Brief Analysis and Visualization of CT Poison Data

The code below was produced to accompany the Cushing/Whitney Medical Library blog post, "Poison Yesterday and Today."

In brief, the code below does the following:

1. Imports necessary libraries (pandas, matplotlib, seaborn)
2. Reads in data, styles columns to snake case, and previews data
3. Slices off data of interest, i.e., where 'accidental poison' is mentioned in 'cause of death'
4. Removes data marked 'unreliable'
5. Combines similar data (such as all accidental poisoning due to drugs) together
6. Visualizes the final results

Data citation: Centers for Disease Control and Prevention, National Center for Health Statistics. National Vital Statistics System, Mortality 2018-2021 on CDC WONDER Online Database, released in 2021. Data are from the Multiple Cause of Death Files, 2018-2021, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative Program. Accessed at http://wonder.cdc.gov/ucd-icd10-expanded.html on Mar 16, 2023.

_See caveats from CDC Wonder below as well.

import pandas as pd
import matplotlib.pyplot as plt
import seaborn

#read in data

cdc_ct_death = pd.read_csv("CDCWONDER_underlying_death_cause_external_causes_2018-2021.csv",
                           dtype={"Cause of death": str,
                                  "Cause of death Code": str,
                                  "Deaths": int,
                                  "Population": int})

#personal style preference here for columns

def make_cols_snake_case(df):
    df.columns = [x.lower() for x in df.columns]
    df.columns = df.columns.str.replace("[ ]", "_", regex=True)

make_cols_snake_case(cdc_ct_death)

#check info and make sure function above took

cdc_ct_death.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 63 entries, 0 to 62
Data columns (total 5 columns):
 #   Column                 Non-Null Count  Dtype 
---  ------                 --------------  ----- 
 0   cause_of_death         63 non-null     object
 1   cause_of_death_code    63 non-null     object
 2   deaths                 63 non-null     int64 
 3   population             63 non-null     int64 
 4   crude_rate_per_100000  63 non-null     object
dtypes: int64(2), object(3)
memory usage: 2.6+ KB

#preview data

cdc_ct_death

	cause_of_death	cause_of_death_code	deaths	population	crude_rate_per_100000
0	Total	not applicable	9192	10727890	85.70
1	Accidental poisoning by and exposure to other ...	X44	2159	10727890	20.10
2	Accidental poisoning by and exposure to narcot...	X42	1669	10727890	15.60
3	Unspecified fall	W19	1215	10727890	11.30
4	Person injured in unspecified motor-vehicle ac...	V89.2	459	10727890	4.30
...	...	...	...	...	...
58	Urinary catheterization	Y84.6	11	10727890	Unreliable
59	Pedestrian injured in collision with car, pick...	V03.1	10	10727890	Unreliable
60	Motorcycle rider injured in noncollision trans...	V28.4	10	10727890	Unreliable
61	Fall from, out of or through building or struc...	W13	10	10727890	Unreliable
62	Contact with other and unspecified machinery	W31	10	10727890	Unreliable

63 rows × 5 columns

#peel off data slice where cause of death references 'accidental poison'

cdc_ct_accidental_poison = cdc_ct_death[cdc_ct_death["cause_of_death"].str.contains("Accidental poison")]

cdc_ct_accidental_poison

	cause_of_death	cause_of_death_code	deaths	population	crude_rate_per_100000
1	Accidental poisoning by and exposure to other ...	X44	2159	10727890	20.10
2	Accidental poisoning by and exposure to narcot...	X42	1669	10727890	15.60
9	Accidental poisoning by and exposure to alcohol	X45	168	10727890	1.60
19	Accidental poisoning by and exposure to antiep...	X41	57	10727890	0.50
39	Accidental poisoning by and exposure to other ...	X47	21	10727890	0.20
50	Accidental poisoning by and exposure to nonopi...	X40	13	10727890	Unreliable

#drop row (index=[50]) with unreliable result

cdc_ct_accidental_poison = cdc_ct_accidental_poison.drop(index=[50])

cdc_ct_accidental_poison

	cause_of_death	cause_of_death_code	deaths	population	crude_rate_per_100000
1	Accidental poisoning by and exposure to other ...	X44	2159	10727890	20.10
2	Accidental poisoning by and exposure to narcot...	X42	1669	10727890	15.60
9	Accidental poisoning by and exposure to alcohol	X45	168	10727890	1.60
19	Accidental poisoning by and exposure to antiep...	X41	57	10727890	0.50
39	Accidental poisoning by and exposure to other ...	X47	21	10727890	0.20

#look at full cause of death strings in a list

ct_poison_cause_of_death_strings_list = cdc_ct_accidental_poison["cause_of_death"].to_list()

ct_poison_cause_of_death_strings_list

['Accidental poisoning by and exposure to other and unspecified drugs, medicaments and biological substances',
 'Accidental poisoning by and exposure to narcotics and psychodysleptics [hallucinogens], not elsewhere classified',
 'Accidental poisoning by and exposure to alcohol',
 'Accidental poisoning by and exposure to antiepileptic, sedative-hypnotic, antiparkinsonism and psychotropic drugs, not elsewhere classified',
 'Accidental poisoning by and exposure to other gases and vapours']

#make other columns lists for tiny df

ct_poison_deaths = cdc_ct_accidental_poison["deaths"].to_list()
ct_population = cdc_ct_accidental_poison["population"].to_list()
ct_crude_rate_per_100K = cdc_ct_accidental_poison["crude_rate_per_100000"].to_list()

#make new tiny df with only rows interested in and make new short titles for better viz

ct_poison = pd.DataFrame({'ct_poison_cause_of_death_short': ['Drugs, etc.', 'Narcotics, etc.', 'Alcohol', 'Antiepileptic drugs, etc.', "Gases and vapours"],
                             'ct_poison_cause_of_death_long': ct_poison_cause_of_death_strings_list,
                             'ct_poison_deaths': ct_poison_deaths,
                             'ct_population': ct_population,
                             'ct_poison_death_rate_per_100K': ct_crude_rate_per_100K
                            })

#preview data again

ct_poison

	ct_poison_cause_of_death_short	ct_poison_cause_of_death_long	ct_poison_deaths	ct_population	ct_poison_death_rate_per_100K
0	Drugs, etc.	Accidental poisoning by and exposure to other ...	2159	10727890	20.10
1	Narcotics, etc.	Accidental poisoning by and exposure to narcot...	1669	10727890	15.60
2	Alcohol	Accidental poisoning by and exposure to alcohol	168	10727890	1.60
3	Antiepileptic drugs, etc.	Accidental poisoning by and exposure to antiep...	57	10727890	0.50
4	Gases and vapours	Accidental poisoning by and exposure to other ...	21	10727890	0.20

#combine all drug rows into single variables to use in even smaller dataframe below

combined_drugs_poison_death_count = ct_poison.ct_poison_deaths[0] + ct_poison.ct_poison_deaths[1] + ct_poison.ct_poison_deaths[3]
combined_drugs_cause_of_death_long = ct_poison.ct_poison_cause_of_death_long[0] + " OR "+ ct_poison.ct_poison_cause_of_death_long[1] + " OR " +  ct_poison.ct_poison_cause_of_death_long[3]

print("Total combined number of deaths from drug poisoning:", combined_drugs_poison_death_count)
print(" ")
print("Total combined text strings (each separated by 'OR') about cause of death from poison --", combined_drugs_cause_of_death_long)

Total combined number of deaths from drug poisoning: 3885
 
Total combined text strings (each separated by 'OR') about cause of death from poison -- Accidental poisoning by and exposure to other and unspecified drugs, medicaments and biological substances OR Accidental poisoning by and exposure to narcotics and psychodysleptics [hallucinogens], not elsewhere classified OR Accidental poisoning by and exposure to antiepileptic, sedative-hypnotic, antiparkinsonism and psychotropic drugs, not elsewhere classified

#making even smaller dataset for ideal viz

ct_poison_min = pd.DataFrame({'ct_poison_cause_of_death_short': ['Drugs', 'Alcohol',  "Gases and vapours"],
                             'ct_poison_cause_of_death_long': [combined_drugs_cause_of_death_long, ct_poison_cause_of_death_strings_list[2], ct_poison_cause_of_death_strings_list[4]],
                             'ct_poison_deaths': [combined_drugs_poison_death_count, ct_poison.ct_poison_deaths[2], ct_poison.ct_poison_deaths[4]],
                             'ct_population': ct_population[0:3]
                            })

#use seaborn to plot - https://seaborn.pydata.org/generated/seaborn.barplot.html

fig = seaborn.barplot(data=ct_poison_min, 
                x="ct_poison_cause_of_death_short", 
                y="ct_poison_deaths", 
                palette="rocket", 
                orient="v")

fig.set_xlabel("Accidental poisoning types")
fig.set_ylabel("Number of deaths")
fig.set_title("Accidental poisoning deaths in Connecticut, 2018-2021*")

plt.savefig("ct_poison_2018-21.png")

#simplified version of combined text above - also see caveats below

print("* See full definition of 'drugs' below, plus caveats below.")
print("Drugs include other and unspecified drugs, medicaments and biological substances; narcotics and psychodysleptics [hallucinogens], not elsewhere classified; antiepileptic, sedative-hypnotic, antiparkinsonism and psychotropic drugs, not elsewhere classified.")

* See full definition of 'drugs' below, plus caveats below.
Drugs include other and unspecified drugs, medicaments and biological substances; narcotics and psychodysleptics [hallucinogens], not elsewhere classified; antiepileptic, sedative-hypnotic, antiparkinsonism and psychotropic drugs, not elsewhere classified.

Caveats & Notes from CDC Wonder data extract:

Caveats: Data are Suppressed when the data meet the criteria for confidentiality constraints. More information.

Death rates are flagged as Unreliable when the rate is calculated with a numerator of 20 or less. More information.

The population figures for year 2021 are single-race estimates of the July 1 resident population, from the Vintage 2021 postcensal series released by the Census Bureau on June 30, 2022. The population figures for year 2020 are single-race estimates of the July 1 resident population, from the Vintage 2020 postcensal series released by the Census Bureau on July 27, 2021. The population figures for year 2019 are single-race estimates of the July 1 resident population, from the Vintage 2019 postcensal series released by the Census Bureau on June 25, 2020. The population figures for year 2018 are single-race estimates of the July 1 resident population, from the Vintage 2018 postcensal series released by the Census Bureau on June 20, 2019.

The population figures used in the calculation of death rates for the age group 'under 1 year' are the estimates of the resident population that is under one year of age. More information.

Beginning with the 2018 data, changes have been implemented that affect the counts for ICD-10 cause of death codes O00-O99 compared to previous practice. In addition, data for the cause of death codes O00-O99 for 2003 through 2017 reflect differences in information available to individual states and probable errors. Caution should be used in interpreting these data. More information can be found at: https://www.cdc.gov/nchs/maternal-mortality/.

On March 11, 2021, the 2019 mortality data on CDC WONDER was updated with the 2019 mortality data updated by NCHS on March 4, 2021 to include corrected information for residents of Texas affecting 5 records previously coded to cause code *U01.4, Terrorism involving firearms (homicide). The underlying and multiple cause of death codes for 5 records were corrected in the 2019 data. Underlying and multiple cause of death codes for those 5 records were recoded to Assault (homicide) by other and unspecified firearm discharge, ICD-10 code X95. The corrected final death records replaces the data released on December 22, 2020.

Changes to cause of death classification affect reporting trends. More information.

Help:
See Underlying Cause of Death, 2018-2021, Single Race Documentation for more information.

Query Date: Mar 16, 2023 2:51:43 PM

Query Criteria: ICD-10 Codes: V01-Y89 (External causes of morbidity and mortality) States: Connecticut (09) Year/Month: 2019; 2020; 2021 Group By: Cause of death Show Totals: True Show Zero Values: False Show Suppressed: False Calculate Rates Per: 100,000 Rate Options: Default intercensal populations for years 2001-2009 (except Infant Age Groups)