Respiratory tract virus infections in the elderly with pneumonia

The number of over 80-year-old patients with weaning immune system is rising rapidly in Western societies, but the clinical significance of respiratory virus infections among this group remains unclear. The burden of pneumonia among the elderly is high as it includes significant morbidity, mortality and costs around the world [1]. In the United States alone, about 85% of the deaths caused by pneumonia or influenza occur in the age group of 65 years and older, and in 2013, in the health care of patients of all ages, more than 16.1 billion dollars were spent on pneumonia [2].

As polymerase chain reaction (PCR) and rapid antigen detection tests are increasingly available for respiratory virus detection [3], currently in 13 to 30% of lower respiratory tract infections among the elderly, a virus has been implicated [4, 5]. In the age group of 65 years and older, influenza virus (Flu), respiratory syncytial virus (RSV) and parainfluenza virus (PIV) are the leading viral causes of respiratory morbidity and mortality [6‐8], while other viral causes include rhinovirus (RV) and coronaviruses (CoV) [9]. However, in contrast to pediatric data, there are only a few reports concerning the usefulness of respiratory virus diagnostics in the elderly, and virus diagnostics have shown only limited value in reducing antibiotic use and the length of hospital stay [10, 11]. While influenza detection is seen as an important diagnostic tool by the physicians, the detection of other viruses is seen less useful [11].

Among the elderly, the all-cause mortality rate associated with respiratory viral infection increases with age and is approximately 6–7% among the over 85-year-old subjects [8]. This may, however, be an underestimate due to a general bias towards predominantly recording influenza as a death course, while neglecting other virus infections. Other common risk factors of a severe respiratory viral infection among the elderly include underlying medical conditions and poor response to influenza vaccine [12, 13]. In adults, mixed infections with a respiratory virus and a bacterial pathogen, especially rhinovirus with pneumococci, have been shown to associate with more severe pneumonia and longer hospitalization period [14‐17]. C-reactive protein (CRP) value and white blood cell count (WBC) seem to be insufficient methods in differentiating sole bacterial, mixed and sole virus pneumonias, although no unambiguous methods exist [18, 19].

All of the aforementioned findings support a clinically meaningful role of virus diagnosis among the frail elderly. Thus, the aim of this study was to investigate how viral pathogens detected in the nasopharynx and conventional inflammatory markers (WBC and CRP) correlate to signs, symptoms and prognosis of pneumonia among the age group of 65 years and over. We hypothesized that virus detection could give clinically relevant information in addition to conventional inflammatory markers and chest radiograph findings in treating frail elderly patients.

The study shows three main findings. Firstly, radiologically confirmed pneumonia was not associated with respiratory virus detection. Moreover, in the studied episodes of hospital care diagnosed with pneumonia, the presence of a respiratory virus was associated neither with clinical outcomes, nor with WBC or CRP values. Against our study hypothesis, all the studied episodes of hospital care in which the patient was diagnosed with one or more respiratory viruses were, in fact, associated with a less severe clinical course in terms of the number of hospital revisits. Secondly, radiologically confirmed pneumonia was associated with the indicators of a severe bacterial infection, WBC over 15 × 10⁹/L and CRP over 100 mg/l. Thirdly, a CRP value over 100 mg/l was associated with death at ward.

Our finding, namely that 30% of the study patients diagnosed with pneumonia also had a respiratory virus present in nasopharynx, is in line with recent studies executed on adults that suggest that even one third of pneumonia cases are associated with a respiratory virus [4, 5, 9]. In our study, rhinovirus was the most common virus present in the study episodes diagnosed with pneumonia, followed by coronavirus and influenza virus. Treanor et al. anticipated in their study that the role of these common cold viruses, coronavirus and rhinovirus, among the elderly will rise in the future, although PIV, RSV and influenza are still considered the most harmful viruses among the elderly [6‐8]. Our findings also support this idea of a rising clinical significance of common cold viruses, especially rhinovirus, among the elderly [6, 24, 25]. An association between elevated disease severity and dual infection, especially rhinovirus/pneumococcal infection, in the adult population have been reported previously [14, 17]. Our analyses, however, showed no differences in the severity of the study episodes diagnosed with pneumonia regardless of whether there was a respiratory virus present or not.

Frailty, immunologic weakening and cardiopulmonary diseases are understood to predispose to pneumonia when a viral infection occurs [26]. In our study, a respiratory virus was found in no less than 40% of the elderly patients who suffered from respiratory symptoms but were not diagnosed with pneumonia. At the same time, the risk of a hospital revisit in all the studied episodes of hospital care seemed to be lower when a virus was present than when no virus was fund. These findings support the idea that respiratory viruses are merely innocent bystanders in patients with pneumonia [27].

Our study strengthens the idea that high CRP and WBC values are associated with pneumonia in patients with respiratory symptoms but have limited value as independent predictors [28]. In adult populations, only relatively high CRP values have been shown useful in predicting the presence of pneumonia, and a cut-off value of 100 mg/l is mentioned in some studies [29]. Krueger et al. concluded in their CAPNETZ-study with 1337 patients aged 62 ± 18 years that WBC and CRP are higher in typical bacterial than in atypical or viral etiology community-acquired pneumonias [30]. Gao et al. showed in their study that high levels of CRP were induced as well as correlated with the complement activation in patients infected with severe influenza A [31]. In our study, we saw no difference in inflammatory markers according to virus etiology.

According to our data, among the elderly, a respiratory disease that elevates CRP to over 100 mg/l could be linked to death on the count of that in such cases pneumonia is probable. Lee et al. showed similar results in their study with 424 patients aged 70.4 +/− 15.6 years [32]. Interestingly, they also showed that in addition to CRP the albumin level was associated with a 28-day mortality in hospitalized patients with a community-acquired pneumonia. On the other hand, Ortqvist et al. saw no association between high CRP and mortality in hospital-treated pneumonia patients and Krueger et al. stated in the CAPNETZ-study that WBC in contrast to CRP increased with the severity of a community-acquired pneumonia [30, 33].

The strengths of our study include prospective design, large sample size and sensitive virus-detection methods. Also, pneumonia was radiologically confirmed. However, there are some limitations to our study as well. As the study observed hospitalization-requiring episodes among frail geriatric patients, the results cannot be generalized as such to treating outpatients. Also, the swab samples were collected from the upper airways and thus infections solely in the lower airways may have been missed. Further, as the virus samples were collected in 2007–2009, they naturally give specific information concerning those years only. The study was carried out ten years ago and in one center which limits the generalizability of the results. Many respiratory viruses exhibit a seasonal variation in temperate climates. However, at species level annual virus epidemics are relatively stable in climates with defined winter seasons like in Finland [34]. We used modern PCR diagnostics that have been in routine use ever since. We believe that nearly a 2-year recruitment period with 438 samples gives a relatively good picture of virus epidemics at species levels in our area. This study has prospective design and gives information about the effect of common respiratory virus infections, even though there is some annual variation in circulating viruses. Due to many chronic diseases in the elderly, virus-induced respiratory symptoms may be difficult to distinguish from other symptoms. However, for dyspnea, we used an objective criterion based on oxygen saturation.

This study gives valuable information about the significance of virus findings in nasopharynx and inflammatory markers among frail elderly patients with respiratory symptoms.

In elderly patients, the presence of respiratory viruses in the nasopharynx seems to have limited value in assessing the severity and the short-time prognosis of the disease. If a virus is found, it may in fact indicate a clinical situation with a better prognosis. In this study, not one respiratory virus correlated with the presence, signs and symptoms or prognosis of radiographically-verified pneumonia among the elderly. However, this study did find that a CRP over 80 mg/l and a WBC over 15 × 10⁹/L were linked to pneumonia and a CRP over 100 mg/l to elevated mortality during hospital stay. All in all, this study shows that pneumonia should be treated in elderly people as a bacterial disease regardless of virus findings.