Weaning Pattern Characteristics, Based on Simplified Acute Physiology Score 3, of Critically Ill Patients Requiring Ventilator Care

Severity scoring systems are used to predict and compare outcomes, to help guide the allocation of limited resources and to evaluate the process of care in intensive care units (ICU). In critically ill patients, several scoring systems have been developed over the last three decades [1,2]. The Acute Physiology and Chronic Health Evaluation (APACHE) and the Simplified Acute Physiology Score (SAPS) are the most widely used scoring systems in ICUs. Recently, the SAPS 3 was developed through a worldwide prospective study to predict hospital mortality in critically ill patients. It is based on 20 different variables, that are easily measured at patient admission, and dissociating patient status from the quality of care in the ICU [3-7]. There has, however, been no investigation into how outcomes differ in patients receiving or not receiving mechanical ventilation. The aim of this study was to evaluate the epidemiology and prognostic performance of the SAPS 3 in a retrospective electric chart review, and to describe the weaning pattern characteristics of patients receiving mechanical ventilation.


Introduction
Severity scoring systems are used to predict and compare outcomes, to help guide the allocation of limited resources and to evaluate the process of care in intensive care units (ICU). In critically ill patients, several scoring systems have been developed over the last three decades [1,2]. The Acute Physiology and Chronic Health Evaluation (APACHE) and the Simplified Acute Physiology Score (SAPS) are the most widely used scoring systems in ICUs. Recently, the SAPS 3 was developed through a worldwide prospective study to predict hospital mortality in critically ill patients. It is based on 20 different variables, that are easily measured at patient admission, and dissociating patient status from the quality of care in the ICU [3][4][5][6][7]. There has, however, been no investigation into how outcomes differ in patients receiving or not receiving mechanical ventilation.
The aim of this study was to evaluate the epidemiology and prognostic performance of the SAPS 3 in a retrospective electric chart review, and to describe the weaning pattern characteristics of patients receiving mechanical ventilation.

Materials and Methods
The study protocol was approved by the institutional review board.

002
an ICU stay < 24 h, and patients who were readmitted after an initial ICU discharge were excluded.

Data collection
One individual retrospectively reviewed the electronic medical records. These records provided all of the data required to predict the mortality rate using the SAPS 3 model. The SAPS 3 score was obtained from the most severe laboratory findings 1 h before or after ICU admission. Predicted hospital mortality rate (PMR) was calculated using the following equation; where score means SAPS 3 admission score [6].
The performance of the model was evaluated in all patients, as well as, in two subgroups of patients who had received mechanical ventilation (MV group) or not (Non-MV group). Based on the ventilator weaning pattern, the MV group was further subdivided into two groups to compare the characteristics and mortality based on SAPS 3. The simple weaning group (MV-SW group) included patients with a successful 1st extubation after the 1st spontaneous breathing trial (SBT); the MV-Other group included patients with difficult, prolonged, or chronic mechanical ventilation weaning.

Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics 21 for Windows. Data were reported as means ± standard deviation (SD) or medians with 25th and 75th quartiles for continuous variables, and percentages for quantitative variables. Student's t-test, chi-squared test, or Fisher's exact test were used depending on whether the variables were continuous or categorical. P-values less than 0.05 were used to indicate statistical significance. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve was used to measure discrimination for hospital mortality. Of the 154 patients admitted to the ICU between October and December 2014, 2 pediatric patients, 4 readmissions, and 7 patients with missing data, mostly due to ICU length of stay < 24 h, were excluded. The study group, therefore, comprised 141 patients: 76 males (53.9%) and mean age 67.7 yr. The characteristics of the study group are shown in (Table 1). There were no significant differences in demographic characteristics between patients in the MV group and the Non-MV group. The SAPS 3 score and ICU mortality were significantly higher in the MV group (p = 0.004 and p < 0.001, respectively). In addition, length of ICU stay was significantly longer (p = 0.007) for the MV group.  The MV group (n = 43; excluding 2 patients with missing weaning protocol data) was subdivided based on weaning pattern. When the reason for the intubation was compared between subgroups, the MV-SW group included patients requiring significantly more postoperative care, while the MV-Other group had significantly more intubations due to hypoxemia (p = 0.001). Observed mortality, SAPS 3 score, and predicted mortality were significantly higher in the MV-Other group (Table 2), and observed mortality (60.0%) was higher than the predicted mortality (39.4%).  Hospital mortality was considerably greater in patients with higher SAPS 3 scores. The highest hospital mortality rate was observed in patients with a SAPS 3 score greater than 90 ( Figure  1). Discrimination, as measured by the AUC, was good (AUCs = 0.871), (Figure 2).

Discussion
In the present study, the mean SAPS 3 score of all patients was 46.1; the score was 10 points higher for the group requiring

Journal of Anesthesia & Intensive Care Medicine
mechanical ventilation compared to the group without. Although there were no significant differences in gender, age, route of admission, and department between the groups, the group requiring mechanical ventilation exhibited longer ICU stays and higher mortality. Among members of the MV group, those capable of simple weaning showed lower severity scores and mortality.
Many previous studies have shown that SAPS 3 is a scoring system model with good discrimination but poor calibration [5,[8][9][10]. In the present study, the AUC value, which indicates discrimination, was 0.871; this is similar to previous studies (0.8-0.89) and indicates favorable discrimination [5,11]. While there were no in-hospital mortalities in patients with SAPS 3 scores of ≤40 points, patients with scores of 41-90 points had a mortality rate under 50%, and the mortality rate increased rapidly for patients with scores >90.
Unlike previous SAPS 3 studies that compared discrimination or calibration to outcomes from other scoring models or investigated regional variations [10,[12][13][14], the present study focused on how outcomes differed in patients receiving or not receiving mechanical ventilation. This is because, among various factors affecting SAPS 3, the effect of applying mechanical ventilation on the score is minimal; however, a significant number of patients in the ICU receive ventilator care and applying mechanical ventilation has a clinically significant impact on the clinical course of critically ill patients.
The patient group requiring mechanical ventilation was divided into two subgroups based on the weaning pattern. The simple weaning (MV-SW) group included patients with successful 1st extubation after the 1st SBT. The other (MV-Other) group included all other weaning groups: Difficult weaning (failed 1st SBT trial, but succeeded within the 3rd SBT trials or successful weaning within 7 days after the 1st SBT); prolonged weaning (failed weaning on the 3rd SBT trial or required more than 7 days on the 1st SBT); and chronic mechanical ventilation weaning (the same as tracheostomy) [15,16].
The majority of patients from our hospital had chronic mechanical ventilation weaning when simple weaning failed; for this reason, we consolidated the three groups into one. Since most of the patients who had simple weaning were those who underwent extubation after maintaining mechanical ventilation for postoperative care due to old age, prolonged operation time, or underlying diseases (19 subjects, 82.6%), they not only showed lower SAPS 3 scores, but also lower mortality rates compared to the MV-Other group. Conversely, most of the patients within the MV-Other group were intubated for mechanical ventilation because of hypoxemia caused by impairment of normal ventilation function (17 subjects, 85%), which may have manifested as an increase in the severity of weaning.
The mean length of hospital stay for the MV-Other group, whose conditions were more severe, was not significantly different from the MV-SW group; this may be attributed to a shortened overall length of hospital stay due to the larger number of "do not resuscitates" (DNRs) and patients who passed away in this group. Moreover, it can be surmised that the observed mortality rate (60.0%) in this group was higher than the predicted mortality (39.4%) because of the influence limited proactive management for patients who were expected to have unfavorable prognosis and had effectuated DNRs in advance.
The limitations of this study include having a small number of participants, which resulted in a low number of patients in the ventilated group and corresponding subgroups. In addition, at the time of data collection, the hospital did not have a standard weaning protocol; weaning was carried out either by applying a T-piece or a pressure support ventilation (PSV) mode after the SBT and the protocol used was determined by the doctor in charge of the department. Consequently, the reason for a patient not having been placed into a weaning subgroup may not have been due to the patient's condition.
Furthermore, while all charts were reviewed by a single person responsible for the ICU, the SAPS 3 scores were inputted by different doctors who were in charge of the department at the time of admission; for this reason, individual evaluator errors cannot be eliminated. We plan to perform future studies with a larger number of patients; furthermore, the hospital plans to implement a standard SBT protocol, therefore data obtained after the protocol is applied may be compared to the results presented to allow the mechanical ventilation subgroups to be more clearly defined to determine any differences.

Conclusion
In conclusion, in the present study, conducted on patients who were hospitalized in the surgical or internal medicine ICU, SAPS 3 score assisted-evaluations showed good discrimination. It is believed that this will be a useful method for predicting weaning difficulties and mortalities in patients requiring mechanical ventilation.