Research Article

Efficacy of Prognostic Scoring Systems and Neutrophil-to-Lymphocyte Ratio (NLR) Among Critically Ill Elder Sepsis Patients


  • Bilgin Bahadır Başgöz
  • Musa Barış Aykan
  • Ramazan Acar
  • İlker Taşçı

Received Date: 14.09.2020 Accepted Date: 23.07.2021 J Ankara Univ Fac Med 2021;74(3):289-295


Several scoring models and biomarkers are available to predict survival among critically ill patients. However, their accuracy among older adults with sepsis has been questioned. In this study, we aimed to evaluate the currently used prognostic scoring scales and neutrophil-tolymphocyte ratio (NLR) in the prediction of survival among older adults with sepsis.

Materials and Methods:

The study prospectively included older adults who were admitted to the intensive care unit (ICU) with sepsis. On admission, Acute Physiology And Chronic Health Evaluation (APACHE)-II, Sequential (sepsis-related) Organ Failure Assessment (SOFA), Glasgow coma scale (GCS) score, and NLR were calculated. By the end of a 28-day follow-up period, survivors and non-survivors were compared for the study parameters.


Overall, the data of 36 patients were analyzed (mean age: 80.00±6.37 years, female: 58.3%). The rate of mortality was 47.3% (n=17). The mean SOFA score and the median NLR on admission were significantly lower in survivors [SOFA: 10.37±2.91, and NLR: 9.64 (11.25)] vs non-survivors [SOFA: 12.82±3.21 and NLR: 14.95 (35.53)] (p<0.05) but the mean APACHE-II and the median GCS score were comparable. Using ROC curve analysis, we determined that only the SOFA score and NLR could predict mortality.


The present study showed that, among older adults with sepsis admitted to the ICU, baseline SOFA score and NLR but not APACHE-II or GCS score could successfully predict mortality. Further studies are required to evaluate the utility of existing prognosis scales in older people.

Keywords: Organ Dysfunction Scores, Sepsis, APACHE, Survival, Intensive Care Units, Aged


Sepsis is characterized by the development of organ failure as a result of the systemic inflammatory response of the host to infection (1). Almost 3% of all hospitalized patients have sepsis, 51% of sepsis patients were admitted to the intensive care unit (ICU), and all-cause mortality rate of sepsis patients is around 28% (2). With the advancing age, the incidence of sepsis is disproportionately increased, and the rate of mortality dramatically increases with age, up to 26.2% and 38.4%, respectively, in those over 85 years old (2,3).

In patients admitted to the ICU, prognostic scoring models are used to assess the severity of disease and to predict the risk of mortality. These include sequential (Sepsis-related) organ failure assessment (SOFA), Acute Physiology and Chronic Health Evaluation (APACHE)-II, and Glasgow coma scale (GCS) (4-6). The predictive power of these scoring systems in the determination of in-hospital mortality has been reported by many authors (7-9). Interestingly, studies comparing SOFA, APACHE-II, and GCS with each other in estimating the probability of adverse events in the ICU yielded contradictory results (10-13).

In addition to scoring models, the predictive power of neutrophil-to-lymphocyte ratio (NLR) on mortality is also studied in sepsis patients (14,15). However, the results of such studies focused on the association of NLR and mortality show discrepancies due to several confounders.

With the increased proportion of aged people in the population, more patients are hospitalized due to sepsis and admitted to the ICU (2). Nevertheless, the utility of existing prognostic scoring systems and NLR in critically ill older adults has not been studied extensively. A study with a small sample size published in 1993, and a large one published recently both suggested that APACHE was a better predictor of survival (16,17). However, many other studies failed to propose the utility of any tool in the clinical practice for patients aged 65 and over to estimate prognosis in the short term or long term (18).

Therefore, in this study, we aimed to evaluate the predictive value of prognostic scoring systems and NLR on mortality among older adult patients admitted to the ICU with the diagnosis of sepsis.

Materials and Methods

Setting and Participants

We prospectively included patients diagnosed with sepsis or septic shock aged 65 and over who were hospitalized in the ICU of a tertiary care hospital. The diagnosis of sepsis was made using the criteria in the 3rd Sepsis Consensus (Sepsis-3) report (19) and the International Sepsis and Septic Shock Management Guidelines (1) published jointly by the Society of Critical Care Medicine and European Society of Intensive Care Medicine. Subjects with a history of advanced dementia, end-stage cancer, cirrhosis, aged under 65 years, requiring urgent surgery, a history of recent trauma and lack of written consent from the patients or caregivers. The Health Sciences University Non-Invasive Research Ethics Committee approved the study protocol (no: 46418926-18/55). Written informed consent was obtained for each participant. All procedures followed the standards of the Turkish Medicine and Medical Devices Agency Good Clinical Practices Guidelines and per the Declaration of Helsinki.

Patient Characteristics and Procedures

On ICU admission, demographic and clinical characteristics, anthropometric variables, and comorbid conditions including diabetes mellitus, dyslipidemia, hypertension, congestive heart failure, coronary artery disease, chronic kidney disease, and chronic obstructive pulmonary disease were noted from electronic health records of the participants. Then, we calculated SOFA, APACHE-II and GCS scores to determine the severity of sepsis. In addition, we recorded neutrophil and lymphocyte counts and calculated the NLR. The predicted mortality rate of each patient was calculated by APACHE-II risk of death equation using the following information: The diagnosis leading to ICU admission, APACHE-II score, and requirement of emergency surgery (5). The duration of follow-up was 28 days, and the primary outcome was mortality from any cause. We divided the patients into two groups as “non-survivors” and “survivors” based on the 28-day death records.

Statistical Analysis

Statistical Package for Social Sciences (SPSS) (Version 23.0, Chicago, Illinois) program was used for statistical analysis. The distribution of the data was evaluated by the Shapiro-Wilk test. Results were expressed as the mean ± standard deviation for continuous variables or as the percentage for categorical variables. Skewed variables were presented as median interquartile range (IQR). The differences between the continuous variables in the non-survivor and survivor groups were compared Student’s t-test or Mann-Whitney U test according to the distribution of variables. Chi-square test was used to compare categoric variables. Pearson correlation coefficients were calculated to evaluate potential correlations. The power of SOFA, APACHE-II and GCS score in the prediction of the 28-day mortality rate was tested with the receiver operator characteristics (ROC) area under curve (AUC) analysis. ROC analysis was also used to determine cut-off values and their sensitivity and specificity to estimate the risk of mortality in the ICU. Statistical significance was accepted at the level of p<0.05.


The study included 36 patients with a mean age of 80.0±6.37 years (65 to 90 years) and female predominance (n=21, 58.3%). The rate of mortality during the 28-day follow-up was 47.3% (n=17). Non-survivor and survivor groups showed no statistical difference in terms of age, gender, body mass index, diabetes mellitus, dyslipidemia, hypertension, congestive heart failure, coronary artery disease, chronic kidney disease, and chronic obstructive pulmonary disease (Table 1). Baseline laboratory findings other than NLR were also similar. However, the NLR was significantly higher among non-survivors (p=0.04) (Table 1).

The non-survivor group had significantly higher mean SOFA score on admission (Table 2). However, APACHE-II score, predictive mortality rate, and GCS score showed no significant difference between the two groups (Table 2).

As shown in Table 3, a higher SOFA score and NLR showed a significant correlation with mortality (p=0.022, and p=0.034, respectively). However, there was no relationship between mortality of older sepsis patients and any of the APACHE-II score, predictive mortality rate, or GCS score (p=0.289 for APACHE-II score; p=0.331 for predicted mortality rate, and p=0.187 for GCS Score) (Table 3).

ROC-AUC analysis of these prognostic scoring models and NLR showed that both SOFA score and NLR have predictive power on the mortality of elderly patients with sepsis in the ICU (p=0.029, and p=0.040, respectively) (Table 3). However, APACHE-II score, predictive mortality rate, GCS score showed no utility in the prediction of mortality (p=0.311; p=0.303; and p=0.138 respectively) (Table 3). ROC curve diagrams are displayed in Figure 1.

The median values of SOFA, APACHE-II, predicted mortality rate, GCS and NLR are given in Table 4. The mortality rate of patients with a SOFA scores higher than the median value was significantly higher compared to those with a SOFA scores below the median (p=0.043). However, the mortality rate of patients with the above-median value of APACHE-II score, predicted mortality rate, GCS score and NLR at the day of admission were not different from the participants having below-median values (p=0.317; p=0.317; p=0.692 and p=0.130 respectively) (Table 4).


Despite improvements in diagnostic approaches and therapeutic interventions in recent years, sepsis and septic shock are still the leading causes of death among adults hospitalized in the ICU (1). Several scoring systems and also NLR are used to determine the severity of illness and to predict in-hospital mortality among critically ill patients (7-9,20). However, their usefulness among older patients has not been thoroughly identified and studies may reveal inconsistent results (14,15,18). Given the growing number of aged individuals in the community, the requirement of robust prediction tools to use in older adults is more critical. In this study, we evaluated the predictive value of prognostic scoring models SOFA, APACHE-II, GCS and NLR, on 28-day mortality among older adults with sepsis admitted to the ICU. We demonstrated that higher SOFA scores and higher NLR on admission were associated with increased mortality in patients with sepsis aged 65 years and over. However, APACHE-II score and GCS score did not show a similar association with mortality. In addition, we did not observe a relationship between mortality and predicted mortality rate derived from the APACHE-II score.

Given the mortality rate of 47.2% in the present work is comparable with the findings in a much larger study published recently with a mortality rate of 48.8% (17). Our patient group can be considered, at least in the regional context, representative of older adults admitted to the ICU with sepsis.

Prognostic scoring systems guide clinicians to determine the severity of patients and also help them to establish a standardized approach in the management of critically ill patients. One of the essential features expected from a scoring system is its ability to discriminate survivors from non-survivors consistently. Also, it is expected to be easy to use by all healthcare professionals. The SOFA score was developed to evaluate organ dysfunctions in sepsis and validated in adult ICU patients (4,21). If the SOFA score is calculated periodically after admission to the ICU, it is called as “SOFA” score, and a 30% increase in the SOFA score during follow-up is associated with an increase in mortality (22). It is easy to calculate the SOFA score by using simple surrogate indicators related to major organ functions. Even though the patient’s age is not a criterion in the calculation of the SOFA score, older people tend to die earlier following sepsis-induced hospitalization, and survivors need longer-term rehabilitation after discharge (23). In addition, although the applicability of the SOFA score has been demonstrated in adults, its relevance in the elderly should be investigated (18,21).

The SOFA score calculated on admission to the ICU had a positive correlation with mortality, suggesting that it can be used effectively to estimate mortality rate (24). However, more the evidence is required for the older adults as only 11% of patients enrolled in that study were aged 65 and over. In our study, all patients were elderly with a mean age of 80.00±6.37 years, and the relationship between the SOFA scoring system and survival was significant. Our results suggest that the SOFA score that is determined on admission can be used to estimate the survival rate among older adult sepsis patients. In another study that included almost 400 elderly patients aged 80 years or older admitted to the ICU due to all causes, authors identified an association between higher SOFA scores and in-hospital mortality rate, along with age, higher SAPS II score, multiple trauma with a head injury, and requirement mechanical ventilation (25).

It has been shown that NLR is a more reliable inflammatory biomarker for predicting mortality than neutrophils or lymphocytes alone (26,27). In addition to its reliability, the calculation of NLR is simple, easy to obtain, and inexpensive. Regardless of different ethnicities, studies conducted with Belgian and South Korean adults to determine a reference value for healthy patients showed a similar result with the value of 1.65 (28,29). In our study, median NLR values were six times higher in the survivors [median=9.64 (11.25)] and nine times higher in the non-survivors [median=14.95 (35.53)]. It has been revealed that NLR can be used as an independent predictor of mortality in several clinical conditions, such as malignancies, fibrotic liver disease, and cardiovascular diseases (30-32). In a meta-analysis that published in 2020, authors evaluated 14 studies (n=11,564) focused on the predictive value of NLR for sepsis, and the results of this meta-analysis showed that NLR was significantly higher in non-survivors than in survivors (random-effects model: Standard mean differences=1.18, 95% confidence interval; 0.42-1.94) (20). Similarly, NLR values of non-survivors in our study are significantly higher than survivors [median (IQR)]=14.95 (35.53) vs. 9.64 (11.25), respectively.

Another widely used and validated predictor of survival is the APACHE-II scoring system. The APACHE-II score is calculated upon ICU admission using various variables such as age, history of severe organ failure or immunocompromise, current vital findings, and laboratory test results to calculate the APACHE-II score (5). Although it is recommended to use the worst values recorded in the first 24 hours in the ICU to calculate the initial score, the variables recorded on admission can be used for practical reasons. The rational use of APACHE-II score is to help determine the patient’s mortality risk. It is not calculated sequentially and does not have a utility to follow clinical improvement or response to interventions. APACHE-II scoring system could be used successfully among the adult population (10,33).

Baseline APACHE-II score was not associated with the risk of 28-day mortality in our study on exclusively older adults. In contrast, a recently published prospective study with a patient population similar to us concluded that the APACHE-II score predicting mortality in addition to the age of patients. However, the mean APACHE-II score was markedly lower than we report in the present study (22.6±7.0 vs 36.03±6.37) (17). Indeed, mean APACHE-II score in both the non-survivor and survivor subjects was marginally high, suggesting an estimated in-hospital mortality rate exceeding 85% (5,34,35). Likely, such extreme scores did not allow successful discrimination between the non-survivors and survivors in the present study.

According to the results of another study including 50 adult sepsis patients, while the SOFA score was successful in predicting mortality, the APACHE II score was not countable for predicting mortality rate (11). Similar to our study, this research was accomplished with limited participants and also the APACHE II score of the non-survivor group were higher than the survivor group, but the difference did not reach a statistical significance.

The mental state evaluated by GCS has an important place in the calculation of scoring systems. Sepsis-associated encephalopathy is a clinical reflection of neurological dysfunction in the host’s impaired response to infection. Although GCS has been reported in some studies as a good predictor for sepsis, it is still considered unreliable and uncertain (36). Mohammad (37) demonstrated that GCS score reliably predicts the outcomes of elder ICU patients. In our study, however, median GCS scores showed no difference between the non-survivor and survivor groups. Nevertheless, our study population consisted of only sepsis patients, and we did not include any ICU patients with a history of trauma or need for emergency surgery. Besides, patients included in our study are not just over 65 years old but also at advanced ages. Thus, the GCS score could have been influenced by age-related cognition disorders other than sepsis, weakening its predictive power in the case of severe illnesses. A previous study that focused on the importance of GCS in assessing the severity of brain injury among trauma patients demonstrated that older moderate brain trauma patients had a higher GCS score than younger ones (38). Thus, it couldn’t determine the disease severity in older adults, unlike younger patients (38). Besides, several other studies have also suggested that more research on GCS are needed to prove its utility in older adults (38-40).

Study Limitations

Our study has several limitations. First, as we did not have a control group without sepsis, it remains unclear as to whether our findings apply to only sepsis patients. Second, the number of participants in our study was limited, and we were not able to perform either subgroup analyses or adjusted analyses. Third, the originating site and responsible microbial source of sepsis are unclear due to the lack of blood culture results, which has a critical impact on survival.


The present study showed that a lower SOFA score and NLR is significantly associated with survival among elderly sepsis patients admitted to the ICU. Unlike the SOFA or NLR, the present study showed no significant differences in baseline APACHE-II score and GCS between survivors and non-survivors of sepsis. Future studies are warranted to confirm the current findings and to validate the use of scoring systems among patients at advance ages.


Ethics Committee Approval: The Health Sciences University Non-Invasive Research Ethics Committee approved the study protocol (code: 46418926-18/55).

Informed Consent: Written informed consent was obtained for each participant.

Peer-reviewed: Externally peer-reviewed.

Authorship Contributions

Surgical and Medical Practices: B.B.B., R.A., Concept: B.B.B., Design: B.B.B., Data Collection or Processing: B.B.B., R.A., İ.T., Analysis or Interpretation: B.B.B., İ.T., Literature Search: B.B.B., M.B.A., Writing: B.B.B., M.B.A.

Conflict of Interest: We declare that there are no conflicts of interest associated with this publication.

Financial Disclosure: We declare that we have not received any financial support to perform this study.

  1. Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Intensive Care Med. 2017;43:304-377.
  2. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med. 2001;29:1303-1310.
  3. Rudd KE, Johnson SC, Agesa KM, et al. Global, regional, and national sepsis incidence and mortality, 1990-2017: analysis for the Global Burden of Disease Study. Lancet. 2020;395:200-211.
  4. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22:707-710.
  5. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13:818-829.
  6. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. A practical scale. Lancet. 1974;2:81-84.
  7. Godinjak A, Iglica A, Rama A, et al. Predictive value of SAPS II and APACHE II scoring systems for patient outcome in a medical intensive care unit. Acta Med Acad. 2016;45:97-103.
  8. Nik A, Sheikh Andalibi MS, Ehsaei MR, et al. The Efficacy of Glasgow Coma Scale (GCS) Score and Acute Physiology and Chronic Health Evaluation (APACHE) II for Predicting Hospital Mortality of ICU Patients with Acute Traumatic Brain Injury. Bull Emerg Trauma. 2018;6:141-145.
  9. Raith EP, Udy AA, Bailey M, et al. Prognostic Accuracy of the SOFA Score, SIRS Criteria, and qSOFA Score for In-Hospital Mortality Among Adults With Suspected Infection Admitted to the Intensive Care Unit. JAMA. 2017;317:290-300.
  10. Naqvi IH, Mahmood K, Ziaullaha S, et al. Better prognostic marker in ICU - APACHE II, SOFA or SAP II! Pak J Med Sci. 2016;32:1146-1151.
  11. Desai S, Lakhani JD. Utility of SOFA and APACHE II score in sepsis in rural set up MICU. J Assoc Physicians India. 2013;61:608-611.
  12. Trancă S, Petrișor C, Hagău N, et al. Can APACHE II, SOFA, ISS, and RTS Severity Scores be used to Predict Septic Complications in Multiple Trauma Patients? J Crit Care Med (Targu Mures). 2016;2:124-130.
  13. Knox DB, Lanspa MJ, Pratt CM, et al. Glasgow Coma Scale score dominates the association between admission Sequential Organ Failure Assessment score and 30-day mortality in a mixed intensive care unit population. J Crit Care. 2014;29:780-785.
  14. Salciccioli JD, Marshall DC, Pimentel MA, et al. The association between the neutrophil-to-lymphocyte ratio and mortality in critical illness: an observational cohort study. Crit Care. 2015;19:13.
  15. Riché F, Gayat E, Barthélémy R, et al. Reversal of neutrophil-to-lymphocyte count ratio in early versus late death from septic shock. Crit Care. 2015;19:439.
  16. Chelluri L, Pinsky MR, Donahoe MP, et al. Long-term outcome of critically ill elderly patients requiring intensive care. JAMA. 1993;269:3119-3123.
  17. Martin-Loeches I, Guia MC, Vallecoccia MS, et al. Risk factors for mortality in elderly and very elderly critically ill patients with sepsis: a prospective, observational, multicenter cohort study. Ann Intensive Care. 2019;9:26.
  18. Minne L, Ludikhuize J, de Jonge E, et al. Prognostic models for predicting mortality in elderly ICU patients: a systematic review. Intensive Care Med. 2011;37:1258-1268.
  19. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315:801-810.
  20. Huang Z, Fu Z, Huang W, et al. Prognostic value of neutrophil-to-lymphocyte ratio in sepsis: A meta-analysis. Am J Emerg Med. 2020;38:641-647.
  21. Vincent JL, de Mendonça A, Cantraine F, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsis-related problems” of the European Society of Intensive Care Medicine. Crit Care Med. 1998;26:1793-1800.
  22. Ferreira FL, Bota DP, Bross A, et al. Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA. 2001;286:1754-1758.
  23. Martin GS, Mannino DM, Moss M. The effect of age on the development and outcome of adult sepsis. Crit Care Med. 2006;34:15-21.
  24. Gupta V, Karnik ND, Agrawal D. SOFA Score and Critically Ill Elderly Patients. J Assoc Physicians India. 2017;65:47-50.
  25. Andersen FH, Flaatten H, Klepstad P, et al. Long-term survival and quality of life after intensive care for patients 80 years of age or older. Ann Intensive Care. 2015;5:53
  26. Kumarasamy C, Sabarimurugan S, Madurantakam RM, et al. Prognostic significance of blood inflammatory biomarkers NLR, PLR, and LMR in cancer-A protocol for systematic review and meta-analysis. Medicine (Baltimore). 2019;98:e14834.
  27. Menges T, Engel J, Welters I, et al. Changes in blood lymphocyte populations after multiple trauma: association with posttraumatic complications. Crit Care Med. 1999;27:733-740.
  28. Forget P, Khalifa C, Defour JP, et al. What is the normal value of the neutrophil-to-lymphocyte ratio? BMC Res Notes. 2017;10:12.
  29. Lee JS, Kim NY, Na SH, et al. Reference values of neutrophil-lymphocyte ratio, lymphocyte-monocyte ratio, platelet-lymphocyte ratio, and mean platelet volume in healthy adults in South Korea. Medicine (Baltimore). 2018;97:e11138.
  30. Zhou M, Li L, Wang X, et al. Neutrophil-to-Lymphocyte Ratio and Platelet Count Predict Long-Term Outcome of Stage IIIC Epithelial Ovarian Cancer. Cell Physiol Biochem. 2018;46:178-186.
  31. Afari ME, Bhat T. Neutrophil to lymphocyte ratio (NLR) and cardiovascular diseases: an update. Expert Rev Cardiovasc Ther. 2016;14:573-577.
  32. Peng Y, Li Y, He Y, et al. The role of neutrophil to lymphocyte ratio for the assessment of liver fibrosis and cirrhosis: a systematic review. Expert Rev Gastroenterol Hepatol. 2018;12:503-513.
  33. Badrinath K, Shekhar M, Sreelakshmi M, et al. Comparison of Various Severity Assessment Scoring Systems in Patients with Sepsis in a Tertiary Care Teaching Hospital. Indian J Crit Care Med. 2018;22:842-845.
  34. Headley J, Theriault R, Smith TL. Independent validation of APACHE II severity of illness score for predicting mortality in patients with breast cancer admitted to the intensive care unit. Cancer. 1992;70:497-503.
  35. Capuzzo M, Valpondi V, Sgarbi A, et al. Validation of severity scoring systems SAPS II and APACHE II in a single-center population. Intensive Care Med. 2000;26:1779-1785.
  36. Alalawi MSM, Aljabran HAM, Alkhamri AM, et al. Glasgow Coma Scale in Anticipation of Sepsis and Septic Shock: Review Article. Egypt J Hosp Med. 2017;69:2663-2666.
  37. Mohammad RJ. Comparison of Glasgow Coma Scale and GCS-Age Prognosis Score in Older Adult Patients. 2019;10:35-40.
  38. Rau CS, Wu SC, Chen YC, et al. Effect of Age on Glasgow Coma Scale in Patients with Moderate and Severe Traumatic Brain Injury: An Approach with Propensity Score-Matched Population. Int J Environ Res Public Health. 2017;14:1378.
  39. Bledsoe BE, Casey MJ, Feldman J, et al. Glasgow Coma Scale Scoring is Often Inaccurate. Prehosp Disaster Med. 2015;30:46-53.
  40. Kehoe A, Rennie S, Smith JE. Glasgow Coma Scale is unreliable for the prediction of severe head injury in elderly trauma patients. Emerg Med J. 2015;32:613 LP -615.