The global population is currently experiencing a decrease in fertility rates alongside increased life expectancy, leading to a rapid aging demographic. A significant aspect of this shift is the growth in the population of very elderly individuals, with projections indicating that by 2030, the number of nonagenarians will surpass 30 million (1). Due to this demographic development and advancements in providing medical care, the proportion of very elderly patients admitted to the intensive care unit (ICU) is expected to rise substantially. Presently, approximately 15% of critically ill patients in ICU present at an advanced age of 80 years and above; 1% was observed to be beyond 90 years of age (2, 3). While studies have reported acceptable outcomes among this age group, debates on limitation of therapy and futility of care in this age group remain subject of considerable debate (3, 4). Nonetheless, age alone should be a determining factor in withholding critical care (3).

The median length of ICU stay in nonagenarians ranges from 1.4 to 6.1 days, depending on the patient groups studied, however, a majority can be discharged within 7 days or less (3, 5, 6). In contrast, a prolonged ICU stay is generally associated with increased mortality and unfavourable functional outcomes (7–9). With continuous advancements in medical treatments, it is anticipated that an increasing number of patients will experience extended ICU stays, including those at advanced age (10). This will pose challenges in decision-making concerning treatment intensity and futility of care, particularly in cases with unanticipated outcomes. Among the general ICU population, age was identified as relevant risk factor in patients with very prolonged ICU stays (11–13). Notably, prolonged mechanical ventilation (MV) as well as length of ICU stay (LOS) are associated with impaired long-term survival and unfavourable quality of life, especially in the very elderly (7–9, 14–16). Additionally, functional outcomes post-prolonged critical illness tend to be poorer and often persistent, necessitating careful consideration in the management of these cases (15). Despite these concerns, there remains a notable scarcity of studies on the characteristics and outcomes of nonagenarians, specifically in those with an extended ICU stay.

Therefore, the primary objective of this study was to investigate the incidence of prolonged ICU stay (≥7 days). Further clinical characteristics, risk factors, and outcomes to identify patients at risk for a prolonged intensive care therapy will be investigated.

In this study of patients ≥90 years admitted to the ICU, we found that 10% of the patients required a prolonged ICU stay (≥7 days) which was associated with a significantly higher short- and long-term mortality. To our knowledge, this is the first study investigating occurrence, outcome and factors associated with presence of prolonged ICU stay among very elderly critically ill patients.

Across all ages, the average ICU length of stay usually ranges between 1.4 to 6.1 days, with the majority of patients requiring a stay below 7 days (3, 5, 6, 21). Additionally, for elderly patients not requiring critical care, prolonged hospital stays have been reported to be associated with an increased rate of complications and unfavourable outcome, including hospital re-admission, mortality, as well as deterioration of functional status (22). In their paper on the syndrome of the “chronically critically ill,” Girard and Raffin (23) described a patient group surviving an initial episode of critical illness, however, failing to recover beyond the point of becoming independent of critical care (“neither dying nor recovering”). Respiratory failure and prolonged dependence on invasive ventilation constitute the hallmarks of chronic critical illness. It is estimated that between five to 10% of patients who undergo invasive ventilation initially will eventually become part of this chronically critically ill group (11, 24). It appears reasonable to expected that this proportion will be significantly higher the group of nonagenarians, especially in those with a prolonged ICU stay.

The population of patients with a prolonged stay in the ICU is consuming a high number of ICU resources (25–29). This issue takes on particular significance in the context of limited bed capacity, which is often exacerbated by constraints in nursing staff availability. Furthermore, while most published studies report to be favourable cost-efficiency ratios (less than $50,000 per quality-adjusted life year), a review by Wilcox et al. (30) found that 17% (n = 20) of studies indicated cost-efficiency-ratios surpassing $100,000 per quality adjusted life year.

Recent trends in population aging, along with advancements in critical care management, have been identified as key factors contributing to the increasing number of patients experiencing prolonged ICU stays (1, 10, 11). However, due to the large heterogeneity within this patient group, a universally accepted definition of “prolonged ICU stay” has yet to be established. In the distinct cohort of nonagenarians investigated in this study, an ICU stay beyond 7 days was observed in 10%. In those, a clear association with an underlying medical condition could be established, affecting approximately half of them.

In this study, the overall severity of illness was low. Nonetheless, patients admitted with a higher SOFA and SAPS II scores were at increased risk for prolonged ICU stays. Notably, during their extended time in the ICU, more than 80% of patients required vasopressor support. It is important to note, though, that this high proportion may be influenced by the inclusion of patients suffering cardiac arrest, with potential high rates of post-cardiac arrest shock or heart failure (31, 32).

Interestingly, we observed a high rate of hypoxic liver injury and cholestasis in patients with a prolonged ICU stay. This might be attributed to the common co-occurrence of hypoxic liver injury in severe illness and might be observed in settings of cardiac, circulatory, or respiratory failure (33, 34).

Acute kidney injury frequently complicates ICU stay, often mirroring severity of illness (35). Indeed, 10% of patients in the prolonged stay group required renal replacement therapy. The occurrence of organ failure was notably influential in the development of prolonged ICU stays in this cohort.

Due to the growing complexity of medical interventions, even in advanced age (such as trans-femoral aortic valve implantation, TAVI), along with the management of ensuing complications, as well as advancements in intensive care medicine, it is anticipated that a greater number of patients will experience prolonged lengths of stay in the ICU (10). This is accompanied with challenges in decision making on the treatment intensity and futility of care. Physical function and cognitive status following (prolonged) critical illness are frequently poorer and more enduring than initially anticipated: Studies reveal that less than 12% of chronically critically ill patients are alive and independent a year after acute illness. This situation places heavy burdens on families, leading to high rates of depression and caregiver burnout. Notably, the decline in caregivers’ physical health and the increased stress are prevalent not just in home care settings but are often at least as severe for families of institutionalized patients (11, 15).

Recent research underscores the significant impact of frailty on mortality among critically ill, very elderly patients (36, 37). Consequently, frailty should be a critical consideration in ICU treatment discussions. Although frailty was not directly assessed in our cohort, we noted that 61% of patients lived independently prior to their ICU admission, while the remainder resided in nursing homes or assisted living facilities, suggesting they might be more vulnerable and presumably frailer. Contrary to expectations, a recent study found the most significant impact of frailty on mortality risk in middle-aged patients, rather than in the very elderly. Consequently, frailty should not be the sole determinant in excluding very elderly patients from critical care therapy, highlighting the need for a more nuanced approach in decision-making (38).

Although, we observed an increased ICU- and 1-year mortality in patients with a prolonged ICU stay, survival rates were acceptable considering the advanced patient age. However, the clinical situation and potential deteriorations should be perceived early and treatment intensity should be adapted according patients and family wishes to optimise compassionate care while avoiding futile life-sustaining interventions.

This study has several limitations: First, we show the results of a retrospective single centre study alongside with its limitations. However, data were documented prospectively by trained staff. Second, we show results of a large tertiary care centre experienced in the management of critically ill patients. Therefore, results and conclusions may not be transferable to other settings. Third, in the logistic regression model we applied a stepwise selection method which can produce an unstable model and biased estimates, particularly in datasets with a limited number of events. However, we used a stepwise selection with the intend to explore associations in an observational dataset rather than to infer causality. Fourth, we acknowledge that some exposures are time-dependent and that competing risks may influence observed associations. The timing of exposure measurement—either at ICU admission or during the ICU stay—may limit causal interpretation and should be considered when applying these findings clinically. Fifth, residual confounding from unmeasured covariables is a matter of concern and cannot be entirely excluded.

Our study demonstrated that prolonged ICU stays of nonagenarians are not uncommon and are associated with little long-term benefit, which, however, may be considered acceptable given the advanced age of the patients. Intra ICU markers associated with prolonged ICU stays included mechanical ventilation, vasopressor therapy, and renal replacement therapy. These findings suggest that prolonged ICU therapy for critically ill nonagenarians can be beneficial and should generally not be pre-emptively withheld. Due to limited physical reserves in this population, patients’ propensity to recover is low even when treated for a relatively short period of time in the ICU. Nevertheless, decisions regarding such therapy must be tailored, taking into account individual patient factors (e.g., individual frailty, premorbid function, and goals of care) for a more personalised approach, also integrating the expertise and realistic judgment of experienced medical professionals.