Epidemiological and clinical management aspects of pneumonias diagnosed in the emergency department in elderly patients in Spain: Results of the EDEN-29 study.

OBJECTIVES: To estimate the incidence of pneumonia diagnosis in elderly patients in Spanish emergency departments (ED), need for hospitalization, adverse events and predictive capacity of biomarkers commonly used in the ED. METHODS: Patients ≥65 years with pneumonia seen in 52 Spanish EDs were included. We recorded in-hospitaland 30-day mortality as adverse events, as well as intensive care unit (ICU) admission among hospitalizedpatients. Association of 10 predefined variables with adverse events was calculated and expressed as odds ratio (OR) with 95% confidence interval (CI), as well as predictive capacity of 5 commonly used biomarkers in the ED (leukocytes, hemoglobin, C-reactive protein, glucose, creatinine) was investigated using area under the receiver operating characteristic curve (AUC-ROC). RESULTS: 591 patients with pneumonia attended in the ED were included (annual incidence of 18,4 per 1000 inhabitants). A total of 78.0% were hospitalized. Overall, 30-day mortality was 14.2% and in-hospital mortality was 12.9%. Functional dependency was associated with both events (OR=4.453, 95%CI=2.361-8.400; and OR=3.497, 95%CI=1.578-7.750, respectively) as well as severe comorbidity (2.344, 1.363-4.030, and 2.463, 1.252-4.846, respectively). Admission to the ICU during hospitalization occurred in 3.5%, with no associated factors. The predictive capacity of biomarkers was only moderate for creatinine for ICU admission (AUC-ROC=0.702, 95% CI=0.536-0.869) and for leukocytes for post-discharge adverse event (0.669, 0.540-0.798). CONCLUSIONS: Pneumonia is a frequent diagnosis in elderly patients consulting in the ED. Their functional dependence and comorbidity is the factor most associated with adverse events. The biomarkers analyzed do not have a good predictive capacity for adverse events.

Deep learning-based lesion subtyping and prediction of clinical outcomes in COVID-19 pneumonia using chest CT.

The main objective of this work is to develop and evaluate an artificial intelligence system based on deep learning capable of automatically identifying, quantifying, and characterizing COVID-19 pneumonia patterns in order to assess disease severity and predict clinical outcomes, and to compare the prediction performance with respect to human reader severity assessment and whole lung radiomics. We propose a deep learning based scheme to automatically segment the different lesion subtypes in nonenhanced CT scans. The automatic lesion quantification was used to predict clinical outcomes. The proposed technique has been independently tested in a multicentric cohort of 103 patients, retrospectively collected between March and July of 2020. Segmentation of lesion subtypes was evaluated using both overlapping (Dice) and distance-based (Hausdorff and average surface) metrics, while the proposed system to predict clinically relevant outcomes was assessed using the area under the curve (AUC). Additionally, other metrics including sensitivity, specificity, positive predictive value and negative predictive value were estimated. 95% confidence intervals were properly calculated. The agreement between the automatic estimate of parenchymal damage (%) and the radiologists' severity scoring was strong, with a Spearman correlation coefficient (R) of 0.83. The automatic quantification of lesion subtypes was able to predict patient mortality, admission to the Intensive Care Units (ICU) and need for mechanical ventilation with an AUC of 0.87, 0.73 and 0.68 respectively. The proposed artificial intelligence system enabled a better prediction of those clinically relevant outcomes when compared to the radiologists' interpretation and to whole lung radiomics. In conclusion, deep learning lesion subtyping in COVID-19 pneumonia from noncontrast chest CT enables quantitative assessment of disease severity and better prediction of clinical outcomes with respect to whole lung radiomics or radiologists' severity score.

Mortality in patients treated for COVID-19 in the emergency department of a tertiary care hospital during the first phase of the pandemic: Derivation of a risk model for emergency departments. / Análisis de mortalidad de los pacientes atendidos por COVID-19 en el servicio de urgencias de un hospital de tercer nivel en la fase incial de la pandemia. Derivación de un modelo de riesgo para urgencias.

OBJECTIVES: To develop a risk model to predict 30-day mortality after emergency department treatment for COVID-19. MATERIAL AND METHODS: Observational retrospective cohort study including 2511 patients with COVID-19 who came to our emergency department between March 1 and April 30, 2020. We analyzed variables with Kaplan Meier survival and Cox regression analyses. RESULTS: All-cause mortality was 8% at 30 days. Independent variables associated with higher risk of mortality were age over 50 years, a Barthel index score less than 90, altered mental status, the ratio of arterial oxygen saturation to the fraction of inspired oxygen (SaO2/FIO2), abnormal lung sounds, platelet concentration less than 100 000/mm3, a C-reactive protein concentration of 5 mg/dL or higher, and a glomerular filtration rate less than 45 mL/min. Each independent predictor was assigned 1 point in the score except age, which was assigned 2 points. Risk was distributed in 3 levels: low risk (score of 4 points or less), intermediate risk (5 to 6 points), and high risk (7 points or above). Thirty-day risk of mortality was 1.7% for patients who scored in the low-risk category, 28.2% for patients with an intermediate risk score, and 67.3% for those with a high risk score. CONCLUSION: This mortality risk stratification tool for patients with COVID-19 could be useful for managing the course of disease and assigning health care resources in the emergency department.

OBJETIVO: Derivar un modelo de riesgo para estimar la probabilidad de mortalidad a los 30 días de la visita a urgencias de pacientes con COVID-19. METODO: Estudio observacional de cohortes retrospectivo de 2.511 pacientes con COVID-19 atendidos en el servicio de urgencias hospitalario (SUH) del 1 de marzo al 30 de abril de 2020. Se realizó análisis de supervivencia mediante Kaplan Meier y regresión de Cox. RESULTADOS: La mortalidad por cualquier causa a los 30 días fue de un 8%. Los factores asociados de forma independiente a mayor mortalidad fueron: edad 50 años, índice de Barthel 90 puntos, alteración del nivel de consciencia, índice de SaO2/FIO2 400, auscultación respiratoria anómala, cifra de plaquetas 100.000/mm3, PCR 5 mg/dL y filtrado glomerular 45 mL/min. A estos factores se les asignó una puntuación de 1, excepto a la edad, que se le asignó un valor de 2 puntos. Se dividió el modelo de riesgo en 3 categorías: riesgo bajo (menor o igual a 4 puntos), riesgo intermedio (5-6 puntos) y riesgo alto (igual o superior a 7 puntos). Para los pacientes clasificados como de bajo riesgo la probabilidad de mortalidad a los 30 días fue del 1,7%, en los casos de riesgo intermedio fue del 28,2% y para los de alto riesgo fue del 67,3%. CONCLUSIONES: Disponer de una herramienta para estratificar el riesgo de mortalidad de los pacientes con COVID-19 que consultan a un SUH podría ser de utilidad para la gestión de los recursos sanitarios disponibles.

Factors associated with revisits by patients with SARS-CoV-2 infection discharged from a hospital emergency department. / Factores asociados a revisita en pacientes con diagnóstico de infección por SARS-CoV-2 dados de alta de un servicio de urgencias hospitalario.

OBJECTIVES: To analyze emergency department (ED) revisits from patients discharged with possible coronavirus disease 2019 (COVID-19). MATERIAL AND METHODS: Retrospective observational study of consecutive patients who came to the ED over a period of 2 months and were diagnosed with possible COVID-19. We analyzed clinical and epidemiologic variables, treatments given in the ED, discharge destination, need to revisit, and reasons for revisits. Patients who did or did not revisit were compared, and factors associated with revisits were explored. RESULTS: The 2378 patients included had a mean age of 57 years; 49% were women. Of the 925 patients (39%) discharged, 170 (20.5%) revisited the ED, mainly for persistence or progression of symptoms. Sixty-six (38.8%) were hospitalized. Odds ratios (ORs) for the following factors showed an association with revisits: history of rheumatologic disease (OR, 2.97; 95% CI, 1.10-7.99; P = .03), digestive symptoms (OR, 1.73; 95% CI, 1.14-2.63; P = .01), respiratory rate over 20 breaths per minute (OR, 1.03; 95% CI, 1.0-1.06; P = .05), and corticosteroid therapy given in the ED (OR, 7.78; 95% CI, 1.77-14.21, P = .01). Factors associated with hospitalization after revisits were age over 48 years (OR, 2.57; 95% CI, 1 42-4.67; P = .002) and fever (OR, 4.73; 95% CI, 1.99-11.27; P = .001). CONCLUSION: Patients under the age of 48 years without comorbidity and with normal vitals can be discharged from the ED without fear of complications. A history of rheumatologic disease, fever, digestive symptoms, and a respiratory rate over 20 breaths per minute, or a need for corticosteroid therapy were independently associated with revisits. Fever and age over 48 years were associated with a need for hospitalization.

OBJETIVO: Analizar las revisitas y los factores asociados a la misma en pacientes con diagnóstico de posible COVID-19 dados de alta de un servicio de urgencias hospitalario (SUH). METODO: Estudio observacional, retrospectivo que incluyó pacientes consecutivos que consultaron al SUH en un periodo de 2 meses y fueron diagnosticados de posible de COVID-19. Se analizaron variables clínico-epidemiológicas, tratamiento administrado en urgencias, destino final, revisita al SUH y motivo de esta. Se hizo un análisis comparativo entre ambos grupos (revisita sí/no) y se identificaron factores asociados a la revisita. RESULTADOS: Se incluyeron 2.378 pacientes (edad media 57 años; 49% mujeres). De los pacientes dados de alta (39% del total; n = 925), 170 (20,5%) reconsultaron al SUH, principalmente por persistencia o progresión de síntomas, y 66 (38,8%) precisaron ingreso. Los factores relacionados con la revisita fueron: antecedentes de enfermedad reumatológica [OR: 2,97 (IC 95%: 1,10-7,99, p = 0,03)], síntomas digestivos [OR: 1,73 (IC 95%: 1,14-2,63, p = 0,01)], frecuencia respiratoria $ 20 [OR: 1,03 (IC 95%: 1,0-1,06, p = 0,05)] y haber recibido tratamiento con esteroides en urgencias [OR: 7,78 (IC 95%: 1,77-14,21, p = 0,01)]. Los factores asociados al ingreso en la revisita fueron la edad $ 48 años [OR: 2,57 (IC 95%: 1,42-4,67, p = 0,002)] y presentar fiebre [OR: 4,73 (IC 95%: 1,99-11,27, p = 0,001)]. CONCLUSIONES: Los pacientes con posible COVID-19 menores de 48 años, sin comorbilidad y con signos vitales normales podrían ser dados de alta desde urgencias sin temor a sufrir complicaciones. Los antecedentes de enfermedad reumatológica, fiebre, sintomas digestivos, frecuencia respiratoria $ 20/min o necesidad de tratamiento con esteroides fueron factores independientes de revisita, y la fiebre y edad $ 48 años de necesidad de ingreso.

Improving home oxygen therapy in patients with interstitial lung diseases: application of a noninvasive ventilation device.

BACKGROUND: Proper adjustment of arterial oxygen saturation (SaO2) during daily activities in patients with interstitial lung disease (ILD) requiring long-term oxygen therapy is challenging. Given the multifactorial nature of the limited exercise tolerance in patients with ILDs, the isolated use of oxygen therapy may not be enough. As demonstrated previously in patients with chronic obstructive pulmonary disease, the use of a noninvasive ventilation (NIV) device combined with oxygen therapy may prevent the falling of oxygen saturation during exercise, due to an improvement of the ventilation-perfusion ratio and a reduction of the respiratory work, thus enhancing exercise tolerance. We sought to assess in patients diagnosed with ILD who are in need of oxygen therapy, the effect of associating an NIV to improve oxygen parameters and the distance covered during the 6 min walking test (6MWT). METHODS: We conducted a prospective observational study in patients with ILDs. After a clinical characterization, we performed a 6MWT in two different situations: using a portable oxygen concentrator with the regular flow used by the patient during their daily life activities and afterwards adding the additional support of a NIV. The oxygen saturation parameters were registered with a portable oximeter. RESULTS: We included 16 patients with different ILDs who have oxygen therapy prescribed. The use of NIV associated with oxygen therapy in comparison with the use of oxygen therapy alone showed an increase of the average SaO2 [91% (88-93) versus 88% (86-90%); p = 0.0005] and a decrease in the percentage of time with oxygen saturation <90%: 36% (6-56%) versus 58% (36-77%); p < 0.0001. There were no changes in the 6MWT distance: 307 m (222-419 m) versus 316 m (228-425 m); p = 0.10. CONCLUSIONS: In our study the use of a NIV system associated with long-term oxygen therapy during exercise showed beneficial effects, especially improvement of oxygen saturation.The reviews of this paper are available via the supplemental material section.

Impact of a systematic evaluation of connective tissue disease on diagnosis approach in patients with interstitial lung diseases.

To date, there is no clear agreement regarding which is the best method to detect a connective tissue disease (CTD) during the initial diagnosis of interstitial lung diseases (ILD). The aim of our study was to explore the impact of a systematic diagnostic strategy to detect CTD-associated ILD (CTD-ILD) in clinical practice, and to clarify the significance of interstitial pneumonia with autoimmune features (IPAF) diagnosis in ILD patients.Consecutive patients evaluated in an ILD Diagnostic Program were divided in 3 groups: IPAF, CTD-ILD, and other ILD forms. Clinical characteristics, exhaustive serologic testing, high resolution computed tomography (HRCT) images, lung biopsy specimens, and follow-up were prospectively collected and analyzed.Among 139 patients with ILD, CTD was present in 21 (15.1%), 24 (17.3%) fulfilled IPAF criteria, and 94 (67.6%) were classified as other ILD forms. Specific systemic autoimmune symptoms such as Raynaud phenomenon (19%), inflammatory arthropathy (66.7%), and skin manifestations (38.1%) were more frequent in CTD-ILD patients than in the other groups (all P < .001). Among autoantibodies, antinuclear antibody was the most frequently found in IPAF (42%), and CTD-ILD (40%) (P = .04). Nonspecific interstitial pneumonia, detected by HRCT scan, was the most frequently seen pattern in patients with IPAF (63.5%), or CTD-ILD (57.1%) (P < .001). In multivariate analysis, a suggestive radiological pattern by HRCT scan (odds ratio [OR] 15.1, 95% confidence interval [CI] 4.7-48.3, P < .001) was the strongest independent predictor of CTD-ILD or IPAF, followed by the presence of clinical features (OR 14.6, 95% CI 4.3-49.5, P < .001), and serological features (OR 12.4, 95% CI 3.5-44.0, P < .001).This systematic diagnostic strategy was useful in discriminating an underlying CTD in patients with ILD. The defined criteria for IPAF are fulfilled by a considerable proportion of patients referred for ILD.

Acute exacerbations of idiopathic pulmonary fibrosis: Does clinical stratification or steroid treatment matter?

Acute exacerbation (AE) of idiopathic pulmonary fibrosis (IPF) is defined as a sudden acceleration of the disease with the appearance of pulmonary infiltrates superimposed on the characteristic pattern of IPF that leads to a significant decline in lung function. It has high in-hospital mortality rates, despite medical treatment with systematic steroids. We sought to investigate whether there were in-hospital mortality differences according to clinical stratification (AE, suspected AE, or AE of known cause) and/or treatment with systemic steroids. We reviewed the clinical characteristics and outcomes of patients with IPF admitted to our hospital during the years 2003-2014 due to a worsening of their clinical status. We identified 50 IPF patients, 9 with AE (18%), 12 with suspected exacerbation (24%), and 29 with AE of known cause (58%), mostly respiratory infections. In-hospital mortality was similar in the three groups (33% vs. 17% vs. 34%, respectively). Likewise, we did not find differences between them with respect to the use of systemic steroids (length of treatment duration or total dose). Nevertheless, there was an independent association between in-hospital mortality and high average daily steroid dose. We did not observe significant differences in prognosis or use of systemic steroids according to current diagnostic stratification groups in patients hospitalized because of an exacerbation of IPF.

Characterization, localization and comparison of c-Kit+ lung cells in never smokers and smokers with and without COPD.

BACKGROUND: c-Kit + lung stem cells have been described in the human healthy lung. Their potential relation with smoking and/or chronic obstructive pulmonary disease (COPD) is unknown. METHODS: We characterized and compared c-Kit+ cells in lung tissue of 12 never smokers (NS), 15 smokers with normal spirometry (S) and 44 COPD patients who required lung resectional surgery. Flow cytometry (FACS) was used to characterize c-Kit+ cells in fresh lung tissue disaggregates, and immunofluorescence (IF) for further characterization and to determine their location in OCT- embedded lung tissue. RESULTS: We identified 4 c-Kit+ cell populations, with similar proportions in NS, S and COPD: (1) By FACS, c-Kithigh/CD45+ cells (4.03 ± 2.97% (NS), 3.96 ± 5.30% (S), and 5.20 ± 3.44% (COPD)). By IF, these cells were tryptase+ (hence, mast cells) and located around the airways; (2) By IF, c-Kitlow/CD45+/triptase- (0.07 ± 0.06 (NS), 0.03 ± 0.02 (S), and 0.06 ± 0.07 (COPD) cells/field), which likely correspond to innate lymphoid cells; (3) By FACS, c-Kitlow/CD45-/CD34+ (0.95 ± 0.84% (NS), 1.14 ± 0.94% (S) and 0.95 ± 1.38% (COPD)). By IF these cells were c-Kitlow/CD45-/CD31+, suggesting an endothelial lineage, and were predominantly located in the alveolar wall; and, (4) by FACS, an infrequent c-Kitlow/CD45-/CD34- population (0.09 ± 0.14% (NS), 0.08 ± 0.09% (S) and 0.08 ± 0.11% (COPD)) compatible with a putative lung stem cell population. Yet, IF failed to detect them and we could not isolate or grow them, thus questioning the existence of c-Kit+ lung stem-cells. CONCLUSIONS: The adult human lung contains a mixture of c-Kit+ cells, unlikely to be lung stem cells, which are independent of smoking status and/or presence of COPD.

Auscultation of Velcro Crackles is Associated With Usual Interstitial Pneumonia.

Auscultation of Velcro crackles has been proposed as a key finding in physical lung examination in patients with interstitial lung diseases (ILDs), especially in idiopathic pulmonary fibrosis (IPF). However, no studies have been carried out to assess the association of Velcro crackles with other clinical variables.We evaluated a cohort of 132 patients, prospectively and consecutively included in our ILD diagnostic program at a tertiary referral center. All patients were auscultated during the physical examination. The patients were divided into 2 groups: "presence" or "nonpresence" of bilateral Velcro crackles.Of all patients assessed, 83 (63%) presented Velcro crackles in the respiratory auscultation. Patients with Velcro crackles usually had more frequently cough and dyspnea at the moment of diagnosis. Forced vital capacity (P = 0.002) and lung diffusion capacity for carbon monoxide (P = 0.04) was lower in these patients. The ILD-GAP index was higher in the group with Velcro crackles (P = 0.01). All patients with usual interstitial pneumonia (UIP) in high-resolution computed tomography and all patients with final IPF diagnosis presented Velcro crackles. In multivariate analysis, the presence of Velcro crackles was independently associated with an UIP pattern.In patients suspected of having ILD, the auscultation of Velcro crackles was associated with UIP, a possibility which must be taken into consideration in early ILD detection in primary care.

Aspiration pneumonia in old patients with dementia. Prognostic factors of mortality.

BACKGROUND: Prognostic factors of mortality in elderly patients with dementia with aspiration pneumonia (AP) are scarcely known. We determined the mortality rate and prognostic factors in old patients with dementia hospitalized due to AP. METHODS: We prospectively studied 120 consecutive patients aged ≥ 75 years with dementia admitted with AP to two tertiary university hospitals. We collected data on demographic and clinical variables and comorbidities. Oropharyngeal swallowing was assessed by the water swallow test. RESULTS: Sixty-one (50.8%) patients were female, and mean age was 86 ± 9 years. The swallow test was performed in 68 patients, revealing aspiration in 92.6%. Patients with repeat AP (28.3%) were more-frequently taking thickeners (61.8% vs.11.6%, p<0.0001) and were less-frequently prescribed angiotensin-converting-enzyme (ACE) inhibitors (8.8% vs. 27.9%, p<0.001) than patients with a first episode. Hospital mortality was 33.3%; these patients had lower lymphocyte counts and higher percentage of multilobar involvement. In the multivariate model, involvement of ≥ 2 pulmonary lobes was associated with hospital mortality (OR 3.051, 95% CI 1.248 to 7.458, p<0.01). Six-month mortality was 50.8%; these patients were older and had worse functional capacity and laboratory data indicative of malnutrition. In the multivariate model, lower albumin levels were associated with six-month mortality (OR 1.129, 95% CI 1.008 to 1.265, p<0.03). CONCLUSION: In-hospital and 6-month mortality were high (one-third and one-half patients, respectively). Multilobar involvement and lower lymphocyte counts were associated with hospital mortality, and older age, greater dependence and malnutrition with six-month mortality.

[Hyponatremia in elderly patients admitted in an acute geriatric care unit. prevalence and prognosis]. / Hiponatremia en pacientes ancianos ingresados en una Unidad de Agudos de Geriatría. Prevalencia y pronóstico.

BACKGROUND AND OBJECTIVE: In order to analyze the prevalence of hyponatremia in in-hospital elderly patients and its prognosis factor value, we performed a transversal prospective study in an Acute Geriatric Unit (AGU). PATIENTS AND METHODS: Two hundred and sixty in-hospital patients were collected prospectively in an AGU. Sociodemographic and lab data were collected as well as the Barthel, Pfeiffer and Charlson tests. A questionnaire was performed in hospital. Hyponatremia was considered when the plasmatic sodium was ≤134 mmol/l. RESULTS: The sample consisted of 137 women (52.7%) and 123 men (47.3%). Mean age was 83.6 years (SD 7.9). Mean plasmatic sodium values were 137.3 mmol/l (range 112-168). Emergency lab tests showed 60 patients with hyponatremia (23.7%), 35 (13.6%) in the AGU. Cardiopulmonary were the most related diseases. Mean hospital length of stay was 12.8 days (SD 12.8). In-hospital mortality was 12.9%. We observed a statistic relationship between the presence of hyponatremia and a greater mean length of stay (15.53 vs 11.99 days, P=.003). No relationship was found between hyponatremia and mortality (15 vs 12.7%, P=.588). CONCLUSION: Hyponatremia in patients admitted in an AGU is a prevalent disorder related with a greater hospital length of stay, but not with mortality.

Serum high-density lipoprotein cholesterol levels, their relationship with baseline functional and cognitive status, and their utility in predicting mortality in nonagenarians.

AIM: Little is known about the role of high-density lipoprotein cholesterol (HDL-C) in oldest-old subjects. The aim of this study is to evaluate the association between HDL-C levels and physical and cognitive performance indicators in nonagenarians, and also to determine the influence of HDL-C levels on the 3-year mortality risk. METHODS: The data analyzed were taken from the NonaSantfeliu Study. Functional status was determined by the Lawton-Brody Index (LI) for instrumental activities of daily living (IADL) and the Barthel Index (BI) for basic activities (BADL). Cognition was assessed using the Spanish version of the Mini-Mental State Examination (MEC). RESULTS: The sample consisted of 49 women (79%) and 13 men, aged 94.3 ± 2.6 years. Mean HDL-C levels were 60 ± 16 mg/dL, and 16 subjects (25.8%) had low HDL-C values. HDL-C levels did correlate with BI (r = 0.28, P = 0.02) and LI (r = 0.32, P = 0.01), but not with MEC (r = 0.18, P = 0.15). Normal HDL-C levels at baseline were significantly associated with higher BI scores (P < 0.006, odds ratio [OR] = 1.03, 95% confidence interval [CI] = 1.01-1.05) and a lower number of prescription drugs used (P < 0.04, OR = 0.71, 95% CI = 0.49-0.99). Baseline HDL-C levels were significantly lower among the group of nonagenarians who died within the 3 years of follow up (P = 0.02). However, after adjusting for potential confounders, the association between HDL-C and mortality lost significance. CONCLUSION: Higher levels of HDL-C correlate with better functional status and less use of prescribed drugs in nonagenarians. However, the relationship between low HDL-C levels and long-term mortality in this population remains unclear.

Predictors of long-term survival in nonagenarians: the NonaSantfeliu study.

BACKGROUND: few studies have prospectively evaluated long-term predictors of mortality in nonagenarians. OBJECTIVE: to determine predictors of death in a nonagenarian cohort after 5 years of follow-up. DESIGN: a prospective community-based study. SETTING: a community-based study. SUBJECTS: one hundred and eighty-six nonagenarians both living in the community and institutionalised. METHODS: functional status was determined by the Lawton-Brody and Barthel Indexes (BI) and cognition by the Spanish version of the mental state examination (MEC). The Charlson Index was used to measure comorbidity. Nutritional status was evaluated by the short version of the Mini-Nutritional Assessment questionnaire. RESULTS: mortality after 5 years was 75.53%. Patients who did not survive were significantly older, with lower cognitive and functional performance, with diminished visual acuity, higher comorbidity, high risk of malnutrition, higher number of drugs taken and a higher percentage of patients with the diagnosis of dyslipidaemia, heart failure or previous stroke. Cox regression analysis, identified the Charlson Index (hazard ratio 1.23, 95% CI 1.09-1.37) and MEC (hazard ratio 0.98, 95% CI 0.97-0.99) as independent predictors of mortality after 5 years. CONCLUSIONS: better cognitive status and lesser comorbidity at baseline are the best predictors to identify which nonagenarians survived after a 5-year follow-up period.