Search

Objectives To describe the effect of implementing a contemporary perioperative pulmonary hypertension (PH)-targeted protocol in patients with pulmonary arterial hypertension (PAH) undergoing noncardiac surgery (NCS). Methods The data of consecutive patients with PAH diagnosed by right heart catheterization who underwent NCS between January 1, 2006 and February 9, 2016 were reviewed. Patient demographics, etiology of PAH, clinical features, diagnostic data, utilization of PH-specific medications, and trend of perioperative complications rate were recorded during the study period. Results In the base cohort of 375 patients, 37 had NCS. The mean age at surgery was 62 years. Most patients were women (78%) classified in group 1 PAH. At the time of the surgery, 86% were New York Heart Association functional class III/IV and 97% had American Society of Anesthesiologists classifications 3 and 4. A larger proportion of patients displayed lower PAH risk scores between 2006 and 2011 (P = 0.045). Conversely, a higher percentage of patients exhibited moderately high to very high PAH risk scores between 2012 and 2016 (P = 0.003). Perioperative and anesthetic-related morbidity was 27%, and no difference was observed between either period (P = 0.944). Most of the complications (70%) were related to general anesthesia. Two deaths (5%) occurred in our study group, both during the 2006-2011 period. Conclusions The combination of a multidisciplinary perioperative approach, utilization of novel pulmonary vascular disease-targeted therapy, adequate perioperative optimization, and thoughtful selection of anesthetic technique seems to be a potential strategy to at least maintain similar perioperative outcomes among higher- and lower-risk patients with PAH undergoing NCS.

In intensive care unit (ICU) patients, seizure or status epilepticus treatment with intravenous benzodiazepines or conventional antiepileptic drugs (AEDs), such as phenytoin, may be accompanied by cardiovascular depression or hypotension. Levetiracetam (LVM) is a novel AED that does not undergo extensive liver metabolism, does not require drug level monitoring, and is not associated with hemodynamic instability. We retrospectively analyzed the use, safety, and efficacy of LVM in ICU patients.Collected data included age, sex, therapy indication and duration, dosing regimen, documented seizure activity, ICU admission diagnoses, length of ICU stay, serum creatinine, liver function tests, adverse reactions, concomitant use of other AEDs, and drug interactions.Fifty-one patients were identified (26 males; mean (SD) age, 58.2 (19.8) years). Most patients (65%) did not receive a loading dose; the most common loading dose was 1,500 mg (50% of 18 patients). The most common maintenance dose was 500 mg twice daily (59% of 51 patients), and average duration of therapy was 13.6 (12.7) days. Approximately 47% of patients had preexisting liver disease, and 25% had elevated serum creatinine. Twenty-two patients received LVM therapy for seizure prophylaxis; 29 for acute seizure treatment. Ninety-three percent of patients treated with LVM for acute seizure had no subsequent seizures; the remaining patients (7%) required additional AEDs. One patient receiving LVM for seizure prophylaxis had documented seizures requiring additional AEDs. No adverse hemodynamic events or cardiac arrhythmias were reported.LVM appears to be safe for ICU patients when dosing is adjusted for renal function.

OBJECTIVE To characterize the frequency and pattern of diffusion-weighted imaging (DWI) abnormalities detected as part of brain magnetic resonance imaging (MRI) and their association with short-term neurologic outcomes in patients successfully resuscitated after cardiopulmonary arrest (CPA). PATIENTS AND METHODS We retrospectively analyzed a case series of patients who experienced CPA between May 1, 2000, and April 29, 2004, at St Luke's Hospital in Jacksonville, Fla. Eligible patients required treatment by the Code Blue team and had 1 DWI study before discharge or death. Two neuroradiologists jointly classified DWI abnormalities by anatomic location. Outcome was measured by Cerebral Performance Category score. RESULTS Resuscitation was performed 628 times during the 48-month study period. Of 514 CPA survivors, 18 (3.5%) had MRI studies. The median age was 62 years (interquartile range [IQR], 49-73), and 10 were men. Median code duration was 16 minutes (IQR, 11-19 minutes), and median code-to-scan time was 72 hours (IQR, 28-229 hours). A DWI abnormality was noted in 9 (50%) of 18 patients. Cortical areas (global and regional) were the most common sites of restricted diffusion. Diffusion-weighted imaging abnormalities were present in 7 (70%) of 10 patients with a poor neurologic outcome at discharge. CONCLUSION Magnetic resonance imaging is performed rarely after survival of CPA. In this study with limited sample size, a greater proportion of patients with normal DWI findings had a good neurologic outcome at the time of hospital discharge vs those with abnormal findings. Prospective studies of early and serial MRI (with DWI) are needed to confirm this association and to clarify the prognostic usefulness of such studies.

OBJECTIVES To determine the provider cost of administering intensive care unit (ICU) services, comparing 3 different staffing models for ICU coverage, and to compare the costs of using house staff vs nonphysician providers (NPPs). METHODS Data were collected on total staff composition and number of beds in ICUs from January 1, 2004, through December 31, 2004, at the 3 Mayo Clinic sites: Rochester, Minn; Jacksonville, Fla; and Scottsdale, Ariz. Institutional or national average staff salaries were used to determine total staffing costs per ICU bed per year at each site. Medicare medical education reimbursements were also taken into account. RESULTS Costs per ICU bed for physician staffing were $18,630 in Rochester, $37,515 in Jacksonville, and $38,010 in Scottsdale. When NPPs were substituted for house staff, the costs per bed were $72,466 in Rochester, $61,291 in Jacksonville, and $49,909 in Scottsdale. Incremental costs per ICU bed using NPPs were $53,836 in Rochester, $23,776 in Jacksonville, and $11,899 in Scottsdale. CONCLUSION Use of residents and fellows in ICU staffing at a major tertiary health center is more cost-efficient than use of NPPs. This finding could have implications for the cost of physician services in nonteaching community hospitals and the methods by which care is provided.

Context: The delivery of end-of-life care (EOLC) in the intensive care unit (ICU) varies widely among medical care providers. The differing opinions of nurses and physicians regarding EOLC may help identify areas of improvement. Objective: To explore the differences of physicians and nurses on EOLC in the ICU and how these differences vary according to self-reported proficiency level and primary work unit. Design: Cross-sectional survey of 69 ICU physicians and 629 ICU nurses. Setting: Single tertiary care academic medical institution. Results: A total of 50 physicians (72%) and 331 nurses (53%) participated in the survey. Significant differences between physicians and nurses were noted in the following areas: ability to safely raise concerns, do not resuscitate (DNR) decision making, discussion of health care directives, timely hospice referral, spiritual assessment documentation, utilization of social services, and the availability of EOLC education. In every domain of EOLC, physicians reported a more positive perception than nurses. Additional differences were noted among physicians based on experience, as well as among nurses based on their primary work unit and self-reported proficiency level. Conclusions: Even with an increased focus on improving EOLC, significant differences continue to exist between the perspectives of nurses and physicians, as well as physicians among themselves and nurses among themselves. These differences may represent significant barriers toward providing comprehensive, consistent, and coordinated EOLC in the ICU.

Insulin-like growth factor-binding protein-1 (IGFBP-1) is one of several related proteins that bind and modulate the actions of IGFs. The liver is the primary source of IGFBP-1, and insulin is a major regulator of hepatic IGFBP-1 production. We report five sets of investigations that further define the characteristics of hepatic IGFBP-1 response to insulin. In normal subjects, a continuous high-dose insulin infusion caused a rapid decrease in plasma IGFBP-1 concentrations, with a rate of 0.24 ± 0.04 μg/L·min−1 and a t 1 2 of 89 ± 4 minutes. Conversely, a 3-hour somatostatin (SRIF) infusion caused a 4.5-fold increase in plasma IGFBP-1 levels. SRIF plus low-dose insulin infusion (to inhibit break-through insulin secretion) resulted in a plateau in IGFBP-1 concentrations at 5 to 8 hours, with a t 1 2 to achieve steady state of 60 to 75 minutes. Under similar conditions, a stepped increase in plasma glucose level from 5 to 9 mmol/L had no effect on the rate of IGFBP-1 increase in plasma, indicating that an acute increase in glucose concentration within a physiologic range has no independent inhibitory effect on IGFBP-1 production in the presence of a nonsuppressive insulin level. Using SRIF plus sequential graded insulin infusions, the threshold peripheral (=portal) plasma insulin concentration for IGFBP-1 suppression was between 65 and 172 pmol/L. Subjects with insulin-dependent diabetes mellitus (IDDM) showed a similar dose-response pattern, suggesting that insulin regulation of IGFBP-1 may be normal in IDDM. The rapid and sensitive response of IGFBP-1 to insulin is consistent with its postulated role to modulate IGF bioaction in relation to substrate availability. Moreover, measurement of plasma IGFBP-1 concentrations may serve as a useful marker of hepatic insulin sensitivity.

The effect of glucocorticoids on adipose tissue lipolysis in animals and humans is controversial. To determine whether a physiological increase in plasma cortisol, similar to that observed in diabetic ketoacidosis and other stress conditions, stimulates lipolysis, palmitate kinetics were measured in seven nondiabetic volunteers on two occasions with [1-14C]palmitate as a tracer. Subjects received a 6-h infusion of either 2 micrograms.kg-1.min-1 hydrocortisone or saline in random order. On both occasions, a pancreatic clamp (0.12 micrograms.kg-1.min-1 somatostatin, 0.05 mU.kg-1.min-1 insulin, and 3 ng.kg-1.min-1 growth hormone) was used to maintain plasma hormone concentrations at desired levels. Plasma cortisol concentrations increased to approximately 970 nM during cortisol infusion. Palmitate rate of appearance (Ra) and concentration increased by approximately 60% during cortisol infusion but did not change during saline infusion. Increments in palmitate Ra and concentration over the 6-h study were significantly greater during cortisol than saline infusion when compared by area-under-the-curve analysis (152 +/- 52 vs. -48 +/- 23 mumol.kg-1 and 12.2 +/- 4.1 vs. -4.9 +/- 4.1 mmol.min-1.L-1, respectively; P less than 0.02). Plasma glucose concentrations did not change significantly during cortisol (5.0 +/- 0.3 vs. 6.1 +/- 0.6 mM, NS) or saline (4.9 +/- 0.2 vs. 4.9 +/- 0.1 mM, NS) infusion. In nondiabetic volunteers, a 6-h cortisol infusion was associated with a 60% increase in palmitate Ra that did not occur with saline infusion. Thus, physiological hypercortisolemia may contribute to the increased rates of lipolysis observed in humans during stress.

We investigated the clinical significance of time of onset, duration, and type of pulmonary edema after orthotopic liver transplantation by retrospectively reviewing 93 consecutive recipients. Pulmonary edema was diagnosed by means of radiographic criteria and Pao(2)/Fio(2) ratio300. Type was identified by pulmonary artery wedge pressure (hydrostatic,18 mm Hg; permeability,or =18 mm Hg). Of 91 evaluable patients, 44 (48%) had no pulmonary edema, 23 (25%) had immediate pulmonary edema resolving within 24 hours, 8 (9%) had late pulmonary edema (developing de novo in the first 16 to 24 hours), and 16 (18%) had persistent pulmonary edema (developing immediately and persisting for at least 16 hours). At 16 to 24 hours, mean arterial pressure was lower with persistent permeability-type edema than without pulmonary edema (75 versus 87 mm Hg, P.01). Patients with persistent permeability-type edema had higher mean pulmonary arterial pressure (23 versus 16 mm Hg, P.01) and higher pulmonary vascular resistance (103 versus 53 dyn. second. m(-5), P.05), consistent with a resistance-dependent mechanism. Patients with persistent hydrostatic-type edema did not differ from those without edema in mean arterial pressure (84 versus 87 mm Hg, P.05) or pulmonary vascular resistance (67 versus 53 dyn. second. m(-5), P.05), but had increased mean pulmonary arterial pressure (27 versus 16, P.01), suggesting a flow volume-dependent mechanism. Duration of mechanical ventilation, intensive care, and hospital stay were prolonged in patients with late or persistent permeability-type edema but not in patients with immediate pulmonary edema of any type. In conclusion, immediate pulmonary edema resolving within 24 hours after liver transplantation had little clinical consequence; persistent permeability-type pulmonary edema portended a worse outcome.

To determine whether the sensitivity of adipose tissue lipolysis to catecholamines is increased in poorly controlled insulin-dependent diabetes, the lipolytic response to epinephrine was measured in seven nondiabetic volunteers and seven poorly controlled diabetic subjects with use of [1-(14)C]palmitate as a tracer. Subjects received sequential 1-h infusions of epinephrine, which produced epinephrine concentrations of approximately 1,000, approximately 1,750, approximately 3,500, and approximately 6,000 pmol/l. A pancreatic clamp was used to maintain constant plasma hormone levels. Concentration-response curves were constructed for each subject from the integrated lipolytic response during each epinephrine infusion. There was no difference in maximal lipolytic response (117 +/- 19 vs. 152 +/- 11 mumol.kg-1.h-1) or in maximally effective (3,171 +/- 267 vs. 3,357 +/- 349 pmol/l) or half-maximally effective (1,081 +/- 109 vs. 1,015 +/- 120 pmol/l) epinephrine concentrations between nondiabetic and diabetic subjects, respectively (all P = NS). In control subjects, maximum beta-hydroxybutyrate concentrations were achieved at lower epinephrine concentrations than those required for a maximum lipolytic effect. Thus, under pancreatic clamp conditions, the lipolytic response to epinephrine in nondiabetic and diabetic subjects was similar.

Insulin-like growth factor binding protein-1 (IGFBP-1) modulates the metabolic and mitogenic actions of the IGF peptides. Previous studies have established insulin as the major regulator of plasma IGFBP-1 in humans, acting to suppress hepatic IGFBP-1 synthesis. In this study, we investigated the regulation of plasma IGFBP-1 by cortisol in humans, independent of insulin. Following an overnight fast, six healthy adult volunteers received a euglycemic pancreatic clamp (somatostatin, 0.12 μg·kg−1·min−1; GH, 3 ng·kg−1·min−1; insulin, 0.05 mU·kg−1·min−1) to block endogenous insulin secretion and to control glucose and plasma hormone concentrations at desired levels. Three hours after the initiation of the pancreatic clamp, each subject received an additional 360 min infusion of either cortisol (2 μg·kg−1·min−1) or saline on separate occasions and in random order. Plasma cortisol concentrations increased from 220 to 970 nmol/1 during the cortisol infusion. Insulin concentrations were maintained at approximately 30 pmol/1 throughout saline and cortisol infusions. Plasma IGFBP-1 concentrations increased threefold in response to hypoinsulinemia, reaching plateau values of ∼ 140 μg/l with saline infusion. During cortisol infusion, IGFBP-1 levels increased to ∼300 μg/I. Over the 360 min study period, the integrated response of plasma IGFBP-1 to cortisol infusion was 314% greater than to saline infusion (p