Diabetes mellitus in living pancreas donors: Use of integrated national registry and pharmacy claims data to characterize donation-related health outcomes

Abstract

Background. Living donor pancreas transplant is a potential treatment for diabetic patients with end-organ complications. Although early surgical risks of donation have been reported, long-term medical outcomes in living pancreas donors are not known. Methods. We integrated national Scientific Registry of Transplant Recipients data (1987-2015) with records from a nationwide pharmacy claims warehouse (2005-2015) to examine prescriptions for diabetes medications and supplies as a measure of postdonation diabetes mellitus. To compare outcomes in controls with baseline good health, we matched living pancreas donors to living kidney donors (1:3) by demographic traits and year of donation. Results. Among 73 pancreas donors in the study period, 45 were identified in the pharmacy database: 62%women, 84%white, and 80%relatives of the recipient. Over a mean postdonation follow-up period of 16.3 years, 26.7% of pancreas donors filled prescriptions for diabetes treatments, compared with 5.9% of kidney donors (odds ratio, 4.13; 95% confidence interval, 1.91-8.93; P = 0.0003). Use of insulin (11.1% vs 0%) and oral agents (20.0% vs 5.9%; odds ratio, 4.50, 95% confidence interval, 2.09-9.68; P = 0.0001) was also higher in pancreas donors. Conclusions. Diabetes is more common after living pancreas donation than after living kidney donation, supporting clinical consequences from reduced endocrine reserve.