Geriatric intensive care is already a very important aspect of our world's healthcare concerns, and it is growing exponentially. In this issue of Intensive Care Medicine, we are fortunate to learn from Somme et al. [1] about the outcomes of older patients (considered by most authors to include those 70–75 years and older) from a well-established critical care program in Paris, France. In this editorial, I will summarize what we have learned about the outcomes of older patients who receive intensive care and consider emerging frontiers both in the intensive care unit (ICU) and during recovery that are of particular importance to successful care of older ICU patients.

A few facts will help frame this subject: (1) 60% of all ICU days are incurred by patients over 65 years of age [2]; (2) the per capita number of days per year spent in the ICU is sevenfold higher above age 75 years compared to those less than 65 years [2]; and (3) the incidence of acute respiratory failure treated with mechanical ventilation rises tenfold from the age of 55 years to 85 years [3], resulting in rising numbers of elderly patients in our ICUs [2, 4, 5]. The reality of our aging society dictates that we must focus on how to best care for elders who develop acute critical illness. In caring for an older critically ill patient, healthcare professionals have several options that should be discussed at the outset of this editorial because of heated ongoing debates regarding resource utilization, patient and physician autonomy, futility, compassionate care, and acceptance of the limitations of medical care [6]. First, the medical team must estimate the benefits of intensive care considering baseline disease state, quality of life, and acuity of illness. Simultaneously, it is imperative to determine the patient's preferences (or the surrogate's best estimate of the patient's wishes) regarding mechanical ventilation and other types of life support commonly offered in the ICU. For some older patients, we should provide the most "peaceful and high quality" dying process by either, not admitting the patient to the ICU or planning a short stay in an ICU but avoiding prolonged life support. The other scenario is that the patient, family, and healthcare team conclude that an older patient has the right mix of baseline health, individual preference for aggressive care, and potentially reversible acute illness. In this latter type of older patient, ICU care is indicated just as it would be for a 40-year-old patient.

The current report by Somme et al. [1] is a descriptive study derived from an observational cohort of 410 patients, 75 years and older, admitted to a single medical ICU in Paris from 1991 to 1996. The primary objective of the authors was to assess the effect of age on ICU and long-term mortality. Particular strengths include its relatively large size (227 patients were 80 years or older and 90 patients were 85 years or older), prospective design, and 2-year follow-up from a European cohort. ICU survival rates for those below 75 years, 75–79 years, 80–84 years, and 85 years and older were 80%, 68%, 75%, and 69%, respectively. Most deaths occurred during the first 3 months following ICU discharge, with survival rates at 3-months for those 75–79 years, 80–84 years, and 85 years and older were 54%, 56%, and 51%, respectively. Across age categories, the Kaplan-Meier survival curves were remarkably similar through the first 250 days of follow-up. Using multivariable analysis, Somme's main finding were that: (1) the APACHE II score, which the authors point out is not modified by age above 75 years, was the only independent predictor of ICU mortality; and (2) independent predictors of mortality at 2 years following the ICU admission were age and pre-existing limitations of activity. Future cohort studies such as this should collect and report pertinent variables such as rates of do-not-resuscitate orders, cause of death, disposition to home vs nursing home, repeat hospitalizations, health-related quality of life, and long-term neurocognitive outcomes [7] among survivors.

Other recent reports have yielded new and important data pertinent to the outcomes of older critically ill patients [8, 9, 10, 11 12]. The primary findings of Somme et al. are complementary to those published by Chelluri and colleagues [9] who showed that independent predictors of mortality at 2 months following mechanical ventilation were age, comorbidities, and prehospital functional status. Considering the growing number of older persons receiving mechanical ventilation [13], we have studied the effect of age on ability to recover from respiratory failure. Using a cohort of patients derived from a randomized controlled trial of a weaning protocol versus standard care [14], we observed that older patients, counter to conventional thinking, recovered their ability to oxygenate and ventilate as rapidly as their younger counterparts [8]. Consistent with these findings, a subsequent investigation of the NHLBI-sponsored ARDSnet database of 902 acute lung injury patients found no age-related differences in the proportion of survivors achieving oxygenation and ventilation recovery landmarks, and the median time to pass a 2-h spontaneous breathing trial was similar between older and younger patients (4 days versus 5 days, respectively) [10]. However, after passing a spontaneous breathing trial, older patients took 3 days longer to leave the ICU. In multivariable Cox proportional hazards analysis, age >70 was a strong predictor of in-hospital mortality with a hazard ratio of 2.5 [10]. Considering that older patients initially recovered from their pulmonary insufficiency as rapidly as younger patients, age-related differences in mortality and outcomes during latter stages of recovery may be due to non-pulmonary organ system failure. Herridge et al. [15] showed the importance of non-pulmonary disabilities such as physical deconditioning and wasting during acute lung injury recovery. Delirium, an often overlooked form of organ dysfunction that can now be readily measured at the bedside [16, 17], is an independent predictor of reintubation [18], prolonged hospital stay [19], and mortality [20]. While many patients come into the ICU because of the need for a ventilator due to respiratory failure, on leaving the ICU, we need to focus our attention on other things as well, such as long-term cognitive impairment and neuromuscular disease. Others have suggested that pulmonary aspects of recovery are not the main determinants of the long-term outcome in acute lung injury survivors [21, 22].

Two very distressing aspects of critical illness (in addition to financial considerations) for older survivors are the prospect of losing their independence and the burden that this may impose on their family (e.g., no longer being able to drive, shop, live in their own house, etc.). In an investigation of 879 mechanically ventilated patients with a median age of 65, it was recently shown that 78% of survivors required a caregiver at 2 months following their illness [9]. As a last point of discussion, I would like to review briefly two important aspects of recovery for older patients (though I acknowledge these as potential issues for younger survivors as well):

  1. 1.

    Long-term cognitive impairment. Angus et al. [23] documented that at 6 months the proportion of acute lung injury survivors demonstrating psychiatric, cognitive, and neurologic symptoms was substantial, 57%, 32%, and 44%, respectively. Hopkins et al. [24] assessed cognitive function in acute lung injury survivors and found at 1-year follow-up that 78% of patients experienced at least one domain deficit in neuropsychological function. In a case-control investigation, acute lung injury survivors had more profound neuropsychological deficits at 1-year follow-up than did controls [25], with problems of attention, visual processing, psychomotor speed, and executive function. In another cohort of ICU survivors, up to 25% had deficits at 6 years following their ICU stay [26]. Most recently, using a very conservative method of diagnosing cognitive impairment and after exclusion of patients with potential baseline deficits, Jackson et al. [27] found at 6 months follow-up that at least one in three survivors of mechanical ventilation from a medical ICU had neuropsychological deficits across a variety of cognitive domains equivalent to mild dementia. The report from the "Surviving Intensive Care" 2002 Roundtable Conference held in Brussels [28] stressed the need for further investigations to discover the predictors of long-term cognitive impairment among ICU survivors and propose new treatment strategies.

  2. 2.

    Physical deconditioning and neuromuscular disease. Most data suggest that profound weakness is due not only to physical deconditioning and malnutrition, but also to critical illness polyneuropathy and several types of critical illness myopathy [29]. Herridge et al. [15] recently showed that weakness was among the most important determinants of well-being during the first year of recovery from acute lung injury. These abnormalities of the muscles and nerves develop in up to 80% of patients with sepsis and multiple organ dysfunction and increase complication rates such as prolonged mechanical ventilation [29]. Persistent neuromuscular and neurophysiologic abnormalities are known to be disabling for up to 5 years following critical illness [30]. While no specific therapies are currently available to prevent or treat these problems, future work will need to determine the role of physical rehabilitation during and following the ICU, attention to nutritional supplementation, and early use of treatment strategies aimed at reducing inflammatory and coagulopathic disturbances associated with sepsis and multiple organ failure.

Both the lay public and healthcare professionals are becoming increasingly concerned not only with survival, but also with the quality of patients' lives. As the medical community strives to improve the outcomes of older patients cared for in the ICU, it is imperative that we better understand the epidemiology and risk factors of poor outcomes including mortality, impaired instrumental activities of daily living, and global measures of health-related quality of life. This work by Somme et al. from a European cohort has added to the fabric of our understanding. One out of every two older patients who received mechanical ventilation was discharged and alive 3 months later, and the mortality rates 2 years after leaving the ICU for those 75–79 years, 80–84 years, and 85 years and older were within 5% of one another. It is inevitable that we will care for increasing numbers of older patients in the ICU (even if we screen better for those most likely to benefit), so now is the time to make real strides toward improving not only survival but also quality of survival. I believe that this can be done through a better understanding of (and interventions toward) non-pulmonary comorbidities that limit our patients' ability to return to their previous livelihood and productivity, such as newly acquired neuromuscular or brain disease.