Search

Background: Critical illness is characterized by hypoferremia, iron-deficient erythropoiesis (IDE), and anemia. The relative risks and benefits of iron supplementation in this setting are unknown., Methods: Anemic, critically ill surgical patients with an expected intensive care unit length of stay (ULOS) >or= 5 days were randomized to either enteral iron supplementation (ferrous sulfate 325 mg three times daily) or placebo until hospital discharge. Outcomes included hematocrit, iron markers (i.e., serum concentrations of iron, ferritin, and erythrocyte zinc protoporphyrin [eZPP]), red blood cell (RBC) transfusion, transfusion rate (mL RBC/study day), nosocomial infection, antibiotic days, study length of stay (LOS), ULOS, and death. Iron-deficient erythropoiesis was defined as an elevated eZPP concentration., Results: Two hundred patients were randomized; 97 received iron, and 103 received placebo. Socio-demographics, baseline acuity, hematocrit, and iron markers were similar in the two groups. No differences were observed between the iron and placebo groups with respect to either hematocrit or iron markers following up to 28 days. However, patients treated with iron were significantly less likely to receive an RBC transfusion (29.9% vs. 44.7%, respectively; p = 0.03) and had a significantly lower transfusion rate (22.0 mL/day vs. 29.9 mL/day; p = 0.03). Subgroup analysis revealed that these differences were observed in patients with baseline IDE only. Iron and placebo groups did not differ with respect to incidence of infection (46.8% vs. 48.9%; p = 0.98), antibiotic days (14 vs. 16; p = 0.45), LOS (14 vs. 16 days; p = 0.24), ULOS (12 vs. 14 days; p = 0.69), or mortality rate (9.4% vs. 9.9%; p = 0.62)., Conclusions: Enteral iron supplementation of anemic, critically ill surgical patients does not increase the risk of infection and may benefit those with baseline IDE by decreasing the risk of RBC transfusion. A trial comparing enteral and parenteral iron supplementation in this setting is warranted (ClinicalTrials.gov number, NCT00450177).

Background: Multiple organ dysfunction syndrome (MODS) remains prevalent and the leading cause of mortality in the surgical intensive care unit (ICU). Improvements in ICU care in the last 10 years (e.g., tight glycemic control, activated protein C, fewer transfusions causing fewer nosocomial infections) may have decreased the incidence, magnitude, and mortality of MODS, as hypothesized in this study., Methods: Longitudinal 17-year prospective study of 11,314 ICU patients (academic/tertiary unit, Level I trauma center), 5,157 (45.5%) of whom developed any degree of MODS (Marshall score, cumulative). Data collected included Admission Acute Physiology and Chronic Health Evaluation (APACHE)-II and APACHE-III scores, MOD score (MODsc), hospital mortality, and the incidence and magnitude of MODS. The ratio of MODsc: APACHE III was calculated. Analyses (X +/- SEM, chi2, repeated-measures ANOVA, linear and polynomial regression, c-statistic) were performed for calendar-year intervals beginning in 1990 through 2006., Results: Among MODS patients, the mean MODsc was 6.3 +/- 0.1 points, and the mortality rate was 22%. The APACHE III score increased significantly (p < 0.0001) over time, but the mortality rate was unchanged (r2 = 0.02). Adjusted for illness severity (MODsc:A3), the magnitude of MODS decreased significantly (p < 0.0001) during the time period., Conclusions: Despite significant increases in admission APACHE III score over 17 years, the adjusted magnitude of MODS (MODsc:A3) decreased. Given the strong association between MODS and mortality for critically ill surgical patients, it is likely that the unchanged risk-adjusted mortality observed over time is due to the reduced magnitude of MODS.

Background: Prospective data addressing end-of-life care in the surgical ICU are lacking. We determined factors surrounding life-sustaining therapy discussions (LSTDs) in our surgical ICU as experienced by housestaff., Study Design: Housestaff were interviewed daily about the occurrence of an LSTD between themselves and either a patient or surrogate. Patients for whom at least one LSTD occurred were compared with patients for whom an LSTD never occurred. Housestaff also completed a standardized questionnaire that captured events surrounding each LSTD., Results: Eighty LSTDs occurred among 50 patients. Lack of decision-making capacity (p = 0.04), age (p = 0.02), and acuity (p = 0.01) predicted independently the occurrence of an LSTD. Housestaff were significantly more likely to both report recent clinical deterioration (p < 0.01) and to assign a worse prognosis (p < 0.01) to patients for whom an LSTD occurred. Housestaff initiated the majority of LSTDs (70.0%) and usually did so because of clinical deterioration (60.7%); patient surrogates were most commonly believed to initiate LSTDs because of lack of improvement (60.1%). In no instance did a patient initiate an LSTD. For 39 of 50 patients (78.0%), changes in end-of-life care plans were eventually enacted as proposed originally. Housestaff reported that the likelihood of enactment depended on both the preexisting end-of-life care plan and the proposed change in end-of-life care plan., Conclusions: Age, acuity, and lack of decision-making capacity were the most important factors involved in the initiation of an LSTD. Housestaff reported that they initiated LSTDs for different reasons and proposed different end-of-life care plans relative to both patients and their surrogates. These disparities can contribute to failed enactment of proposed changes in end-of-life care plans.

Recent interest in cutting cost and improving utilization and delivery of perioperative services has prompted surgeons to identify patient populations that would benefit from care in an intensive care unit as opposed to intermediate or standard care. The purpose of this study was to evaluate patients admitted to the surgical intensive care unit (SICU) after major plastic/reconstructive surgical procedures in order to determine appropriate perioperative management strategies for these patients. We reviewed retrospectively the data from 2,805 consecutive admissions to the SICU between 1990 and 1996. Forty-two patients (1.5%) who had undergone major plastic/reconstructive procedures were identified. Outcomes (mortality, length of stay in the SICU and hospital, and the degree of organ dysfunction) were compared between this population, an illness severity-matched (Acute Physiology and Chronic Health Evaluation [APACHE]-II and APACHE III) population of patients recovering from vascular surgical procedures, and a similarly matched population of SICU patients who were randomly assigned to serve as a second control group. The hospital mortality of the plastic surgical patient population (9.5%) was significantly higher than the zero mortality of the random cohort (p < 0.05). A second analysis compared the SICU plastics group to a case-controlled group of patients who were admitted to the postanesthesia care unit (PACU) for at least 24 hours of perioperative monitoring. SICU patients had significantly higher APACHE II scores (10.9) when compared to PACU patients (7.2; p < 0.01). Based on severity of illness scoring and eventual mortality, patients admitted to our SICU after major reconstructive surgery were selected appropriately for that setting. In contrast, the patients who stayed in the PACU for perioperative monitoring did not require life-supporting therapy and, therefore, were overmonitored. Care could be provided in a specialized unit with dedicated nursing specifically trained for that purpose.

Objective: Compelling internal and external influences are stimulating global re-evaluations of care standards for efficacy and cost. Critical care uses huge amounts of resources despite widespread shortages of beds and nurses. This study tested the hypothesis that ancillary expenditures can be decreased without compromising care., Methods: Costs for laboratory tests, radiographs, blood products, nutritional supplements, and drugs were compared prospectively for all surgical intensive care unit care for two 4-month periods (January 1 to April 30, 1994 and January 1 to April 30, 1995) at a urban university center. A systematic, multidisciplinary cost-reduction program began May 1, 1994, with emphasis on laboratory and radiographic testing and procedures, and blood product, nutritional, and drug therapies. Cohorts were compared by age, Acute Physiology and Chronic Health Evaluation (APACHE) II and III admission scores, and case mix. Outcome variables were hospital mortality, days in the intensive care unit and hospital, the development of multiple organ dysfunction syndrome, and expenditures. Cost data were taken weekly from the hospital's clinical information system. No new equipment was introduced during the study period except for pumps for patient-controlled analgesia, and there were no new critical pathways or other patient care guidelines., Results: Case mix and all noncost variables were identical. Overall costs were reduced by 29% when normalized by the number of patient-days in each period. Laboratory testing was reduced in frequency by 24 to 32%, and cost by 26 to 28%. Comparable reductions in the cost of blood products (32%) were exceeded by the reductions in expenditures for nutritional supplements (49%) and pharmaceuticals (45%) (all, p < 0.01 or less). Modestly increased (2%) x-ray charges in 1995 were owing entirely to insertion of prophylactic inferior vena cava filters (each, $2,800, n = 5) and computed tomography scans for sinusitis (each, $350, n = 5), although the 7% reduction in portable chest radiographs that was achieved did not meet expectations., Conclusions: Substantial reductions in physician-ordered ancillary expenditures are possible without compromising the standard of care of critically ill patients, or the support of an elaborate framework of defined care plans. With additional experience, incremental savings may accrue from refinement of successful strategies and new approaches to intractable problems.