PURPOSE:To test the agreement of the Clinical Frailty Scale (CFS) and the Tilburg Frailty Indicator (TFI), their association with 3, 6 months and 1-year mortality and the trajectory of frailty in a mixed population of ICU survivors. MATERIAL AND METHODS:This is a prospective, multicenter, longitudinal study on ICU survivors ≥18 years old with an ICU stay >72 h. For each patient, sociodemographic and clinical data were collected. Frailty was assessed during ICU stay and at 3, 6, 12 months after ICU discharge, through both CFS and TFI. RESULTS:124 patients with a mean age of 66 years old were enrolled. The baseline prevalence of frailty was 15.3% by CFS and 44.4% by TFI. Baseline CFS and TFI correlated but showed low agreement (Cohen's K = 0.23, p < 0.001). Baseline CFS score, but not TFI, was significantly associated to 1 year mortality. Moreover, CFS score during the follow-up was independently associated 1-year mortality (OR = 1.43; 95% CI: 1.18-1.73). CONCLUSIONS:CFS and TFI identify different populations of frail ICU survivors. Frail patients before ICU according to CFS have a significantly higher mortality after ICU discharge. The CFS during follow-up is an independent negative prognostic factor of long-term mortality in the ICU population.

BACKGROUND:The Oxford Elbow Score (OES) is an English-language questionnaire specifically designed to evaluate surgical elbow outcomes. This scoring system has been translated into other languages. Given the lack of an Italian version of the OES, the present study was designed to establish, culturally adapt, and validate the Italian version.METHODS:The OES questionnaire was culturally adapted to Italian patients in accordance with the literature guidelines with a pilot phase including seven patients with elbow problems and seven healthy subjects. The study includes 110 participants from three hospitals, who underwent elbow surgery for acute (70%) or chronic diseases. At least one month after elbow surgery, at the "index visit", the physician completed the Mayo Elbow Performance Index and patients completed the following questionnaires: the Italian OES, the shortened version of the Disability of Arm, Shoulder and Hand Questionnaire (QuickDASH) and the Short-Form 36 Health Survey. Internal consistency was evaluated using Cronbach's alpha. Reproducibility was assessed using the intraclass correlation coefficient in ten patients who completed the OES again two-three days after the index visit. Construct validity was assessed using Spearman correlation coefficients. Responsiveness was evaluated in 68 patients who answered the questionnaires four months after the index visit, using the Wilcoxon signed-rank test, the effect size and the standardized response mean calculation.RESULTS:Cronbach's alpha was excellent: 0.86 (0.82-0.90) for OES pain, 0.92 (0.90-0.94) for OES function, and 0.90 (0.87-0.93) for OES social/psychological. The intraclass correlation coefficient was 0.94 (0.78-0.98) for OES pain, 0.91 (0.71-0.97) for OES function, 0.95 (0.83-0.98) for OES social-psychological and 0.93 (0.76-0.98) for OES total. The Spearman ρ was >0.7 for OES pain and QuickDASH, for OES function and both QuickDASH and Mayo Elbow Performance Index, and for OES social-psychological and QuickDASH. Regarding responsiveness, the mean of the changes between the two visits ranged from 33.9 for OES pain, to 44 points for OES function and OES social/psychological. The effect size and the standardized response mean were >0.8 for all OES domains.CONCLUSION:This study demonstrates that the Italian version of the OES, translated in accordance with the international standardized guidelines, is reliable, valid, and responsive in patients who have undergone elbow surgery.

FOR RELATED ARTICLE, SEE PAGE 999In the last decades, the world of critical care is becoming increasingly focused not only on survival rates, but also on the well-being of critical care survivors. ARDS continues to represent an important public health problem despite an improvement in survival of approximately 60%.1Schenck E.J. Oromendia C. Torres L.K. et al.Rapidly improving ARDS in therapeutic randomized controlled trials.Chest. 2019; 155: 474-482Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 2Bellani G. Laffey J.G. Pham T. et al.Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries.JAMA. 2016;23; 315: 788-800Crossref PubMed Scopus (1872) Google Scholar, 3Spadaro S. Park M. Turrini C. et al.Biomarkers for acute respiratory distress syndrome and prospects for personalised medicine.J Inflamm (Lond). 2019; 16: 1Crossref PubMed Scopus (69) Google Scholar The long-lasting impairment after ARDS has been suggested as a more meaningful outcome measure than mortality.4Hopkins R.O. Weaver L.K. Collingridge D. et al.Two-year cognitive, emotional, and quality-of-life outcomes in acute respiratory distress syndrome.Am J Respir Crit Care Med. 2005;15; 171: 340-347Crossref PubMed Scopus (495) Google Scholar Accordingly, the research has focused on the postcritical sequelae that include physical, psychological, and emotional aspects. Patients who survive ARDS are at risk for physical dysfunction, neurocognitive disorders, and worsened quality of life 6 to 12 months after discharge from the ICU.4Hopkins R.O. Weaver L.K. Collingridge D. et al.Two-year cognitive, emotional, and quality-of-life outcomes in acute respiratory distress syndrome.Am J Respir Crit Care Med. 2005;15; 171: 340-347Crossref PubMed Scopus (495) Google Scholar FOR RELATED ARTICLE, SEE PAGE 999 Some years ago, a consensus workshop involving a spectrum of stakeholders identified research priorities in the field of integrative oncology through a Delphi survey.5Weeks L.C. Seely D. Balneaves L.G. et al.Canadian integrative oncology research priorities: results of a consensus-building process.Curr Oncol. 2013; 20: 289-299Crossref PubMed Scopus (10) Google Scholar The first-ranked research topic was the "impact of integrative oncology on symptom management and quality of life (for example, dyspnea, nausea, and fatigue)."5Weeks L.C. Seely D. Balneaves L.G. et al.Canadian integrative oncology research priorities: results of a consensus-building process.Curr Oncol. 2013; 20: 289-299Crossref PubMed Scopus (10) Google Scholar This finding emphasizes how patients and stakeholders value these symptoms. Among these symptoms, there is fatigue, perceived by the individual as an overwhelming sense of tiredness during rest.6Finsterer J. Mahjoub S.Z. Fatigue in healthy and diseased individuals.Am J Hosp Palliat Care. 2014; 31: 562-575Crossref PubMed Scopus (114) Google Scholar The clinical expression of fatigue is multidimensional making evaluation challenging. Patients with ARDS experience a long-term stay in ICU along with prolonged mechanical ventilation, which may lead to an impairment of muscular/cardiorespiratory functions, increasing the risk of chronic fatigue. As long-term sequelae, patients might suffer from a chronic state of exhaustion, which may interfere with their functional capacity. Different types of measurement tools have been used in the reporting of functional impairment in critically ill patients to explore various domains within the different cognitive components of cognitive, physical, emotional, and social functioning. In this field, the use of tools assessing patient-reported outcomes is widespread, and as far as fatigue is concerned, a dedicated tool to appraise the perceived fatigue in detail has been developed.7Cella D. Lai J.S. Chang C.H. Peterman A. Slavin M. Fatigue in cancer patients compared with fatigue in the general United States population.Cancer. 2002; 94: 528-538Crossref PubMed Scopus (496) Google Scholar In this issue of CHEST, Neufeld et al8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar investigated the occurrence of fatigue assessed by the Functional Assessment of Chronic Illness Therapy (FACIT-F), after 6 and 12 months, in ARDS survivors. Interestingly, 70% and 66% of former patients with ARDS reported clinically significant symptoms of fatigue at 6 and 12 months, respectively. In a previous study, our research group validated the FACIT-F in a mixed population of ICU survivors 1 year after hospital discharge.9Spadaro S. Capuzzo M. Valpiani G. et al.Fatigue in intensive care survivors one year after discharge.Health Qual Life Outcomes. 2016; 14: 148Crossref PubMed Scopus (16) Google Scholar The mean transformed FACIT-F score ± SD that we found in the patients was 66 ± 12, which is quite similar to the 62 ± 18 found by Neufeld et al.8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar This finding may be surprising considering that the patients we presented were less critically ill than those of Neufeld et al8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar (mean ICU length of stay, 10 ± 11 and 14 ± 11 days, respectively). Considering that the transformed FACIT-F score for the normal US population is 74.7 ± 15,7Cella D. Lai J.S. Chang C.H. Peterman A. Slavin M. Fatigue in cancer patients compared with fatigue in the general United States population.Cancer. 2002; 94: 528-538Crossref PubMed Scopus (496) Google Scholar both studies8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar,9Spadaro S. Capuzzo M. Valpiani G. et al.Fatigue in intensive care survivors one year after discharge.Health Qual Life Outcomes. 2016; 14: 148Crossref PubMed Scopus (16) Google Scholar suggest that, generally, ICU survivors at 1 year suffer from symptoms of fatigue, especially younger patients with ARDS (mean age, 49 ± 15 years8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar and 68 ± 11 years, respectively).9Spadaro S. Capuzzo M. Valpiani G. et al.Fatigue in intensive care survivors one year after discharge.Health Qual Life Outcomes. 2016; 14: 148Crossref PubMed Scopus (16) Google Scholar Furthermore, the authors pointed out that the prevalence of fatigue at 6 months was greater than impaired physical function, impaired cognition, anxiety, or depression. Interestingly, fatigue was less severe in men and in those employed before ARDS. Specifically, patients employed prior to ARDS reported less fatigue, with a mean difference of 6 points. This association between employment prior to ARDS and fatigue over 6 and 12 months may suggest that other factors might play a role in both. We may hypothesize that employment is a proxy of social integration, which may affect both perceived fatigue and health-related quality of life (HRQoL). Orwelius et al10Orwelius L. Backman C. Mats F. et al.Social integration: an important factor for health-related quality of life after critical illness.Intensive Care Med. 2011; 37: 831-838Crossref PubMed Scopus (20) Google Scholar showed that the level of social integration significantly affected HRQoL of former ICU patients. Indeed, social integration affected HRQoL to a larger extent than age, sex, and the ICU-related factors examined, but to a lower extent than the preexisting diseases.10Orwelius L. Backman C. Mats F. et al.Social integration: an important factor for health-related quality of life after critical illness.Intensive Care Med. 2011; 37: 831-838Crossref PubMed Scopus (20) Google Scholar The strengths of the study of Neufeld et al8Neufeld K.J. Leoutsakos J.-M.S. Yan H. et al.Fatigue symptoms during the first year following ARDS.Chest. 2020; 158: 999-1007Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar include the multicenter longitudinal prospective design, the follow-up at 6 and 12 months after discharge, the high rate of responders, and the use of comprehensive measures of neurocognitive outcome in a well-defined ARDS population. Nevertheless, the high number of hospitals involved in the study (n = 38) and the long period taken to enroll patients (7 years) may reflect a selection bias, possibly testified by the quite young mean age of participants (49 years of age) and the high percentage of those living independently prior to admission in the ICU (91%). This selection bias may reduce the generalizability of the study. However, a high proportion of these ARDS survivors, who suffer from fatigue, require higher attention. What is the next step? We first need to know whether fatigue occurs in survivors of other critical illnesses. For instance, the patients admitted to the ICU because of severe sepsis or septic shock could also perceive long-lasting fatigue, and those admitted to ICU for neurologic disease. Interestingly, Goërtz et al11Goërtz Y.M.J. Spruit M.A. Van 't Hul A.J. et al.Fatigue is highly prevalent in patients with COPD and correlates poorly with the degree of airflow limitation.Ther Adv Respir Dis. 2019; 13 (1753466619878128)Crossref PubMed Scopus (22) Google Scholar found that fatigue is more prevalent in patients with COPD with airflow limitation. The latter is common among ICU patients and correlates with adverse outcome.12Volta C.A. Dalla Corte F. Ragazzi R. et al.Expiratory flow limitation in intensive care: prevalence and risk factors.Crit Care. 2019; 23: 395Crossref PubMed Scopus (9) Google Scholar Future studies should focus on the influence of physical, psychological, behavioral, and systemic factors that could potentially perpetuate fatigue in patients with COPD. Malnutrition could be another factor in the complex mechanism of the development of fatigue in ARDS survivors. We also need to understand whether the use of optimal support nutrition may represent another way in which daily practice in the ICU could affect patients' fatigue. Moreover, it would be useful to assess any relationship between fatigue and social integration. If there is a relationship, we might investigate whether interventions that promote interpersonal relationships affect fatigue and possibly introduce these interventions in the clinical practice. Fatigue Symptoms During the First Year Following ARDSCHESTVol. 158Issue 3PreviewDuring the first year following ARDS, more than two-thirds of survivors reported clinically significant fatigue symptoms. Due to frequent co-occurrence, clinicians should evaluate and manage survivors' physical, cognitive, and mental health status when fatigue is endorsed. Full-Text PDF

As highlighted by European statistics, the employment of donor oocytes is a growing option for women who cannot make use of their own gametes. As the potential recipients are continuously increasing in number, a donor programme which satisfies this demand is mandatory. Improvements in cryopreservation techniques, like oocyte and embryo vitrification, have led to the overcoming of the sequence of stimulation-retrieval-transfer both from a spatial and a temporal point of view, with the development of cryobanks of oocytes permitting crossborder donation. However, while some studies report comparable success when using vitrified and fresh oocytes we still need to investigate whether the use of fresh oocytes give higher live birth rate than cryopreserved ones, when the same number of oocytes are given. The performance of embryo cryopreservation, conversely, seems to be more reliable. A novel approach based on the shipment of frozen sperm from the recipient's country to the oocyte donor's one, where fresh oocytes are inseminated and the resulting embryos frozen and transported back to the referring IVF centre to perform a frozen embryo transfer may be a good strategy. We believe that the use of frozen embryos from fresh oocytes could be associated with a higher cumulative live birth rate per cycle, while favouring personalised oocyte recipient care with a flexible number of oocytes assigned and limiting the burden of travelling abroad.

The aim of the Simplified Acute Physiology Score (SAPS) II and SAPS 3 is to predict the mortality of patients admitted to intensive care units (ICUs). Previous studies have suggested that the calibration of these scores may vary across countries, centers, and/or characteristics of patients. In the present study, we aimed to assess determinants of the calibration of these scores.

Fatigue has not been investigated in long-term Intensive Care Unit (ICU) survivors. This study aimed to assess fatigue through a specific instrument, namely the Functional Assessment of Chronic Illness Therapy Fatigue (FACIT-F) scale, in ICU survivors one year after hospital discharge. A secondary aim was to compare the findings of FACIT-F with those of the Vitality domain (VT) of the 36-item Short-Form Health Survey (SF-36).

In a recent article in Intensive Care Medicine, Hernu and colleagues report the results of a survey in which they asked intensivists to estimate the costs of common intensive care unit (ICU) drugs and treatments, and then compared those estimates to actual costs [1]. Somewhat depressingly, they found that over two-thirds of estimates were wildly off the mark, with many estimates coming either well below or well above the actual values. On the basis of these results the authors call for efforts to better educate intensivists about the cost of critical care so that we can incorporate cost-control into daily practice. However, this recommendation begs a larger question: should cost considerations be included in medical decisions in the ICU? In this commentary, part of an invited pro–con debate on the topic of cost containment, we argue ‘‘no’’. We do not take this position out of opposition to cost-containment in critical care, which all would agree is important given rising healthcare costs worldwide [2]. Instead, we make the case that the potential savings to be had through cost-conscious decision-making are very small, while the potential risks, both to patient outcomes and the patient– physician relationship, are large. Additionally, we argue that there are other, better ways to reduce ICU spending without emphasizing cost-cutting measures at the bedside.

In the early 1990s, the in-hospital mortality rate of intensive care unit (ICU) patients dropped, and interest in the quality of life (QOL) of ICU survivors increased. In 1996, we published a study to investigate 1-year survival after hospital discharge and 6-month QOL after intensive care. Now, we compare our previous results with those reported in the recent literature to appraise any changes, and new knowledge in the area.

BACKGROUND:Administration of vancomycin in critically ill patients needs close regulation. While subtherapeutical vancomycin serum concentration (VSC) is associated with increased mortality, accumulation is responsible for nephrotoxicity. Our study aimed to estimate the efficacy of a vancomycin-dosing protocol in reaching appropriate serum concentration in patients with and without kidney dysfunction.METHODS:This was a retrospective study in critically ill patients treated with continuous infusion of vancomycin. Patients with creatinine clearance > 50 ml/min (Group A) were compared to those with creatinine clearance ≤ 50 ml/min (Group B).RESULTS:348 patients were enrolled (210 in Group A, 138 in Group B). At first determination, patients with kidney dysfunction (Group B) had a statistically higher percentage of vancomycin in target range, while the percentage of patients with a VSC under the range was almost equal. These percentages differed at the subsequent measurements. The number of patients with low vancomycin concentration progressively decreased, except in those with augmented renal clearance; the percentage of patients with VSC over 30 mg/L was about 28 %, irrespective of the presence or absence of kidney dysfunction. Patients who reached a subtherapeutic level at the first VSC measurement had a significant correlation with in-hospital mortality.CONCLUSIONS:Our protocol seems to allow a rapid achievement of a target VSC particularly in patients with kidney dysfunction. In order to avoid subtherapeutical VSC, our algorithm should be implemented by the estimation of the presence of an augmented renal clearance.

BACKGROUND:The aim of this study was compare the Aldrete score at 5 minutes of two groups of patients undergoing carotid endarterectomy with intravenous anesthesia, receiving either anesthetic superficial cervical plexus block or intravenous morphine as transition analgesia.METHODS:After Ethics Committee approval, this randomized, controlled, double-blind, single-center study was performed on patients undergoing carotid endarterectomy, who received total intravenous anesthesia with propofol and remifentanil infusion. After intubation, each patient was randomly assigned to Block (superficial cervical block with levobupivacaine before the surgical incision), or Morphine group (standardized dose of morphine 30 minutes before the end of surgery). In the recovery room, an investigator unaware of the patient randomization recorded time to extubation, Aldrete score, pain, nausea/vomiting, and shiver at T0 (time of extubation), and at 5 (T5), 10 (T10), and 30 minutes (T30) after extubation.RESULTS:Sixty-four patients were studied. Time to extubation was 11 ± 6 min for Block and 20 ± 10 min for Morphine group (P>0.001). Median Aldrete score at T0 was 9 in the Block and 6.5 in the Morphine group (P<0.001), at T5 it was 8.5 and 7.0 (P<0.001), respectively. At T5, 29 (90.6%) Block and 8 (25%) Morphine patients had an Aldrete score ≥ 8 (P<0.001). Systolic blood pressure was higher in the Block than in Morphine patients at T0 and T5.CONCLUSION:This study demonstrates shorter time to extubation and better emergence from anesthesia when total intravenous anesthesia is associated with superficial cervical block than with morphine as transition analgesia.

Introduction The aim of the study was to assess whether adults admitted to hospitals with both Intensive Care Units (ICU) and Intermediate Care Units (IMCU) have lower in-hospital mortality than those admitted to ICUs without an IMCU. Methods An observational multinational cohort study performed on patients admitted to participating ICUs during a four-week period. IMCU was defined as any physically and administratively independent unit open 24 hours a day, seven days a week providing a level of care lower than an ICU but higher than a ward. Characteristics of hospitals, ICUs and patients admitted to study ICUs were recorded. The main outcome was all-cause in-hospital mortality until hospital discharge (censored at 90 days). Results One hundred and sixty-seven ICUs from 17 European countries enrolled 5,834 patients. Overall, 1,113 (19.1%) patients died in the ICU and 1,397 died in hospital, with a total of 1,397 (23.9%) deaths. The illness severity was higher for patients in ICUs with an IMCU (median Simplified Acute Physiology Score (SAPS) II: 37) than for patients in ICUs without an IMCU (median SAPS II: 29, P <0.001). After adjustment for patient characteristics at admission such as illness severity, and ICU and hospital characteristics, the odds ratio of mortality was 0.63 (95% CI 0.45 to 0.88, P = 0.007) in favour of the presence of IMCU. The protective effect of the IMCU was absent in patients who were admitted for basic observation, for example, after surgery (odds ratio 1.15, 95% CI 0.65 to 2.03, P = 0.630) but was strong in patients admitted to an ICU for other reasons (odds ratio 0.54, 95% CI 0.37 to 0.80, P = 0.002). Conclusions The presence of an IMCU in the hospital is associated with significantly reduced adjusted hospital mortality for adults admitted to the ICU. This effect is relevant for the patients requiring full intensive treatment. Trial registration Clinicaltrials.gov NCT01422070 . Registered 19 August 2011.

July 2013 • Volume 41 • Number 7 In the last two centuries, scientific knowledge has increased enormously. As a consequence, complete knowledge in all fields is precluded to human beings, and only a superficial part of general knowledge and a deeper insight into few specific fields can be retained. Medicine has become such a wide discipline that expert clinicians are only able to reach a high level of knowledge in limited fields. Therefore, all clinicians, including intensivists, have to rely on experts in other fields, and often, they are just simple users of complex solutions prepared by others. In this issue of Critical Care Medicine, Johnson et al (1), who are experts in Engineering Science, describe the technique used to develop a new score (Oxford Acute Severity of Illness Score [OASIS]) requiring only 10 variables collected on the first day of ICU admission for assessing severity of illness in critically ill patients and predicting mortality. The predictive ability of the new parsimonious score is compared with that of Acute Physiology and Chronic Health Evaluation (APACHE) IV. Patient data consist of a retrospective cohort of 72,474 ICU admissions without missing values from 2007 to 2011 on 86 ICUs at 49 U.S. hospitals. Data were sent electronically without involvement from intensivists at each unit. However, patients, families, intensivists, insurance companies, and the entire society are interested in predictive models that may allow riskadjusted assessment of ICU performance and benchmarking. Johnson et al (1) started from 37 available predictor variables and used a genetic algorithm (GA) capable of solving highly nonlinear problems to determine which subset of these variables should be used for creating the severity score. Each of the 37 potential measurements, called gene, was included or not included in the model. There were many collections (vectors) of genes, each called chromosome, and each chromosome had its fitness determined by a severity score developed using the particle swarm optimization technique. The particle swarm optimization was used to select the weights for the components of OASIS without any assumption about increasing a component’s weight with increasing deviation from normal values. The reader can appreciate this effect looking at Figure 1 where, for instance, less OASIS age points are given to the patients aged more than 90 years than to the patients aged 78–90 years (7 vs 9, respectively). Clinicians will agree that the oldest-old patients may be less frail than the younger, their long survival testifying their strength or adaptive capacity (2). In the external validation sample (1), the discrimination was very good for both the OASIS and the recalibrated APACHE IV models either to predict ICU or to hospital mortality. The calibration for ICU mortality (Table 3) showed a similar Brier Score and standardized mortality ratio (SMR) for the OASIS and the recalibrated APACHE IV and a better Hosmer-Lemeshow statistic for OASIS. The calibration for hospital mortality (Table 4) showed better Brier Score and SMR for the recalibrated APACHE IV. The authors emphasize that OASIS was developed using ICU mortality, and further studies are needed to evaluate OASIS calibration for predicting hospital mortality. Indeed, we need reliable risk-adjusted assessment of hospital mortality, which is a recognized quality indicator (3). Some new approaches for mortality prediction in mixed ICU populations have already been presented, but not extensively assessed. The GA applied on EURICUS-I database from 84 European ICUs provided an accuracy similar to that of commonly used statistically derived formulas (4). In a singlecenter study, the data mining technique was used to predict ICU mortality, but again, only discrimination was assessed (5). The artificial neural network architecture was used to develop a 15-variable model, which had calibration and discrimination superior to the APACHE II, but the study was performed on only 2,062 admissions to a single ICU and used the old APACHE model (6). In comparison with the currently available general scoring systems, OASIS offers the major advantage of reducing the number of variables. The consequent reduced burden of data collection may reduce costs and possibly increase the number of ICUs collecting data for performance analysis and benchmarking, which are used in only 10% to 15% of ICUs in the United States (7). On the other hand, the reduction of the number of variables makes the quality of data collection due to the increased proportional weight of each variable on the final score even more important. This issue may particularly concern the Glasgow Coma Score, which has been shown to be a variable with poor agreement (kappa statistic 0.36) between previously collected and reabstracted data (8). Furthermore, in a model with a reduced number of variables, missing values can strongly influence the final score, and future users of OASIS must be aware of this. A limitation of the study could result from the data used to develop OASIS. These data were collected in ICUs having a Patient Data Management System (PDMS), and the authors suggest that OASIS can be calculated without any PDMS. Nevertheless, any model working on data collected by PDMS with a high sampling rate should be expected to overestimate How Can We Trust the Experts?*