Risk and Prognostic Factors in Very Old Patients with Sepsis Secondary to Community-Acquired Pneumonia

Background: Little is known about risk and prognostic factors in very old patients developing sepsis secondary to community-acquired pneumonia (CAP). Methods: We conducted a retrospective observational study of data prospectively collected at the Hospital Clinic of Barcelona over a 13-year period. Consecutive patients hospitalized with CAP were included if they were very old (≥80 years) and divided into those with and without sepsis for comparison. Sepsis was diagnosed based on the Sepsis-3 criteria. The main clinical outcome was 30-day mortality. Results: Among the 4219 patients hospitalized with CAP during the study period, 1238 (29%) were very old. The prevalence of sepsis in this age group was 71%. Male sex, chronic renal disease, and diabetes mellitus were independent risk factors for sepsis, while antibiotic therapy before admission was independently associated with a lower risk of sepsis. Thirty-day and intensive care unit (ICU) mortality did not differ between patients with and without sepsis. In CAP-sepsis group, chronic renal disease and neurological disease were independent risk factors for 30-day mortality. Conclusion: In very old patients hospitalized with CAP, in-hospital and 1-year mortality rates were increased if they developed sepsis. Antibiotic therapy before hospital admission was associated with a lower risk of sepsis.

In the overall population, 455 patients (37%) were treated with a single antibiotic, with fluoroquinolones and -lactams being the most commonly used monotherapies. We treated 769 patients (63%) with combination therapy, with the most frequent combinations being a -lactam plus a macrolide (360 patients (29%)) or a -lactam plus a fluoroquinolone (297 patients (24%)).

Comparison between Old (65-79 years) and Very Old (≥80 years) Patients
We compared patients aged 65-79 years (the old) and patients aged ≥80 years (the very old). The main characteristics and clinical outcomes by age groups are presented in the online data (Table S1). Focusing on statistically significant differences, we observed that very old patients were less frequently males, current smokers, and alcohol consumers. They were also less likely to have received previous inhaled or systemic corticosteroid, but were more likely to have received influenza vaccination. In addition, the very old group presented a higher proportion of neurological and chronic renal diseases, but a lower proportion of chronic respiratory and liver diseases. The rate of nursing home residents was also higher among very old patients. The uncommon presentation of pneumonia was frequent in very old patients: compared with patients aged 65-79 years, they presented a lower rate of fever, cough, pleuritic pain and a higher frequency of altered mental status at admission. Also, this uncommon presentation of pneumonia was observed in the subgroup of patient with and without sepsis. Very old patients also experienced severe CAP at a higher frequency. Although there was no significant difference in the percentage of sepsis between the two groups, the rates of septic shock and ICU admission were significantly higher in the old group. Notably, inhospital mortality, ICU mortality, 30-day mortality, and 1-year mortality were significantly higher in the very old group compared with the old group.

Comparison of Very Old Patients from Home and Nursing Homes Residents
A total of 227 patients (19%) were nursing home residents. Compared with patients from home, nursing home residents were significantly older and more frequently females, but were less frequently smokers and alcohol consumers. They presented a significantly higher rate of neurological disease, but lower rates of chronic respiratory disease and diabetes mellitus. Severe CAP was more frequent in the nursing home group, but the groups had comparable sepsis presentations. The main characteristics of the two groups, including the main clinical outcomes, are presented in Table S2.

Very Old ICU and Non-ICU Patients
A total of 136 patients (11%) were admitted to ICU during the study period. Compared to those not admitted to ICU, these patients were significantly younger, more frequently male, and less frequently nursing home residents. They presented a significantly lower rate of neurological disease but a higher rate of diabetes mellitus. Severe CAP was more common in ICU patients. The percentages of sepsis and septic shock were significantly higher in the ICU group. Similarly, the length of stay and the in-hospital, 30-day, and 1-year mortality rates were significantly higher in the ICU group. The main characteristics of the ICU and non-ICU groups are presented in Table S3.

Changes in Sepsis Over Time in Very Old Patients
To investigate the changes in sepsis presentation over time, we divided the study population into four cohorts by period: 2005-2007, 2008-2010, 2011-2013, and 2014-2017. This revealed that the prevalence of sepsis in very old patients did not change significantly over time (p = 0.051), with rates ranging from 77% in the first period to 75% in the last period ( Figure S1).

Septic Patients with Septic Shock
Of the 856 patients with sepsis, 61 (7%) developed septic shock (these patients were significantly younger than the group without septic shock). In-hospital mortality, ICU mortality, 30-day, and 1year mortality were significantly higher in the patients with septic shock than in those without septic shock. The main characteristics of these groups are presented in Table S4. Table S4. Characteristics and clinical outcomes of very old patients with and without septic shock.