Order sets are a CDS function which presents multiple orders for a particular clinical purpose as a set (such as an `Admission Order Set') for clinicians to select2. Order sets are expected to improve patient safety by reducing prescribing variations and errors, and also facilitate efficient order placement based on best practices and guidelines4,5. The creation of order sets has been considered a requirement for a successful CPOE implementation2,5,6. However, the association among order sets, their expected benefits, and barriers for their usage, is understudied.4,7,8 The goal of this pilot study is twofold. First, we will identify potential scenarios where the use of order sets leads to better outcomes than non-use of order sets, specifically opioid prescribing, by applying advanced data analytics to historical data. Data on order placement are extracted from electronic health records (EHR). In parallel, we will use a survey method to understand perceptions on order sets, in order to remove barriers to the use of this CDS tool which are often reported to not be utilized to its maximum benefit. In particular, we are interested in the role of order sets in the prescription of opioid medications during patients inpatient and emergency room (ER) visits. Our overarching hypothesis for this project is that order set use is associated with improved quality of care (i.e., fewer unexplained variations in care, and reduced opioid prescribing overall), yet clinician-level barriers are limiting uptake of this CDS modality. Analyses will be conducted on ordering data from the Department of Internal Medicine, Surgery, and Emergency Medicine.
In Aim 1, we will assess the relationship between order set use and care variation within Internal Medicine, Surgery, and Emergency Medicine, respectively, while controlling for principal diagnoses, patient complexity, and campus locations. We hypothesize that more frequent use of order sets is associated with reduced care variations while controlling for principal diagnoses, patient complexity, and campus locations.
In Aim 2, we will compare the number of opioids prescribed from order sets and prescribed as standalone orders. We hypothesize that more frequent use of order sets is associated with reduced opioid prescriptions while controlling for principal diagnoses, patient complexity, and campus locations.
In Aim 3, a survey will be conducted with Internal Medicine, Surgery, and Emergency Medicine clinicians in 3 campuses associated with NYP Hospital. We hypothesize that trust and self-efficacy about order sets are associated with an increased order set use while controlling for satisfaction, level of experience, IT training, department, and campus location. Findings from this study may lay the groundwork for prospective large-scale and interventional studies to strategize safe and efficient opioid prescriptions through order sets that have sufficient clinician uptake.
This proposed study aims to add to existing evidence about the value of order sets, a clinical decision support for computerized provider order entry, while uncovering clinician perceptions that hinder their efficient use. We will design and conduct a survey to clinicians to identify potential barriers to order set. We will assess the relationships between order set use, care process, and care outcome, through analyses of electronic health record data.
