• Users Online: 158
  • Print this page
  • Email this page


 
 
Table of Contents
ORIGINAL ARTICLE
Year : 2020  |  Volume : 69  |  Issue : 1  |  Page : 148-154

Predicting acute respiratory distress syndrome in high-risk trauma and surgical patients: validation of previous scores


1 Intensive Care Unit Department, Al-Haram Hospital, Cairo University, Cairo, Egypt
2 Critical Care Medicine Department, Cairo University, Cairo, Egypt

Date of Submission29-Mar-2019
Date of Acceptance12-Sep-2019
Date of Web Publication31-Jan-2020

Correspondence Address:
Khaled M Taema
Ass. Professor of Critical Care Medicine Department, Cairo University, Cairo
Egypt
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/ejcdt.ejcdt_79_19

Get Permissions

  Abstract 


Aim We validated lung injury prediction scores (LIPS) in predicting acute respiratory distress syndrome (ARDS) and in-hospital mortality in high-risk major trauma and surgical ICU patients.
Patients and methods A total of 79 patients admitted to the ICU for major trauma and/or major surgery were included in our prospective observational study. Only patients with acute physiology and chronic health evaluation-II score more than 15 were included. LIPS was subsequently calculated using two formulas (score-1 and score-2) that were derived by other investigators. C-reactive protein was measured on admission and on day 2 (CRP-1 and CRP-2). Our end points included ARDS development and in-hospital mortality.
Results A total of 33 (41.8%) patients developed ARDS after a median (Q1Q3) of 4 (2–7) days. The score-1 and score-2 were 5 (4–6) and 4.5 (3–6.3) in ARDS versus 3 (1.8–4.3) and 3 (1.5–4.6) in non-ARDS patients (P=0.000 and 0.001, respectively). CRP-2 was 96 (57–192) mg/l and 48 (24–96) mg/l in both groups (P=0.000). The area under curve was 0.741 and 0.712 for score-1 and score-2, respectively. Score-1 of 3.5 had a sensitivity and specificity of 79 and 59%, respectively, whereas score-2 of 2.25 was 91% sensitive and 41% specific for ARDS prediction. CRP-2 was the only significant predictor of in-hospital mortality (P=0.000). CRP-2 of 94 mg/l was 73% sensitive and 69% specific in predicting in-hospital mortality.
Conclusion Despite LIPS being validated in general hospitalized patients as significant predictors for ARDS in high-risk trauma and/or surgery ICU patients, their accuracy seems to be questionable.

Keywords: acute respiratory distress syndrome, lung injury prediction score, major surgery, major trauma


How to cite this article:
El-Hady Ahmed M, Hamed G, Fawzy S, Taema KM. Predicting acute respiratory distress syndrome in high-risk trauma and surgical patients: validation of previous scores. Egypt J Chest Dis Tuberc 2020;69:148-54

How to cite this URL:
El-Hady Ahmed M, Hamed G, Fawzy S, Taema KM. Predicting acute respiratory distress syndrome in high-risk trauma and surgical patients: validation of previous scores. Egypt J Chest Dis Tuberc [serial online] 2020 [cited 2020 Jun 4];69:148-54. Available from: http://www.ejcdt.eg.net/text.asp?2020/69/1/148/277314




  Introduction Top


Despite the recent developments in the management of acute respiratory distress syndrome (ARDS), its mortality rate remains high [1],[2]. Being with limited specific therapeutic options [3], identifying patients at risk for ARDS is important for searching for prevention strategies [4]. Many scores have been studied for predicting ARDS [5],[6] in several patient categories. The use of these scores in the subgroups of trauma and surgery patients is not well validated.

The exposure to stresses of trauma or surgery rendered those patients an important subgroup who are more liable to ARDS [7]. ARDS occurred in 52.6% of surgical patients [8] and in 12–25% of patients with trauma [1],[9],[10], with mortality rates as high as 50–80% [11],[12],[13].

We intended to validate two previously established scores in predicting ARDS and in-hospital mortality in critically ill patients [with acute physiology and chronic health evaluation-II (APACHE-II) score ≥15] with major trauma and/or surgeries.


  Patients and methods Top


We recruited patients admitted to ICU with major surgery and/or trauma and APACHE-II score more than or equal to 15 during the period from January 2016 to May 2017 in this prospective observational study. We excluded from the study patients diagnosed to have ARDS on admission, patients who had a history of previous hospital admission within 7 days, patients in whom cardiac cause may explain their hypoxia, and those with age less than 18 years old.

We reported the predisposing conditions of ARDS and hemodynamic parameters including heart rate, systolic blood pressure, and diastolic blood pressure. Dividing the heart rate by the systolic blood pressure represents the shock index [14]. We measured C-reactive protein (CRP) levels on admission (CRP-1) and 2 days later (CRP-2). Lung injury prediction scores (LIPS) were subsequently calculated using two previously studied scores that will be referred in our study as score-1 [5] and score-2 [6].

The primary and secondary outcomes were the ARDS development and the in-hospital mortality, respectively. ARDS diagnosis was established by two different ICU physicians blinded to score results according to Berlin definition [15]. Patients or first-degree relatives were consented before enrollment. The study protocol was approved by the institutional review board at Cairo University.

Statistical methods

The study variables were studied in terms of normality. Shapiro–Wilk’s test with P value more than 0.05 [16],[17] and z value of skewness and kurtosis between −1.96 and +1.96 [18] were required to consider a variable as normally distributed. Being non-normally distributed, we expressed the continuous variables as median (Q1Q3). Categorical variables were expressed as frequency and proportion. When two groups were studied, nonparametric test (Mann–Whitney U test) was used for quantitative variables and χ2 test was used for qualitative data. Receiver operator characteristic analysis was performed to define test accuracy. The accuracy of different variables was expressed as area under curve (AUC) for the receiver operator characteristic. The highest Youden’s index was used to identify the best cutoff values. The statistical significance was identified at P value less than or equal to 0.05. The software used was the statistical package of social science (IBM Corporation, Released 2013, IBM SPSS Statistics for Windows, Version 22.0, Armonk, NY: IBM Corporation).


  Results Top


We recruited 93 patients in our study. Fourteen patients were excluded later owing to suspected cardiac cause in eight patients and admission by ARDS in six patients. The remaining 79 patients represented the study sample. The demographic data and predisposing conditions are seen in ([Table 1]).
Table 1 Comorbidities and predisposing conditions in study groups

Click here to view


Thirty-three (41.8%) patients developed ARDS during hospital stay and 46 (58.2%) patients did not develop ARDS. ARDS was developed after 4 (2–7) days. The presence of predisposing conditions in both groups is seen in ([Table 1]).

Both LIPS were significant predictors for ARDS development. The score-1 was 5 (4–6) in patients who developed ARDS versus 3 (1.8–4.3) in those who did not develop ARDS (P=0.000) and the score-2 was 4.5 (3–6.3) and 3 (1.5–4.6) in those who developed and did not develop ARDS, respectively (P=0.001). The CRP-1 was 48 (24–48) mg/l in both groups (P=0.96), whereas CRP-2 was 96 (57–192) mg/l and 48 (24–96) mg/l in both groups, respectively (P=0.000) ([Figure 1]). The AUC was 0.741 for score-1, 0.712 for score-2, and 0.740 for CRP-2 ([Figure 2]). The best cutoff values of the different variables are shown in ([Table 2]).
Figure 1 The different scores and CRP measures in relation to ARDS development. ARDS, acute respiratory distress syndrome; CRP, C-reactive protein.

Click here to view
Figure 2 The ROC curve for score-1, score-2, and CRP-2 in predicting ARDS. ARDS, acute respiratory distress syndrome; CRP, C-reactive protein; ROC, receiver operator characteristic.

Click here to view
Table 2 The best cutoff values of the different variables in predicting acute respiratory distress syndrome

Click here to view


The in-hospital mortality was reported in 30 (38% mortality rate) patients, with significantly higher mortality in patients with ARDS. The patients with ARDS had mortality rate of 66.7% (22 patients), whereas those without ARDS had mortality rate of 17.4% (eight patients) (P=0.000). Score-1 and score-2 were 3 (2–5) and 3.5 (1.5–5.3) in survivors compared with 4 (3–6) and 4.5 (2.4–5.5) in nonsurvivors (P=0.12 and 0.47 for both scores, respectively). The CRP-1 was 48 (24–48) mg/l in both survivors and nonsurvivors (P=0.58). CRP-2 was 48 (24–96) mg/l in survivors versus 96 (62–192) mg/l in nonsurvivors (P=0.000). The AUC for CRP-2 in predicting in-hospital mortality was 0.747. CRP-2 of 94 mg/l was seen to have sensitivity, specificity, positive predictive value, and negative predictive value of 73, 70, 60, and 81%, respectively ([Figure 3]).
Figure 3 The ROC curve for the scores and CRP-2 in predicting in-hospital mortality. CRP, C-reactive protein; ROC, receiver operator characteristic.

Click here to view



  Discussion Top


Surgical and traumatic critically ill patients have different inflammatory pathways rendering them unique pathophysiologic subset of patients with ARDS. Early recognition of those patients at higher risk for ARDS can help in identifying subset of patients who can be included in preventive studies and to modify the strategies of treatment before the disease becomes too severe. Some investigators derived prediction scores for ARDS development in general hospital admissions not specified to surgery and/or trauma patients [5],[6]. The confirmed validity and simplicity of these scores increased their use in predicting patients at risk for ARDS. We intended to validate two previously studied scores [5],[6] in predicting ARDS and in-hospital mortality in critically ill surgical and/or trauma patients.

ARDS developed in 41.8% of our sample. The ARDS incidence was 5.5% in one study [5] and was 17% in the other [6]. Higher incidence of 52.6%, which is closer to ours, was reported by Bauman et al. [8] in surgical critical care patients. Bauman and colleagues speculated this higher incidence to the different ARDS diagnostic criteria. We should also consider the more critically ill patients enrolled in Bauman’s study and in our study. They enrolled ventilated patients and we enrolled patients with high APACHE-II, whereas the other studies enrolled all emergency department (ED) patients with one risk factor for ARDS.

We found that both scores are significant predictors for ARDS development. The accuracy of these scores were however not high in predicting ARDS. The AUC was 0.741 for score-1 and 0.712 for score-2. In general ED patients, score-1 [5] and score-2 [6] had higher AUC of 0.85 and 0.84 for predicting ARDS. Bauman et al. [8] found, however, lower AUC of 0.79 when applying the LIPS in surgically ventilated patients. Other authors showed that a surgical lung injury prediction score is a poor ARDS predictor in high-risk surgical patients with AUC of 0.56. They modified this score to a surgical lung injury prediction-2 score, which had an AUC of 0.84 [19].

In the validation study of score-2, the authors identified a cutoff value of 3 to be 69% sensitive and 84% specific. In trauma and/or surgery patients, we identified a score of 2.25 to be 91% sensitive but with only 41% specificity. Bauman et al. [8] showed that every one unit increase in the score is associated with 50% increase in the ARDS risk with much higher cutoff value of 7 in surgical cohort of patients.

The enthusiasm associated with the use of biomarkers in clinical medicine had directed our attention to its use in predicting ARDS. Despite that a biomarker that is specifically involved in ARDS pathogenesis is expected to be more specific, we used the CRP which is in common clinical use in a variety of ICU conditions. The CRP-2 but not CRP-1 was found as a significant ARDS predictor in trauma and/or surgery patients with an AUC of 0.740. The CRP-2 of 57 mg/l was seen to be 76% sensitive and 67% specific. Zheng et al. [20] also identified the CRP as a predictor for ARDS occurrence in patients with trauma. The incorporation of CRP with other more specific predisposing conditions in one score was initially derived but needs to be validated [21].

We reported a mortality rate of 38% with significantly higher mortality in patients with ARDS. More than 60% of patients with ARDS were nonsurvivors. Many studies showed mortality rates for ARDS in patients with trauma of 24% [1],[12],[22] and in postoperative patients to be between 15 and 45% [7],[23],[24],[25]. We postulated the higher mortality rates in our study to the enrollment of critically ill patients with APACHE-II more than or equal to 15. Salim et al. [26] identified the occurrence of ARDS as an independent mortality predictor in nontrauma surgical patients.We could not elucidate any relation between the studied LIPS and mortality. The CRP-2 was the only mortality predictor in our study. Contrary to these findings, Bauman et al. [8] found that the LIPS is a significant mortality predictor in surgical patients with 22% increase in the 30-day mortality for every unit increase in the score. Even in general ICU patients, Soto et al. [27] identified the LIPS as a mortality predictor. We found an AUC of 0.747 for CRP-2 to predict in-hospital mortality with a cutoff value of 94 mg/l to be 73% sensitive and 70% specific. Sharma et al. [28] identified the high CRP as a mortality predictor in ARDS, whereas Luecke et al. [29] found no relation between CRP and survival. Irrelevant on the presence of ARDS, CRP was seen to be a mortality predictor in all spectrum of critically ill patients [30].

The absence of some predisposing conditions for ARDS in the score items, such as the large volume of blood transfusion and fluids [31],[32], might affect the score accuracy. Using those additional variables which are specifically important in the subsets of trauma and/or surgery patients might improve the accuracy of these scoring systems. This will however affect the easy and fast nature of these scores.

Being a single-center study with small sample size is considered a limitation of this study. The enrollment of patients with higher APACHE-II scores who eventually have higher ARDS risk might explain the small sample size.


  Conclusion Top


In conclusion, the LIPS that were validated in general hospitalized patients are shown to be significant predictors for ARDS in high-risk trauma and/or surgery ICU patients but with lower accuracy.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Rubenfeld GD, Caldwell E, Peabody E, Weaver J, Martin DP, Neff M et al. Incidence and outcomes of acute lung injury. N Engl J Med 2005; 353:1685.  Back to cited text no. 1
    
2.
Doyle RL, Szaflarski N, Modin GW, Wiener-Kronish JP, Matthay MA. Identification of patients with acute lung injury. Predictors of mortality. Am J Respir Crit Care Med 1995; 152:1818.  Back to cited text no. 2
    
3.
Duggal A, Ganapathy A, Ratnapalan M, A N. Pharmacological treatments for acute respiratory distress syndrome: systematic review. Minerva Anestesiol 2015; 81:567.  Back to cited text no. 3
    
4.
Spragg RG, Bernard GR, Checkley W, Curtis JR, Gajic O, Guyatt G et al. Beyond mortality: future clinical research in acute lung injury. Am J Respir Crit Care Med 2010; 181:1121.  Back to cited text no. 4
    
5.
Cartin-Ceba R, Trillo-Alvarez C, Li G, Kashyap R, Kojicic M, Dong Y et al. Derivation of a lung injury prediction score (LIPS) to identify patients at high risk of ARDS at the time of hospital admission. Am J Respir Crit Care Med 2009; A4653.  Back to cited text no. 5
    
6.
Trillo-Alvarez C, Cartin-Ceba R, Kor DJ, Kojicic M, Kashyap R, Thakur S et al. Acute lung injury prediction score: derivation and validation in a population-based sample. Eur Respir J 2011; 37:604.  Back to cited text no. 6
    
7.
Kor DJ, Warner DO, Alsara A, Fernandez-Perez ER, Malinchoc M, Kashyap R et al. Derivation and diagnostic accuracy of the surgical lung injury prediction model. Anesthesiology 2011; 115:117–128.  Back to cited text no. 7
    
8.
Bauman ZM, Gassner MY, Coughlin MA, Mahan M, Watras J. Lung injury prediction score is useful in predicting acute respiratory distress syndrome and mortality in surgical critical care patients. Crit Care Res Pract 2015; 2015:157408.  Back to cited text no. 8
    
9.
Treggiari MM, Hudson LD, Martin DP, Weiss NS, Caldwell E, Rubenfeld G. Effect of acute lung injury and acute respiratory distress syndrome on outcome in critically ill trauma patients. Crit Care Med 2004; 32:327–331.  Back to cited text no. 9
    
10.
Poole GV, Ward EF, Griswold JA, Muakkassa FF, Hsu HS. Complications of pelvic fractures from blunt trauma. Am Surg 1992; 58:225–231.  Back to cited text no. 10
    
11.
Sauaia A, Moore FA, Moore EE, Haenel JB, Read RA, Lezotte DC. Early predictors of postinjury multiple organ failure. Arch Surg 1994; 129:39–45.  Back to cited text no. 11
    
12.
Hudson LD, Milberg JA, Anardi D, Maunder RJ. Clinical risks for development of the acute respiratory distress syndrome. Am J Respir Crit Care Med 1995; 151:293–301.  Back to cited text no. 12
    
13.
Durham RM, Moran JJ, Mazuski JE, Shapiro MJ, Baue AE, Flint LM. Multiple organ failure in trauma patients. J Trauma 2003; 55:608–616.  Back to cited text no. 13
    
14.
Berger T, Green J, Horeczko T, Hagar Y, Garg N, Suarez A et al. Shock index and early recognition of sepsis in the emergency department: pilot study. West J Emerg Med 2013; 14:168.  Back to cited text no. 14
    
15.
Force* TADT. Acute respiratory distress syndrome: The berlin definition. JAMA 2012; 307:2526.  Back to cited text no. 15
    
16.
Shapiro SS, Wilk MB. An analysis of variance test for normality (complete samples). Biometrika 1965; 52:591.  Back to cited text no. 16
    
17.
Razali NM, Wah YB. Power comparisons of Shapiro-Wilk, Kolmogorov-Smirnov, Lilliefors and Anderson-Darling tests. J Stat Model Anal 2011; 2:21.  Back to cited text no. 17
    
18.
Doane DP, Seward LE. Measuring skewness: a forgotten statistic. J Stat Educ 2011; 19:1.  Back to cited text no. 18
    
19.
Kor DJ, Lingineni RK, Gajic O, Park PK, Blum JM, Hou PC et al. Predicting risk of postoperative lung injury in high-risk surgical patients: a multicenter cohort study. Anesthesiology 2014; 120:1168–1181.  Back to cited text no. 19
    
20.
Zheng Y, Huang X, Liang H. Observation on CRP levels in severe trauma patients with ARDS and its significance. J Trauma Surg 2005; 4:274.  Back to cited text no. 20
    
21.
Ahmed MEH, Hamed G, Fawzy S, Taema KM. Lung injury prediction scores: Clinical validation and C-reactive protein involvement in high risk patients. Med Intensiva 2019; 13:S0210–5691.  Back to cited text no. 21
    
22.
Hoyt DB, Simons RK, Winchell RJ, Cushman J, Hollingsworth-Fridlund P, Holbrook T et al. A risk analysis of pulmonary complications following major trauma. J Trauma 1993; 35:524–531.  Back to cited text no. 22
    
23.
Fernández-Pérez ER, Sprung J, Afessa B, Warner DO, Vachon CM, Schroeder DR et al. Intraoperative ventilator settings and acute lung injury after elective surgery: a nested case-control study. Thorax 2008; 64:121–127.  Back to cited text no. 23
    
24.
Alam N, Park BJ, Wilton A, Seshan VE, Bains MS, Downey RJ et al. Incidence and risk factors for lung injury after lung cancer resection. Ann Thorac Surg 2007; 84:1085–1091.  Back to cited text no. 24
    
25.
Sen S, Sen S, Senturk E, Kuman NK. Postresectional lung injury in thoracic surgery pre and intraoperative risk factors: a retrospective clinical study of a hundred forty-three cases. J Cardiothorac Surg 2010; 5:62.  Back to cited text no. 25
    
26.
Salim A, Martin M, Constantinou C, Sangthong B, Brown C, Kasotakis G et al. Acute respiratory distress syndrome in the trauma intensive care unit: morbid but not mortal. Arch Surg 2006; 141:655–658.  Back to cited text no. 26
    
27.
Soto GJ, Kor DJ, Park PK, Hou PC, Kaufman DA, Kim M et al. Lung injury prediction score in hospitalized patients at risk of acute respiratory distress syndrome. Crit Care Med 2016; 44:2182.  Back to cited text no. 27
    
28.
Sharma SK, Gupta A, Biswas A, Sharma A, Malhotra A, Prasad KT et al. Aetiology, outcomes & predictors of mortality in acute respiratory distress syndrome from a tertiary care centre in north India. Indian J Med Res 2016; 143:782.  Back to cited text no. 28
    
29.
Luecke T, Muench E, Roth H, Friess U, Paul T, Kleinhuber K et al. Predictors of mortality in ARDS patients referred to a tertiary care centre: a pilot study. Eur J Anaesthesiol 2006; 23:403.  Back to cited text no. 29
    
30.
Zhang Z, Ni H. C-reactive protein as a predictor of mortality in critically ill patients: a meta-analysis and systematic review. Anaesth Intensive Care 2011; 39:854.  Back to cited text no. 30
    
31.
Gong MN, Thompson BT, Williams P, Pothier L, Boyce PD, Christiani DC. Clinical predictors of and mortality in acute respiratory distress syndrome: potential role of red cell transfusion. Crit Care Med 2005; 33:1191.  Back to cited text no. 31
    
32.
Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, Hayden D, DeBoisblanc B et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006; 354:2564.  Back to cited text no. 32
    


    Figures

  [Figure 1], [Figure 2], [Figure 3]
 
 
    Tables

  [Table 1], [Table 2]



 

Top
 
  Search
 
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

 
  In this article
Abstract
Introduction
Patients and methods
Results
Discussion
Conclusion
References
Article Figures
Article Tables

 Article Access Statistics
    Viewed214    
    Printed9    
    Emailed0    
    PDF Downloaded24    
    Comments [Add]    

Recommend this journal