RESUMO
BACKGROUND: Patients with cardiogenic shock or end-stage heart failure can be maintained on mechanical circulatory support (MCS) devices. Once a patient undergoes placement of a device, obtaining and maintaining therapeutic anticoagulation is vital. Guidelines recommend the use of institutional protocols to assist in dosing and titration of anticoagulants. OBJECTIVE: The purpose of this study was to characterize the use of bivalirudin before and after the implementation of a standardized titration protocol in patients with MCS. METHODS: A retrospective review of patients who received bivalirudin for MCS (VA ECMO [veno-arterial extracorporeal membrane oxygenation], Impella, or LVAD [left ventricular assist device]) before and after the implementation of the titration protocol into the electronic health record (EHR) was conducted. The primary outcome was to compare the proportion of therapeutic activated partial thromboplastin time (aPTT). Secondary outcomes included number of subtherapeutic and supratherapeutic aPTTs, incidence of bleeding and clotting events, bivalirudin titrations per day, and percentage of patients with therapeutic aPTT level. RESULTS: A total of 100 patients were included (precohort = 67; postcohort = 33). The proportion of therapeutic aPTTs was significantly higher in the postcohort than that in the precohort (62% vs 48%; P < 0.001). The postcohort had 0% of patients failing to achieve therapeutic aPTT levels. The number of titrations per day was significantly lower in the postcohort, with 1.20 titrations per day versus 1.93 in the precohort (P < 0.001). CONCLUSIONS: Implementation of the bivalirudin titration nomograms within the EHR significantly increased the number of therapeutic aPTTs, reduced the number of patients who never achieved a therapeutic aPTT, and reduced the required number of titrations per day.
RESUMO
Acute kidney injury (AKI) is defined by changes in serum creatinine and urine output (UO). Significant limitations exist regarding accurate ascertainment of urine output even within the intensive care unit. We sought to evaluate an automated urine output collections system and compare it to nursing measurements. We prospectively collected urine output using an electronic urine monitoring system and compared it to charted hourly UO in 44 patients after cardiac surgery at a single university hospital ICU. We calculated UO and oliguria rates and compared them to data from the sensor and from nursing charting. A total of 187 hourly UO measurements were obtained and on average, UO was reported 47 min late, with a median of 18 min, and a maximum of almost 6 h. Patients had a mean hourly UO of 76.3 ml over the observation period. Compared to manual measurements by study personnel, nurses significantly overestimated hourly UO by 19.9 ml (95% CI: 10.3; 29.5; p = < 0.001). By contrast, the mean difference between the UO measured with the sensor and by study personnel was 2.29 ml (95% CI: - 6.7; 11.3), p = 0.61. Electronic UO monitoring is significantly more accurate than nurse-performed manual measurements in actual intensive care patients. Furthermore, timely ascertainment of UO is difficult to achieve with manual technique, resulting in important delays in detecting oliguria perhaps leading to missed cases of AKI.