Risk Models to Improve Long-term Care Medication Safety

摘要

a) Purpose: To determine if socio-technical probabilistic risk assessment (ST-PRA) can create statewide risk models identifying combinations of medication delivery system and behavioral elements producing wrong drug, wrong dose, wrong resident, and omission medication errors in nursing and community based care (CBC) facilities. b) Scope: Long-term care (LTC) providers and state agencies, disappointed by the failure of educational and regulatory interventions to improve medication delivery system safety, designed this study to focus on system risks within the control of nursing, assisted living, and residential care facilities. c) Methods: This developmental study uses four tools-process mapping, control system mapping, modified failure modes and effects analysis (FMEA), and socio-technical probabilistic risk assessment (ST-PRA) to construct risk models. Multidisciplinary teams from a convenience sample often nursing and eight assisted living/residential care facilities created the models, with input from pharmacists, community physicians, and state surveyors. A stratified, random sample of 20 nursing and CBC facilities were surveyed to determine if critical elements in the models generally represent medication delivery systems in similar facilities across Oregon. d) Results: Nursing and CBC risk models weresuccessfully completed. Prescribing and administration errors are the models' dominant risks. Multiple single failure path errors were identified. Validation survey data confirmed 89.1% of selected exposure and error rates in the models were comparable to values from a statewide sample of nursing and CBC facilities. Oregon LTC providers, state agencies, pharmacy companies, and medical providers are collaborating on strategies to address the risks identified in these models.