Effects of Sepsis on Morbidity and Mortality in Critically Ill Patients 2 Years After Intensive Care Unit Discharge

Livia Biason, MD; Cassiano Teixeira, MD, PhD; Jaqueline Sangiogo Haas, RN, MS; Cláudia da Rocha Cabral, RN; Gilberto Friedman, MD, PhD

Disclosures

Am J Crit Care. 2019;28(6):424-432. 

In This Article

Results

Of 1219 patients admitted to the ICU during the study period, 34 did not meet the inclusion criteria and 35 met the exclusion criteria, for a total of 1150 patients analyzed in the study. Our sample consisted of 426 patients with a diagnosis of sepsis/septic shock and 724 critically ill patients without sepsis (Figure 1). The characteristics of patients in the 2 groups before and during the ICU stay are listed in Table 1, showing that these populations are very different. Data collected during the 24 hours preceding ICU discharge and during room hospitalization after ICU discharge are presented in Table 2, showing that patients with sepsis remained hospitalized after ICU discharge longer than patients without sepsis.

Figure 1.

Patients enrolled in the study.

During the study period, 859 patients (74.7%) were discharged from the ICU (242 with sepsis and 617 without sepsis), 21 were lost to follow-up (6 with sepsis and 15 without sepsis), and 509 were interviewed (103 with sepsis and 406 without sepsis) after 2 years. Patients with sepsis had a higher mortality rate than did patients without sepsis during the 2 years after ICU discharge (57.4% vs 34.2%, P < .001). Even after adjustment for potential confounders, the Cox regression result was still significant, especially in the first 3 months (41.7% vs 28.4%, P = .004) (Figures 2–4). In the multivariate analysis, sepsis, age, and APACHE II score at ICU discharge were independently associated with mortality (Table 3).

Figure 2.

Survival after admission to the intensive care unit.

Figure 3.

Survival after discharge from the intensive care unit.

Figure 4.

Survival after adjusting for disease severity.

The 2 participating institutions had different rates of patients with and without sepsis. After a site-adjusted analysis, patients with sepsis still had a higher mortality rate than did patients without sepsis (P = .05).

"Patients with sepsis had a higher mortality rate, were more likely to die, and had a greater degree of functional loss than patients without sepsis."

Compared with patients without sepsis, patients with sepsis reported subjectively more pain (48.5% vs 35.2%, P = .003) and were more frequently readmitted to the hospital (65.5% vs 55.0%, P = .02) during the 2 years of follow-up (Table 4). Analysis of the changes in Karnofsky and IADL scores from the time of ICU admission to 2 years after ICU discharge showed that patients with sepsis were less functional both at admission and 2 years after discharge (Figures 5 and 6). However, after adjustment for age, APACHE II score, and Sequential Organ Failure Assessment score at ICU discharge, sepsis remained related to IADL score worsening (P = .03) but only marginally associated with the Karnofsky score (P = .08).

Figure 5.

Change in score on the Karnofsky scale from intensive care unit (ICU) admission to 2 years after ICU discharge in patients with sepsis versus patients without sepsis. For patients with sepsis, the Karnofsky score decreased a mean of 7.8 (SD, 12.9) points; for patients without sepsis, the Karnofsky score decreased a mean of 5.9 (SD, 13.9) points. A Student t test for independent samples yielded a P value of .20 for the difference in scores between patients with and patients without sepsis. Analysis of covariance, with adjustment for baseline scores, yielded a P value of .15 for the difference in scores between patients with and patients without sepsis.

Figure 6.

Change in score on the Lawton instrumental activities of daily living (IADL) scale from intensive care unit (ICU) admission to 2 years after ICU discharge in patients with sepsis versus patients without sepsis. For patients with sepsis, the IADL score decreased a mean of 4.0 (SD, 8.0) points; for patients without sepsis, the IADL score decreased a mean of 3.4 (SD, 8.2) points. A Student t test for independent samples yielded a P value of .52 for the difference in scores between patients with and patients without sepsis. Analysis of covariance, with adjustment for baseline scores, yielded a P value of .31 for the difference in scores between patients with and patients without sepsis.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....