Average Number of Years Lost For People Who Died of Coronavirus in China

We illustrate the use of the package by estimating the average number of years by which people’s lives are shortened due to coronavirus. Using data from Table 1 of the paper that gives us the distribution of ages of people who died from COVID-19 in China, with conservative assumptions (assuming gender of the dead person to be male, taking the middle of age ranges) we find that people’s lives are shortened by about 11 years on average. These estimates are conservative for one additional reason: there is likely an inverse correlation between people who die and their expected longevity. And note that given a bulk of the deaths are among older people, when people are more infirm, the quality adjusted years lost is likely yet more modest. Given that the last life tables from China are from 1981, we estimate the average number of years lost if people had the same profile as Americans. Using the most recent SSA data, we find that number to be 16. Compare this to deaths from road accidents, the modal reason for death among 5-24 and 25-44 ages in the US. Male life expectancy in the US at 25 is another ~ 52 years.

table1.png 
[1]:

import pandas as pd

from lost_years import lost_years_hld

Prepare example input in DataFrame

Please look at country codes here:- https://www.lifetable.de/cgi-bin/country_codes.php

[2]:

df = pd.read_csv("data/table_1.csv")
df["year"] = 2020
df

[2]:
lowest_age highest_age middle_age n_deaths sex country year
0 10 19 15 1 M CHN 2020
1 20 29 25 7 M CHN 2020
2 30 39 25 18 M CHN 2020
3 40 49 45 38 M CHN 2020
4 50 59 55 130 M CHN 2020
5 60 69 65 309 M CHN 2020
6 70 79 75 312 M CHN 2020
7 80 89 85 208 M CHN 2020

Get Human Life Table data columns from HLD dataset

[3]:

highest_ldf = lost_years_hld(
    df, {"age": "lowest_age", "country": "country", "sex": "sex", "year": "year"}
)
highest_ldf.head()

[3]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 10 M 1981 60.38
1 20 29 25 7 M CHN 2020 CHN 20 M 1981 50.89
2 30 39 25 18 M CHN 2020 CHN 30 M 1981 41.56
3 40 49 45 38 M CHN 2020 CHN 40 M 1981 32.33
4 50 59 55 130 M CHN 2020 CHN 50 M 1981 23.54

Note that the year we are matching to is 1981.

[4]:

# From table 1, 95.7% deaths in Hubei region == 110 (https://www.lifetable.de/cgi-bin/regions.php)
# Note: Current function version doesn't return region data, using all available data for China
highest_ldf = highest_ldf.copy()
highest_ldf

[4]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 10 M 1981 60.38
1 20 29 25 7 M CHN 2020 CHN 20 M 1981 50.89
2 30 39 25 18 M CHN 2020 CHN 30 M 1981 41.56
3 40 49 45 38 M CHN 2020 CHN 40 M 1981 32.33
4 50 59 55 130 M CHN 2020 CHN 50 M 1981 23.54
5 60 69 65 309 M CHN 2020 CHN 60 M 1981 15.73
6 70 79 75 312 M CHN 2020 CHN 70 M 1981 9.57
7 80 89 85 208 M CHN 2020 CHN 80 M 1981 5.31

Assuming all the people who died were at the bottom of the age ranges

[5]:

highest_ldf["years_lost"] = (
    highest_ldf["hld_life_expectancy"] * highest_ldf["n_deaths"] / highest_ldf["n_deaths"].sum()
)
highest_ldf

[5]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy years_lost
0 10 19 15 1 M CHN 2020 CHN 10 M 1981 60.38 0.059022
1 20 29 25 7 M CHN 2020 CHN 20 M 1981 50.89 0.348221
2 30 39 25 18 M CHN 2020 CHN 30 M 1981 41.56 0.731261
3 40 49 45 38 M CHN 2020 CHN 40 M 1981 32.33 1.200919
4 50 59 55 130 M CHN 2020 CHN 50 M 1981 23.54 2.991398
5 60 69 65 309 M CHN 2020 CHN 60 M 1981 15.73 4.751290
6 70 79 75 312 M CHN 2020 CHN 70 M 1981 9.57 2.918710
7 80 89 85 208 M CHN 2020 CHN 80 M 1981 5.31 1.079648 
[6]:

highest_ldf["years_lost"].sum().round()

[6]:

np.float64(14.0)

[7]:

lowest_ldf = lost_years_hld(
    df, {"age": "highest_age", "country": "country", "sex": "sex", "year": "year"}
)
lowest_ldf.head()

[7]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 19 M 1981 51.83
1 20 29 25 7 M CHN 2020 CHN 29 M 1981 42.50
2 30 39 25 18 M CHN 2020 CHN 39 M 1981 33.24
3 40 49 45 38 M CHN 2020 CHN 49 M 1981 24.39
4 50 59 55 130 M CHN 2020 CHN 59 M 1981 16.46

Assuming all the people who died were at the top of the age ranges

[8]:

# From table 1, 95.7% deaths in Hubei region == 110 (https://www.lifetable.de/cgi-bin/regions.php)
# Note: Current function version doesn't return region data, using all available data for China
lowest_ldf = lowest_ldf.copy()
lowest_ldf

[8]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 19 M 1981 51.83
1 20 29 25 7 M CHN 2020 CHN 29 M 1981 42.50
2 30 39 25 18 M CHN 2020 CHN 39 M 1981 33.24
3 40 49 45 38 M CHN 2020 CHN 49 M 1981 24.39
4 50 59 55 130 M CHN 2020 CHN 59 M 1981 16.46
5 60 69 65 309 M CHN 2020 CHN 69 M 1981 10.11
6 70 79 75 312 M CHN 2020 CHN 79 M 1981 5.68
7 80 89 85 208 M CHN 2020 CHN 89 M 1981 3.29 
[9]:

lowest_ldf["years_lost"] = (
    lowest_ldf["hld_life_expectancy"] * lowest_ldf["n_deaths"] / lowest_ldf["n_deaths"].sum()
)
lowest_ldf

[9]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy years_lost
0 10 19 15 1 M CHN 2020 CHN 19 M 1981 51.83 0.050665
1 20 29 25 7 M CHN 2020 CHN 29 M 1981 42.50 0.290811
2 30 39 25 18 M CHN 2020 CHN 39 M 1981 33.24 0.584868
3 40 49 45 38 M CHN 2020 CHN 49 M 1981 24.39 0.905982
4 50 59 55 130 M CHN 2020 CHN 59 M 1981 16.46 2.091691
5 60 69 65 309 M CHN 2020 CHN 69 M 1981 10.11 3.053754
6 70 79 75 312 M CHN 2020 CHN 79 M 1981 5.68 1.732317
7 80 89 85 208 M CHN 2020 CHN 89 M 1981 3.29 0.668935 
[10]:

lowest_ldf["years_lost"].sum().round()

[10]:

np.float64(9.0)

[11]:

middle_ldf = lost_years_hld(
    df, {"age": "middle_age", "country": "country", "sex": "sex", "year": "year"}
)
middle_ldf.head()

[11]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 15 M 1981 55.61
1 20 29 25 7 M CHN 2020 CHN 25 M 1981 46.24
2 30 39 25 18 M CHN 2020 CHN 25 M 1981 46.24
3 40 49 45 38 M CHN 2020 CHN 45 M 1981 27.85
4 50 59 55 130 M CHN 2020 CHN 55 M 1981 19.43

Assuming all the people who died were at the middle of the age ranges

[12]:

# From table 1, 95.7% deaths in Hubei region == 110 (https://www.lifetable.de/cgi-bin/regions.php)
# Note: Current function version doesn't return region data, using all available data for China
middle_ldf = middle_ldf.copy()
middle_ldf

[12]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy
0 10 19 15 1 M CHN 2020 CHN 15 M 1981 55.61
1 20 29 25 7 M CHN 2020 CHN 25 M 1981 46.24
2 30 39 25 18 M CHN 2020 CHN 25 M 1981 46.24
3 40 49 45 38 M CHN 2020 CHN 45 M 1981 27.85
4 50 59 55 130 M CHN 2020 CHN 55 M 1981 19.43
5 60 69 65 309 M CHN 2020 CHN 65 M 1981 12.46
6 70 79 75 312 M CHN 2020 CHN 75 M 1981 7.26
7 80 89 85 208 M CHN 2020 CHN 85 M 1981 4.02 
[13]:

middle_ldf["years_lost"] = (
    middle_ldf["hld_life_expectancy"] * middle_ldf["n_deaths"] / middle_ldf["n_deaths"].sum()
)
middle_ldf

[13]:
lowest_age highest_age middle_age n_deaths sex country year hld_country hld_age hld_sex hld_year hld_life_expectancy years_lost
0 10 19 15 1 M CHN 2020 CHN 15 M 1981 55.61 0.054360
1 20 29 25 7 M CHN 2020 CHN 25 M 1981 46.24 0.316403
2 30 39 25 18 M CHN 2020 CHN 25 M 1981 46.24 0.813607
3 40 49 45 38 M CHN 2020 CHN 45 M 1981 27.85 1.034506
4 50 59 55 130 M CHN 2020 CHN 55 M 1981 19.43 2.469110
5 60 69 65 309 M CHN 2020 CHN 65 M 1981 12.46 3.763578
6 70 79 75 312 M CHN 2020 CHN 75 M 1981 7.26 2.214194
7 80 89 85 208 M CHN 2020 CHN 85 M 1981 4.02 0.817361 
[14]:

middle_ldf["years_lost"].sum().round()

[14]:

np.float64(11.0)

Assume the Longevity is the Same as People in the US

[15]:

from lost_years import lost_years_ssa

[16]:

ssa_middle_ldf = lost_years_ssa(df, {"age": "middle_age", "sex": "sex", "year": "year"})
ssa_middle_ldf.head()

[16]:
lowest_age highest_age middle_age n_deaths sex country year ssa_age ssa_year ssa_life_expectancy
0 10 19 15 1 M CHN 2020 15 2022 60.39
1 20 29 25 7 M CHN 2020 25 2022 51.03
2 30 39 25 18 M CHN 2020 25 2022 51.03
3 40 49 45 38 M CHN 2020 45 2022 33.32
4 50 59 55 130 M CHN 2020 55 2022 24.94 
[17]:

ssa_middle_ldf["years_lost"] = (
    ssa_middle_ldf["ssa_life_expectancy"]
    * ssa_middle_ldf["n_deaths"]
    / ssa_middle_ldf["n_deaths"].sum()
)
ssa_middle_ldf

[17]:
lowest_age highest_age middle_age n_deaths sex country year ssa_age ssa_year ssa_life_expectancy years_lost
0 10 19 15 1 M CHN 2020 15 2022 60.39 0.059032
1 20 29 25 7 M CHN 2020 25 2022 51.03 0.349179
2 30 39 25 18 M CHN 2020 25 2022 51.03 0.897889
3 40 49 45 38 M CHN 2020 45 2022 33.32 1.237693
4 50 59 55 130 M CHN 2020 55 2022 24.94 3.169306
5 60 69 65 309 M CHN 2020 65 2022 17.48 5.279883
6 70 79 75 312 M CHN 2020 75 2022 10.92 3.330440
7 80 89 85 208 M CHN 2020 85 2022 5.75 1.169110 
[18]:

ssa_middle_ldf["years_lost"].sum().round()

[18]:

np.float64(15.0)

Assume Everyone Lives Till 90

[19]:

y90_middle_ldf = df.copy()
y90_middle_ldf["y90_life_expectancy"] = 90 - y90_middle_ldf["middle_age"]
y90_middle_ldf.head()

[19]:
lowest_age highest_age middle_age n_deaths sex country year y90_life_expectancy
0 10 19 15 1 M CHN 2020 75
1 20 29 25 7 M CHN 2020 65
2 30 39 25 18 M CHN 2020 65
3 40 49 45 38 M CHN 2020 45
4 50 59 55 130 M CHN 2020 35 
[20]:

y90_middle_ldf["years_lost"] = (
    y90_middle_ldf["y90_life_expectancy"]
    * y90_middle_ldf["n_deaths"]
    / y90_middle_ldf["n_deaths"].sum()
)
y90_middle_ldf

[20]:
lowest_age highest_age middle_age n_deaths sex country year y90_life_expectancy years_lost
0 10 19 15 1 M CHN 2020 75 0.073314
1 20 29 25 7 M CHN 2020 65 0.444770
2 30 39 25 18 M CHN 2020 65 1.143695
3 40 49 45 38 M CHN 2020 45 1.671554
4 50 59 55 130 M CHN 2020 35 4.447703
5 60 69 65 309 M CHN 2020 25 7.551320
6 70 79 75 312 M CHN 2020 15 4.574780
7 80 89 85 208 M CHN 2020 5 1.016618 
[21]:

y90_middle_ldf["years_lost"].sum().round()

[21]:

np.float64(21.0)

[ ]: