Characteristics of Donors and Modelling of the Characteristics to Possible Forecast the Repeat Donors Profile at Thammasat University Hospital.

Blood donations are essential to the blood supply available for patients in life-saving treatments. We aimed to identify characteristics affecting repeat donations, and to model a tool to forecast repeat donation among Thammasat University Hospital (TUH) donors. A retrospective study for 4 years of donations at TUH was conducted to identify characteristics affecting continuous donations and model a scoring tool, as well as pilot test it, prospectively. Data concerning age, sex, ABO grouping, Rh(D) typing, and collection site were included. The outcome was dichotomized as controls and cases based on first time and repeat donations. Receiver operating characteristic curve was used to obtain the cut-off, while odds ratio was used to assign the score. During the study, 37,736 donations comprised 6,305 controls and 31,431 cases. Characteristics that positively predicted repeat donation included male, age ≥ 30 years, AB blood group and on-site donation, and they were chosen to model the score. The total score value of 3 was chosen as the rounded cut-off. A pilot study, the score was observed to have an accuracy of 67.5%. In conclusion, 4 significant characteristics appeared to positively influence repeat donation. The predictive scoring model is a simple reliable and valid tool exhibiting good accuracy.