Analysis for Patients in Brazil Medical appointments and centered on the question of whether or not patients show up for their appointment
The Dataset can be obtained from Kaggle. It contains 100K records of medical appointments in Brazil and is focused on if patients will show up or not.
It has the following patient charecteristics:
- PatientId: Identification of a patient
- AppointmentID: Identification of each appointment
- Gender: Male or Female.
- AppointmentDay: The day of the actual appointment, when they have to visit the doctor.
- ScheduledDay: The day someone called or registered the appointment, this is before appointment of course.
- Age: How old is the patient.
- Neighbourhood: Where the appointment takes place.
- Scholarship: True of False . Observation, on Bolsa Familia
- Hipertension: True or False
- Diabetes: True or False
- Alcoholism: True or False
- Handcap: True or False
- SMS_received: 1 or more messages sent to the patient.
- No-show: True or False.
To create an environment from an environment file, use the following command:
conda env create -f environment.yaml
Among the selected features, age, gender, and all diagnostic diseases features(hipertension, diabetes, alcoholism) and also patients on the Bolsa Familia scholarship, tend to have no significantly dominant diagnosis group with the target no_show as they have the same attendance rate like the rest of the population with a a difference between .
Patients who have received sms messages surprisingly have a higher chance of not showing up for their appointment with ~7% increase compare with the rest of the population while the latter had a no-show rate reduction of 3.49%. Among the 35 out of 81 neighborhoods with atleast 1% patient population, 10 with a lower no-show rate ranged 15.5% - 18.5%.