Presentation
Scheduled Medication Administration in a Pediatric Intensive Care Unit: It’s Harder Than You Think
SessionPoster Session 1
DescriptionBackground: Medication administration is a core process in healthcare and critical to the delivery of quality patient care. A central principle of medication administration is the five rights or “five R’s”: right patient, medication, route, dose, and time (Hanson and Haddad, 2023). To help achieve the five R’s, technologies including electronic health records with computerized order entry, controlled access cabinets, “smart” infusion pumps, and medication barcoding have been implemented. Despite these technologies, adverse drug events (ADEs) cause over 770,000 deaths or injuries annually in the US (Slight et al., 2018). ADEs impact all areas of care but are particularly challenging to prevent and manage in pediatric intensive care units (PICUs), where patients are medically complex, depend on weight-based dosing, and rely on high-risk therapies (Alshammari et al., 2022; Holdsworth et al., 2003; Kaushal et al., 2001).
One potential explanation for this continued preventable harm from medications is the difference between ‘work as imagined’ and ‘work as done’ (Hollnagel, 2017). Work as imagined reflects how individuals believe the process of administering medications occurs, whereas work as done represents the reality of medication administration by healthcare professionals in complex clinical environments. Strikingly, little research has examined the actual work done by nurses during medication administration (Hawkins and Morse, 2022; Hermanspann et al., 2019; Huynh et al., 2016), which may help explain the limited progress in improving medication safety. To address this gap in the literature, our study characterized the observable steps and structure of scheduled medication administration by nurses in the PICU, providing an initial framework to understand its complexity.
Methods: A prospective, observational study was conducted in a 44-bed academic PICU within a free-standing, tertiary-care children’s hospital. The institutional review board deemed the study non-human research. Two observers (AC and AB) were trained by recording observations in a non-clinical setting and reviewing their findings with three human factors engineering (HFE) professionals. Nurses provided verbal consent to be observed during scheduled medication administration. Initially, both observers shadowed the same nurse to align data collection processes and terminology. Early scheduled medication administration observations were subsequently reviewed with HFE professionals and a nursing expert. Observations focused exclusively on workflow; no patient or nurse identifiers were collected. Observations were first recorded in writing and then translated into flowcharts of the core processes of scheduled medication administration. Flowcharts were further annotated to capture observed problems, variation in nursing methods, workflow irregularities, and other notable findings. Processes were grouped into three subcategories of the scheduled medication administration: pre-care (obtaining patient and medication information), medication preparation (obtaining and preparing medications), and administration of medication (administering medication and documentation).
Results: Observers recorded 54 hours of scheduled medication administration, generating 40+ pages of flowcharts covering pre-care (7), preparation (16), and administration (18) processes. Findings revealed that administration of medications could not be studied in isolation from pre-care and medication preparation work, as all phases were integral to the overall process. Compilation of the flowcharts provided high-level yet detailed documentation of the breadth of medication administration work.
Importance/Take Away: The observational data revealed considerable variation between nurses and within the practice of individual nurses, resulting in non-linear flowcharts where steps/processes were often repeated. The numerous permutations demonstrated that the work as done of medication administration is far more complex and less algorithmic than previously understood. This complexity requires nurses to rely on their expertise to navigate patient- and situation-specific variations. These findings highlight the complexity of PICU scheduled medication administration work as done and provides a framework to guide targeted interventions to improve safety and efficiency.
Limitations/Future Work: Limitations of this study include capturing only the observable physical tasks of nurses; therefore, the flowcharts do not reflect the many cognitive steps and decision-making processes involved in medication administration. Additionally, only scheduled medications were observed during weekday day shifts, which may not fully represent the scope of medication administration. Future work will examine workflows across different nursing shifts and during urgent/emergent medication administration to provide a more comprehensive understanding.
One potential explanation for this continued preventable harm from medications is the difference between ‘work as imagined’ and ‘work as done’ (Hollnagel, 2017). Work as imagined reflects how individuals believe the process of administering medications occurs, whereas work as done represents the reality of medication administration by healthcare professionals in complex clinical environments. Strikingly, little research has examined the actual work done by nurses during medication administration (Hawkins and Morse, 2022; Hermanspann et al., 2019; Huynh et al., 2016), which may help explain the limited progress in improving medication safety. To address this gap in the literature, our study characterized the observable steps and structure of scheduled medication administration by nurses in the PICU, providing an initial framework to understand its complexity.
Methods: A prospective, observational study was conducted in a 44-bed academic PICU within a free-standing, tertiary-care children’s hospital. The institutional review board deemed the study non-human research. Two observers (AC and AB) were trained by recording observations in a non-clinical setting and reviewing their findings with three human factors engineering (HFE) professionals. Nurses provided verbal consent to be observed during scheduled medication administration. Initially, both observers shadowed the same nurse to align data collection processes and terminology. Early scheduled medication administration observations were subsequently reviewed with HFE professionals and a nursing expert. Observations focused exclusively on workflow; no patient or nurse identifiers were collected. Observations were first recorded in writing and then translated into flowcharts of the core processes of scheduled medication administration. Flowcharts were further annotated to capture observed problems, variation in nursing methods, workflow irregularities, and other notable findings. Processes were grouped into three subcategories of the scheduled medication administration: pre-care (obtaining patient and medication information), medication preparation (obtaining and preparing medications), and administration of medication (administering medication and documentation).
Results: Observers recorded 54 hours of scheduled medication administration, generating 40+ pages of flowcharts covering pre-care (7), preparation (16), and administration (18) processes. Findings revealed that administration of medications could not be studied in isolation from pre-care and medication preparation work, as all phases were integral to the overall process. Compilation of the flowcharts provided high-level yet detailed documentation of the breadth of medication administration work.
Importance/Take Away: The observational data revealed considerable variation between nurses and within the practice of individual nurses, resulting in non-linear flowcharts where steps/processes were often repeated. The numerous permutations demonstrated that the work as done of medication administration is far more complex and less algorithmic than previously understood. This complexity requires nurses to rely on their expertise to navigate patient- and situation-specific variations. These findings highlight the complexity of PICU scheduled medication administration work as done and provides a framework to guide targeted interventions to improve safety and efficiency.
Limitations/Future Work: Limitations of this study include capturing only the observable physical tasks of nurses; therefore, the flowcharts do not reflect the many cognitive steps and decision-making processes involved in medication administration. Additionally, only scheduled medications were observed during weekday day shifts, which may not fully represent the scope of medication administration. Future work will examine workflows across different nursing shifts and during urgent/emergent medication administration to provide a more comprehensive understanding.
Authors
Event Type
Poster Presentation
TimeMonday, March 234:45pm - 6:15pm EDT
LocationRhinelander Gallery
Patient Safety Research and Initiatives