A comparative study of PSI and Curb-65 scoring systems in predicting ICU admissions and mortality in cases of community-acquired pneumonia

Background :Community-acquired pneumonia (CAP) is a leading cause of morbidity and mortality worldwide. Despite advanced diagnostic modalities and treatment options, CAP is the fourth leading cause of death in developing countries.Several severity scores have been proposed to guide initial decision making on hospitalization and to predict the outcome. Pneumonia Severity Index (PSI) and CURB 65 are the two most widely used scoring systems to prognosticate pneumonia. Aim : To compare the efficacy of PSI and CURB 65 scoring systems inprognosticating the ICU admission and outcome in cases of CAP. Methodology: This wasan observational study conducted at a tertiary care hospital in westernMaharashtra.A hundred patients of CAP fulfilling the inclusion criteria were enrolled in the study, classified as per CURB 65 and PSI system and their outcome compared . Result: The study subjects comprised of 100 patients (64 men and 36 women) of CAP. Twenty-four patients needed ICU admission.In both PSI and CURB-65 risk scoring systems, the need for intensive care unit (ICU) admission and mortality rates increased progressively with increasing scores.PSI class ≥ IV and CURB 65 ≥ III had 77.52% and 40.24% sensitivity and 88.46% and 69.48% specificity respectively in predicting ICU admissions. The PSI class ≥ IV had more sensitivity and specificity in predicting ICU admission than CURB-65.CURB 65 class III and IV had sensitivity86.59% and 89.64% and specificity 89.64% and 97.54% respectivelyin predicting mortality, while PSI class IV and Vhadsensitivity68.92% and 72.58% and specificity 24.74% and 54.86% respectively. CURB 65 had more sensitivity and specificity than PSI in predicting mortality. Conclusion: The PSI is better in predicting the need for ICU admission and CURB 65 is a better predictor of mortality in cases of community-acquired pneumonia.


Introduction
Community-acquired pneumonia (CAP) is a leading cause of morbidity and mortality worldwide.It is the third leading cause of death in the world and the fourth leading cause of death in developing countries [1].CAP is one of the most common infectious diseases needing hospitalization.The reported incidence rate of CAP in India is 4 million cases per year [2].India accounts for 23% of the global pneumonia burden and 36% of the WHO regional burden [2,3].The clinical presentation of CAP is variable and because of the wide spectrum of associated clinical features, CAP is a part of the differential diagnosis of nearly all respiratory illnesses [3].The presentation of CAP may range from mild pneumonia characterized only by fever and productive cough to severe fulminant pneumonia leading to respiratory distress and sepsis syndrome requiring management in ICU.Any delay in ICU admission has been shown to be associated with increased mortality [1,2,3].Also,unnecessary admission to ICU increases the treatment cost and leads to depletion of precious hospital resources.It is therefore important for physicians to identify patients at low risk of complications who are suitable for outpatient management [4].Multiple serum biomarkers and several established risk scores such as CURB 65, CRB 65, Pneumonia Severity Index (PSI), Infectious Diseases Society of America/American Thoracic Society (IDSA/ATS) score and Extended CURB 65 have been used to assess the severity of CAP to optimize the management of CAP patients [5].Out of these scoring systems, PSI and CURB-65 have been most extensively used.[7].Each risk factor scores one point, for a maximum score of 5 (Table 2).CURB-65 is easier to remember and simpler to calculate in clinical practice.CURB 65 and PSIscoring systems have not been validated in developing countries where population demographics and health-care delivery systems are totally different from the developed world [8].To date, only a few studies have been conducted in India comparing various prognostic scores.This study was conducted to compare PSI and CURB-65 scores in an Indian context and compared the two scoring systems.

Aim
The present observational study was aimed to compare the sensitivity and specificity of CURB-65 and PSI scoring systemsin predicting ICU admission and mortality in cases of CAP.

Material and Methods
Thisobservational study was conducted in a tertiary care teaching hospital in western Maharashtra from August 2018 for a period of six months after due approval from the institutional ethical committee and scientific committee.One hundred consecutiveconsentingCAP patients admitted to the hospital were included in the study.
Inclusion criteria: Allconsenting patients of ≥18 years of age reporting with clinical features and laboratory/radiological evidence of pneumonia, were included in the study.
Exclusion criteria: All cases where clinical and/or radiological findings could be explained by any respiratory pathology other than CAP were excluded from the study.Immunosuppressed patients (HIV patients, solid organ transplant, post-splenectomy on steroids or chemotherapy)and those withhospital-acquired or healthcare-associated pneumonia were also excluded from the study.

Methodology
A detailed preapproved proforma was filled up for each patient including history, clinical findings,routine blood investigations, chest x-ray, sputum for Gram stain, culture and antibiotic sensitivity pattern, sputum AFB, bloodculture and other investigations on as required basis.CURB-65 and PSI scores were calculated for each patient and they were categorized as per score at the time of admission.All patients were reassessed daily clinically and radiographically for improvement or development of complications.Treatment of the patients including the decision for ICU admission, mechanical ventilation, and inotropic/vasopressor support was by the treating physician who was blinded to the prognostic score of the patient.Observations-The present study included one hundred cases of CAP;the mean age of patients was 54.33±16.87years, ranging from 18 to 90 in males and 18 to 82 in females.Twenty-nine percent of patients were aged more than 65 years.Males (64%) were affected more than females (36%) almost in the ratio of 2:1.Twenty-five patients had more than one co-morbid condition.Eighteen patients expired during the hospital stay.

Analysis of scoring systems
ICU Admissions:Inthe present study, 24 patients out of 100 cases with CAP required ICU admission.There was a statistically significant difference (P < 0.005) when the sensitivity and specificity of scoring systems were compared.CURB 65 has higher sensitivity and specificity than PSI in predicting mortality in CAP.PPVfor mortality was higher for CURB 65 class III and IV (79.56% and 90.60%) as compared to PSI class IV and V (12.48%, 58.24%).NPVfor mortality was higher for CURB 65 III and IV (92.56% and 95.24%) as compared to PSI class IV and V (58.62% and 74.86%).Both PPV and NPV werea better predictor of mortality for CURB 65 class ≥III than PSI class ≥IV.

Discussion
In cases of CAP,the majority of admissions to the ICU occur within the first 24 h of illness.Delayed transfer to the ICU is associated with increased mortality, and therefore early recognition of these patients is important.The mean age in the present study was 54.33±16.87(38 to 71).Mean age in studiesby Babu et al and Dey et al was 53(±17) years and 50.6 years respectively [10,11].In the present study, the incidence of pneumonia increased with age; this finding was consistent with the study by Mohanty S et al [9].This is because the PSI relies more on age and comorbidities, and therefore, the young patients without any comorbidities may be placed in a lower PSI class and may not receive the care they actually required.In contrast, the CURB-65 approach may be ideal for identifying high mortality risk patients with severe illness due to CAP.However, one clear shortfall of the CURB-65 approach is that it does not account for comorbid illness, and thus may not be realistic in older patients who may have considerable mortality risk even with low CURB-65 score.
Even a mild form of CAP may destabilize a chronic, but compensated disease process.Both the prognostic systems offer a valuable assessment of patient illness, but from different perspectives, and are therefore complementary to each other at identifying low risk and high-risk CAP patients [6].

Conclusion
In this study, an attempt has been made to compare the prognostic capability of the two commonly used scoring systems for assessing severity in cases of CAP.The study has its limitation in that the patients with comorbidities like malignancy, chemotherapy, and steroid therapy were excluded from the study and therefore PSI score may not have been a true reflection of the severity of CAP cases in the community.
What does the study add to the existing knowledge?
In the current study, PSI was found to be better in the predictor of the need for ICU admission; the CURB 65 was found to be a better predictor of mortality.By applying the knowledge of these two scoring systems, patients of CAP can be better prognosticated regarding the severity of their illness and the need for intensive monitoring.More realistic triaging of patients will ultimately result in providing appropriate timely interventionforoptimalutilization of hospital resources and favorable outcomes.

Table 1 )
[6].CURB-65 was developed by Lim et al at the University of Nottingham, the UK in 2002for predicting mortality in cases of CAP.The score is an acronym for each of the risk factors measured:new-onsetConfusion, Blood Urea nitrogen greater than 7 mmol/l (19 mg/dl), Respiratory rate of ≥ 30 breaths per minute, Blood pressure ≤ 90 mmHg systolic or diastolic blood pressure ≤60 mmHg and age 65 or older Statistical Analysis-A χ2 statistics test (with Yates correction when applicable) was used to evaluate the statistical significance of categorical variables.The results were presented as mean (SD).Odds ratios with 95% confidence intervals were computed using a univariate logistic regression model with ICU outcome as the dependent variable.Using EpiInfo1.4.3 all statistical tests were 2-tailed and a P-value <0.05 was considered significant.
International Journal of Medical Research and ReviewAvailable online at: www.medresearch.in243 | P a g e

Table - 4: Sensitivity, specificity, NPV, and PPV for different CURB 65 classes in predicting ICU admissions.
Table 3shows patients of various PSI classes who required ICU admission and the majority of them belonged to class IV and V. PSI Class IV hadnine (26.47%) admissions with sensitivity and specificity of 77.52% and 88.46% respectively.As the severity of CAP increased, the number of ICU admissions increased; with PSI class Vshowingthe highest number of admissions, ten out of 14 patients (71.42%) with sensitivity and specificity of 84.16% and 91.58% respectively.CURB 65 III and IV had a sensitivity of 40.84% and 38.84% and specificity 69.48%and 72.08% respectively for ICU admissions (Table 4).The PSI class ≥IV was found to have higher sensitivity and specificity in predicting ICU admission than CURB-65 class ≥III.Positive Predictive Value (PPV) for ICU admission was highest for PSI class IV and V (67.24% and 69.40%) as compared to CURB 65 class III and IV (48.42% and 60.52).The Negative Predictive Value (NPV) for ICU admission in PSI class IV and V (77.61% and 82.24%) was higheras compared to CURB 65 III and IV (68.17% and 78.26%).Both PPV and NPV werefound better for PSI class ≥IV than CURB 65 class ≥III.Table-3: Sensitivity, specificity, NPV, and PPV for different PSI classes in predicting ICU admissions.International Journal of Medical Research and ReviewAvailable online at: www.medresearch.in244 | P a g e

Table - 6: Sensitivity, Specificity, NPV, and PPV for different PSI classes in predicting mortality.
[5,9,11]admission was 24% almost similar to the current study[12].In the present study, the overall mortality was 18%.Ina study by Mohanty S et al, Dey et al and Shah et almortality was 13.28%, 25.38%, and 10.7% respectively[5,9,11].The mortality increased as the PSI and CURB 65 severity increased.This finding is similar to that found by International Journal of Medical Research and ReviewAvailable online at: www.medresearch.in245 | P a g e