Clinical evolution of adult, elderly and very elderly patients admitted in Intensive Care Units

This study compared clinical outcomes among adult, elderly and very elderly patients admitted to Intensive Care Units (ICUs) located in São Paulo, Brazil. This retrospective, longitudinal and comparative study included 279 adult (≥18 and <60 years), 216 elderly (≥60 and <80 years) and 105 very elderly (≥80 years) patients. Adult patients differed from other groups regarding the unit to which they were referred and severity, according to the Simplified Acute Physiology Score II. Adults were most frequently sent to hospitalization wards; elderly and very elderly patients who survived hospitalization in critical units showed sharper improvement before discharge. There were differences in relation to mortality between adult and elderly patients, with a higher rate in the elderly group; however, the mortality rate of very elderly and adult patients was similar. In general, the results indicated that older age was not associated with undesirable outcomes in ICUs.


Introduction
Intensive Care Units (ICUs) care for severe and recoverable patients requiring continuous care, offering physical, material and specialized human resources to enable the reversal of disorders that put the lives of patients at risk (1) . These resources have become increasingly sophisticated in recent years and, consequently, increasingly expensive.
The fact that the world population is aging brings with it medical and socioeconomic challenges to governments and societies. As demographic changes occur with increased expectancy of life, the increase of the elderly population is also observed, including within ICUs (2) .
Studies indicate that whatever the health indicator analyzed, results show a greater proportion of disorders among those older than 60 years of age accruing from chronic diseases that should be identified early, treated and monitored to reduce mortality and costs from healthcare (3)(4) .
Even though the aging process is considered to be physiological, it is depicted by a progressive reduction in organic functional reserve. In situations when the body is overloaded, the homeostatic balance is harmed, which makes an elderly individual more susceptible to disorders and diseases (3,5) . Additionally, elderly patients' responses to treatment, when compared to younger patients, may be harmed given a low physiological reserve, inherent to the aging process.
Therefore, identifying how these individuals, increasingly more expressive in the population, presenting distinct physiological characteristics and at a greater risk of worsening conditions or of death, progresses within ICUs. Such information can show outcomes, the quality of care delivered in these units and enable the supply of the resources necessary for continuing treatment after patients are discharged from the ICU.
Many parameters can be used to characterize this evolution. In this study we chose mortality, readmission to the ICU during the same hospitalization, the patient's condition worsening while in the ICU, and unit to which the patient is referred after discharge from the ICU.
Mortality is a frequent indicator of quality of care in an ICU, which in the analysis process cannot be detached from the severity of the patient's condition (6)(7) .
In this study the following severity indicators were used: Simplified Acute Physiology Score II (SAPS II) and the Logistic Organ Dysfunction Score (LODS) (9) ; both are frequently used in the literature and easy to apply.
Early readmissions are an acknowledged indicator of the quality of intensive care associated with premature discharges from hospitalization units (10) .
The "unit to which the patient is referred" and "evolution of severity of the patient in the ICU" are also important parameters to facilitate the management of services to provide indications on how to proceed with a patient's treatment.
In summary, this study compares the clinical evolution of adult, elderly and very elderly patients hospitalized in an ICU. This proposal intends to seek support to qualify the treatment and structural planning of intensive care under the light of potential specificities of these age groups.

Methods
This is a retrospective-longitudinal study of the comparative type. This study's primary source was the electronic databases of patients hospitalized in four ICUs of two public and private hospitals. The patients included in these databases were monitored from admission to the hospital through discharge in order to obtain data from the first and last 24 hours of hospitalization in these units and identify any readmissions.
The study's settings met the following criteria: medium size, large or extra large hospitals located in the city of São Paulo, SP, Brazil with ICUs and Semiintensive Units with more than five active beds; general ICUs with a number of beds above 6% of the total number of hospital beds (11) . The units caring for specifically pediatric clientele were excluded from this selection.
The sampling size was based on the literature addressing readmissions and mortality in ICUs in the city of São Paulo, SP, Brazil. Aiming to obtain a sample of at least 40 cases in which patients were readmitted, we estimated the need for 400 survivals with a rate of 10% of readmission (12) . The average mortality of 33% observed in ICUs in the city of São Paulo (13)  The severity or Risk of Death (RD) of the analyzed patients was measured by SAPS (8) and LODS (9) . The severity of all patients' conditions was evaluated according to these two indexes and this evolution was estimated by the result of the difference between the RD on the first and last day of hospitalization of the patient in the ICU (RD SAPS II admission -RD SAPS II discharge or death and RD LODS admission -RD LODS discharge or death).
Scores in SAPS II are assigned to 12 physiological variables (heart rate, systolic blood pressure, temperature, blood oxygen pressure divided by the fraction of inspired oxygen, urine output, urea serum, leukocytes, serum potassium, serum sodium, serum bicarbonate, bilirubin and Glasgow Coma Scale) in addition to age, time of admission (scheduled surgery, unscheduled surgery, or medical hospitalization) and chronic disease (AIDS, metastatic carcinoma and hematologic malignancy) (8) .
LODS evaluates RD and the organic function of patients in ICU using physiological variables and identifies 1347 www.eerp.usp.br/rlae Oliveira VCR, Nogueira LS, Andolhe R, Padilha KG, Sousa RMC. from one to three levels of dysfunction for six organic systems: neurological, cardiovascular, renal, pulmonary, hematologic and liver (9) .

In the computation of SAPS II and LODS, each
variable is scored (which varies according to the patient's data) and at the end these scores are summed up and a total score is obtained; the higher the score the greater the condition's severity. The total sum can be translated in RD.
The Chi-square test was used to analyze the nominal variables (mortality, readmission into ICU and the unit where the patient was referred to). This test was initially applied to jointly analyze the three groups. When the results showed statistically significant differences

Results
Age among the 600 studied patients ranged from 18 to 97 years old and when the groups of patients 60 years old or older were totaled we found out that most  According to what is observed in Table 1, there was no significant difference among the age groups in relation to readmission into an ICU and the incidence of these readmissions was very similar among the groups.  were referred to other places (ten to other hospitals and one was sent home). More than half of patients in all age groups were sent to the semi-intensive unit.
Data presented in Table 2 shows a significant difference among age groups and the unit to which patients were sent after being discharged from the ICU (p=0.00). The comparative analysis of pairs of groups through the Chi-square test showed that adult patients were more frequently sent to the hospitalization unit than elderly (p=0.00) and very elderly patients (p=0.01).

Discussion
The ages of the participants in this study corroborate with other studies that indicate a predominantly elderly clientele in ICUs (14)(15) . Expectancy of life in Brazil has progressively increased (4,16) and showed that the average age of patients significantly increased in the period (17) .
This study showed that elderly patients present a greater incidence of death (25.5%) in relation to adults (15.8%) and very elderly (20.0%) patients though a significant difference was observed only between adults and elderly patients. One study that analyzed the rate of mortality between two groups, case (>90 years elderly) and control (20-69 years elderly), corroborate this observation revealing that there was no significant difference between these groups (18) . These results suggest that extreme ages do not differ in relation to mortality in ICUs. Additionally, other studies have shown that mortality is not isolated associated with age but also with the clinical conditions of the patients and associated factors such as: severity of acute dysfunction, comorbidities, and the patients' functional condition before being admitted in ICU (19)(20) .
A total of 9.1% of patients from the studied sample were readmitted in ICU during the same hospitalization and no differences were found among the groups in relation to this variable. The percentage of readmission found in this study was similar to the study that investigated 4,684 patients admitted in a clinical ICU during five years (9.6%) (21) and also close to that observed in a clinical ICU in the city of São Paulo, SP, Brazil 10.7% (12) .
Various factors can contribute to the readmission of patients in the ICU and among them the following stand out: evolution of the disease, inappropriate care delivered to patients in hospitalization units and inappropriately early discharges. Age was not a factor associated with readmission in ICU in this study; 64.8% of the survivors were sent to the semi-intensive unit and 32.9% to the hospitalization unit after being discharged from ICU.
In the comparative analysis of groups, the adults differed from the elderly patients and the very elderly patients in relation to the unit to where they were refereed to, while the percentage of hospitalization in the semi-intensive unit was higher in the last two groups. This higher percentage is probably associated with a greater need of surveillance among these groups as a consequence of the aging process itself.
One study that analyzed patients older than 60 years old admitted in the ICU of a hospital without The scores obtained in LODS at admission increased as age increased, similar to SAPS II. The aging process may be seen as a dynamic and progressive process; functional changes that follow morphological and structural changes progressively interfere in the body and make it become susceptible to intrinsic and extrinsic aggression (5) .  (20,23) . However, it is important to consider that the elderly and very elderly patients were admitted to ICUs with an indication of greater RD and required a sharper improvement in order to achieve a level of physiological stability that allowed being discharged from ICU.
In this context, it is worth noting that the average time of hospitalization in an ICU of the three groups was between the 8 and 9 days and older ages were not associated with prolonged stays. Therefore, the results indicate that advanced age was not an essential factor for undesirable outcomes of intensive care. The low physiological reserve of elderly patients probably worked in concert with the greater severity of their condition at the time they were admitted into an ICU, however the resources and quality of care provided in these units compensated for the physiological imbalances that were initially observed. LODS and SAPS II, shows that with the advance of age, people tend to significantly benefit from intensive care.