Machine Learning to avoid “No-Shows” at medical appointments

Problem Statement: One of the largest healthcare networks in Chile needed to reduce the quantity of “No-Shows” at their offices and understand the factors that caused an increase in 30% to “No-Shows” medical appointments. 

Solution: Zenta developed data mining and machine learning techniques with google cloud technology to interpret and learn from historical data in order to predict patient behavior and estimate the probability that a patient would attend the appointment. 

Results:This system helped decrease the number of “No-Shows” appointments in the health centers by sending reminders to the patients and also allowing them to reschedule with ease when needed. As a result, medical “No-Shows” decreased to 12% and a 5% overbooking was achieved. This saved + US $500,000 p/month – US 6M year. In addition, algorithms and models were created that allow the prioritization of doctors according to the needs of the organization. Zenta developed M

Red Salud is one of the largest health networks in Chile, located in every region with over 200 hospitals and medical attention centers.

On many occasions, the patients request appointments but do not end up going. Red Salud wanted to understand the main factors that provoked the no show tendency in order to identify the patients who were most likely to not attend their appointments. 

Zenta was able to develop the system of machine learning through google cloud to predict no-shows using 4 years of data corresponding to medical appointments and their follow up policies. 

These tools created a decrease in no shows and with the help of appointment reminders via email and confirmation messages Red Salud was able to avoid lost appointments and double bookings. Today the solution is still being implemented by Red Salud. 

Do you have a project  in mind?