ANNALS OF MEDICINE 2022, VOL. 54, NO. 1, 3052–3059 https://doi.org/10.1080/07853890.2022.2138529 ORIGINAL ARTICLE Community-acquired pneumonia in hospitalised patients: changes in aetiology, clinical presentation, and severity outcomes in a 10-year period lia Sellares-Nadala,b,c , Joaquın Burgosa,b,c, Marıa Teresa Martın-Go Ju mezd, Andres Antond, Roger Sordee, Daniel Romero-Herrero d a,b , Pau Bosch-Nicolau , Anna Falco -Roget , Cristina Kirkegaardb, b Dolors Rodrıguez-Pardo b b , Oscar Len and Vicenç Falco  a,b,c a Departament de Medicina, Universitat Autonoma de Barcelona, Bellaterra, Spain; bInfectious Diseases Department, Vall d’Hebron Barcelona Hospital Campus, Vall d’Hebron Hospital Universitari, Barcelona, Spain; cMalalties Infeccioses Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron Barcelona Hospital Campus, Vall d’Hebron Hospital Universitari, Barcelona, Spain; dMicrobiology Department, Vall d’Hebron Barcelona Hospital Campus, Vall d’Hebron Hospital Universitari, Barcelona, Spain; eInternal Medicine Department, Hospital de Trauma Manuel Giagni, Ministerio de Salud P ublica y Bienestar Social, Asuncion, Paraguay ABSTRACT ARTICLE HISTORY Background and objective: Community-acquired pneumonia (CAP) is a frequent cause of hos- Received 3 March 2022 pitalisation. Several factors, such as pandemics, vaccines and globalisation may lead to changes Revised 12 October 2022 in epidemiology, clinical presentation, and outcomes of CAP, which oblige to a constant actual- Accepted 17 October 2022 isation. We performed this study to analyse how these factors have evolved over a 10- KEYWORDS year period. Pneumonia; virology; Materials and methods: Patients diagnosed with CAP for two 1-year periods that were 10 years Staphylococcus aureus; apart (2007–2008 and 2017–2018) were included. We compared microbiological information, MRSA; mortality clinical data and evolutive outcomes in the two periods. A mortality analysis was performed. Results: 1043 patients were included: 452 during the first period (2007- 2008), and 591 during the second period (2017–2018). Bacterial aetiology did not change during the 10-year period, besides a slight increase in Staphylococcus aureus (0.9% vs 2.9%, p ¼ 0.026). There was a decline in the proportion of bacteraemia in the second period (14.8% vs 9.6%, p ¼ 0.012). The incidence of complicated pleural effusion and septic shock declined too (6.4% vs 3.6%, p ¼ 0.04 and 15.5% vs 6.3%, p < 0.001). Respiratory failure and Intensive care unit (ICU) admission were similar in both periods. Variables independently associated with mortality were age and septic shock. Influenza vaccine was a protective factor against mortality in the second period. Conclusions: We have not found relevant differences in the bacterial aetiology of CAP over this 10-year period. There has been a decline in septic complications of CAP such as septic shock, bac- teraemia, and complicated pleural effusion. Influenza vaccination is an important tool to reduce mortality. KEY MESSAGES  There were no differences in the bacterial pathogens causing CAP among the 10-year study period. There has been a decline in septic complications of CAP such as septic shock, bacter- aemia, and complicated pleural effusion. Introduction CAP. Therefore, periodic survey studies are needed to Community-acquired pneumonia (CAP) is one of the ensure that clinical guidelines remain adequate [4,5]. With the development of new diagnostic methods main challenges in infectious diseases, as it represents an important cause of morbidity and mortality, and it such as molecular testing, there has been an improve- is a frequent cause of hospitalisation worldwide [1–3]. ment in the identification of etiologic causes of CAP Moreover, infectious diseases are a dynamic field, in [6,7]. This has led to an increase of viral detection in which epidemiology is constantly evolving. Changes in patients with CAP, reaching up to 24.5–30% of all environmental factors, the introduction of new vac- cases [8,9]. Some bacteria, such as Staphylococcus aur- cines and globalisation itself may lead to changes in eus, have traditionally been related to viral co-infec- epidemiology, clinical presentation, and severity of tion. Recent guidelines recommend the investigation CONTACT Joaquın Burgos jburgos@vhebron.net Infectious Diseases Department, Vall d’Hebron Barcelona Hospital Campus, Vall d’Hebron Hospital Universitari, Passeig Vall d‘Hebron 119-129, Barcelona 08035, Spain ß 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ANNALS OF MEDICINE 3053 of respiratory viruses mainly during the flu-season [4], (solid organ transplantation, haematopoietic trans- however, the impact on bacterial aetiology and clinical plantation, chemotherapy, long-term use of corticoste- outcomes of patients has been barely evaluated. With roids, and HIV infection). We also registered clinical such diagnostic approach, we wonder if there has information, laboratory results, radiological findings, been changes in the pathogens detected and in the microbiological information, and severity data (septic outcomes of patients with CAP in the last years. shock and respiratory failure). Empirical treatment was Another relevant change affecting the management recorded. Evolutive variables, such as admission at the of respiratory-tract infections in the latest years is the Intensive Care Unit (ICU) and in-hospital mortality expansion of vaccination programs with higher vaccin- were collected. CURB-65 score and Pneumonia ation targets [10]. On the one hand, it is known that Severity Index (PSI) were calculated. influenza vaccine reduces the rates of influenza infec- tion as well as improves the prognosis of patients [11]. Microbiologic procedures On the other hand, in our setting the 13-valent conju- gate pneumococcal vaccine was introduced in 2010, Microbiologic diagnostic procedures were performed and there is little information on its impact on clinical according to the hospital protocol and included (1) presentation and severity outcomes of adult patients two sets of blood cultures, (2) when available, qualita- with CAP [12]. We wonder if vaccination policies have tive and semi-quantitative culture of a good quality had an impact on the severity and outcomes of sputum sample as previously defined [13], (3) in patients with CAP in the last decade. patients who required orotracheal intubation, an We designed this study to analyse if there have endotracheal aspirate or bronco-alveolar-lavage sam- been changes in the epidemiology, clinical presenta- ples (4), urinary antigen for Streptococcus pneumoniae tion, and outcomes of CAP over a 10-year period. We in all patients, (5) urinary antigen for Legionella pneu- included patients diagnosed with CAP for two 1-year mophila if there was clinical or epidemiological suspi- periods that were 10 years apart (2007–2008 and cion and in all cases of severe CAP. Other 2017–2018). Between these 2 periods detection of microbiological techniques such as PCR for S. pneumo- respiratory virus, mainly during the flu season, was niae in pleural fluid or serologic determinations for implemented in our hospital. The objectives of the antibodies against atypical pathogens were performed study were to study if there have been changes in the according to clinical or epidemiological suspicion. We clinical presentation, bacterial aetiology, and outcomes considered a positive sputum culture when the semi- of patients with CAP between these two periods. quantitative culture yielded > 1.000.000 colony-form- ing units (CFU) or when the microorganism was found Materials and methods to be predominant in qualitative cultures. The only difference in the microbiologic procedures Study design and inclusion criteria performed in the two periods was viral detection. In the This is an observational study of all consecutive adult second period (2017–2018) a multiplex real-time PCR inpatients (age 18 years) diagnosed with CAP in two determination of respiratory virus (Influenza A and B, periods separated by 10 years: 2007–2008 and including Flu A-H1pdm09, Respiratory Syncytial Virus A 2017–2018. The study was conducted at the Hospital and B, Adenovirus, Enterovirus, Metapneumovirus, Universitari Vall d’Hebron, a 1100 beds-tertiary teach- Parainfluenza 1–4 Virus, Rhinovirus, Bocavirus 1–4, and ing hospital in Barcelona. Patients with nosocomial Coronavirus NL63, OC43 and 229E) in a nasopharyngeal pneumonia and those with evidence of aspiration swab was performed if requested by the attending phys- pneumonia (dysphagia, altered gag reflex, low level of ician in case of clinical or epidemiological suspicion consciousness) were excluded. (AllplexTM Respiratory Panels 1, 2 and 3, Seegene Inc., Korea). During the flu season (from November 2017 to March 2018) a rapid narrow-range real-time PCR for Data collection influenza virus was also performed (Xpert Flu/RSV!. We collected epidemiologic information (age, sex, resi- Cepheid, Sunnyvale, CA, USA). dency in nursing home, smoking, alcohol consumption and vaccination status), comorbidities (hypertension, Definitions chronic obstructive pulmonary disease (COPD), dia- betes mellitus, chronic renal failure, neurological disor- Pneumonia was defined as the presence of signs or ders, and neoplasms) and immunosuppressive factors symptoms of respiratory-tract infection (cough, fever, 3054 J. SELLARÈS-NADAL ET AL. purulent sputum, pleuritic chest pain or pulmonary (forward onwards) by binary logistic regression to semiology compatible with lung consolidation) associ- identify variables independently associated with mor- ated with the presence of a newly visualised infiltrate tality. Secondly, we performed a multivariate analysis in the chest radiography. of mortality in each period in order to know the evo- Bacterial pneumonia was diagnosed in patients lution of risk factors in both periods. Significant when (1) a microorganism likely to cause bacterial (p < 0.05) variables from the univariate analysis were pneumonia was isolated in blood, pleural fluid, accept- included in the multivariate analysis. Statistical analy- able-quality sputum, endotracheal aspirate or bronco- ses were performed using IBM SPSS Statistics for alveolar-lavage samples, (2) a PCR for S. pneumoniae Windows, Version 20.0. Armonk, NY; IBM Corp. was positive in pleural fluid, (3) a urinary antigen test Released 2011. for S. pneumoniae or L. pneumophila was positive, (4) seroconversion of L. pneumophila, M. pneumoniae, C. Ethics statement pneumoniae, C. psittaci and C. burnetii antibody titres was documented. Viral infection was considered in all The study was approved by the Ethics Committee of patients presenting a positive real PCR test for respira- Vall d’Hebron Research Institute (registration code tory virus, regardless of the detection of a bacterial PR(AG)345/2018). Need for informed consent was microorganism. Respiratory failure was defined as pO2 waived, as data and samples were analysed retrospect- in arterial blood lower or equal to 60 mmHg or periph- ively, and the study was non-interventional in nature. eral pulse oximetry lower than 90%. Septic shock was defined as the need of vasopressors to maintain Results medium blood pressure over 65 mmHg despite an adequate volume status. Overall, 1043 patients were included: 452 during the first period (2007–2008) and 591 during the second period (2017–2018). Monthly distribution of cases was Treatment similar in both periods (Figure 1). Six hundred and Patients received antibiotic treatment according to sixty-two (63.5) were men, and the mean age was current local protocols, which did not change between 65 years (SD 18.8). In Table 1, we present the demo- the two study periods. Hospitalised patients with graphic characteristics and comorbid conditions of the non-severe CAP received amoxicillin-clavulanic acid study population. In the second period, patients were and in those with severe CAP, a third-generation slightly older (63.4 vs 66.1 years; p ¼ 0.02) and there cephalosporin (ceftriaxone or cefotaxime) associated was a higher proportion of women (33% vs 39.3%; with a macrolide (azithromycin) was prescribed. p ¼ 0.038). The proportion of patients with any under- Fluoroquinolones were recommended in cases of peni- lying condition was similar, despite there were some cillin allergy or as an alternative treatment at the differences in the type of comorbidities. A higher pro- attending physician discretion when there was a clin- portion of patients were vaccinated against influenza ical suspicion for atypical pneumonia. In the second during the flu season in the second period. period, a 5-day course of oseltamivir was recom- Table 2 shows the pathogens documented in both mended for inpatients with confirmed influ- periods. Bacterial aetiology did not change over the enza infection. 10-years period. The only difference observed was a slight but significant increase in S. aureus CAP in the second period (p ¼ 0.026). Methicillin-resistant S. aur- Statistical analysis eus (MRSA) represented 1 out of 4 S. aureus isolates in We performed a descriptive analysis of basal charac- the first period and 4 out of 13 in the second period teristics, clinical and microbiological information of the (25% vs 23.5%, p ¼ 1.0). There was a decline in the study population. Categorical variables are expressed proportion of bacteraemic CAP in the second period as percentages and numerical data as mean and (14.8% vs 9.6%, p ¼ 0.012). standard deviation (SD) in cases of normal distribution Regarding severity and clinical presentation, CURB- or as median and interquartile range (IQR) for non- 65 and PSI scores were similar in both periods (Table normally distributed data. Variables were compared 3). The incidence of septic complications (complicated between the two periods using the Chi-square test for pleural effusion and septic shock) declined in the qualitative variables and T-student test for quantitative second period (6.4% vs 3.6%, p ¼ 0.004 and 15.5% vs variables. We performed first a multivariate analysis 5.3%, p < 0.001, respectively). On the other hand, ANNALS OF MEDICINE 3055 Figure 1. Monthly distribution of patients included in the two periods. Table 1. Basal characteristics and underlying conditions of patients included in the two study periods. Characteristics Total, n ¼ 1043 (%) 2007–2008, n ¼ 452 (%) 2017–2018, n ¼ 591(%) p Mean (SD) age (years) 65.0 (18.8) 63.4(19.5) 66.14 (18.1) 0.02 Gender Male 662 (63.5) 303 (67) 359 (60.7) 0.038 Female 381 (36.5) 149 (33) 232 (39.3) Smoking (>5 cig/day) 229 (22) 108 (23.9) 121 (20.5) 0.200 Alcohol (>60 gr/d) 84 (8.1) 39 (8.6) 45 (7.6) 0.567 Nursing home 56 (5.4) 21 (4.6) 35 (5.9) 0.407 Any chronic underlying conditiona 576 (55.2) 250 (55.3) 326 (55.2) 1.00 Any immunosuppressive conditionb 201 (19.3) 88 (19.5) 113 (19.1) 0.937 Previous vaccination Pneumococcal vaccination (previous 5 years) 110 (10.5) 46 (10.2) 64 (10.8) 0.761 Influenza vaccination 263 (25.2) 76 (16.8) 187 (31.6) <0.01 Data are expressed as numbers and percentages unless otherwise indicated. SD: standard deviation; IQR: interquartile range; HIV: human immunodefi- ciency virus. aUnderlying conditions: heart disease, chronic lung disease, chronic renal disease, neurological disease (vascular or degenerative) or cirrhosis. b Immunosupressive conditions: HIV infection, solid organ transplant, haematologic malignancies, solid organ neoplasm. respiratory failure and ICU admission were similar in Respiratory failure was a risk factor but only in the first both periods. During the second period, hospitalisa- period (OR 7.57; 95%CI 2.83–20.2). Finally, we tion length was shorter than in the first period. observed that in the second period, in which the rate Regarding mortality, although there were no signifi- of influenza vaccination was significantly higher (Table cant differences between the two periods, there was a 1), influenza vaccination was an independent protect- slight trend to lower 30-day mortality in the second ive factor against mortality (OR 0.15; 95%CI 0.05–0.44). period (8.8% vs 5.8%, p ¼ 0.067). In the multivariate analysis, the independent factors associated with mor- Discussion tality were age (OR 1.05; 95%CI 1.3–1.7), septic shock (OR 5.38; 95%CI 3.00–9.67) and respiratory failure (OR The epidemics and the bacterial aetiology of CAP did 3.97; 95%CI 2.14–7.35). In order to know the variables not suffer a relevant clinical change along a 10-year associated with mortality in each period we performed period. Bacterial pathogens did not differ over the a mortality analysis in each period which is shown in years, although there was a slight but significant Table 4. Age and septic shock were variables inde- increase in the presence of S. aureus during the pendently associated with mortality in both periods. second period. During that period, the prevalence of 3056 J. SELLARÈS-NADAL ET AL. Table 2. Viral and bacterial pathogens documented in both periods. Total, n ¼ 1043 (%) 2007–2008, n ¼ 452 (%) 2017–2018, n ¼ 591 (%) p Bacteria documented Streptococcus pneumoniae 382 (36.6) 171 (37.8) 211 (35.7) 0.517 Staphylococcus aureus 21 (2.0) 4 (0.9) 17 (2.9) 0.026 Legionella pneumophila 31 (3.0) 15 (3.3) 16 (2.7) 0.585 Other atypical bacteria 15 (1.4) 7 (1.5) 8 (1.4) 0.799 Haemophilus influenzae 28 (2.7) 13 (2.9) 15 (2.5) 0.847 Pseudomonas aeruginosa 17 (1.6) 11 (2.4) 6 (1.0) 0.086 Other bacteria 36 (3.5) 16 (3.5) 20 (3.4) 1.00 More than one bacterium 39 (3.7) 10 (2.2) 29 (4.9) 0.031 No bacteria detected 474 (45.4) 205 (45.4) 269 (45.5) 1.00 Virus documented Influenza A – – 31 (5.2) – Influenza B – – 34 (5.8) – Respiratory syncytial virus – – 16 (2.7) – Rhinovirus – – 20 (3.4) – Adenovirus – – 3 (0.5) – Coronavirus (not SARS-CoV2) – – 9 (1.5) – Metapneumovirus – – 5 (0.8) – Parainfluenza virus – – 1 (0.2) – Any bacterial or viral pathogen 623 (59.7) 247 (54.6) 376 (63.6) 0.004 Positive blood cultures 124 (11.9) 67 (14.8) 57 (0.6) 0.12 Table 3. Clinical information of CAP in the two periods. Total, n ¼ 1043 (%) 2007–2008, n ¼ 452 (%) 2017–2018, n ¼ 591 (%) p Bilateral consolidation 131 (12.6) 59 (13.1) 72 (12.2) 0.706 Complicated pleural effusiona 50 (4.8) 29 (6.4) 21 (3.6) 0.04 Severity scores PSI high mortality classes (IV or V) 580 (55.7) 255 (56.7) 325 (55) 0.615 CURB-65 high mortality risk group (3 points) 238 (22.8) 92 (20.4) 146 (24.7) 0.102 ICU admission 94 (9.0) 41 (9.1) 53 (9.0) 1.0 Septic shockb 107 (10.3) 70 (15.5) 37 (6.3) <0.001 Respiratory failurec 454 (43.5) 196 (43.4) 258 (43.7) 0.95 Invasive mechanical ventilation 54 (5.2) 30 (6.6) 24 (4.1) 0.068 Mean of hospitalisation days (SD) 10.2 (12.9) 11.9 (14.0) 8.9 (11.8) <0.001 Mortality Mortality during hospitalisation 83 (8.0) 41 (9.1) 42 (7.1) 0.251 48-h mortality 26 (2.5) 11 (2.4) 15 (2.5) 1.0 30-day mortality 74 (7.1) 40 (8.8) 34 (5.8) 0.067 Data are expressed as numbers and percentages unless otherwise indicated. ICU: intensive care unit, SD: standard deviation, PSI: pneumonia severity index. aComplicated pleural effusion: pH < 7.2, low glucose levels or evidence of microorganism by culture or gram stain. bSeptic shock: need of vaso- active drugs. cRespiratory failure: pO2 in arterial blood lower or equal to 60 mmHg or peripheral pulse-oximetry lower than 90%. Table 4. 30-Day mortality analysis (univariate and multivariate) in the two periods. Univariate analysis Multivariate analysis Alive at 30 days, Death at 30 days, Odds ratio n ¼ 412 (%) n ¼ 40 (%) p (95% CI) p First period (2007–2008) Mean (SD) age (years) 62.2 (19.7) 73.4(15.2) 0.001 1.037 (1.01–1.06) 0.002 Nursing home 15 (3.6) 6 (15) 0.007 – – Any underlying condition 222 (53.9) 28 (70) 0.066 – – Influenza vaccination 71 (17.2) 5 (12.5) 0.657 – – Bilateral consolidation 110 (11.4) 21 (28.4) <0.001 – – Respiratory failure 161 (39.1) 35 (87.5) <0.001 7.571 (2.83–20.2) <0.001 Septic shock 53 (12.9) 17 (42.5) <0.001 3.735 (1.74–8.02) 0.001 Positive blood cultures 56 (13.6) 11 (27.5) 0.032 – – Alive at 30 days, Death at 30 days, Odds ratio n ¼ 557 (%) n ¼ 34 (%) p (95% CI) p Second period (2017–2018) Mean (SD) age (years) 65.3 (18.2) 79 (10.8) <0.001 1.092 (1.05–1.13) <0.001 Nursing home 29 (5.2) 5 (17.6) 0.001 – – Any underlying condition 298 (53.5) 28 (82.4) 0.001 – – Influenza vaccination 183 (32.9) 4 (11.8) 0.012 0.146 (0.049–0.436) 0.001 Bilateral consolidation 60 (10.8) 12 (35.3) <0.001 – – Respiratory failure 233 (41.8) 25 (73.5) 0.001 – – Septic shock 28 (5.0) 9 (26.5) <0.001 8.676 (3.3–22.83) <0.001 Positive blood cultures 49 (8.8) 8 (23.5) 0.011 – – SD: Standard deviation, CI: confidence interval. Septic shock: need of vasoactive drugs; Respiratory failure: pO2 in arterial blood lower or equal to 60 mmHg or peripheral pulse-oximetry lower than 90%. ANNALS OF MEDICINE 3057 S. aureus CAP was 2.9% which is similar to that its introduction and reported stability in the rates of reported in other studies [2]. Staphylococcus aureus septic shock in 2010–2014 [12]. In comparison, in our has classically been related with influenza infection study we have observed a decrease in septic compli- and several studies have demonstrated an increase in cations in the second period, which could be S. aureus detection during the flu-season [2,14]. explained by changes in the population pattern of Microbiological procedures did not differ between the serotypes in adults due to the implementation of the two periods except for the determination of respira- 13-valent vaccine in children. Further studies with tory viruses, which was implemented in the second pneumococcal serotype determinations are needed to period. Therefore, we do not have epidemiological make more concise statements on this subject. data on viral infection during the first period and it is Finally, in the mortality analysis we found that fac- not possible to concrete if the slight increase in S. aur- tors independently associated with 30-day mortality eus detection in the second could be explained by a were age and septic shock. These findings are rise in influenza virus infections. expected since they represent risk factors for mortality In the era of antibiotic resistance, antibiotic suscep- in other diseases as well [22,23], and this fact has not tibility should constantly be assessed. Studies per- changed in the 10-year period. formed in other settings show a high proportion of Interestingly, influenza vaccination was a protective MRSA, reaching 40–60% of all S. aureus isolates factor against mortality in the second period, in which [14,15]. In our study, CAP caused by MRSA represents vaccination rates were significantly higher than in the less than 0.5%, with no changes between the two first period. With the increase in vaccination rates, periods. Considering the stability in the proportion of influenza vaccine became a protective factor of mor- MRSA detection, we consider that empirical treatment tality. It is known that an increase in the population should not be modified to offer MRSA coverage in vaccination rates can imply a herd effect and promote our setting. community protection [24]. Our results are concordant Concerning clinical and severity outcomes, despite with other studies in which influenza vaccine is a pro- the population in both periods had similar rates of tective factor against in-hospital death [25]. Therefore, underlying conditions and initial severity scores we support that influenza vaccination is an important (CURB-65 and PSI), in the second period there was a tool to prevent mortality in patients with CAP. decline of septic complications such as septic shock Our study has some limitations. First, the retro- and complicated pleural effusion. The proportion of spective nature of the study entails design limitations. patients with positive blood-cultures decreased as Second, it presents data from only one centre, so it well. These results go in the opposite direction than might not be representative of other geographical previous studies, in which an increase of septic com- areas, where some factors could differ. Third, the study plications over the years was reported [12,16]. was performed before the emergence of the SARS- Pneumococcal serotypes may be related with these CoV-2 pandemic, so we have not addressed the inter- changes since they are one of the determinants of action between bacterial pneumonia and SARS-CoV-2 severity in patients with pneumococcal disease and infection. However, in contrast to what occurs in therefore in patients with CAP. Over the years, patients with severe influenza in which bacterial co- changes in the serotype prevalence have led to pathogens are commonly identified, the overall pro- changes in the clinical presentation and severity of portion of bacterial coinfection among patients with subjects with CAP. The implementation of the 7-valent COVID-19 seems to be only 6.9–8% [26,27]. Finally, we vaccine brought a decline in the number of hospital- do not have data on pneumococcal serotypes in isations for pneumonia in children and adults [17]. patients with pneumococcal pneumonia so we cannot Despite this, in the years after its implementation, con- make concise statements on the effect of pneumococ- cerns raised regarding the increase of non-vaccine cal vaccination in the decrease of septic complications. serotypes [16,18,19] which came along with an increase of suppurated complications such as empy- Conclusion ema [20,21]. In our study, between the two study peri- ods, 13-valent conjugate pneumococcal vaccine was We have not found relevant differences in the bacter- implemented, which could explain the changes men- ial aetiology of CAP over this 10-year period besides a tioned above. Clinical changes after the introduction slight increase in S. aureus infections. Moreover, there of the 13-valent vaccine have barely been reported. has been a decline in septic complications of CAP Simonetti et al assessed clinical outcomes 3 years after such as septic shock, bacteraemia, and complicated 3058 J. SELLARÈS-NADAL ET AL. pleural effusion. Influenza vaccination is an important incidence, epidemiology, and mortality. Clin Infect strategy to decrease mortality in patients with CAP. Dis. 2017;65(11):1806–1812. [4] Metlay JP, Waterer GW, Long AC, et al. Diagnosis and Therefore, in our opinion, the current management for treatment of adults with community-acquired pneu- patients with CAP stated in recent guidelines remains monia. Am J Respir Crit Care Med. 2019;200(7): adequate in our setting. E45–E67. [5] Naucler P, Henriques-Normark B, Hedlund J, et al. Changing epidemiology of community-acquired Author contributions pneumonia: nationwide register-based study in s-Nadal, J Burgos, V. Sweden. J Intern Med. 2019;286(6):689–701. Conception and design: J. Sellare [6] Katsurada N, Suzuki M, Aoshima M, et al. The impact Falco; Analysis and interpretation of the data: J. of virus infections on pneumonia mortality is complex Sellar es-Nadal, R. Sorde, J Burgos, V. Falco ; Drafting of in adults: a prospective multicentre observational the paper: J. Sellares-Nadal, J Burgos, V. Falco ; Critical study. BMC Infect Dis. 2017;17(1):755. revision for intellectual content; MT. Martin-Go mez, R. [7] Johansson N, Kalin M, Tiveljung-Lindell A, et al.  Etiology of community-acquired pneumonia: Sorde, A. Anton, D. Romero-Herrero, P. Bosch-Nicolau, increased microbiological yield with new diagnostic A. Falco-Roget, C. Kirkegaard, D. Rodrıguez-Pardo, O. methods. Clin Infect Dis. 2010;50(2):202–209. Len; Final approval of the version to be published: J. [8] Burk M, El-Kersh K, Saad M, et al. Viral infection in Sellar es-Nadal, J. Burgos MT. Martin-Go mez, R. Sorde, community-acquired pneumonia: a systematic review A. Anton D. Romero-Herrero, P. Bosch-Nicolau, A. and meta-analysis. Eur Respir Rev. 2016;25(140): Falco-Roget, C. Kirkegaard, D. Rodrıguez-Pardo, O. Len, 178–188. [9] Shoar S, Musher DM. Etiology of community-acquired . V. Falco pneumonia in adults: a systematic review. Pneumonia. 2020;12(1):11. [10] Healthy People 2020. United States Department of Disclosure statement Human and Health Services. IID-12.12. Increase the No potential conflict of interest was reported by percentage of noninstitutionalized adults aged 18 the author(s). and older who are vaccinated annually against sea- sonal influenza. n.d. [cited 2021 Dec 17]. www. healthypeople.gov/node/6360/data_details. Funding [11] Castilla J, Godoy P, Domınguez A,  et al. Influenza vac- cine effectiveness in preventing outpatient, inpatient, The author(s) reported there is no funding associated with and severe cases of laboratory-confirmed influenza. the work featured in this article. Clin Infect Dis. 2013;57(2):167–175. [12] Simonetti AF, Garcia-Vidal C, Viasus D, et al. Declining ORCID mortality among hospitalized patients with commu- nity-acquired pneumonia. Clin Microbiol Infect. 2016; lia Sellares-Nadal Ju http://orcid.org/0000-0002-0209-6494 22(6):567.e1-567–e7. Daniel Romero-Herrero http://orcid.org/0000-0002- [13] Murray PR, Washington JA. Microscopic and bacerio- 1476-0815 logic analysis of expectorated sputum. Mayo Clin Dolors Rodrıguez-Pardo http://orcid.org/0000-0001- Proc. 1975;50(6):339–344. 6781-5405 [14] Self WH, Wunderink RG, Williams DJ, et al. Staphylococcus aureus community-acquired pneumo- nia: prevalence, clinical characteristics, and outcomes. Data availability statement Clin Infect Dis. 2016;63(3):300–309. The authors confirm that the data supporting the findings of [15] Moran GJ, Krishnadasan A, Gorwitz RJ, et al. this study are available within the article. Prevalence of methicillin-resistant Staphylococcus aur- eus as an etiology of community-acquired pneumo- nia. Clin Infect Dis. 2012;54(8):1126–1133. References [16] Burgos J, Falco  V, Borrego A, et al. Impact of the emergence of non-vaccine pneumococcal serotypes [1] Prina E, Ranzani OT, Polverino E, et al. Risk factors on the clinical presentation and outcome of adults associated with potentially antibiotic-resistant patho- with invasive pneumococcal pneumonia. Clin gens in community-acquired pneumonia. Ann Am Microbiol Infect. 2013;19(4):385–391. Thorac Soc. 2015;12(2):153–160. [17] Griffin MR, Zhu Y, Moore MR, et al. U.S. hospitaliza- [2] Jain S, Self WH, Wunderink RG, et al. Community- tions for pneumonia after a decade of pneumococcal acquired pneumonia requiring hospitalization among vaccination. N Engl J Med. 2013;369(2):155–163. U.S. adults. N Engl J Med. 2015;373(5):415–427. [18] Ardanuy C, Tubau F, Pallares R, et al. Epidemiology of [3] Ramirez JA, Wiemken TL, Peyrani P, et al. Adults hos- invasive pneumococcal disease among adult patients pitalized with pneumonia in the United States: in Barcelona before and after pediatric 7-valent ANNALS OF MEDICINE 3059 pneumococcal conjugate vaccine introduction, 1997- population based cohort study. BMJ. 2019;364: 2007. Clin Infect Dis. 2009;48(1):57–64. l525–12. [19] Lepoutre A, Varon E, Georges S, et al. Impact of the [24] Switzer C, Babiuk L, Loeb M. Determining optimal pneumococcal conjugate vaccines on invasive community protection strategies for the influenza pneumococcal disease in France, 2001-2012. Vaccine. vaccine. Expert Rev Vaccines. 2019;18(8):755–764. 2015;33(2):359–366. [25] Arriola C, Garg S, Anderson EJ, et al. Influenza vaccin- [20] Grijalva CG, Nuorti JP, Zhu Y, et al. Increasing inci- ation modifies disease severity among community- dence of empyema complicating childhood commu- dwelling adults hospitalized with influenza. Clin Infect nity-acquired pneumonia in the United States. Clin Dis. 2017;65(8):1289–1297. Infect Dis. 2010;50(6):805–813. [26] Langford BJ, So M, Raybardhan S, et al. Bacterial co- [21] Burgos J, Lujan M, Falco  V, et al. The spectrum of infection and secondary infection in patients with pneumococcal empyema in adults in the early 21st century. Clin Infect Dis. 2011;53(3):254–261. COVID-19: a living rapid review and meta-analysis. [22] Hespanhol V, Barbara C. Pneumonia mortality, comor- Clin Microbiol Infect. 2020;26(12):1622–1629. bidities matter? Pulmonology. 2020;26(3):123–129. [27] Rawson TM, Zhu N, Ranganathan N, et al. Bacterial [23] Gharbi M, Drysdale JH, Lishman H, et al. Antibiotic and fungal co-infection in individuals with corona- management of urinary tract infection in elderly virus: a rapid review to support COVID-19 antimicro- patients in primary care and its association with bial prescribing timothy. Clin Infect Dis. 2020;71: bloodstream infections and all cause mortality: 2459–2468.