Covid 19 epidemics in Chile: how much excess and competitive mortality?

Aims

1. To describe the excess mortality 2020 compared to the 2015-2019 period according to the week of occurrence.

2. To compare the differences in excess mortality according to sex, age and region

3. To compare the distribution of deaths by causes between 2020 and the previous five-year period

4. To study the magnitude of the effect of competitive mortality due to COVID with respect to other causes of death.

Methods

Data was obtained from the Ministry of Health Department of statistics (http://deis.minsal.cl) with the cumulative mortality registries from 1 Jan 2016 till 31 Dec 2020. We combined historical counts of mortality data from the Chilean Vital Statistics Death Database 1990-2015 and 2016-2020.

Data from Jan 2016 till Dec 2019 was used to estimate the weekly death count expected value for 2020. We used an overdispersed Poisson regression models with 11 spline terms to account for seasonal patterns and its upper bound of the 95% prediction interval.

In order to analyse comparable weeks across the years, the week number was obtained by defining week 1 as starting on 1 January each year, week 2 on 8 January and so on. We used weeks from 1 to 52 (the first 364 days of each year).

We estimated the number of excess deaths (i.e. observed numbers above each threshold) and percentage excess, also known as p-score, calculated as excess deaths divided by average expected number of deaths. We stratified the p-score analyses by age and sex.

Using the mortality codes (i.e. U07.1 and U07.2 ICD-10 codes) we estimate excess with and without confirmed and suspected COVID-19.

All analyses were conducted in R version 4.0.4 (R Development Core Team, Vienna, Austria).

Results

Table 1: Number of deaths from week 1 to 52 (included) and from 53 onwards (excluded). Chile, 1 Jan -29 Dec of every year 2015 - 2020

h1

Year	Deaths included	Death excluded	Date min	Date max
2015	103054	267	2015-01-01	2015-12-30
2016	103648	378	2016-01-01	2016-12-29
2017	106143	245	2017-01-01	2017-12-30
2018	106516	280	2018-01-01	2018-12-30
2019	110050	286	2019-01-01	2019-12-30
2020	125259	573	2020-01-01	2020-12-29

Figure 1: Weekly numbers of deaths from all causes and from all causes excluding COVID-19 relative to the average expected number and the upper bound of the 95% prediction interval. Chile, 1 Jan -29 Dec of every year 2015 - 2020

Figure 2: Weekly numbers of deaths from all causes and from all causes excluding COVID-19 above the average expected number and the upper bound of the 95% prediction interval . Chile, 1 Jan -29 Dec of every year 2015 - 2020

Figure 3: Weekly P-score (relative excess mortality observed vs expected) by gender and age. Chile, 1 Jan -29 Dec of every year 2015 - 2020

Figure 3: Weekly P-score (relative excess mortality observed vs expected) by gender and age. Chile, 1 Jan -29 Dec of every year 2015 - 2020

Figure 4: Causes of death. Chile, 1 Jan -29 Dec of every year 2015 - 2020

Table 2: Observed, expected and P-score by ICD-10 cause of mortality Chile, 1 Jan -29 Dec of every year 2015 - 2020

exp_	Title_s	value	expected	p_score
A00-B99	Infectious	2159	2413.1	-10.5
C00-D48	Neoplasms	26676	28759.7	-7.2
D50-D89	Blood	616	647.7	-4.9
E00-E90	Metabolic	4529	4109.4	10.2
F00-F99	Mental	2352	2679.4	-12.2
G00-G99	Nervous	3936	4382.9	-10.2
I00-I99	Circulatory	27160	28157.1	-3.5
J00-J99	Respiratory	10216	14678	-30.4
K00-K93	Digestive	7800	8076	-3.4
L00-L99	Skin	864	999.1	-13.5
M00-M99	Musculoskeletal	561	623.9	-10.1
N00-N99	Genitourinary	3423	3561.5	-3.9
O00-O99	Pregnancy	55	42.4	29.7
P00-P96	Perinatal	528	680.5	-22.4
Q00-Q99	Malformations	603	687.6	-12.3
R00-R99	No-classified	3974	3059.1	29.9
S00-T98	External	7682	7906.2	-2.8

Note: Not possible to estimate expected value for “H60-H95” Ear and “H00-H59” Eye using a monthly model (infrequent outcomes: a total of 22 and 21 between 2015 and 2020, respectively)