About this Dataset

Predict Survival Of Patients With Heart Failure

Quick Start 🚀: If you're not up for reading all of this, head straight to the file section. There, you'll find detailed explanations of the files and all the variables you need.

This dataset contains the medical records of 299 patients who had heart failure, collected during their follow-up period, where each patient profile has 13 clinical features.

Dataset Characteristics: Multivariate

Subject Area: Health and Medicine

Associated Tasks: Classification, Regression, Clustering

Feature Type: Integer, Real

Instances: 299

Features: 12

Dataset Information

A detailed description of the dataset can be found in the Dataset section of the following paper:

Title:
Machine learning can predict survival of patients with heart failure from serum creatinine and ejection fraction alone

Authors:

Davide Chicco
Giuseppe Jurman
Source:
BMC Medical Informatics and Decision Making 20, 16 (2020)

DOI:
https://doi.org/10.1186/s12911-020-1023-5

Dataset Details

Feature	Explanation	Measurement	Range
Age	Age of the patient	Years	[40,..., 95]
Anaemia	Decrease of red blood cells or hemoglobin	Boolean	0, 1
High blood pressure	If a patient has hypertension	Boolean	0, 1
Creatinine phosphokinase	Level of the CPK enzyme in the blood	mcg/L	[23,..., 7861]
(CPK)
Diabetes	If the patient has diabetes	Boolean	0, 1
Ejection fraction	Percentage of blood leaving the heart at each	Percentage	[14,..., 80]
	contraction
Sex	Woman or man	Binary	0, 1
Platelets	Platelets in the blood	kiloplatelets/mL	[25.01,..., 850.00]
Serum creatinine	Level of creatinine in the blood	mg/dL	[0.50,..., 9.40]
Serum sodium	Level of sodium in the blood	mEq/L	[114,..., 148]
Smoking	If the patient smokes	Boolean	0, 1
Time	Follow-up period	Days	[4,...,285]
(target) death event	If the patient died during the follow-up period	Boolean	0, 1

Statistical quantitative description of the category features

number of patients. %: percentage of patients. Full sample: 299 individuals. Dead patients: 96 individuals. Survived patients: 203 individuals.

Category feature	Full sample	Dead patients	Survived patients
Anaemia (0: false)
	#	%	#
	170	56.86	50
Anaemia (1: true)
	#	%	#
	129	43.14	46
High blood pressure (0: false)
	#	%	#
	194	64.88	57
High blood pressure (1: true)
	#	%	#
	105	35.12	39
Diabetes (0: false)
	#	%	#
	174	58.19	56
Diabetes (1: true)
	#	%	#
	125	41.81	40
Sex (0: woman)
	#	%	#
	105	35.12	34
Sex (1: man)
	#	%	#
	194	64.88	62
Smoking (0: false)
	#	%	#
	203	67.89	66
Smoking (1: true)
	#	%	#
	96	32.11	30

Statistical quantitative description of the category features

Full sample: 299 individuals. Dead patients: 96 individuals. Survived patients: 203 individuals. σ: standard deviation

Here's the organized table:

Numeric feature	Full sample	Dead patients	Survived patients
	Median	Mean	σ
Age	60.00	60.83	11.89
Creatinine phosphokinase	250.00	581.80	970.29
Ejection fraction	38.00	38.08	11.83
Platelets	262.00	263.36	97.80
Serum creatinine	1.10	1.39	1.03
Serum sodium	137.00	136.60	4.41
Time	115.00	130.30	77.61

Deep dive into the dataset

Dataset Overview:

Medical records of 299 heart failure patients collected at Faisalabad Institute of Cardiology and Allied Hospital in Faisalabad, Punjab, Pakistan, during April–December 2015.
Patients had left ventricular systolic dysfunction and were classified as NYHA classes III or IV.
Consisted of 105 women and 194 men, aged between 40 and 95 years old.

Features:

13 features including clinical, body, and lifestyle information.
Binary features: anaemia, high blood pressure, diabetes, sex, and smoking.
Anaemia defined as haematocrit levels lower than 36%.
Definition of high blood pressure not provided in the dataset.
Creatinine phosphokinase (CPK) indicates the level of CPK enzyme in blood, possibly indicating heart failure or injury with high levels.
Ejection fraction measures the percentage of blood pumped out by the left ventricle with each contraction.
Serum creatinine indicates kidney function; high levels may suggest renal dysfunction.
Serum sodium test checks sodium levels in the blood, abnormal levels may indicate heart failure.
Death event feature used as target in binary classification study, indicating if the patient died or survived during the follow-up period (130 days on average).

Dataset Characteristics:

Dataset represented as a table with 299 rows (patients) and 13 columns (features).
Imbalance in the dataset with 203 survived patients and 96 dead patients.
Survival rate: 67.89% negatives (survived), 32.11% positives (died).
Further details and changes to feature names available in the original dataset curator's publication.

Note 📝: If you find this dataset useful, please consider giving it an upvote! Your support is appreciated.

Tables

Heart Failure Clinical Records Dataset

@kaggle.rabieelkharoua_predict_survival_of_patients_with_heart_failure.heart_failure_clinical_records_dataset

14.46 KB
299 rows
13 columns


CREATE TABLE heart_failure_clinical_records_dataset (
  "age" DOUBLE,
  "anaemia" BIGINT,
  "creatinine_phosphokinase" BIGINT,
  "diabetes" BIGINT,
  "ejection_fraction" BIGINT,
  "high_blood_pressure" BIGINT,
  "platelets" DOUBLE,
  "serum_creatinine" DOUBLE,
  "serum_sodium" BIGINT,
  "sex" BIGINT,
  "smoking" BIGINT,
  "time" BIGINT,
  "death_event" BIGINT
);