The prognostic nutritional index is associated with mortality of patients in intensive care unit

Abstract

Background/Aim: It has been reported that the prognostic nutritional index (PNI) is ), an immunonutritional index, associated with poor prognosis, especially in cardiovascular and malignant diseases. However, the clinical significance of PNI in intensive care (ICU) patients remains unclear. In this study, we aimed to measure the predictive value of the PNI in predicting mortality in patients hospitalized in the ICU. Methods: A total of 80 patients hospitalized in the internal medicine ICU of our hospital between January 2021 and September 2021 were included in this observational cohort study. The patients' demographic characteristics, comorbidities, laboratory parameters, need for and duration of mechanical ventilation, length of stay in ICU, and mortality rates were retrospectively analyzed. The patients were divided into two groups according to their survival; the first group comprised of survivors while the second group comprised of those who died in the ICU. The two groups were compared in terms of all variables. Results: The mean age of all subjects included in the study was 63 (18.2) years and 50% (n=40) were female and 50% (n=40) were male. When patients are grouped as survivors and non-survivors, the mean age and sex distribution were similar (P=0.23, P=0.27, respectively). The median follow-up period of the patients was 5 (IQR 3-11) days and the mortality rate was 38.7% (n=31). Those in the non-survivor group had higher APACHE II and SOFA scores (P=0.02, P<0.001, respectively), and a lower PNI level (P=0.01). In the multivariate regression analysis, PNI value [OR: 1.210 (95%CI: 1.048-1.396) P=0.009] was the negative independent risk factor and SOFA score [OR: 1.697 (95%CI: 1.201-2.398) P=0.03] was a positive independent risk factor. Conclusion: Despite our small cohort, we believe our findings corroborate our hypothesis that as a simple and inexpensive test, PNI is a useful biomarker to assess mortality risk in ICU patients.

Comparison of nutritional risk indices (PNI, GNRI, mNUTRIC) and HALP score in predicting adverse clinical outcomes in older patients staying in an intensive care unit

“…However, both the GNRI and PNI scores were reported as independent risk factors for short-term mortality in earlier studies. [24,33] In our study, HALP score showed no significant relationship with ICU mortality (P = .78), which is inconsistent with earlier studies reporting that HALP is good at predicting mortality with either high or low sensitivity and specificity in geriatric patients with acute ischemic stroke. [29,34] In brief, the key finding of our study was the presence of an association between the indices used in the study, except for HALP score, ICU LOS, and ICU mortality.…”

“…The median scores of APACHE-II, SOFA, and mNUTRIC were significantly higher and the mean scores of PNI and GNRI were significantly lower in the non-survivor group (P > .05 for all) and were consistent with those reported in several studies on different patient populations. [22][23][24]27] In the present study, the HALP score did not differ between the survivor and non-survivor groups (P = .29), which was in line with the results of another study by Türkiye (P = .380). [28] However, in another study, the HALP score was significantly lower in the non-survivor group than in the survivor group of patients with ischemic stroke (P = .00).…”

“…In the present study, PNI (AUC = 0.636) predicted ICU mortality with a sensitivity of 51.2% and a specificity of 71.26%, and GNRI (AUC = 0.633) predicted ICU mortality with a sensitivity of 57.58% and a specificity of 65.52%, which is in line with those reported by Soylu et al [ 22 ] from Türkiye (a sensitivity of 56.1% and a specificity of 56.9% for PNI and a sensitivity of 60.6% and a specificity of 60.2% for GNRI). In another study by Türkiye, Dal et al [ 24 ] reported higher sensitivity and specificity for the PNI (at a cutoff value of 32.01, the sensitivity and specificity for predicting mortality were 82.9% and 95.6%, respectively; AUC = 0.957; P < .001). The difference between the results of the studies is probably due to differences in the age of the study population and the reason(s) for ICU admission.…”

“…In our study population, the ICU mortality rate was 43.1%, which was lower than that found in a retrospective observational study (65.1%) [ 22 ] but higher than that of 2 other retrospective observational studies (16.1% and 38.7%, respectively) from Türkiye. [ 23 , 24 ] This difference might be due to the reason for admission to the ICU. Consistent with the literature, [ 25 , 26 ] we found that severe malnutrition was more prevalent in non-survivors according to all nutritional indices, but the difference was significant only with the mNUTRIC (98.5% vs 77%, P = .001).…”