Many studies have demonstrated that increased age is associated with pneumonia-induced mortality 11 12 14 15. In this prospective observational study, which was the subject of pneumonia in the elderly, the patients with pneumonia were stratified into younger (18 to 64 years) and older (≥65 years) groups. The poor prognostic indicators were determined and compared in both age groups. In our study, we determined three variables that could be potential independent risk factors for poor prognosis in the elderly with pneumonia: previous antibiotic use within the last 3 months (OR:2.97, CI:2.51–3.43, p = 0.02), acute renal failure (OR:2.51, CI:2.06–2.96, p = 0.04) and dyspnea (OR:5.85, CI:5.18–6.52, p = 0.01). Also, we found that serum procalcitonin (p<0.001) and CRP levels (p = 0.008) were valuable indicators of poor prognosis in the elderly. In addition, mental status changes, the CURB–65 score, and the pneumonia severity index (PSI) as well as the independent risk factors were associated with poor prognosis of those that were 65 years and older.
Antibiotic exposure is one of the main reasons for increased pneumonia cases with resistant microorganisms, and leads to a lack of response to empirical antimicrobial therapies. There are several studies showing that previous antibiotic use is a risk factor for infection with drug-resistant Streptococcus pneumoniae 16 17 18. In our study, the rate of previous antibiotic use within the last 3 months was quite high in the elderly (n = 71, 55.9%) and in the younger group (n = 30, 52.6%). Also, previous antibiotic use within the last 3 months was an independent risk factor for poor prognosis in both age groups.
In our study, the rate of acute renal failure was 52.6% in the elderly with poor prognosis. Acute renal failure was also an independent risk factor for poor prognosis. This finding was consistent with other studies 19 20. In the study of Murugan et al., acute renal failure was associated increased mortality risk, and also, an increased severity of acute renal failure was correlated with the increased mortality rates 20.
In this study, dyspnea was found to be an independent risk factor for poor prognosis. The diagnosis of pneumonia in the elderly is delayed due to the fact that the signs and symptoms are infrequent 1 21 22. Although dyspnea was seen as less frequent in the elderly, it is vital for the prognostic evaluation. However, due to the weakness of the compensation mechanisms, multiple organ failure and mental status changes develop more easily in the elderly 1. In our study, mental status changes were found to be more frequent in the elderly (p = 0.035). This finding was consistent with other studies 22 23 24. That is why mental status changes should be considered one of the most important findings in the early diagnosis of pneumonia in the elderly. Also, an alteration in mental status may be the first clue in the diagnosis of pneumonia in this group.
In our study, fever and hypothermia were less frequent in the elderly group than in the younger group, but no statistically significant difference was found between the two groups. In addition, when the fever and hypothermia were considered together as body temperature changes in the elderly group, there was no statistically significant relationship with poor prognosis (p = 0.157). Also, in the univariate and multivariate regression analysis, there was no relationship between hypothermia and poor prognosis in the elderly group (p = 0.19). Some studies have shown that fever and hypothermia contribute to the diagnostic and prognostic evaluation of pneumonia 25 26 27. It is known that fever development is less frequent, especially in the elderly population due to the reduced host immune response. In this reduced response, the decrease in the production of endogenous pyrogens such as interleukin–1, interleukin–6 and tumor necrosis factor and the reduced response to these pyrogens has been thought to play a role 28 29 30 31 32 33.In addition, hypothalamic changes occurring with aging, and changes in thermogenic brown fat tissue may also play a role in decreased fever response to infections observed in the elderly 28 29 30.
In our study, sputum culture positivity was 40.4% in the elderly group and 25.8% in the younger group. In the study of Saltoglu et al., microbiological evidence was obtained in 44% of the cases 34. In Gutierrez’s study, the rate was quite high (50.7%) 35. On the contrary, the microbiological evidence was obtained in 20.1% of the cases in our study. The rate of obtaining the sputum sample in elderly patients was significantly lower than in the younger group (p = 0.037). The low rates in our study can be explained for a number of reasons such as antibiotic use before inpatient treatment, and sputum production and collection problems in the elderly.
Although the results of microbiological examinations are generally not obtained during diagnosis and empirical treatment, but these results are very important for the reassessment of the initial treatment. Sputum Gram stain and culture and other microbiological examinations including polymerase chain reaction assay in respiratory samples are crucial to providing the most appropriate empirical treatment. Owing to these microbiological evaluations, antimicrobials can properly be tailored. Furthermore, collateral damage including antimicrobial resistance and even mortality can be reduced 1.
In our study, Pseudomonas spp. isolated from clinical specimens were significantly higher compared to the other isolates. This may be because of the previous antibiotic use and multiple comorbid diseases. Among cases with Pseudomonas spp., the rate of previous antibiotic use and multiple comorbid diseases (≥ 2 chronic comorbidities) were 81.8% and 63.6% respectively. von Baum et al. reported that age >65 years, congestive heart failure and cerebrovascular disease were indicators for Enterobacteriaceae (36).Also, chronic respiratory disease and enteral tube feeding were indicators for Pseudomonas aeruginosa. However, other studies have demonstrated that increased age is not an indicator for gram negative microorganisms 37 38. In our study, there was no significant relationship between Pseudomonas spp. as a causative agent and poor prognosis (p = 0.573). The rates of carbapenem resistance were also quite high in Pseudomonas spp. (45.4%) and in Acinetobacter spp. (50%). And 25% of Staphylococcus aureus strains were methicillin resistant.
In this study, we found that serum procalcitonin and CRP levels were valuable indicators of poor prognosis in the elderly. There are various studies showing the contribution of complete blood count, CRP and procalcitonin used in the diagnosis and follow-up of pneumonia. However, there are fewer studies evaluating elderly patients with pneumonia in terms of these parameters 39 40 41 42. In our study, the mean CRP value on D0 was 181.68 ± 15.86 mg/L in the younger group and 118.11 ± 8.34 mg/L in the elderly (p = 0.001). This difference between the elderly group and the younger group showed that the initial CRP values on D0 may be lower in the elderly group than in the younger group. If the cut-off value is evaluated independent of age, it should be considered that CRP value may be less sensitive in the diagnostic and prognostic evaluation of the elderly group. In order to evaluate poor prognosis, the optimal cut-off value of CRP on D0 was 91.5 mg/L in the age-independent group, and 79 mg/L in the elderly group. In the study of 70 patients by Zhang et al., most of them were with pneumonia, examined the relationship of leukocyte, CRP and procalcitonin with sepsis/septic shock 41. They showed that CRP can predict poor prognosis at least as accurately as procalcitonin. They found that the cut-off value for CRP was 74.2 mg/L, sensitivity and specificity were 78% and 75% respectively.
In our study, procalcitonin was found to be the best prognostic indicator in the ROC curve in both the age-independent and the elderly group. For poor prognosis, the cut-off value was 0.295 ng/mL in the elderly group, with a sensitivity of 83% and a specificity of 69% (p<0.001). The cut-off value was 0.265 ng/mL in age-independent analysis, with a sensitivity of 77% and a specificity of 65% (p <0.001). In the meta-analysis of Liu et al., the prognostic cut-off value of procalcitonin was less than 0.5 ng/mL in only two studies 43. However, our study was consistent with the studies showing that procalcitonin is a reliable prognostic indicator 40 44 45 46 47 48 49 50 51. On the other hand, in a total of 667 cases evaluated by Akagi et al., including 436 pneumonia cases aged 75 years and over, procalcitonin was not an independent predictor of mortality in the elderly and in the young group, but was associated with the severity of pneumonia 39. In Zhang’s study, when the cut-off value of procalcitonin was 0.250 ng/ml, the sensitivity and specificity were 88% and 65%, respectively 41. This finding was consistent with our study.
In the elderly, the immune response to infections is decreased due to immunosenescence, and a chronic, low-grade systemic inflammation occurs. In addition, subclinical inflammation caused by exposure to various antigens in elderly patients manifests with relatively lower CRP and procalcitonin release in response to exogenic antigens. However, decreased procalcitonin levels in elderly patients can also be due to various etiologies of pneumonia with varying cytokine release patterns 39 52.
The modified Charlson comorbidity score was not correlated with poor prognosis in both age
groups. These findings suggest that the CURB–65 and the PSI are still consistent pneumonia cases and are superior to the modified Charlson comorbidity classification to accurately predict the prognosis.
Various studies have demonstrated that mortality rates are high in the elderly population 39 53 54. In our study, the 30-day mortality rates were found to be higher in the elderly group (11.8%) compared to the younger group (7%), but no statistically significant (p = 0.435). In the study of Saltoglu et al., the mortality rate of 130 patients with CAP was 3% and the mean age was 40±13.6 years 34. The high mortality rate in our study may be because of the high mean age of the patients (69.27±1.23) and the inclusion of HAP with severe infection. This may be due to the relatively high (49.63±1.68) mean age of our younger group.
Our study has several limitations. First, it was conducted in a single-center. Second, the rate of microbiologically confirmed cases was low, and we did not consider the causative pathogens other than Pseudomonas spp. as a risk factor. This study has also several strengths. First, it is a prospective study. Second, multiple comorbidities and different types of variables were included in the multivariate regression analysis. Also, this study has good generalizability because the results are broadly applicable to many different types of people and situations.