%)15,657 (64.6%)38,859 (65.8%)
    Female7,394 (34.3%)4,262 (32.0%)8,575 (35.4%)20,230 (34.2%)
Race
    White16,299 (75.6%)11,894 (89.5%)20,721 (85.5%)48,912 (82.8%)
    Non-white5,266 (24.4%)1,402 (10.5%)3,511 (14.5%)10,181 (17.2%)
Age group (years)
    30–40448 (2.1%)241 (1.8%)362 (1.5%)1,051 (1.8%)
    40–501,590 (7.4%)970 (7.3%)1,403 (5.8%)3,962 (6.7%)
    50–604,137 (19.2%)2,761 (20.8%)4,105 (16.9%)10,999 (18.6%)
    60–7015,390 (71.4%)9,324 (70.1%)18,362 (75.8%)43,081 (72.9%)
Figure 1

Spatial distribution of IHD mortality rates in the 1,191 municipalities of southern Brazil, and their clusters. 1A. IHD mortality rates per adjusted population/100,000 inhabitants in Southern Brazil, 2018 to 2022; 1B. Analysis of Local Indicators of Spatial Association (LISA) indicating clusters according to distribution patterns of high and low IHD mortality rates.

Figure 2

Bivariate Local Moran of ischemic heart disease (IHD) in the south region of Brazil.

Table 3

Results of global and local multivariate analysis of IHD mortality rates of municipalities in southern Brazil.

OLSGWRRFGWRF
VARIABLECOEFFICIENTStd. ERRORT-VALUEPROBABILITY (>|t|)MIN1 st QUMEDIAN3RD QUMAXMINMAXMEANStdMINMAXMEANStd
Electrocardiograph rate–1.0620.217–4.8790.000***–2.387–1.035–0.4070.2202.7453744.67599887.7799959.00019040.083  0.3643239.136270.614425.543
Cardiac catheterization rate  0.2290.065  3.5000.000***–1.441  0.021  0.2030.4101.2324755.09230693.3614577.5695217.347  0.2467015.351319.890598.804
Access index to hemodynamics–1.5940.478–3.3300.000***–23.272–2.723–0.959.0.6136.2484150.98230591.5412584.1635614.0434  0.0004039.209277.454460.952
Access index of pre-hospital mobile units–0.0900.042–2.1480.031*–2.275–0.320–0.0340.1070.9103988.51230506.9713369.5775293.7682  0.1564500.410307.706510.182
Cardiologists rate–6.4092.429–2.6390.008**–35.999–8.610–2.7991.23016.0511988.9139959.4485098.8211665.442  0.0004643.060140.773339.669
Myocardial scintigraphy rate  0.1890.462  0.4100.681–25.294–2.034  0.1371.46420.759370.27911722.933081.4512522.169  0.0002510.19356.666184.920
Stress test rate  3.0651.197  2.5600.010*–273.639–6.438  0.9125.158400.81662.1186995.9042214.9651598.3471  0.0001426.45936.062122.344
Stress echocardiogram rate–1.0820.893–1.2120.225–50.319–1.381  0.4783.01526.689683.7925482.2942125.966964.8613  0.0002581.38638.119177.638
R2  0.066  0.6560.058  0.983
Aicc  10034.81  9074.024550.3  2298.4
Moran Residues  0.694  0.4080.500–0.151

[i] Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1.

Figure 3

Importance of the variable according to the incMSE.

Figure 4

Spatial distribution of predictor variables’ importance in the GWRF model.

Figure 5

Ischemic heart disease mortality rates predicted using GWRF.

Figure 6

Determination coefficients of validation and test data using Geographically Weighted Random Forest to predict IHD Mortality Rate.

DOI: https://doi.org/10.5334/gh.1371 | Journal eISSN: 2211-8179
Journal RSS Feed
Language: English
Submitted on: Aug 6, 2024
Accepted on: Nov 5, 2024
Published on: Nov 27, 2024
Published by: Ubiquity Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year
Keywords:
Myocardial Ischemia,
Spatial Analysis,
Supervised Machine Learning,
Epidemiology

© 2024 Amanda de Carvalho Dutra, Lincoln Luis Silva, Isadora Martins Borba, Amanda Gubert Alves dos Santos, Diogo Pinetti Marquezoni, Matheus Henrique Arruda Beltrame, Rogério do Lago Franco, Ualid Saleh Hatoum, Juliana Harumi Miyoshi, Gustavo Cezar Wagner Leandro, Marcos Rogério Bitencourt, Oscar Kenji Nihei, João Ricardo Nickenig Vissoci, Luciano de Andrade, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.

Volume 19 (2024): Issue 1
Previous article
Next article