Introduction

SARS-CoV-2 dramatically revealed the sudden impact and the utmost importance of respiratory viruses in our lives. Influenza and respiratory syncytial virus (RSV) infections are associated with high rate of morbidity, mortality, and an important burden on healthcare systems worldwide, especially among elderly patients [1,2,3,4,5,6]. Oldest-old is a term meant to include people aged 85 years and older that represent a growing population in the old European continent. Despite being a non-homogeneous group, they are professionally retired individuals that usually experience multimorbidity and disability and may face a limited life expectancy. Recent evidence suggests that there is a high research interest towards addressing their needs and establishing the best standard of care [7, 8]. Especially in epidemiological research, oldest-old patients are considered to be a part of the wider age group of the elderly (≥ 65 years old), and there is limited published evidence about predictors of severity of illness and mortality in viral infections, such as caused by influenza and RSV.

The aim of this study was to describe the clinical features of an oldest-old population admitted with influenza and/or RSV infections in three southern European hospitals over two consecutive winter seasons and identify predictors of pneumonia, non-invasive ventilation (NIV), and in-hospital death (IHD). Such knowledge might provide insight to assist healthcare policymakers managing chronic conditions, improving patient satisfaction, and reducing hospital utilization.

Methods

This is a multicenter, retrospective study of all oldest-old patients (≥ 85 years old) who were either admitted to the hospital for laboratory-confirmed influenza and/or RSV infection or developed it during the course of admission for other causes, from 1 October 2017 to 30 April 2018 and from 1 October 2018 to 30 April 2019 in three tertiary hospitals in Portugal, Italy and Cyprus. The laboratory confirmation was based on a positive Xpert Flu/RSV PCR (Cepheid Diagnostics, Sunnyvale, CA, USA) and/or Allplex Respiratory Panel (Allplex, Seegene, Republic of Korea) on naso/oropharyngeal swabs obtained from patients with signs or symptoms of viral infection. For patients with more than one positive PCR in a seasonal period, the first episode was considered for study purposes. The infection was characterized as hospital-acquired if symptoms pertaining to viral infection began after 72 h from admission. Variables assessed included demographics, smoking status, co-morbidities, virus type, nosocomial acquisition, pulmonary infiltrate on chest X-ray taken when symptoms were observed, neuraminidase inhibitor use, length of stay (from admission to discharge), NIV, mechanical ventilation, and IHD.

This study was conducted in accordance with the Declaration of Helsinki. Formal ethical approval was obtained by the institutional review board of the coordinating center (Central Lisbon Hospital Center, no. 762_2019). Informed consent was not deemed required for the purposes of this study.

Statistical analysis

Descriptive data are shown as absolute (n) and relative (%) frequencies for categorical data and as mean ± standard deviation (SD) and median and interquartile range (IQR), as appropriate, for continuous variables. On univariate analysis, chi-square test for categorical variables and Student’s t test or Wilcoxon rank-sum test, as appropriate, for continuous variables were carried out to identify factors associated with pneumonia, NIV, and IHD. Odds ratios (OR) and their 95% confidence intervals (95% CI) were also calculated to estimate the strength of those associations. Multivariable analysis models were then fitted to investigate the independent effects of type of virus infection and clinical variables that turned out to be significantly associated with the outcomes at univariate analysis, adjusting for possible confounders like age and gender.

For all tests, a p value ≤ 0.05 was considered significant.

All analyses were performed with Stata 14.

Results

A total of 1151 patients aged ≥ 18 years old were admitted for influenza A/B and/or RSV infections during the study period in the three centers. Oldest-old patients were 251 (21.8%), of which 30 (12%) had hospital-acquired influenza A/B and/or RSV infections.

Clinical features of oldest-old patients included in the study were reported in Table 1. Mean age was 89.4 ± 3.9 (range, 85 to 103) years, 79 (31.5%) were men, and 3.6% was current active smoker. The co-morbidities mainly observed were diabetes (22.7%), COPD or asthma (24.7%), CHF (47%), and CKD (19.5%). The viral agents identified were influenza A (56.6%), influenza B (15.9%), RSV (25.9%), influenza A + influenza B (1.2%), and influenza A + RSV (0.4%) co-infections. Among influenza A infections, H3N2 was the most common (68.5%) followed by H1N1 (13%), 18.5% not having been subtyped. Radiological signs of pneumonia were present on the chest X-ray exams of 32.3% (n = 81) following laboratory diagnosis of viral infection; 8.8% (n = 22) were submitted to NIV and only one patient (0.4%) was invasively mechanically ventilated. A total of 35 patients (13.9%) did not survive admission. Among patients submitted to NIV, 72.7% (n = 16) survived admission. Antiviral treatment with a neuraminidase inhibitor was started in 60.6% of patients. Mean length of stay of patients with community- and hospital-acquired infections was 12 ± 11.1 (median 9, IQR 6–14) and 27.8 ± 28.7 (median 20, IQR 8–30) days, respectively (p < 0.01). Overall, IHD was 13.9%, being 14% and 13.3% for community- and hospital-acquired infections, respectively, with no significant difference.

Table 1 Clinical features of oldest-old population included in the study
Full size table

Results of univariate and multivariable analyses were shown in Table 2.

Table 2 Uni- and multivariable analyses for factors associated with pneumonia, non-invasive ventilation (NIV), and intra-hospital death in oldest-old adults admitted with influenza A/B and/or RSV infection
Full size table

At univariate analysis, COPD or asthma was significantly associated with pneumonia (OR 1.93; 95% CI 1.06–3.49) and use of NIV (OR 3.49; 95% CI 1.43–8.51); RSV infection turned out to be another significant factor associated with NIV use (OR 3.51; 95% CI 1.27–9.68), while CKD was the only clinical feature significantly associated with IHD (OR 2.52; 95% CI 1.15–5.52). Finally, we considered three logistic regression models, where, for each outcome, age, gender, COPD or asthma (for pneumonia and use of NIV models), CKD (for IHD model), and type of virus infection were the independent variables.

Among all patients, multivariable analyses revealed that COPD or asthma was significantly associated with radiologically confirmed pneumonia (OR 1.86; 95% CI 1.02–3.43; p = 0.045); COPD or asthma (OR 4.4; 95% CI 1.67–11.6; p = 0.003), influenza B (OR 3.77; 95% CI 1.06–13.5; p = 0.041), and RSV infections (OR 3.12; 95% CI 1.09–8.92; p = 0.023) were associated with NIV use; CKD turned out to be the only predictor significantly associated with IHD (OR 2.50; 95% CI 1.14–5.51; p = 0.023).

Discussion

SARS-CoV-2 pandemic and its unsustainable burden supplanted every hierarchy of interest in medical research but highlighted how viral infections’ knowledge is crucial in clinical practice. Among the aged population, influenza and RSV infections are important causes of hospital admission during autumn and winter months.

The highlights of this study are the following findings: (1) the proportion of oldest-old patients among total hospitalizations was remarkable; (2) radiological pneumonia, use of NIV, and IHD were considerable; (3) influenza A H3N2 infection was the most prevalent; influenza B and RSV infection were significantly associated with NIV use; (4) COPD or asthma was associated with both pneumonia and NIV use; (5) CKD was a predictor of IHD.

To the best of our knowledge, our line of research is quite novel and barely comparable to previous reports given the higher mean age of patients involved.

Overall, the number of hospitalizations over the study period was remarkable comparing to recent reports [9], showing that over two years one out of five admissions with influenza and/or RSV infections involved oldest-old. Moreover, in our study, hospital-acquired influenza and/or RSV infections were not identified as predictors of pneumonia, use of NIV, and IHD for patients aged 65 years and older, moving away from evidence available so far [9, 10].

Pneumonia on infection presentation was very frequent, in line with more recent reports [9,10,11,12,13,14,15,16,17], revealing how an important proportion of patients showed a direct viral injury in lung parenchyma and/or bacterial co-infection. However, despite being a life-threatening condition, it was neither predictive of NIV use nor of IHD.

The use of NIV was considerable when comparing with available evidence in a cohort of younger patients [9]. This finding might be related to both the type of respiratory failure on infection presentation and a remarkable rate of diagnosed or likely under-diagnosed chronic obstructive lung disease [18]. Similarly, in our population, NIV seemed to be effective since data about its use in respiratory viral infections are limited and uncertain, especially in the presence of pneumonia, hypoxemic respiratory failure, and SOFA ≥ 5, and no COPD and/or cardiogenic pulmonary edema [19].

On the other hand, invasive mechanical ventilation rate was performed only once, probably according to ethical and prognostic considerations, such as coexistence of frailty and patient end-of-life preferences. The discrepancy between the use of these two modalities of ventilation could be presumably due to accept NIV as a ceiling of therapeutic effort.

IHD was also remarkable. Studies suggest mortality rate ranging from 4.1 to 9.8% [9,10,11,12,13, 15,16,17, 20] and up to 24% in ICU patients [14]. Nevertheless, these studies were performed on elderly people with lower median age while our data should be comprehensible for an oldest-old cohort of patients.

Infection due to influenza A H3N2 was the most prevalent, RSV infection was also very frequent, confirming that epidemic viral subtypes and their affinity for the lower respiratory tract differ according to the study period [21]. Severity of illness and IHD due to RSV infection were similar as compared to influenza viruses, but according to literature, these can vary from season to season [9, 22].

Therefore, the role of virus type in morbidity and mortality remains controversial. Our findings might suggest that RSV and influenza B probably caused infections with a clinical scenario that benefited from application of NIV [23] more than influenza A.

COPD or asthma was independently associated with pneumonia on infection presentation and NIV use, suggesting the importance of the aged lung [18] and the attempt to overcome respiratory failure through this widespread and generally well-tolerated ventilation technique.

CKD was the only significant predictor IHD, and it could represent an interesting clinical tool. Indeed, previous studies included acute kidney injury and renal disease as predictors of mortality and disease severity, respectively [9, 12], but no prognostic factors have been identified for the oldest-old.

Our study had limitations. A 72-h period might have led to misclassify the community- vs. hospital-acquisition of the viral infection since evidence suggests longer incubation periods [24]. Several factors contributing to disease severity and mortality, including nursing home residency, frailty scores, bedridden status, immunization status, malnutrition, sarcopenia, presence of mixed viral and bacterial pneumonia, respiratory failure, occurrence of systemic complications, and physicians’ attitude towards more intensive care, were not assessed. Moreover, our study lacks an assessment of post-discharge disability and follow-up.

In conclusion, this study provided one of the largest assessments available so far of clinical features and factors contributing to severity of illness in the oldest-old admitted with influenza and/or RSV infections in Southern Europe. Chronic organ failure, such as COPD or asthma and CKD, predicted pneumonia and IHD, respectively, surpassing the importance of viral virulence. These findings could impact on public health policies, such as fostering influenza immunization campaigns, home-based care programs [25], and end-of-life care. Filling knowledge gaps is crucial to set priorities and advise on transition model of care that best fits the oldest-old.