Performance of CURB-65 in predicting mortality of patients with community- acquired pneumonia in Saudi Arabia

Introduction: Various objective scoring systems were developed to standardize the approach to the designation of severity of communityacquired pneumonia (CAP). There is limited data on the use of CURB-65 among admitted CAP patients in Saudi Arabia. Methodology: The retrospective study included CAP patients, admitted to a general hospital in Eastern Saudi Arabia. The CURB-65 was extracted from the available medical records. Results: During the study period, from 2013 to 2016, a total of 1786 adults were admitted with a mean age of 63.9 ± 21.7 (range 14-108 years). The majority of the patients (51.7%) had CURB-65 score 0 or 1 followed by the score 2, 3 and 4/5 (29%, 15.2%, and 4.1%, respectively). The mean CURB-65 was 1.4 ± 1.12 for those who survived and 2.27 ± 1.03 for those who died (p < 0.001). The mean age was 63.01± 21.9 years for survived patients and 75.1 ± 15.58 years for fatal cases (p < 0.001). The overall 30-day crude mortality rate was 7.6%. The mortality rates for CURB-65 scores 0, 1, 2, 3, and 4/5 were 1.8%, 4.3%, 10.2%, 14%, and 21.9%, respectively. Conclusions: The mortality rates of admitted patients with CAP did not differ from those reported in the literature. However, the utilization of CURB-65 score was low and there is a need for wider implementation of pneumonia severity index for patients presenting with CAP.


Introduction
Community-acquired pneumonia (CAP) is a common diagnosis leading to admissions throughout the world and remains a cause of substantial morbidity and mortality worldwide.CAP is also important as it may lead to respiratory failure and may result in death.Various objective scoring systems were developed to standardize the approach to designation of the severity of CAP.These criteria help in the calculation of the expected mortality rate and thus inform the clinician on the need for hospital admission.The recent CAP guidelines state that major decisions regarding diagnostic and treatment issues of CAP spin around initial assessment of severity using a scoring system [1].There are four proposed pneumonia severity scores: the Pneumonia Severity Index (PSI) [2], the Infectious Diseases Society of America (IDSA)/American Thoracic Society (ATS) severe pneumonia criteria [1], CURB-65 [3] and the Japanese Respiratory Society (JRS) scores [4].The JRS scores relies on physiological and radiological criteria (radiological extent, temperature, pulse rate, respiratory rate and dehydration) and laboratory data (WBC count, Creactive protein value and PaO2 or SpO2 value) [4].Another scoring system is the expanded-CURB-65 (CURB-65, lactate dehydrogenase, platelet, and albumin) [5].The CURB-65 score components include scoring patient based on confusion, blood urea, respiratory rate, blood pressure, and age ≥ 65 years [3,6].The score had been associated with mechanical ventilation, rate of hospital admission, and duration of hospital stay among hospitalized patients [7].Based on local guidelines, CURB-65 should be used for hospitalized patients and that hospitalization is recommended for patients with CURB-65 score of ≥ 2 [8,9].However, there is a limited data on the use of CURB-65 among CAP patients in Saudi Arabia [10].Thus, in this study we analyze the pattern of the CURB-65 severity score for CAP in patients hospitalized in one medical center in Saudi Arabia.We also tried to elucidate the admission pattern and mortality among admitted patients.

Methodology
The medical record of patients admitted with CAP were obtained from the health information unit from 2013 to 2016.The patients' data were collected using a standard Microsoft Excel data collection sheet and the data were obtained from paper charts and electronic medical records.
The Excel sheet contained information regarding the CURB-65 and the mortality within 30 days of the admission.CURB-65 score was calculated as having a score of one for the presence of each one of the following items at the time of admission: confusion, Blood urea ≥ 19 mg/dL, respiratory rate of ≥ 30/minute, a systolic blood pressure (BP) <90 mmHg or diastolic BP ≤ 60 mmHg, age ≥ 65 years, as described previously [3,6].The study was approved by the Institutional Review Board (IRB).

Statistical Analysis
analysis was done using Excel.Descriptive analyses were done for demographic, results of the tests and the monthly number of cases.Minitab (Minitab Inc.Version 17, PA 16801, USA; 2017) was used to calculate the mean age (± SD) of patients and the comparison between those who died and those who survived.One-way ANOVA was used for the Age versus CURB-65 score comparison, and Chi-Square test for association between CURB-65 and death.A significant p value was considered for p < 0.05.

Discussion
This study describes the CURB-65 score among 1786 admitted adults with CAP in a Saudi Arabian hospital.The majority (51.7%) of the patients had CURB-65 score of 0 or 1.The overall mortality rate was 7.1% and was similar to a recent study describing the mortality rate of 6.7% among 1834 CAP patients [11].
The specific mortality rates for CURB-65 scores of 0, 1, 2, 3 and for 4/5 were 1.4%, 4.1%, 9.8%, 13.2%, and 20.5%, respectively.Thus, the CURB-65 score and calculated mortality rates mirror those described  previously [3,6].However, almost half of the admitted patients had a score of 0 or 1.Those patients are recommended to be treated as outpatient as the associated mortality rate is low [3].Patients with a score of 2 needs regular ward admission and patients with a score of 3-5 would require intensive care unit admission [3].There might be other reasons for the admission of patients with low CURB-65 score in the current study.However, these reasons were not specifically sought but could be related to the routine screening for Middle East Respiratory Syndrome Coronavirus (MERS-CoV).Our hospital was of the first hospitals in the region to adopt a standardized screening for MERS-CoV [12][13][14].Such screening may had then resulted in routine admissions of those patients.The Saudi Arabian Ministry of Health guidelines allow home isolation of patients suspected to have mild MERS-CoV infection [15][16][17].However, the logistics of home isolation and the fear of spread of MERS-CoV influence decisions regarding the admission of such patients [12,18].Previously published studies did not show differentiating factors among patients with MERS and those without MERS [12,18].Thus, the lack of predictors of MERS on presentation makes this distinction difficult to achieve.The current study did not evaluate other factors influencing admissions such as ability to safely and reliably take oral medication and the availability of outpatient support resource as suggested by recent guidelines [1].Routine use of CURB-65 score is advised, however, the actual practice in this part of the World is not well documented.In one study from Oman, CURB-65 severity score was documented for only 2.3% of hospitalized patients [19].
In a study from Nigeria, none of 249 CAP patients had CURB-65 score documentation in hospital notes [20].Thus, there is a need to have more education with audit and feedback to utilize CAP severity scores in order to make informed decisions about the need for admission.Pneumonia severity index (PSI) was thought to be superior to the British Thoracic Society's CURB-65 and the modified American Thoracic Society criteria in predicting CAP severity [21].Nevertheless, CURB-65 score remains an easy score to obtain with excellent prediction ability.The CORB score (acute Confusion, Oxygen saturation ≤90%, Respiratory rate > 30/minute, and Systolic Blood pressure < 90 mm Hg or a diastolic blood pressure < 60 mm Hg was proposed for elderly patients.In one study, the CORB score was a useful tool for hospitalized elderly patients [22].

Conclusion
The mortality rates of admitted patients with CAP did not differ from those in the medical literature.However, the utilization of CURB-65 score seems to be low and there is a need for wider implementation of pneumonia severity scores for patients presenting with CAP in our hospital.There is a need for further prospective studies to elucidate the features and characteristics of patients with low CURB-65 scores needing admission.This approach would then enhance the optimal utilization of services and proper placement of patients.Further studies should also be directed towards comparing low and high CURB-65 score patients in relation to length of stay and antibiotic utilization.

Table 1 .
Percentage of different CURB-65 score and the mortality rate in relation to CURB-65 Score.

Table 2 .
CURB 65 Score and Mean Age with Standard Deviation (SD) and 95% Confidence Interval (95% CI) in relation to outcome.