Recency Bias in Organ Transplantation: Does Your Doctor Suffer From the “Hot Hand” Fallacy?

Recency bias is a cognitive bias that may generate irrational decision making. It is based on the belief that someone who has experienced favorable draws from a random distribution will continue to receive favorable draws in the future. This cognitive bias is referred to as the “hot hand” fallacy because it has been applied to explain gambling behavior as well as financial investments. To date, this behavioral response has not been observed in the medical care literature. Using data on deceased donor kidney offers made to transplant surgeons for their patients, we investigate whether or not transplant surgeons suffer from the “hot hand” fallacy. Each deceased donor kidney offer can be viewed as a random draw from a broader distribution of deceased donor organs. However, given the high number of kidney offers received by each patient before transplantation it is possible that favorable draws from the distribution may bias a transplant surgeon’s perception of the underlying distribution. Our central hypothesis is that if a transplant surgeon receives a series of high quality deceased donor organ offers for their patient they may be less likely to accept subsequent high quality offers because they expect for this to continue. This would generate recency bias in organ utilization. In our analysis donor quality is calculated using the kidney donor profile index (KDPI), a measure commonly used in kidney transplantation to quantify donor quality. In addition to our measure of donor quality we utilize patient and donor-specific information to control for other mitigating factors that may influence the decision to accept an organ. Additional controls are used for transplant center fixed effects and temporal trends in the data as well as the regulatory standing of the transplant center. Using a series of different time windows (i.e., prior 30-day, 60-day, 90-day, etc..) to define recent decision processes we observe the presence of recency bias in organ utilization. Furthermore, we observe a parabolic response function with regard to the extent of the recency bias present in the decision process. The bias rises then falls as longer time windows are utilized. Knowledge of this cognitive bias in organ transplantation will help to better understand the decision process of transplant surgeons and perhaps inform them of the need to mitigate this bias in order to better utilize our short supply of transplantable organs.