| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
* From the Service de réanimation médicale polyvalente, Hôpital Bretonneau, Centre hospitalier universitaire de Tours; and the
Laboratoire de biostatistiques, épidémiologie et informatique médicale, Faculté de médecine de Tours, Tours, France.
Address correspondence to: Dr. Stephan Ehrmann, Service de réanimation médicale polyvalente, Hôpital Bretonneau, Centre hospitalier universitaire de Tours, 37 044 Tours cedex 9, France. Phone: + 33 (0) 6 71 10 33 02; Fax: + 33 (0) 2 47 39 65 36; E-mail: stephanehrmann{at}yahoo.co.uk
 |
Abstract |
|---|
 TOP Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Methods: One hundred fifty-four patients were included. Exclusion criteria were: discharge from the ICU or TL before the 72nd hr. Ninety-three patients remained for evaluation. The LOD was calculated on the day of admission (LOD1) and between the 72nd and 96th hr (LOD4). The 
LOD = LOD4 – LOD1 index was calculated for survivors and non-survivors; sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated.
Results: Sixteen patients died in the ICU, they had a higher LOD (0 vs –2; P = 0.0046) than the survivors. After logistic regression, a high LOD was associated with a higher risk of death in the ICU independent of the initial severity of disease. The PPV concerning death in the ICU was 0.66 for a LOD 
4 cut-off. The NPV was 0.89 for a cut-off of 1.
Conclusion: LOD appears to be a predictor of death in the ICU, independent of the initial severity of disease. The PPV is not high enough to assist with making individual TL decisions. The NPV can help to identify patients at low risk of death. The LOD deserves to be evaluated in a population exhibiting greater severity of disease.
|
Introduction |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In order to have a tool to assist with making early TL in the ICU, we evaluated the performance of the logistic organ dysfunction score (LOD) evolution between the first and the fourth day in the ICU to predict death in the ICU.
|
Methods |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The index LOD = LOD4 – LOD1 was used to measure the evolution of the LOD between the first and the fourth day in the ICU. Thus a positive LOD indicated that the patient’s condition was worsening, and a negative LOD indicated improvement. Other variables measured were: time and day of admission and discharge from the ICU, main admission diagnosis, Mc Cabe and Knaus score. All variables were recorded prospectively. The study endpoint was death in the ICU.
Sample size calculation assuming a 25% ICU mortality with a LOD standard deviation of 3.8 (preliminary data) showed that 98 patients had to be analyzed in order to detect a 2.5-difference in LOD between survivors and non-survivors with an 
risk of 0.05 and a power of 80%. The study was therefore conducted over a three-month period to include about 150 patients in order to take into account exclusions and possible incomplete data collection.
Statistical analysis was performed by comparing LOD between the patients who survived and the ones who died in the ICU using a univariate non-parametric test (Mann Whitney). To take into account the initial severity of disease two logistic regressions were performed, one using the SAPSII of the first day in the ICU and one using LOD1 as independent variables measuring the initial severity of disease. The prognostic performance of LOD was evaluated by calculation of the specificity, the sensitivity, the positive (PPV) and negative (NPV) predictive values concerning death in the ICU for different cut-offs and construction of a receiver operating characteristics (ROC) curve. The different scores were calculated using the software Excel® (Microsoft Corporation, USA), the statistical analysis was performed using the software Statview® (SAS Institute Inc., USA).
Results are expressed as median [10th; 90th percentile]. A P value < 0.05 was considered significant.
|
Results |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
). The major characteristics of the final study patients are indicated in Table I. Sixteen patients (17.7%) died in the ICU, their SAPSII was significantly higher than that of survivors (47 [31; 61] vs 39 [19; 59]; P = 0.05). So was LOD4 (6 [2; 13] vs 2 [0; 6]; P = 0.0005). LOD1 values were not significantly different from values in the patients who died (6 [2; 10] vs 5 [2; 10]; NS).
|
|
LOD was significantly higher in the patients who died in the ICU: 0 [–4; 4] vs –2 [–7; 0]; P = 0.0046.
After logistic regression, LOD was associated with a higher risk of death in the ICU independent of the initial severity of disease, whether measured by the SAPSII of the first day (odd ratio [OR] = 1.41 [95% confidence interval [95CI] = 1.15–1.73]; P = 0.0009) or LOD1 (OR = 1.52 [95CI: 1.20–1.93]; P = 0.0005). A ROC curve was constructed by calculation of the sensitivity and specificity of LOD for the prediction of death in the ICU (Figure 2). The area under the ROC curve was graphically small, the most informative cut-off was 1. Positive predictive value increased with higher values of cut-offs. For a cut-off of LOD 1 the PPV was 53% (95CI: 23–77), the NPV was 89% (95CI: 62–99). At a cut-off of LOD 0 the PPV was 33% (95CI: 15–49), the NPV was 88% (95CI: 64–96).
|
LOD 4, the highest NPV (100%; 95CI: 87–100) was observed for a cut-off of LOD –4. |
Discussion |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
LOD was significantly higher in patients who died in the ICU, and thus was independent of the initial severity of disease. According to logistic regression, when initial severity of disease was measured by LOD1, each 1 point increase in LOD multiplied the risk of death in the ICU by a factor of 1.5. In the studied population the PPV appeared to be low, with the relatively small number of patients resulting in large 95CIs. At a cut-off of 1, the NPV was high (89%) indicating that patients with a LOD of < 1 (inferring that their condition is improving or stable), have a probability of death of 11%, which is much lower than the probability of death of the whole study population (17%). A simple way of using LOD is to consider the cut-off of 0: patients with a negative LOD (whose condition is improving) have a probability of surviving their ICU stay of 88%. The usual severity of disease scores (SAPS,3 Acute Physiology And Chronic Health Evaluation [APACHE],4 LOD)5 do predict accurately the risk of death on the level of a population of patients admitted to the ICU. However, they poorly predict the individual likelihood of death,6–8 they are not well suited for patients staying more than three days in the ICU,9 and they do not take into account the patient’s response to treatment. It has been shown that the assessment of scores (APACHE, LOD, Sequential Organ Failure Assessment, Mortality Probability Model) on any day during the first week in the ICU provides a better prediction of the individual likelihood of death,10–14 but this method of evaluating the probability of death does not take into account the response to treatment. The evaluated indices which take into account the patient’s response to treatment are based on the maximum value of a score during the first week in the ICU,15,16 or on the calculation of composite scores with weighting coefficients.17–19 The available objective methods to measure a patient’s prognosis (taking into account the response to treatment) are therefore difficult to use in clinical practice, as they require relatively complex calculation. The indices which require determination of the maximal value of a score can only be used retrospectively, in general after one week, and therefore are of little help for ethical discussion early on during the ICU stay.
The LOD is an interesting prognostic factor in the ICU, as it is easy to calculate, it can be used after 72 hr in the ICU, allowing physicians to consider early TL, and it takes into account patient response to treatment. The LOD is an easy way to objectively assess the patient’s response to treatment, and therefore could be used by physicians to complement subjective evaluation of the clinical condition. When communicating to families and caregivers the rational of TL, LOD can serve as a numerical objective support to the overall ethical discussion, with a LOD value 
0 being associated with an overall good prognosis.
This study presents several limitations. First, the final study population was a selected subgroup of the initial population. Patients with TL before the 72nd hr, due to their survival probability being considered very low by the physician in charge, were excluded. Exclusion of these patients, who would have had high LOD values and probably have died, does not influence the positive results of the study, but could have contributed to the low PPV observed. Another limitation could have been that patients underwent implicit TL before the 72nd hr without notification in the study protocol. This was unlikely to happen as the study ICU has a standard procedure for making TL that has been in use for many years, and includes a staff discussion and the completion of a standard form. The patients who left the ICU between the 72nd and 96th hr were a possible source of bias, as they were susceptible to show very extreme LOD values, particularly in case of death. This was not the case as the LOD of these patients were close to the values in the overall sample: one patient died within this period (LOD = 2) and ten were discharged (LOD = –2.5 [–7; 0]). This study was performed in only one ICU and included a limited number of patients, therefore the results may not necessarily apply to other ICU settings.
The LOD, which is a score constructed objectively through logistic regression, has been chosen for this study because it quantifies organ dysfunction and is therefore well suited for sequential assessment during the ICU stay, in contrast to scores which include age or chronic health status. Furthermore, there is a growing body of evidence suggesting that acute organ failures are the major determinant of prognosis in the ICU, chronic health status being of less prognostic value.20–22 However, TL decisions will always be made considering the whole patient, and particularly his/her health status prior to ICU admission. We chose to measure the evolution of the LOD in an additive model (LOD = LOD4 – LOD1), as preliminary analysis showed that in our population LOD4 and LOD1 exhibited a linear relationship (data not shown).
Accordingly to the Bayes’ theorem the PPV for a given issue is linked positively to the prevalence of this issue in the population. The prevalence of death in the study sample was low (17.7%) and explains partly the low PPV for death observed. This low mortality rate reduced the power of the study as compared to the sample size calculation.
In conclusion, a high LOD is associated with death in the ICU. The LOD can be used in daily clinical practice to assess objectively the patient’s response to treatment. A negative LOD or no change is associated with a low probability of death in the ICU, and therefore could constitute an argument against TL within a global ethical discussion. However LOD should only be considered as one tool to help within the global clinical and ethical evaluation of critically ill patients.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
None of the authors is involved in any commercial or non commercial affiliations or consultancies that are, or may be perceived to be, a conflict of interest with the work.
Accepted for publication August 12, 2005. Revision accepted October 10, 2005.
|
References |
|---|
|
TOPAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
2 Cook D, Rocker G, Marshall J, et al. Withdrawal of mechanical ventilation in anticipation of death in the intensive care unit. N Engl J Med 2003; 349: 1123–32.
3 Le Gall JR, Loirat P, Alperovitch A, et al. A simplified acute physiology score for ICU patients. Crit Care Med 1984; 12: 975–7.[Medline]
4 Knaus WA, Zimmerman JE, Wagner DP, Draper EA, Lawrence DE. APACHE–acute physiology and chronic health evaluation: a physiologically based classification system. Crit Care Med 1981; 9: 591–7.[Medline]
5 Le Gall JR, Klar J, Lemeshow S, et al. The Logistic Organ Dysfunction system. A new way to assess organ dysfunction in the intensive care unit. ICU Scoring Group. JAMA 1996; 276: 802–10.[Abstract]
6 Teres D, Lemeshow S. Why severity models should be used with caution. Crit Care Clin 1994; 10: 93–110.[Medline]
7 Lemeshow S, Klar J, Teres D. Outcome prediction for individual intensive care patients: useful, misused, or abused? Intensive Care Med 1995; 21: 770–6.[Medline]
8 Anonymous. Predicting outcome in ICU patients. 2nd European Consensus Conference in Intensive Care Medicine. Intensive Care Med 1994; 20: 390–7.[Medline]
9 Sicignano A, Carozzi C, Giudici D, Merli G, Arlati S, Pulici M. The influence of length of stay in the ICU on power of discrimination of a multipurpose severity score (SAPS) ARCHIDIA. Intensive Care Med 1996; 22: 1048–51.[Medline]
10 Wagner DP, Knaus WA, Harrell FE, Zimmerman JE, Watts C. Daily prognostic estimates for critically ill adults in intensive care units: results from a prospective, multicenter, inception cohort analysis. Crit Care Med 1994; 22: 1359–72.[Medline]
11 Pettila V, Pettila M, Sarna S, Voutilainen P, Takkunen O. Comparison of multiple organ dysfunction scores in the prediction of hospital mortality in the critically ill. Crit Care Med 2002; 30: 1705–11.[Medline]
12 Timsit JF, Fosse JP, Troche G, et al.; for the OUTCOMEREA Study Group, France. Calibration and discrimination by daily logistic organ dysfunction scoring comparatively with daily sequential organ failure assessment scoring for predicting hospital mortality in critically ill patients. Crit Care Med 2002; 30: 2003–13.[Medline]
13 Ferreira FL, Bota DP, Bross A, Melot C, Vincent JL. Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA 2001; 286: 1754–8.
14 Lemeshow S, Klar J, Teres D, et al. Mortality probability models for patients in the intensive care unit for 48 or 72 hours: a prospective, multicenter study. Crit Care Med 1994; 22: 1351–8.[Medline]
15 Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995; 23: 1638–52.[Medline]
16 Moreno R, Vincent JL, Matos R, et al. The use of maximum SOFA score to quantify organ dysfunction/failure in intensive care. Results of a prospective, multicentre study. Working Group on Sepsis related Problems of the ESICM. Intensive Care Med 1999; 25: 686–96.[Medline]
17 Timsit JF, Fosse JP, Troche G, et al. Accuracy of a composite score using daily SAPS II and LOD scores for predicting hospital mortality in ICU patients hospitalized for more than 72 h. Intensive Care Med 2001; 27: 1012–21.[Medline]
18 Chang RW, Jacobs S, Lee B. Predicting outcome among intensive care unit patients using computerised trend analysis of daily Apache II scores corrected for organ system failure. Intensive Care Med 1988; 14: 558–66.[Medline]
19 Metnitz PG, Lang T, Valentin A, Steltzer H, Krenn CG, Le Gall JR. Evaluation of the logistic organ dysfunction system for the assessment of organ dysfunction and mortality in critically ill patients. Intensive Care Med 2001; 27: 992–8.[Medline]
20 Ferraris VA, Propp ME. Outcome in critical care patients: a multivariate study. Crit Care Med 1992; 20: 967–76.[Medline]
21 Rue M, Quintana S, Alvarez M, Artigas A. Daily assessment of severity of illness and mortality prediction for individual patients. Crit Care Med 2001; 29: 45–50.[Medline]
22 Jacobs S, Zuleika M, Mphansa T. The multiple organ dysfunction score as a descriptor of patient outcome in septic shock compared with two other scoring systems. Crit Care Med 1999; 27: 741–4.[Medline]
This article has been cited by other articles:
 |
|
 |
|
  Y. Skrobik and B. P. Kavanagh Scoring systems for the critically ill: use, misuse and abuse/Les systemes de notation pour les grands malades : usage, mesusage et abus. Can J Anesth, May 1, 2006; 53(5): 432 - 436. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
