Medication errors are associated with increased incidence of infection, rejection, and graft loss in kidney transplant recipients. This study aims to create a model to predict patients at highest risk of medication errors to streamline workflow in a kidney transplant clinic caring for long-term recipients.

This was a prospective, observational study in adult kidney transplant recipients who were ≥ 90 days post-transplant. Prior to their clinic visit, patients were administered a survey assessing medication adherence, perception of health status, and current medication regimen. Subsequently, pharmacists conducted a blinded medication history encounter and documented the medication errors discovered during the visit. A predictive model was then created using binary logistic regression with backward elimination.

This study included 237 patients (Table 1). The logistic model had good predictability, with 61.5% and 66.3% specificity and sensitivity, respectively, for identifying patients likely to have ≥ 6 medication errors. This model had an AUC of 0.724 (Figure 1) and includes the following 12 variables: lack of or unknown baseline calcineurin inhibitor (CNI), use of Social Security disability as income (SSDI), use of Medicaid for prescription insurance, poor health status rating, number of daily scheduled medications, current anti-diabetic regimen, number of antihypertensive medications, and medication adherence, affordability, etc. (Table 2). A more parsimonious model, which included 9 variables, was also created for ease of use in a clinic setting. This model had a sensitivity and specificity of 62.5% and 66.7%, and an AUC of 0.72 (Figure 2).

These results demonstrate that a simple 5 minute patient survey conducted in kidney transplant recipients may be capable of accurately predicting patients at high risk of medication errors, which can be used as a screening tool for transplant pharmacist's interventions. These results require external validation prior to implementation.

.

CITATION INFORMATION: Covert K, Mardis C, Fleming J, Mardis A, Meadows H, Pilch N, Mohan P, Posadas M, Salazar M, Srinivas T, Mour G, Taber D. Predictive Model for Medication Errors in a Chronic Kidney Transplant Clinic. Am J Transplant. 2016;16 (suppl 3).

« Back to 2016 American Transplant Congress