We will test novel methods from complexity science to examine the impact of care team structure on medication safety and errors. An analytic approach will be applied in which newborns, the clinicians caring for them and patient/clinician connections are modeled as networks. Using network visualizations and quantitative metrics we will characterize care teams composed of bedside and advanced practice nurses, pharmacy staff and physicians. The relationship between team characteristics and medication errors will be explored. Medication safety events will be identified using three existing data sources. First, the institution's Patient Safety Reporting System will be queried. Second, we will use the pharmacy's intervention tracking system to identify events requiring staff action. Finally, we will apply a validated NICU trigger tool to the electronic health record (EHR). We will categorize identified events using an adaptation of the NCC MERP index. Patients and their connections to other patients and providers will constitute the networks of interest. Networks will be constructed by detecting patient-patient, patient-provider and provider-provider connections using methods developed by our group that utilize routinely collected EHR based information. This approach allows identification and characterization of the individuals and teams that participate in care for individual patients or groups of patients. Organizing and analyzing data in this way will allow rapid quantification of an individual clinician's workload at a specific point n time, the geographic dispersal of their patient care responsibilities, and the importance of the clinician to information flow within the care team, as well as that clinician's prior experience wih index patients and medications of interest. In addition, this approach allows characterization of the handoff sequences as well as the fault tolerance of team communication existing within the NICU and pharmacy microsystems. These individual and team characteristics will then be entered into multivariate models attempting to predict medication errors and preventable harm. The proposed work will be done in a tertiary care NICU service located in a general hospital with a high risk obstetric service. The institution is served by a robust home-grown EHR including computerized provider order entry and a comprehensive pharmacy information system. The population includes newborns requiring the full range of care up to and including intensive care.
Our specific aims will be to: " Identify/characterize medication safety events that occur in the NICU and other newborn care areas " Describe the team and handoff structures that exist within medication processes in newborn care " Examine the impact care team characteristics on medication errors and related preventable harm. This project represents an innovative approach to understanding how the complex interactions existing in large inter-disciplinary teams impact medication safety. Identification of team structures that are associated with lower rates of medication errors and preventable harm could lead to design of important safety interventions.

Public Health Relevance

This project will examine how the complex teams and systems of care present within the NICU impact the medication errors. It will use innovative methods from network science to understand how the structure of health care teams influences the rates of these events.

Agency
National Institute of Health (NIH)
Institute
Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21HD072880-01
Application #
8334863
Study Section
Health Services Organization and Delivery Study Section (HSOD)
Program Officer
Raju, Tonse N
Project Start
2012-08-13
Project End
2014-07-31
Budget Start
2012-08-13
Budget End
2013-07-31
Support Year
1
Fiscal Year
2012
Total Cost
$217,500
Indirect Cost
$92,500
Name
Beth Israel Deaconess Medical Center
Department
Type
DUNS #
071723621
City
Boston
State
MA
Country
United States
Zip Code
02215