Norepinephrine corrected shock index as an innovative hemodynamic based prognosticator in critically ill patients

Background : Hemodynamic instability in critically ill patients can be assessed by various hemodynamic valid indicators. One of these approved hemodynamics’ prognosticator is the Shock Index (SI), which integrates both the heart rate and systolic blood pressure in one composite indicator. Arbitrarily, higher vasopressor rate indicates poorer prognostic. Whatever, Norepinephrine, the preferred vasopressor in most shock related scenarios, has a tendency to increase heart rate beside its vasoconstriction associated systolic blood pressure augmentation. Aim : In this study, we primarily aim to investigate the predictive power of a newer proposed composite hemodynamic prognosticator which integrates the average Norepinephrine infusion rates with the assessed patients’ Shock Indexes, and to explore its Sensitivity Indices regarding the critically ill patients’ major outcomes. Methods : This study trial was a non-funded


Introduction
In contrast to stable patients, a several possible of Intensive Care Unit (ICU) related potential confounder factors, particularly the vasopressors and the corresponding shock statuses, insulin infusion management therapy, and others fewer impacting confounders (e.g., Oxygenation, acid-base, temperature, and hematocrit statuses) or possible interacting drugs (e.g., Vitamin C, mannitol, and paracetamol), may interfere with the BG_Glk reliability in BG monitoring against the reference BG_Lab. Notably, the BG_Glk precision is often considered clinically accepted, even in ICU admitted patients, as long as the gap against the BG_Lab doesn't exceed 15% or drop into the negative gap direction 1-6 .
In this study, we primarily aim to investigate the predictive power of a newer proposed composite hemodynamic prognosticator which integrates the average Norepinephrine infusion rates with the assessed patients' Shock Indexes, and to explore its Sensitivity Indices regarding the critically ill patients' major outcomes.

Material and methods
This study trial was a non-funded, non-sponsored, observational study, which was conducted retrospectively on the Intensive Care Unit (ICU) at King Hussein Medical Center, Royal Medical Services, Amman, Jordan, over 60 months between Jan 2018 and Dec 2022.
Exclusion criteria including but not excluded to, admission days less than 24 hours and missing data more than 80%. The Age-adjusted Charlson Co-Morbidity Index (AACCI) was used for the co-morbidity burden assessment, and a dichotomized value of 8 was adopted in the comparison analysis across Cohort I-II. All eligible investigated critically ill patients were dichotomously divided into 2 comparative NE×SI products' cohorts; lower product's (NE×SI<14.835 µg.bpm/min. mmHg) cohort (Cohort I) versus higher product's (NE×SI≥14.835 µg.bpm/min. mmHg) cohort (Cohort II).
A Chi Square test were conducted across these 2 dichotomized cohorts to express the comparison results as Number (Percentages), strength of associations (odd ratios), Pearson chi-square statistic (χ 2), Goodness of Fit (G-Test of independence), and Pearson (r) and Spearman (ρ) correlations.
The admitted ICU patients' mortality statuses were defined in our study as Survivors versus Non-Survivors, Survivors with LOS <3 weeks versus Survivors with LOS ≥ 3 weeks, and Early Mortality if LOS ≤2 weeks versus Late Mortality if LOS >2 weeks. The NE infusion was stocked at concentration of 60 mcg/ml and the infusion rate was labelled in mcg/min. The NE rates were categorized to 3 mcg/min incremental rate from 0-≥18 mcg/min. The Binary Logistic Regression (BLgR) analysis was conducted for the tested patients' prognosticator (NE×SI) against the probability of being allocated to the Non-Survivors State (The Positive State) rather than to the Survivors State (The Negative State). The BLgR analysis was primarily conducted to abstract the necessary coefficients to present the corresponding logistic model. The Receiver Operating Characteristic (ROC) and Sensitivity Analysis were processed on a total of 5745 processed cases for the investigated prognosticator against the higher probability of 28-day overall mortality (Positive State and assigned as 1) versus the lower probability (Negative State and assigned as 0) to explore the area under the ROC curves (AUROC±SEM), optimal cut-off points, sensitivities (TPRs), specificities (TNRs), positive and negative predictive values (PPVs and NPVs), positive and negative likelihoods ratios (PLRs and NLRs), and the Youden and accuracy indices (YIs and AIs). For our tested prognosticator, the higher values of the NE× SI indicate stronger evidences for the Positive State (Higher %Prob of mortality). While the lower values of the tested independent variable indicate stronger evidences for the Negative State (lower %Prob of mortality). Statistical analysis was performed using Statistical Package for Social Science (SPSS) software version 23.0. Statistical significance was set at 5%.

Results
The probability of critically ill patients' mortality can be mathematically proposed according to their tracked NE×SI products. In this study, we constructed a BLgR model to prognosticate the admitted ICU patients' mortality and was formulated as [e (-6.986+0.516× NE.SI) /1+ e (-6.986+0.516× NE.SI)]. The explained variations in the dependent variable based on the adopted independent investigated composited variable ranged significantly from 49.1%-65.8% depending on whether you reference the Cox & Snell R2 or Nagelkerke R2 methods, respectively, and correctly classified 83.8% of the cases, χ 2 (8) = 315.67.
Actually, 2528 and 3217 cases were processed as positive actual states (Positive OI, Non-Survivors) and as negative actual states (Negative OI, Survivors) respectively. The AUC±SEM for the NE×SI was significantly determined at 0.944±0.003 (95% CI; 0.938-0.949). The probability of the positive OI was 66.12% at the optimal operating cutoff point of 14.84 µg.bpm/min. mmHg.
The overall tested gender's ratio (Male: Female ratio, M: F) in this study was assigned to 2 All the tested patients' analysis results and illustrations were clearly and fully presented in Table 1   The sensitivity analysis was processed on a total of 5745 processed cases for the investigated prognosticator in the Jordanian investigated patients, against the higher probability of 28-day overall mortality (Positive state and assigned as 1) versus the lower probability (Negative state and assigned as 0) to explore the optimal cut-off points, sensitivities (TPRs), specificities (TNRs), positive and negative predictive values (PPVs and NPVs), positive and negative likelihoods ratios (PLRs and NLRs), and the Youden and accuracy indices (YIs and AIs). 2528 and 3217 cases were processed as positive actual states and as negative actual states, respectively. For our tested prognosticator, the higher values of the NE. SI indicate stronger evidences for the Positive State (Higher %Prob of mortality). While the lower values of the tested independent variable indicate stronger evidences for the Negative State (lower %Prob of mortality).

Discussion
Our observational retrospective study was pursued for Jordanian admitted critically ill patients, including both gender and wide investigated age's ranges, at the King Hussein Medical Center.
An innovative proposed prognosticator was tested in this study to explore its sensitivity utilities in predicting the probability of being on the Non-Survivors' Cohort rather than being on the Survivors' Cohort. This explored mortality predictor integrates dual hemodynamic related independent variables; the Norepinephrine rate in mcg per min (NE) and the shock index (SI) in bpm per mmHg, in one composited tested variable.
A two contrarily cohorts were yielded after we opted the optimal operating point of the NE ×SI (14.835 µg.bpm/min. mmHg) as a dichotomized level. In this study, we compare between the comparative yielded cohorts; the lower composited prognosticator product (Cohort I) versus the higher composited prognosticator product (Cohort II). At this optimal product level, we explored that the probability of mortality for critically ill patients in our critical unit was 66.12%. Also, we determined at this abstracted optimal point that the sensitivity, specificity, accuracy index and positive/negative predictive values were 87.2%, 86.85%, 87.00%, 83.90%, and 89.61%, respectively.

Conclusion
Our results revealed that our proposed an innovative hemodynamic composited product had a reasonable constructed area under the curve with an interesting sensitivity index for prognosticating the admitted critically ill patients' mortality rate. This study is limited by its retrospective design, single-center, and relatively small sample size.

Acknowledgement
Our appreciation goes to staff of the department of King Hussein Medical Center for their enormous assistance and advice.

Disclosure of conflict of interest
There is no conflict of interest in this manuscript.

Statement of ethical approval
There is no animal/human subject involvement in this manuscript.