Awakening and withdrawal of life-sustaining treatment in cardiac arrest survivors treated with therapeutic hypothermia*.

OBJECTIVES: To characterize the prevalence of withdrawal of life-sustaining treatment, as well as the time to awakening, short-term neurologic outcomes, and cause of death in comatose survivors of out-of-hospital resuscitated cardiopulmonary arrests treated with therapeutic hypothermia. DESIGN: Single center, prospective observational cohort study of consecutive patients with out-of-hospital cardiopulmonary arrests. SETTING: Academic tertiary care hospital and level one trauma center in Minneapolis, MN. PATIENTS: Adults with witnessed, nontraumatic, out-of-hospital cardiopulmonary arrests regardless of initial electrocardiographic rhythm with return of spontaneous circulation who were admitted to an ICU. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The study cohort included 154 comatose survivors of witnessed out-of-hospital cardiopulmonary arrests who were admitted to an ICU during the 54-month study period. One hundred eighteen patients (77%) were treated with therapeutic hypothermia. The mean age was 59 years, 104 (68%) were men, and 83 (54%) had an initial rhythm of ventricular tachycardia or fibrillation. Only eight of all 78 patients (10%) who died qualified as brain dead; and 81% of all patients (63 of 78) who died did so after withdrawal of life-sustaining treatment. Twenty of 56 comatose survivors (32%) treated with hypothermia who awoke (as defined by Glasgow Motor Score of 6) and had good neurologic outcomes (defined as Cerebral Performance Category 1-2) did so after 72 hours. CONCLUSIONS: Our study supports delaying prognostication and withdrawal of life-sustaining treatment to beyond 72 hours in cases treated with therapeutic hypothermia. Larger multicenter prospective studies are needed to better define the most appropriate time frame for prognostication in comatose cardiac arrest survivors treated with therapeutic hypothermia. These data are also consistent with the notion that a majority of out-of-hospital cardiopulmonary arrest survivors die after a decision to withdrawal of life-sustaining treatment and that very few of these survivors progress to brain death.

Randomized controlled evaluation of an insulin pen storage policy.

PURPOSE: Results of a quality-improvement project to enhance safeguards against "wrong-pen-to-patient" insulin pen errors by permitting secure bedside storage of insulin pens are reported. METHODS: A cluster-randomized controlled evaluation was conducted at an academic medical center to assess adherence with institutional policy on insulin pen storage before and after implementation of a revised policy allowing pen storage in locking boxes in patient rooms. In phase 1 of the study, baseline data on policy adherence were captured for 8 patient care units (4 designated as intervention units and 4 designated as control units). In phase 2, policy adherence was assessed through direct observation during weekly audits after lock boxes were installed on intervention units and education on proper insulin pen storage was provided to nurses in all 8 units. RESULTS: Phase 1 rates of adherence to insulin pen storage policy were 59% in the intervention units and 49% in the control units (p = 0.56). During phase 2, there was no significant change from baseline in control unit adherence (67%, p = 0.26), but adherence in intervention units improved significantly, to 89% (p = 0.005). Common types of observed nonadherence included pens being unsecured in patient rooms or nurses' pockets or left in patient-specific medication drawers after patient discharge. CONCLUSION: An institutional policy change permitting secure storage of insulin pens close to the point of care, paired with nurse education, increased adherence more than education alone.

Comparison of free fraction serum valproic acid concentrations between inpatients and outpatients.

PURPOSE: The differences in free fraction serum valproic acid concentrations between inpatients and outpatients were compared, and factors associated with therapeutic discordance were evaluated. METHODS: This retrospective, single-center, cross-sectional study included patients with both a free and a total valproic acid concentration drawn within 30 minutes of each other between January 7, 2003, and June 1, 2011. Serum valproic acid concentrations were stratified by admission status (inpatient versus outpatient). In the primary analysis, for patients who had multiple paired concentrations drawn, one free valproic acid level and one total valproic acid level (i.e., one pair) were chosen at random for each inpatient. Information regarding patient demographics, laboratory data, and concomitant medication therapy was collected. Single and multivariable logistic regressions were performed to determine the odds of therapeutic discordance. RESULTS: During the study period, inpatient concentrations were measured in 220 patients, and outpatient concentrations were measured in 41 patients. The median total valproic acid concentration in the inpatient group was significantly lower than that in the outpatient group (54 µg/mL versus 83 µg/mL, p < 0.001). The median free fraction of inpatient levels was nearly twice as high as that of outpatient levels (28.8% versus 15.5%, p < 0.001). Levels drawn in inpatients were therapeutically discordant 63% of the time compared to only 19% in outpatients (p < 0.001). Multivariable logistic regression identified low albumin as an independent risk factor for therapeutic discordance. CONCLUSION: The median free fraction of valproic acid was significantly higher in inpatients than in outpatients. Low albumin concentration was a predictor of discordance between free and total valproic acid concentrations.