Clinical Usefulness Of C-Reactive Protein Versus Shock Index In Predicting Mortality In Septic Critically Ill Patients Who Are Taking Nor-epinephrine

ABSTRACT

Objective: C-reactive protein (CRP) and shock index (SI) have been previously shown to identify high risk septic shock patients.Our objective was to compare the ability of SI and CRP to predict the primary outcome of overall 28-day mortality, and the secondary outcomes of early mortality (≤ 14 days), late mortality (>14 days) in septic critically ill patients who are taking norepinephrine as a vasopressor.

Methods:  We performed a retrospective analysis of patientsadmitted to our adult ICU between April 2017 and Sep 2018 who were meet the inclusion criteria. Independent T-test, Mann Whitney U test, and χ² test were used to express all patient variables.A receiver operating char­acteristic (ROC) curve followed by sensitivity analysis was generated to determine the predictive performances, and the optimal cut-off values for CRP and SI. The binary logistic regression model was used to generate CRP and SI predictive equations and correlation plots for early, late, and overall 28-day ICU mortality.

Results : A total of 163 critically ill patients were finally included in this study. The mean overall age was 58.37±9.96 years, and 112 subjects (68.71%) were male. The early, late, and overall 28-day ICU mortality rate were9.82%, 29.45%, and 39.82%, respectively. SI and CRP were significantly higher in non-survivors (1.29±0.17 bpm/mmHg and 43.09±19.28 mg/dl, p<0.05) than in survivors (1.12±0.03 bpm/mmHg, and 28.38±14.38 mg/dl).

Conclusion: SI is an effective, no-cost bedside modality, which is a realistic, reliable, and discriminative prognosticator with high sensitivity, specificity, performance, and accuracy when compared with CRP.

Key words: C-reactive protein, Shock index, Critically ill patients, Norepinephrine.

JRMS December 2019;26(3):12-27/DOI: 10.12816/0054814 

Introduction 

Sepsis is a complex syndrome caused by the body’s systemic response to an infection with a major cause of high treatment cost, single or multiple organ dysfunction, morbidity, and mortality [1]. The importance of having reliable, cost effective, and easily attainable clinical prognostic indicators that would help to predictand differentiate the mortality risk of these critically ill patientsis invaluable within an Intensive Care Unit (ICU). C-reactive protein (CRP) is a useful positive acute-phase reactant marker that can predict morbidity and mortality among critically ill patients. However,the results of CRP and other severity indices of sepsis may not be immediately available upon request [2], potentially delaying effective dynamic risk stratification and goal directed management in these unstable studied cohort. Also, Shock index (SI), defined as heart rate (HR) over systolic blood pressure (SBP)is a readily and affordable attained comprehensive parameter that combines two physiological variables (HR and SBP) into a single ratio that has previously been shown to stratify and served as an early warning indicator of high risk haemorrhagic shock from various aetiologies when compared to other conventional vital signs [3-8].From previous studies, SI has never been compared with CRP for its value in predicting early mortality (≤ 14 days), late mortality (> 14 days), and overall 28-day mortality among septic critically ill patients who are taking nor-epinephrine. Our objective was to compare the mortality risk stratifying ability of SI and CRP regarding the primary outcome of overall 28-day mortality, and the secondary outcomes of early and late mortality and ICU length of stay.

Methods

Study design and setting

This was a single-centreobservational retrospective study conducted in the department of adult ICU of King Hussein Medical Centre (KHMC) at Royal Medical Services (RMS) in Jordan. This study was approved by our Institutional Review Board (IRB), and a requirement for consent was waived owing to its retrospective design. This study included a cohort of critically ill patients admitted to our adult ICU via the emergency department (ED) or via other hospital wards with any medicalor surgical problems. Flow chart of critically ill patient’s selection and data collection process is fully illustrated in (Figure1).

Statistical analysis

All patient continuous variables were expressed as mean± standard deviation by using the independent samples T-test while categorical and ordinal variables were expressed as numbers with percentages by using the χ² test or as median (interquartile range) by using the Mann-Whitney U test, respectively. Analysis values were compared for the two tested groups (survivors vs. non-survivors) and the non-survival group was further analysed after being divided into 2 subgroups, early (≤14 days) and late (>14 days) mortality. Univariate analysis was conducted first followed by multivariate logistic regression for the most possible affected patient’s variables associated with ICU mortality. A receiver operating char­acteristic (ROC) curve followed by sensitivity analysis wasused to determine the area under the ROC curves (AUROCs), predictive performances, and the optimal cut-off values for CRP and SI. Youden indices, sensitivities, specificities, positive and negative predictive val­ues, and accuracy indices were also calculated. The binary logistic regression model was used to generate CRP and SI predictive equations and correlation plots for early, late, and overall 28-day ICU mortality. Statisticalanalyses were performed using IBM SPSS ver. 25 (IBM Corp., Armonk, NY, USA) and P-values ≤0.05 were considered sta­tistically significant.

Results

Characteristics of the subjects

The mean overall age was 58.37±9.96 years, and 112 subjects (68.71%) were male. The overall 28-day ICU mortality rate was 39.26% (64 patients); in16 patients (9.82%), this was early mortality and in 48 patients (29.45%) it was late mortality. Demographics, admission co-morbidities and class, anthropometrics, and follow-up comparison data of the study’s critically ill patients are fully summarised in (Table I) and (Table II) , respectively.

 Mortality was significantly higher in medical than surgical critically ill patients. Baseline pre-ICU admission days and number of co-morbidities were also significantly higher in non-survivors than survivors. There were insignificant differences between the two groups regarding average child-Pugh score, average Glasgow coma scale (GSC), and average norepinephrine infusion rate. Despite baseline albumin levels (ALB1)being significantly higher in non-survivors (2.94±0.39 g/dl)than survivors (2.63±0.20 g/dl), survivors had significantly higher average administered human albumin dosesand average protein density (PD) inputs (18.89±3.16 g/day and 3.72±0.74 g/100 Cal, respectively)than non-survivors (14.06±6.09 g/day and3.50±0.36 g/100 Cal). SI, HR, and CRP were significantly higher in non-survivors (1.29±0.17 bpm/mmHg, 120.88±5.55 bpm, and 43.09±19.28 mg/dl, respectively) than in survivors (1.12±0.03 bpm/mmHg, 114.76±1.19 bpm, and 28.38±14.38 mg/dl). In contrast, all haemodynamic parameters of SBP, diastolic blood pressure (DBP), and mean arterial pressure (MAP) had significantly higher values in survivors (102.29±2.19 mmHg, 52.29±2.19 mmHg, and 68.95±2.27 mmHg, respectively) than in non-survivors (94.39±6.14 mmHg, 44.39±6.14 mmHg, and 61.11±6.18 mmHg, respectively). SI and CRP were significantly higher in early non-survivors (1.49±0.23 bpm/mmHg and 50.55±21.88 mg/dl, respectively) than in late non-survivors (1.22±0.04 bpm/mmHg and 40.61±17.89 mg/dl, respectively).

Logistic regression analysis

In the univariate analyses, child-Pugh score, CRP, SBP, DBP, MAP, and SI showed statistically significant associations with early (OR, 0.07, 0.47,0.45, 0.45, 0.45, and 5.0*10²⁸, respectively), late (OR, 0.31,0.87,0.91, 0.91, 0.92, and 11.25, respectively), and overall 28-day ICU mortality (OR, 0.17, 0.78, 0.42, 0.42, 0.45, and 1.6*10³⁸, respectively),while total calorie (TC) inputs and HR showed only statistically significant associations with early mortality (OR, 0.99) and late mortality (OR, 1.08), respectively. Average albumin levels (ALB) during the first week of ICU admission showed astatistically significant association with both early (OR, 248.65) and overall 28-day ICU mortality (OR, 47.06). After adjusting for these variables, only SI still showed a statistically significant association with early, late, and overall 28-day ICU mortality,while CRP still showed statistically significant association with only overall 28-day ICU mortality. The odd ratios (ORs) of early, late, and all-cause 28-day ICU mortality events are shown in (Table III).

Prognostic value of the CRP and SI

(Table IV) shows the optimal cut-off point, sensitivity (TPR), specificity (TNR), Youden index (YI), positive and nega­tive predictive values (PPV and NPV), accuracy index (AI), and expected mortality rate of both tested prognostic indicators. The best cut-off values for CRP and SI in our study were 29.5 mg/dl and 1.29 bpm/mmHg, respectively, for early mortality, 34 mg/dl and 1.18 bpm/mmHg, respectively, for late mortality, and 33.60 mg/dl and 1.18 bpm/mmHg, respectively, for overall 28-day ICU mortality. The AUROCs of SI in this study were significantly greater than those of CRP in all mortality groups with 0.996; 95% CI, 0.990-1.00 vs. 0.764; 95% CI, 0.675-0.853, respectively, for early mortality, 0.844; 95% CI, 0.780-0.907 vs. 0.671; 95% CI, 0.583-0.907, respectively, for late mortality, and 0.984; 95% CI, 0.970-0.998 vs. 0.747; 95% CI, 0.672-0.822, respectively, for overall 28-day ICU mortality.The ROC curve analyses and mortality correlations of CRP and SI for early, late, and 28-day ICU mortality are shown in ( Fig 2-10 )

Discussion 

The present study included septic mechanically ventilated critically ill patients who are taking norepinephrine as a vasopressor at an overall average rate of 9.53±1.79 mcg/min. To the best of our knowledge, this is the first study to address the correlations between SI, CRP, and mortality. In the context of ever-shrinking resources, early stratification with fast, affordable, valid, reliable, and discriminativepredictive tools are critically needed in this unstable, high acuity, and high uncertainty status of the septic critically ill to avoid any potentialdelay or under-triaging while appropriately assigning a higher priority to sicker patients [9].SI emphasises current physiologic no-cost bedside triage dynamic rather than static toolsthat can be used at any time for triage decisions regarding septic patients while waiting for the results of other diagnostics, especially white blood cells (WBCs) with differential, CRP, and procalcitonin (PCT) [10-19]. After careful analysis of the data, SI shows higher sensitivity, performance, specificity, positive and negative predictive value, and accuracy than CRP in both late (91.70% vs. 64.60%,72.60% vs. 33.30%, 80.90% vs. 68.70%, 75.63% vs. 57.16%, 93.78% vs. 75.01%, and 85.14% vs. 67.09%, respectively) and 28-day ICU mortality (93.80% vs. 68.80%,87.70% vs. 42.50%,93.90% vs. 73.70%,90.86% vs. 62.84%, 95.91% vs. 78.51%, and 93.86% vs. 71.78%, respectively),while in the case of early mortality, SI shows higher performance, specificity, positive predictive value, and accuracy than CRP (98.60% vs. 55.10%, 98.60% vs. 55.10%, 97.88% vs. 59.01%, and 99.15% vs.72.73%, respectively).This study demonstrates a vast difference insignificance and predictive values of SI when compared with CRP, possibly due to the fact that norepinephrine, which was used as a vasopressor in these septic mechanically ventilated critically ill studied patients, givesrise to alterations of physiological parameters of HR and SBP, making the SI indicator a realistic reflection of the septic patients and a more reliable predictive prognosticator compared to CRP. 

In summary, SI is an effective, no-cost bedside modality, realistic, reliable, and discriminativeprognosticator with high sensitivity, specificity, performance, and accuracy when compared with CRP to predict early, late, and overall 28-day ICU mortality in septic mechanically ventilated critically ill patients who are taking norepinephrine as a vasopressor. SI may be used as an additional or readily available red flag bedside assessment tool for severe disease. This study is limited by its retrospective design, using single-centre data, including only septic mechanically ventilated ICU patients. Nonetheless, our centre isan experienced and high-volume unit, so our data may be useful in other centres.A larger, multisite, and prospective study is needed to control for multiple confounders and to clarify the causation between SI, CRP, and mortality and therole in resource utilisation, and risk stratification of septic patients.

     REFERENCES

1.Shapiro NI, Howell MD, Bates D, et al. The association of sepsis syndrome and organ dysfunction with mortality in emergency department patients with suspected infection. Ann Emerg Med. 2006; 48:583-590. 

2.Dellinger RP, Carlet JM, Masur H, et al. Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock. Crit Care Med 2004; 32: 858-73.

3.Birkhahn Robert H, Theodore J Gaeta, Shawn K Van Deusen, John Tloczkowski. The ability of traditional vital signs and shock index to identify ruptured ectopic pregnancy. American Journal of Obstetrics and Gynecology 2003; 189(5): 1293-6.

4.Zarzaur Ben L., Martin A. Croce, Peter E. Fischer, Louis J. Magnotti, Timothy C. Fabian.New vitals after injury: Shock index for the young and age x shock index for the old. Journal of Surgical Research 2008;147(2): 229-36. 

5.Brasel KJ, Guse C, Gentilello LM, et al. Heart rate: Is it truly a vital sign? J Trauma 2007; 62: 812. 12. 

6.Victorino GP, Battistella FD, Wisner DH. Does tachycardia correlate with hypotension after trauma? J Am CollSurg 2003; 196: 679.

7.Rady M, Smithline H, Blake H, et al. A comparison of the shock index and conventional vital signs to identify acute, critical illness in the emergency department. Ann Emerg Med. 1994; 24:685-690. 

8.Cannon CM, Braxton CC, Kling-Smith M, Mahnken JD, Carlton E, Moncure M. Utility of the shock index in predicting mortality in traumatically injured patients. JTrauma. 2009;67(6):1426–1430. 

9.Horeczko T, Enriquez B, McGrath NE, et al. The paediatric assessment triangle: Accuracy of its application by nurses in triage. J EmergNurs. 2012. In press. 

10.Becker KL, Snider R, Nylen ES. Procalcitonin assay in systemic inflammation, infection, and sepsis: clinical utility and limitations.Crit Care Med. 2008;36:941-52.

11.Castelli GP, Pognani C, Cita M,Paladini R. Procalcitonin as a prognostic and diagnostic tool for septic complications after major trauma. Crit Care Med. 2009;37:1845-9.

12.Becker KL, Snider R, Nylen ES. Procalcitonin in sepsis and systemic inflammation: a harmful biomarker and a therapeutictarget. Br J Pharmacol. 2010;159:253-64.

13.Castelli GP, Pognani C, Meisner M, Stuani A, Bellomi D, Sgarbi L. Procalcitonin and C-reactive protein during systemic inflammatory response syndrome, sepsis and organ dysfunction. Crit Care 2004;8:234–42.

14.Meidani M, Khorvash F, Abolghasemi H, Jamali B. Procalcitonin and quantitative C-reactive protein role in the early diagnosis of sepsis in patients with febrile neutropenia. South Asian J Cancer 2013;2:216–30.

15.Magrini L, Travaglino F, Marino R, Ferri E, De Berardinis B, Cardelli P, et al. Procalcitonin variations after Emergency Department admission are highly predictive of hospital mortality in patients with acute infectious diseases. Eur Rev Med PharmacolSci 2013;17(Suppl 1):133–42.

16.Tang H, Jing J, Bo D, Xu D. Biological variations of leukocyte numerical and morphologic parameters determined by UniCelDxH 800 hematologyanalyzer. Arch Pathol Lab Med 2012;136:1392–6.

17.surgerySapin F, Biston P, Piagnerelli M. Predictive value of C-reactive protein in critically ill patients after abdominal surgery. Clinics (Sao Paulo). 2017;72(1):23-29. Published 2017 Jan. doi:10.6061/clinics/2017(01)05

18.Oh TK, Song IA, Lee JH. Clinical usefulness of C-reactive protein to albumin ratio in predicting 30-day mortality in critically ill patients: A retrospective analysis. Sci Rep. 2018;8(1):14977

19.Christophe Lelubre,Sophie Anselin et al . Interpretation of C-Reactive Protein Concentrations in Critically Ill Patients, BioMed Research International 2013:124021, 11