CC BY
3
УДК 617.5
Miteva D.N., PhD, chief assistant chief assistant Department of Pulmonology and Allergology
Medical University Bulgaria, Varna Dimitrow Y.R., PhD, assoc. professor assoc. professor Department of Pulmonology and Allergology
Medical University Bulgaria, Varna
ROLE OF CHARLSON COMORBIDITY INDEX IN PREDICTING
SEVERITY AND PROGNOSIS OF COMMUNITY-ACQUIRED
PNEUMONIA
Abstract: Many patients with community acquired pneumonia (CAP) are polymorbid and this requires a comprehensive evaluation of comorbid illnesses to determine the risk of unfavorable outcome of pneumonia. Aim: To evaluate the complex impact of comorbidity, assessed by Charlson Comorbidity Index (СС1), on the severity and prognosis of CAP. Materials and Methods: 1292 patients hospitalized in the Clinic of Pneumonology and Phthisiatrics of MHAT "Saint Marina" - Varna were retrospectively studied for the period from 2012 to 2015. For each patient СС1 was calculated and patients were divided into three levels of comorbidity: low (СС1=0); intermediate (СС1=1-2) and high (СС1>3). The assessment of the severity of CAP was made according to PSI, CURB-65 and IDSA/ATS criteria.
Results: Patients were at average age of59,94±17,03, 57,2% of them were male. The mean СС1 was 1,58 ± 1,84. It was significantly higher in non-survivors versus survivors (3,28 vs. 1,36; р< 0,001). For patients treated in the intensive care unit, CCI was also higher compared to patients in the general unit (2,65 ± 2,09 vs. 1,26 ± 1,63, р<0,001). Patients with СС1>3 more often developed complications, they needed intensive care more frequently and had higher inhospital mortality. CCI>3 had OR for in-hospital mortality 6,82 (4,74-9,80, p< 0,001). PSI and IDSA/ATS criteria were the best predictors for the need of intensive care and in-hospital mortality followed by CURB-65 and CCI. Conclusion: СС1 is a good predictor of the development of complicated pneumonia, need of intensive care and in-hospital mortality.
Keywords: Community acquired pneumonia, comorbidity, Charlson Comorbidity Index, prognosis
Introduction: Community acquired pneumonia (CAP) is a common disease with high hospitalization rate and still high mortality. To assess its severity and prognosis, different scales are used, most often these are PSI, CURB-65 and IDSA/ATS criteria [4, 1, 6]. The impact of comorbid illnessess on the course and outcome of CAP often remains underestimated. Many patients are polymorbid and this requires a comprehensive evaluation of comorbid illnesses to determine the risk of unfavorable outcome of pneumonia. Charlson Comorbidity Index (CCI) is
the most widely used comorbidity index to assess the cumulative burden of comorbidities on the outcome of a particular disease. The index was published in 1987 by M. Charlson et al. to assess comorbidities that could change the mortality risk for the purpose of longitudinal studies [3]. The index is presented in table 1.
Table 1. Charlson Comorbidity Index (CCI)
1 point > Myocardial infarction > Congestive heart failure > Peripheral vascular disease > Cerebrovascular disease > dementia > chronic pulmonary disease > connective tissue disease > peptic ulcer disease > diabetes without end-organ damage > mild liver disease
2 points > diabetes with end-organ damage > moderate or severe renal disease > hemiplegia > leukemia > malignant lymphoma > solid tumor without metastasis
3 points > moderate or severe liver disease
6 points > metastatic solid tumor > AIDS
The total index of each patient is calculated by summing the points of all diseases the patient has. Its use to assess the risk of short-term mortality in acute diseases is yet to be approbated.
Material and Methods: 1292 patients hospitalized in the Clinic of Pneumonology and Phthisiatrics of MHAT "Saint Marina" -Varna were retrospectively studied from 2012 to 2015. For each patient CCI was calculated and patients were divided into three levels of comorbidity: low (CCI=0); intermediate (CCI=1-2) and high (CCI>3). The assessment of the severity of CAP was made according to PSI, CURB-65 and IDSA/ATS criteria. The results were analyzed with SPSSv20.
Results: The patients were at average age of 59,94±17,03, 57,2% of them were male and 42,8% - female. 987 (76,40%) of them had comorbidities. 267 patients (20,70 %) were treated in intensive care unit (ICU). The total number of deceased patients was 148 (11,50%). 39,8% of the patients fell into the low level of comorbidity; 35,4% - in intermediate level and 24,8% - in high level. The mean value of CCI for the entire sample was 1,58±1,84, as values varied from 0 to 10.
A moderate correlation between the level of comorbidity assessed by CCI and the hospitalization in ICU was established (r=0,3 p<0,001), as the higher the level of comorbidity was, the greater was the proportion of patients admitted to ICU (fig.1). The average CCI for patients in ICU was significantly higher compared to that of the patients admitted in the general unit (2,65±2,09 vs. 1,26±1,63, p < 0,001).
Fig. 1. Distribution of patients admitted to ICU according to CCI
comorbidity level
It was also proved that the level of comorbidity was associated with development of complications (p<0,001), as the higher the comorbidity level was, the greater was the proportion of patients with complicated CAP. Complications included acute respiratory failure, pleural effusion, cavitations, need for mechanical ventilation, septic shock (fig. 2).
80,00% 60,00% 40,00°% 20,00% 0,00%
74,20%
55,20%
43,00%
CCI=0
CCI=1-2
CCI>3
Fig. 2. Proportion of patients with complicated CAP according to CCI
The average CCI for non-survivors was significantly higher than for survivors (3,28 vs. 1,36; p<0,001). In the low level of comorbidity the in-hospital mortality was only 3,8% and in the high level it reached up to 28,9% (p<0,001) (fig. 3). We established also a positive moderate correlation between the outcome of the disease and CCI (r = 0,3; p<0,001). CCI>3 was able to predict the risk of inhospital mortality with OR 6,82 (4,74-9,80, p<0,001).
40,00% 30,00% 20,00% 10,00% 0,00%
CCI=0
CCI=1-2
CCI>3
Fig. 3. Proportion of deceased patients according to CCI comorbidity
level
We compared the predictive value of CCI for the need of intensive care and in-hospital mortality with the predictive values of the main scales. For the purpose we created ROC-curves, including PSI, CURB-65, IDSA/ATS criteria and CCI and compared the area under the curves (AUC). PSI and IDSA/ATS criteria had the highest predictive value for the need of intensive care, followed by CURB-65 and CCI. AUC for the four scales was 0,800; 0,799; 0,787; 0,683 respectively. Although the predictive value of CCI was lower than that of the other scales, its AUC was also sufficiently high and made it a good predictor for the need of intensive care (fig.4a)
+ 1 - SpteiFlelw
Fig. 4a and b. Predictive value of severity scales and CCI for the need of intensive care fa) and in-hospital mortality (b)
CCI had also a good predictive value for in-hospital mortality. Again, by creating a ROC- curve, its predictive value was compared with the one in the main scales. The best predictors of the in-hospital mortality were PSI and IDSA/ATS criteria, followed by CURB-65 and CCI. The AUC for each one of the scales was respectively 0,859; 0,852; 0,849 and 0,769. The AUC of CCI was quite close to the one of the main scales, making its predictive value for in-hospital mortality similar to that of the main scales ( fig. 4b).
Discussion: Most authors also emphasize the role of CCI as a predictor of severe CAP with poor prognosis. A number of authors have established significantly higher CCI in deceased patients compared to survivors what indeed our results were [2, 5, 7]. Pereira et al., for example, established mean CCI for deceased patients - 5, which was significantly higher than that of the survivors -3; p<0,001)[7]. Our study found significant increase in mortality in the high level of comorbidity according to CCI, where it reached 28,9%. Thomsen [8] has also established such increase in mortality in patients with high CCI. According to his study, in low-level comorbidity 30-day mortality was 10,5%, in the intermediate level - it was 15,8%, and in higher level it increased up to 21%. According to our study, the predictive value of CCI for the need of intensive care and for mortality corresponds to that of the PSI, CURB-65 and IDSA/ATS criteria, although it was slightly lower than theirs. Cabrera et al. also found that the three scales CURB-65, PSI and CCI are good predictors of 30-day mortality (The AUC for each one was
0.735. 0,770; 0,765 respectively), and for transferring to the emergency care unit (the AUC were 0,603; 0,652; 0,647 respectively)[2]. In our study the AUC for both events were even higher, making the scales, including CCI, a very good predictors of the need for intensive care and mortality. Assessing the risk of in-hospital-mortality, CCI>3 also showed a high predictive value with OR 6,82 (4,74-9,80, p<0,001).
Our study established that patients with high level of comorbidity are at higher risk for developing complicated pneumonia and dying in the hospital. They need increased attention, more active monitoring and more aggressive therapeutic measures.
Conclusion: CCI is a good predictor of the development of complicated pneumonia, the need of intensive care and in-hospital mortality. The level of comorbidity should be included in the formation of the risk profile of patients with CAP.
References:
1. British Thoracic Society Standards of Care Committee BTS Guidelines for the Management of Community Acquired Pneumonia in Adults. Thorax 2001;56. Suppl 4: IV 1-64.
2. Cabrera R. , Shakeel Q. , Uduman A., et al. Evaluation of Charlson comorbidity index as a predictor of adverse outcomes in patients admitted with community-acquired pneumonia. J Hosp Med 2013;8,Suppl. 1: 655
3. Charlson M.E., Pompei P., Ales K.L. et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chron Dis 1987 Vol. 40, Issue 5, P. 373-383
4. Fine MJ, Auble TE, Yealy DM, Hanusa BH, et al. A prediction rule to identify low-risk patients with community- acquired pneumonia. N Engl J Med 1997; 336:243-50.
5. Hon Ming Ma, Wing Han Tang , Jean Woo Predictors of in-hospital mortality of older patients admitted for community-acquired pneumonia. Oxford Journals Medicine Age and Ageing 2011 Vol. 40 Issue 6: 736-741.
6. Mandell LA, Wunderink RG, Anzueto A, et al Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007;44( 2):27-72.
7. Pereira JM, Paiva JA, Froes F et. Al. Outcome of severe community-acquired pneumonia: the impact of comorbidities. Critical Care 2013,17 (2):P41
8. Thomsen RW, Riis A, N0rgaard M, et al. Rising incidence and persistently high mortality of hospitalized pneumonia: a 10-year population-based study in Denmark. J Intern Med. 2006;259 (4):410-7.
