Assessing the Relationship Between Care Processes and Clinical Decision Support for Order Entry

Adherence to best practices and clinical evidence varies across hospitals due to multiple barriers, including hospital resources, patient characteristics, and emergency department crowding. These barriers can affect clinical decision making, as reflected in safe and efficient order entries, and hospitals have implemented clinical decision support (CDS) tools within the electronic health record (EHR) to reduce care variation. Order sets, a CDS tool within the EHR, suggests recommended best practices and orders as bundles to standardize care and improve efficiencies in ordering medications and care management. While assumed to provide improvements, the benefits and barriers to usage of order sets are understudied.

This research looked at the relationship between the use of order sets, order variation at the patient level, patient outcomes, and provider attitudes toward the usage of order sets.

The specific aims were as follows:

• Identify potential barriers to order set use involving clinician attitudes and perceptions.
• Assess the relationships between order set use and care processes.
• Assess the relationship between order set use and attitudes and perceptions of clinicians.
• Assess the relationship between order set use and healthcare variation.

A retrospective cohort study was conducted on data from three sites of the New York-Presbyterian Hospital system, focused on patients with sepsis. The researchers carried out three investigations: first they assessed the relationship between order set use and care variation while controlling for diagnosis, patient complexity, and location; second, they examined the relationship between usage and outcomes while controlling for the same items; and third, they surveyed clinicians in internal medicine, surgery, and emergency medicine regarding their attitudes toward order sets. A metric called longest common subsequence (LCS) was used to look at order variation. LCS considers the type, frequency, and general trend in the sequences of orders.

Order set usage was associated with less order variation and better outcomes for patients hospitalized with sepsis: among those with a high proportion of order set orders, there was a significantly shorter length of stay and lower inpatient mortality versus those with low proportion of order set orders. In addition, this group had a shorter time to first antibiotics ordered in the emergency department. The clinician surveys showed that, whether the order sets for pain management were acceptable for use was related to several factors, including: the ability of the order set to serve as decision support; leadership and peer influence toward use; endorsement by the organization; sufficient training; and how well the sets are integrated into the electronic health record and, thus, a clinician’s workflow.

The researchers hope that their study will improve the understanding of how to design and maintain order sets from users’ perspectives to drive better care processes and outcomes.