Critically Ill Elderly Adults With Infection
Critically Ill Elderly Adults With Infection
The EPIC II study was conducted to provide a worldwide snapshot of the frequency of infection, along with the patterns of infection of individuals in the ICU. This analysis of the database compared older and younger individuals in the ICU with infection in terms of severity, comorbidities, and patterns of infection. By including a follow-up period and assessing outcome at ICU and hospital discharge, it was also possible to compare survival rates between groups. This study highlights the large number of elderly adults with infections in ICUs throughout the world and also that individuals aged 85 and older represent an important proportion of the ICU population with infection (4.7%). These results are similar to those reported previously in ICUs in the United States; 43% were aged 65 to 84, and 4% were aged 85 and older. Projections by the U.S. Census Bureau suggest that the population aged 85 and older is expected to grow from approximately 4 million in 2000 to 19 million by 2050. In the United Kingdom, life expectancy at age 80 increased from 5.8 years in 1981 to 7.2 years in 2002 for men and from 7.5 to 8.7 years for women, and similar data have been reported in other European countries and in Japan. Hence, the number of individuals aged 85 and older in the ICU will continue to increase, supporting the need for more information about their specific characteristics.
The results of this analysis are similar to those from previous reports showing that advanced age is associated with greater mortality in hospitalized individuals with infection. One study reported that, in mechanically ventilated individuals with infection as the principal diagnosis, hospital mortality for those aged 80 to 84 was 85%, which is much higher than the 32% ICU mortality and 45% hospital mortality in the current study. Another study reported hospital mortality of 53% in individuals in the ICU aged 75 and older with nosocomial bacteremia, and a third study reported that, in critically ill individuals with community-acquired pneumonia, mortality increased with age (<60, 19%; 60–69, 22%; 70–79, 48%; and ≥80, 57%) and that increasing age was independently associated with risk-adjusted short- and long-term mortality.
Some authors have suggested that it is not age per se, but associated factors, such as severity of illness and premorbid functional status, that are responsible for the worse outcome, but a recent study that compared the effect of advanced age per se with that of severity of chronic and acute diseases on the short- and long-term survival of older adults admitted to the ICU, concluded that advanced age should be regarded as a significant independent risk factor for mortality, especially for individuals in the ICU aged 75 and older. The current study of individuals with infection in the ICU found that severity of illness, as assessed according to SOFA and SAPS II (without the age component) scores, was similar in individuals aged 85 and older and other participants. In unadjusted and adjusted multivariate analyses, with the group aged 85 and older as the reference category, age was an independent risk factor for ICU and hospital mortality, with hazard ratios for mortality gradually increasing with advancing age, results that are in accordance with a previous study.
Previous studies have reported that age has no or minimal effect on duration of ICU and hospital stay, but data are limited and from studies that did not focus just on critically ill individuals with infection. In the current analysis, median ICU and hospital LOS were shorter in elderly adults, but when analyzing only survivors, the differences lost significance, except when comparing participants aged 85 and older with those aged 65 to 74. A possible explanation for why individuals aged 85 and older had a lower rate of renal replacement therapy is that they received less-aggressive supportive therapy, a suggestion that is in accordance with the results of a study in which critically ill elderly adults received less therapy than their younger counterparts. Participants aged 85 and older were also more likely to decide to withdraw or withhold life-supporting therapy before death than their younger counterparts, although without further data regarding end-of-life goals and the decision-making process, these results are difficult to interpret.
A pathogen was less likely to be isolated in individuals aged 85 and older with infection in the ICU than in younger participants. One possible explanation is a less-aggressive approach to diagnosis for individuals aged 85 and older. gram-negative microorganisms were more frequently isolated in individuals aged 85 and older than in other participants, information that should be taken into account in the management of this group of critically ill individuals with infection. Elderly adults residing in nursing homes or long-term care facilities or receiving home care have high rates of colonization with potentially resistant gram-negative microorganisms. A higher proportion of healthcare-associated infections (more individuals from nursing homes and long-term care facilities or receiving home care) may explain the larger numbers of gram-negative isolates in participants aged 85 and older in the current study, but this information was not collected in EPIC II, so it was not possible to investigate this hypothesis further.
Bloodstream infections were less frequent in participants aged 85 and older than in the other participants, a result that is in accordance with the findings of a previous study that reported a lower incidence of nosocomial bloodstream infections in older adults (<65, 8.4/1,000 patient-days; 65–75, 5.5/1,000 patient-days; ≥ 75, 4.6/1,000 patient-days; P< .001), although it cannot be excluded that the lower rate of bloodstream infections in individuals aged 85 and older may simply reflect a less-frequent rate of blood culture sampling in this group. Intravascular catheter–related infections were also less frequent in individuals aged 85 and older than in all other groups except for participants aged 18 to 44.
An important strength of the current analysis is the large number of critically ill individuals with infection included; to the best of the authors' knowledge, this represents the largest cohort on this subject, including an international and diverse population of individuals in the ICU. Another strength is that the comparison between elderly and younger adults reflects the whole population of individuals with infection in the ICU and not just specific subgroups, as was the case in several previous studies that examined individuals with specific infections (e.g., nosocomial bacteremia or severe community-acquired pneumonia). Finally, not only elderly and younger age groups, but also subgroups of elderly adults were compared, focusing on those aged 85 and older, which, although it is an increasingly important group of individuals with infection in the ICU, has been poorly studied.
A limitation of the current analysis is that EPIC II was a point-prevalence study, with all the inherent drawbacks of such studies, such as inclusion of a large number of individuals and the concomitant "cost" of obtaining a limited data set. For example, there are no data on the percentages of infections that were ICU acquired, hospital acquired outside the ICU, healthcare associated (e.g., nursing home residents), or community acquired. In addition, because of the high proportion of university hospitals among the participating centers, the participant case-mix may not be representative of community hospitals and other models of ICU care. Moreover, a point-prevalence study gives a snapshot of events at a particular point in time (index date). Therefore, individuals with medical conditions who require shorter ICU stays or conditions with higher mortality may be underrepresented, although unless there is something unique about the index date, EPIC II participants should be a representative sample of the ICU population (individuals with short and longer stays, low vs high mortality). Finally, the facts that there are limited studies on critically ill elderly adults and that the studies that are available mostly consider the whole ICU population or just individuals with specific infections and the different cutoffs used to define "elderly" make comparisons with previous studies difficult. Furthermore, the small absolute percentage differences in some of the analyses make it difficult to assess the clinical effect of these findings. Further well-designed prospective studies focusing on elderly adults with infection in the ICU are needed to determine in detail the patterns of infection in elderly adults, to analyze community-acquired and hospital-acquired infections separately, and to identify possible modifiable risk factors that may affect outcomes.
Discussion
The EPIC II study was conducted to provide a worldwide snapshot of the frequency of infection, along with the patterns of infection of individuals in the ICU. This analysis of the database compared older and younger individuals in the ICU with infection in terms of severity, comorbidities, and patterns of infection. By including a follow-up period and assessing outcome at ICU and hospital discharge, it was also possible to compare survival rates between groups. This study highlights the large number of elderly adults with infections in ICUs throughout the world and also that individuals aged 85 and older represent an important proportion of the ICU population with infection (4.7%). These results are similar to those reported previously in ICUs in the United States; 43% were aged 65 to 84, and 4% were aged 85 and older. Projections by the U.S. Census Bureau suggest that the population aged 85 and older is expected to grow from approximately 4 million in 2000 to 19 million by 2050. In the United Kingdom, life expectancy at age 80 increased from 5.8 years in 1981 to 7.2 years in 2002 for men and from 7.5 to 8.7 years for women, and similar data have been reported in other European countries and in Japan. Hence, the number of individuals aged 85 and older in the ICU will continue to increase, supporting the need for more information about their specific characteristics.
The results of this analysis are similar to those from previous reports showing that advanced age is associated with greater mortality in hospitalized individuals with infection. One study reported that, in mechanically ventilated individuals with infection as the principal diagnosis, hospital mortality for those aged 80 to 84 was 85%, which is much higher than the 32% ICU mortality and 45% hospital mortality in the current study. Another study reported hospital mortality of 53% in individuals in the ICU aged 75 and older with nosocomial bacteremia, and a third study reported that, in critically ill individuals with community-acquired pneumonia, mortality increased with age (<60, 19%; 60–69, 22%; 70–79, 48%; and ≥80, 57%) and that increasing age was independently associated with risk-adjusted short- and long-term mortality.
Some authors have suggested that it is not age per se, but associated factors, such as severity of illness and premorbid functional status, that are responsible for the worse outcome, but a recent study that compared the effect of advanced age per se with that of severity of chronic and acute diseases on the short- and long-term survival of older adults admitted to the ICU, concluded that advanced age should be regarded as a significant independent risk factor for mortality, especially for individuals in the ICU aged 75 and older. The current study of individuals with infection in the ICU found that severity of illness, as assessed according to SOFA and SAPS II (without the age component) scores, was similar in individuals aged 85 and older and other participants. In unadjusted and adjusted multivariate analyses, with the group aged 85 and older as the reference category, age was an independent risk factor for ICU and hospital mortality, with hazard ratios for mortality gradually increasing with advancing age, results that are in accordance with a previous study.
Previous studies have reported that age has no or minimal effect on duration of ICU and hospital stay, but data are limited and from studies that did not focus just on critically ill individuals with infection. In the current analysis, median ICU and hospital LOS were shorter in elderly adults, but when analyzing only survivors, the differences lost significance, except when comparing participants aged 85 and older with those aged 65 to 74. A possible explanation for why individuals aged 85 and older had a lower rate of renal replacement therapy is that they received less-aggressive supportive therapy, a suggestion that is in accordance with the results of a study in which critically ill elderly adults received less therapy than their younger counterparts. Participants aged 85 and older were also more likely to decide to withdraw or withhold life-supporting therapy before death than their younger counterparts, although without further data regarding end-of-life goals and the decision-making process, these results are difficult to interpret.
A pathogen was less likely to be isolated in individuals aged 85 and older with infection in the ICU than in younger participants. One possible explanation is a less-aggressive approach to diagnosis for individuals aged 85 and older. gram-negative microorganisms were more frequently isolated in individuals aged 85 and older than in other participants, information that should be taken into account in the management of this group of critically ill individuals with infection. Elderly adults residing in nursing homes or long-term care facilities or receiving home care have high rates of colonization with potentially resistant gram-negative microorganisms. A higher proportion of healthcare-associated infections (more individuals from nursing homes and long-term care facilities or receiving home care) may explain the larger numbers of gram-negative isolates in participants aged 85 and older in the current study, but this information was not collected in EPIC II, so it was not possible to investigate this hypothesis further.
Bloodstream infections were less frequent in participants aged 85 and older than in the other participants, a result that is in accordance with the findings of a previous study that reported a lower incidence of nosocomial bloodstream infections in older adults (<65, 8.4/1,000 patient-days; 65–75, 5.5/1,000 patient-days; ≥ 75, 4.6/1,000 patient-days; P< .001), although it cannot be excluded that the lower rate of bloodstream infections in individuals aged 85 and older may simply reflect a less-frequent rate of blood culture sampling in this group. Intravascular catheter–related infections were also less frequent in individuals aged 85 and older than in all other groups except for participants aged 18 to 44.
An important strength of the current analysis is the large number of critically ill individuals with infection included; to the best of the authors' knowledge, this represents the largest cohort on this subject, including an international and diverse population of individuals in the ICU. Another strength is that the comparison between elderly and younger adults reflects the whole population of individuals with infection in the ICU and not just specific subgroups, as was the case in several previous studies that examined individuals with specific infections (e.g., nosocomial bacteremia or severe community-acquired pneumonia). Finally, not only elderly and younger age groups, but also subgroups of elderly adults were compared, focusing on those aged 85 and older, which, although it is an increasingly important group of individuals with infection in the ICU, has been poorly studied.
A limitation of the current analysis is that EPIC II was a point-prevalence study, with all the inherent drawbacks of such studies, such as inclusion of a large number of individuals and the concomitant "cost" of obtaining a limited data set. For example, there are no data on the percentages of infections that were ICU acquired, hospital acquired outside the ICU, healthcare associated (e.g., nursing home residents), or community acquired. In addition, because of the high proportion of university hospitals among the participating centers, the participant case-mix may not be representative of community hospitals and other models of ICU care. Moreover, a point-prevalence study gives a snapshot of events at a particular point in time (index date). Therefore, individuals with medical conditions who require shorter ICU stays or conditions with higher mortality may be underrepresented, although unless there is something unique about the index date, EPIC II participants should be a representative sample of the ICU population (individuals with short and longer stays, low vs high mortality). Finally, the facts that there are limited studies on critically ill elderly adults and that the studies that are available mostly consider the whole ICU population or just individuals with specific infections and the different cutoffs used to define "elderly" make comparisons with previous studies difficult. Furthermore, the small absolute percentage differences in some of the analyses make it difficult to assess the clinical effect of these findings. Further well-designed prospective studies focusing on elderly adults with infection in the ICU are needed to determine in detail the patterns of infection in elderly adults, to analyze community-acquired and hospital-acquired infections separately, and to identify possible modifiable risk factors that may affect outcomes.
Source...