Related Experiment Videos
Predicting blood donor arrival.
Vidar Bosnes1, Magne Aldrin, Hans Erik Heier
1Department of Immunology and Transfusion Medicine, Ullevål University Hospital, Oslo, Norway. vidar.bosnes@uus.no
Transfusion
|January 22, 2005
Summary
Predicting blood donor arrival using statistical modeling significantly reduces prediction intervals by 43%. This helps blood banks better plan sessions and minimize donor wait times for improved experiences.
Area of Science:
- Biomedical Informatics
- Public Health
- Operations Research
Background:
- Short waiting times enhance donor experience and encourage repeat donations.
- Blood banks face challenges with variable donor arrival rates despite using fixed appointments.
- Predictive methods are needed to manage donor flow and reduce wait times.
Purpose of the Study:
- To develop a statistical model for predicting blood donor arrival.
- To identify key factors influencing donor attendance.
- To improve the efficiency of blood donation sessions.
Main Methods:
- Collected data on 179,121 appointments over 971 days.
- Utilized logistic regression to model blood donor arrival prediction.
- Analyzed 18 candidate explanatory variables.
Main Results:
- Key predictors include time to appointment, contact medium, donor age, donation history, and past attendance/no-show rates.
- The model reduced prediction intervals by 43% compared to using average arrival rates.
- Identified significant variables influencing donor no-shows and attendance.
Conclusions:
- Statistical modeling offers valuable insights into blood donor arrival patterns.
- Predictive models enable better planning and resource allocation for blood donation sessions.
- Improved planning can lead to reduced donor waiting times and enhanced satisfaction.