Analysis of dosing-button compliance.

PURPOSE: A project was undertaken at an academic medical center to assess use of available dosing buttons within the computerized provider-order-entry (CPOE) system in order to identify opportunities for optimization of medication builds. METHODS: A retrospective observational study was conducted to identify medication records within a CPOE system meeting prespecified inclusion and exclusion criteria. A report capturing all inpatient adult medication orders associated with the identified medication records over a 6-month period was generated. The primary endpoint was percent dosing-button compliance, calculated as the number of orders with doses consistent with existing dosing-button options divided by the total number of orders during the study period. Secondary study objectives included a comparison of high- and low-performing medication record samples and identification of potential reasons for lack of dosing-button use. RESULTS: A total of 2,506 CPOE medication records associated with a total of 694,877 medication orders entered during the study period were analyzed. Median percent dosing-button compliance was 99.92% (interquartile range, 83.33-100%). High-performing records (n = 1243) were more likely to be associated with anti-infective medications (p = 0.041) and medications not on formulary at the study institution (p < 0.001). Medications in the sample of poor-performing CPOE records (n = 614) were more likely to be agents delivered via the i.v. route (p < 0.001). There were 45 records for which poor dosing-button compliance was attributed to lack of a clinically reasonable dosing option. CONCLUSION: A high level of dosing-button compliance was demonstrated despite the lack of routine revalidation of dosing buttons after initial medication builds. Some opportunity for optimization was identified during the project, which established a quality assurance method to facilitate future auditing of medication builds.

Developing an Anti-Xa-Based Anticoagulation Protocol for Patients with Percutaneous Ventricular Assist Devices.

Because of the complexities associated with anticoagulation in temporary percutaneous ventricular assist device (pVAD) recipients, a lack of standardization exists in their management. This retrospective analysis evaluates current anticoagulation practices at a single center with the aim of identifying an optimal anticoagulation strategy and protocol. Patients were divided into two cohorts based on pVAD implanted (CentriMag (Thoratec; Pleasanton, CA) / TandemHeart (CardiacAssist; Pittsburgh, PA) or Impella (Abiomed, Danvers, MA)), with each group individually analyzed for bleeding and thrombotic complications. Patients in the CentriMag/TandemHeart cohort were subdivided based on the anticoagulation monitoring strategy (activated partial thromboplastin time (aPTT) or antifactor Xa unfractionated heparin (anti-Xa) values). In the CentriMag/TandemHeart cohort, there were five patients with anticoagulation titrated based on anti-Xa values; one patient developed a device thrombosis and a major bleed, whereas another patient experienced major bleeding. Eight patients received an Impella pVAD. Seven total major bleeds in three patients and no thrombotic events were detected. Based on distinct differences between the devices, anti-Xa values, and outcomes, two protocols were created to guide anticoagulation adjustments. However, anticoagulation in patients who require pVAD support is complex with constantly evolving anticoagulation goals. The ideal level of anticoagulation should be individually determined using several coagulation laboratory parameters in concert with hemodynamic changes in the patient's clinical status, the device, and the device cannulation.