sxtforreal/ehr_data_analysis

Electronic Health Record Analysis using SQLite

For end user:

Setup/installation instructions:

$ git clone https://github.com/sxtforreal/ehr_analysis.git $ cd ../path/to/the/file

The expected input file formats

• parse_data() takes text files with the first row as variable names as input.
• Remaining functions takes the output of parse_data as input.

API description:

This repo contains modules working with electronic health record data. ** Defined classes:

• class:Patient has 6 instance attributes: ID, gender, DOB, race, marital, language, PPPBP 1 property: Age
• class:Lab has 5 instance attributes: PatientID, Admission, Name, Value, Units, Time 1 method: visit_age ** Defined functions:
• parse_data() parse the input ehr text file into a list of class:Patient or class:Lab.
• num_older_than() takes a list of class:Patient and return the number of patients older than a given age (in years).
• sick_patients() take a list of class:Lab and return a unique list of patients who have a given test with value above (">") or below ("<") a given level.
• first_admission() take a list of class:Patient, a list of class:Lab and computes the age at first admission of any given patient id and lab name.

Examples:

• A = parse_data('/Users/sunxiaotan/Desktop/BIO821/ehr/PatientCorePopulatedTable.txt', 'patient')
• B = parse_data('/Users/sunxiaotan/Desktop/BIO821/ehr/LabsCorePopulatedTable.txt', 'lab')
• num_older_than(50, A) -> 3
• sick_patients("METABOLIC: GLUCOSE", ">", 103.0, B) -> ["FB2ABB23-C9D0-4D09-8464-49BF0B982F0F"]
• first_admission(B, A, "FB2ABB23-C9D0-4D09-8464-49BF0B982F0F", "METABOLIC: GLUCOSE") -> 44

For contributor:

Testing instructions:

• Test: pytest test_ehr_analysis.py
• Coverage: pytest --cov=..