The impact of age on intensive care

Abstract

Caring for the elderly has always been challenging for the intensive care unit (ICU) physician. Concerns like frailty, comorbidities, polypharmacy and advanced directives come up even before admission into the unit. The COVID-19 pandemic has put forward a variety of issues concerning elderly populations, making the topic more relevant than ever. Admittance to the ICU, an unequivocally multifactorial decision, requires special consideration from the side of the physician when caring for an elderly person. Patients’ wishes are to be respected and thus given priority. Triage assessment must also account for age-related physiological alterations and functional status. Once in the ICU, special attention should be given to age-related specificities, such as therapeutic interventions’ controversial role, infection susceptibility, and post-operative care, that could potentially alter the course of hospitalization and affect outcomes. Following ICU discharge, ensuring proper rehabilitation for both survivors and their caregivers can improve long-term outcomes and subsequent quality of life. The pandemic and its implications may limit the standard of care for the elderly requiring ICU support. Socioeconomic factors that further perplex the situation must be addressed. Elderly patients currently represent a vast expanding population in ICU. Tailoring safe treatment plans to match patients’ wishes, and personalized needs will guide critical care for the elderly from this time forward.

1. Introduction

Vulnerability is a public health concept used to describe the need for protection. Nowadays, senior citizens are exposed to multiple challenges and risks that most healthcare policy managers designate them as a vulnerable population from a socioeconomic standpoint. Measures are taken on different layers so as to ensure their safeguarding and well-being. Nonetheless, due to the inescapable ramifications of aging and the inevitable nature of life itself, emerging threats to humankind's health, including novel pathogens, are likely to affect them the most.

A precise threshold for “elderly” has long been a matter of debate since aging is a continuous and multifactorial process influenced by both genetic and environmental parameters. Many studies have used several different arbitrary cutoff points, such as over 60, 65, or 70 years old. However, most prospective epidemiological studies define the “elderly population” as the age group > 65 years (Ferrucci et al., 2008).

Caring for the elderly has always been challenging for the intensive care unit (ICU) physician. Concerns like frailty, comorbidities, polypharmacy and advanced directives come up even before admission into the unit. The pandemic has added further ethical, organizational, and practical burdens to the mix, thereby shifting perspectives of optimal quality of care for the elderly and calling into question some former practices. We aim to provide an outline by highlighting emerging trends and focusing on urgent topics of attention.

2. Discussion

2.1. Before the ICU

2.1.1. Physiological alterations in the elderly

Aging as a medical concept represents a complex transition that includes numerous physiological changes, which can impact treatment plans and overall clinical course in the ICU. When managing a multi-organ disease like COVID-19, for instance, these alterations inadvertently will manifest in the form of either a complication or a lack of response to treatment. It is a well-known fact that the number of comorbidities increases with age. The mean number of comorbidities per patient is 2.6 ± 2.2 versus 3.6 ± 2.3 in patients 65–84 and over 85 years old, respectively, and is associated with an increase in hospitalization rate, higher in-hospital and long-term term mortality (Stavem et al., 2017). Physiological changes, including immunosenescence -a term used to describe age-related immune dysregulation- ( Fig. 1) and alterations in respiratory capacity (loss of lung elasticity, reduced muscle strength, altered sensitivity of respiratory centers to hypoxia and/or hypercapnia) contribute to increased risk of susceptibility to pathogens, viral reactivation, respiratory failure, and finally death (McDermid et al., 2011).

Fig. 1.

Immunosenescence; physiologic changes in the elderly.

On top of that, frail patients have reduced resilience, making their recovery more difficult and prolonging their hospital length of stay (Brummel et al., 2017). Malnutrition also plays a significant role in the frailty of this population affecting 12–45 % of hospitalized older patients and linked to poor outcomes (Guidet et al., 2018a). Similarly, common pre-existing cognitive impairment represents a decisive risk factor for delirium within the ICU, associated with increased mortality and further cognitive decline (Guidet et al., 2018a). On top of that, inappropriately continued antipsychotics to deal with sleep disturbances, agitation, or other psychological symptoms, exacerbated in demented elderly, further contribute to long-term mortality (Guidet et al., 2018a). To this end, the use of these drugs is strongly discouraged by the American Geriatrics Society and NICE (Workgroup AGSCW, 2013), while different criteria have been made available to support physicians in medication evaluation in order to prevent adverse drug events and improve geriatric care in the critically ill (Blanco-Reina et al., 2012).

2.1.2. Triage assessment

Admittance to the ICU, an unequivocally multifactorial decision, in the times of the pandemic has been subject mainly to organizational difficulties, such as prolonged hospitalization times, prioritization of care decision-making, and unavailability of transitional and rehabilitation premises (Foley et al., 2022). Unavoidably, the harmonization of regulatory matters falls within the scope of responsibility of the ICU physician and, to an extent, may affect clinical decision-making. In order to attenuate the influence of non-medically relevant factors, the approach heading into admission should include assessment and quantification of potential benefits of advanced care as well as consideration of the patient’s and family’s wishes. However, the direct involvement of the intensivist in triaging patients-candidates for intensive care may pertain to a certain degree of bias, especially towards extremely old individuals that could have otherwise benefitted from advanced care (Guidet et al., 2018a, Sprung et al., 2012). In recent cohorts, it is remarkable to notice that the long-term mortality in patients considered too sick for an ICU admission was inferior to 100 %, while on the contrary, the mortality was far from 0 % in patients considered too well to be admitted in ICU (Andersen et al., 2017). This fact underlines the possible over- and under-utilization of ICUs. Maximizing survival by itself, though, is not the goal of advanced care and should not drive resource allocation or ICU bed rationing. In the strive against ageism, medically driven evidence, rather than demographics, must serve as indicators. Utilizing multidisciplinary support, including that of social workers and palliative care, is recommended, if not warranted, in most situations (Garrouste-Orgeas et al., 2009).

The Society of Critical Care Medicine (SCCM) guidelines for ICU admission, discharge, and triage recommend that admission decisions for patients older than 80 not be based on their chronological age but rather on their illness severity, comorbidities, baseline functional status, and personal preferences (Nates et al., 2016). In line with SCCM, the World Federation of Societies of Intensive Care Medicine state that ICU triage decisions should never be based solely on a patient's age, and the final decision to admit a patient to the ICU should be made by the ICU physician (Blanch et al., 2016). These guidelines emphasize that while scoring systems may aid decision-making, ultimately, the decision to admit an older patient to the ICU relies on the clinical judgment and experience of the ICU physician, with input from other healthcare team members such as nurses (Blanch et al., 2016).

Frailty, a surrogate index of the multifactorial age-related physiologic decline, is usually interpreted as a measure of organic vulnerability, primarily utilized in quantifying the risk of adverse outcomes and mortality during ICU stay (De Biasio et al., 2020). Frailty is identified in approximately 30–40 % of elderly patients admitted to the ICU (Xia et al., 2021, Guidet et al., 2020). Frailty is associated with an increased risk of short-term mortality (RR: 1.70), in-hospital mortality (RR: 1.73), and long-term mortality (RR: 1.86) (Xia et al., 2021). In line with these findings, a stronger correlation between frailty and hospital length of stay has been observed (Xia et al., 2021). Association with mortality has also been noted in the context of COVID-19-positive patients. In fact, in this subgroup of patients, frailty has been related to increased odds of delirium and reduced odds of intensive care unit admission (Dumitrascu et al., 2021). Even after ICU discharge, frailty assessments have proven helpful in predicting longer-term outcomes. ICU- survivors pre-defined as frail are found to have significantly increased functional decline and score much lower in health-related quality-of-life evaluations compared to their robust counterparts at 6 and 12 months following discharge (Bagshaw et al., 2014).

Frailty could thus be intuitively deployed for ICU admission benefit assessment based on potential risk stratification (Guidet et al., 2018a). In some countries, a high clinical frailty score- calculated by multiple instruments and associated scoring scales (Buta et al., 2016) ( Table 1)- in conjunction with advanced age was considered a relative contra-indication for intubation during the first wave of the pandemic. In order to objectify frailty assessments for this purpose, it is commonly argued that the patient's baseline should be determined in conjunction with their functional status long before ICU admission, ideally in an outpatient setting in a premorbid condition (Piers et al., 2021). Recording frailty in a patient's history is strongly advised, even in an outpatient setting. However, such a premature appraisal fails to account for physiologic reserve deficit in the context of ongoing critical disease and does not incorporate accumulated functional disability secondary to prolonged hospitalization. No frailty-based recommendations for ICU triage exist at the moment. Further research into outcome evaluation for this population and maintenance of a case-to-case approach towards triaging by the physician’s side could further alleviate bias and age discrimination.

Table 1.

Frailty Assessment Scales according to category of measurement.

Measurement Category	Reference	Frailty Instrument	Description
Physical Frailty	Fried et al., J Gerontol A Biol Sci Med Sci, 2001;	Physical Frailty Phenotype (PFP)	Measures weight loss, exhaustion, strength of grip strength, gait speed, and physical activity level. Frail if 3 or more criteria are met.
Performance-based	Gill TM, et al., N Engl J Med, 2002	Gill Frailty Measure	Gait speed and chair stand are assessed and timed
Cumulative Deficit	Mitnitski et al., Scientific World Journal, 2001	Deficit Accumulation Index (DAI), aka Frailty Index (FI)	The sum of physical deficits, defined as symptoms or signs of disability or comorbidities (scored as binary or graded on a scale of 1) divided by the absolute number of deficits. Higher ratio relates to frailty with a cut-off of 0.2.
Judgment-based	Rockwood et al., CMAJ, 2005.	Clinical Frailty Scale (CFS)	Assignment to one of 7 consecutive categories of frailty, ranging from very fit to severely frail based on clinical observation and patient records, utilizing a pictograph
Multidimensional	Rolfson et al., Age Ageing, 2006	Edmonton Frail Scale (EFS)	Scores 10 categories, utilizing verbal and non-verbal tools, e.g., clock-drawing and timed up-and-go tests (0–5;non-frail, 6–11; vulnerable, 12–17; severely frail)
	Abellan van Kan G, et al., J Am Med Dir Assoc, 2008	FRAIL scale	Assesses self-reported fatigue, resistance (climbing stairs), ambulation, illness, and loss of weight. Frail if 3 or more criteria are met.
	Steverink N, Gerontologist, 2001	Groningen Frailty Indicator (GFI)	Questionnaire of baseline physical, cognitive, and social characteristics. Range from 0 to 15 with a cut-off of 4 for moderate frailty
Other measurement scales		Activities of Daily Living (ADL)	Fundamental skills for independent living act as predictors of function in older individuals
	M E Charlson et sl., J Chronic Dis, 1987	Charlson Comorbidity Index (CCI)	Incorporating weighted comorbidity scoring to age to predict ten-year mortality

2.2. Within the ICU

In a recent metanalysis including 129 studies of older (>75 y/o) patients admitted to the ICU, conducted after the year 2000, ICU mortality ranged from 1 % to 51 %, in-hospital mortality from 10 % to 76 %, 6-month mortality from 21 % to 58 %, and 1-year mortality from 33 % to 72 % (Vallet et al., 2021). The prognostic factors most commonly identified included severity score and mechanical.

ventilation for short-term mortality and comorbidities, age, and functional status for long-term mortality (Vallet et al., 2021). Variations could be attributed to different admission criteria, the intensity of treatment, and life-sustaining measures, as well as organizational differences across countries (Sprung et al., 2012, Guidet et al., 2017). Similarly, data remains heterogeneous regarding reporting of comorbidities and functional status, including frailty, as part of a comprehensive geriatric assessment to alleviate potential bias and allow standardization of potential prognostic factors. ( Table 2). Nonetheless, the survival benefit of admitted versus non-admitted patients seems to increase with age (Sprung et al., 2012). Interestingly, despite ICU patients' increasing age and disease severity, ICU mortality rates have remained relatively stable over time (Vincent et al., 2018).

Table 2.

Studies assessing mortality in elderly ICU populations by incorporating frailty.

Study / Design	Population				Outcomes	Mortality rate
	N	Age	Frail	Scale
Bagshaw SM et al., CMAJ, 2014, Multicenter Prospective cohort	421	> 50	33 %	CFS> 4	In-hospital mortality: higher in the frail subgroup (OR 1.81) Mortality at 1 year: significantly higher (HR 1.82). ICU mortality: no significant difference for frail individuals. Among survivors, frailty was associated with longer ICU and hospital LOS and with an increased incidence of major AEs Long-term outcomes: the frail scored significantly lower in health related QoL questionnaires compared to fit survivors at 6 and 12 months	ICU mortality: 10 % In-hospital mortality: 21 % (32 % v.16 % frail vs non-frail)
Flaatten et al., Intensive Care Med, 2017 Multicenter Prospective cohort	5021	≥ 80	43 %	CFS> 4	Frailty: independently associated with mortality (HR 1.54 for frail vs. fit). Treatment withdrawal was more frequent in frail subgroup	ICU mortality: 22.1 % Overall 30-day mortality: 32.4 %
Heyland et al., Intensive Care Med, 2015 Multicenter Prospective cohort	610	≥ 80	59 %	CFS> 4 FI-43(CGA)> 0.2	Frail vs fit patients: significantly less likely to survive at 12 months (OR 0.53 per 0.2 FI). In-hospital mortality: higher in the frail population (33 % vs 23 %). ICU mortality: did not differ significantly between the two groups. Limitations of treatment were more frequent in frail patients at ICU admission or afterwards (25 % vs 15 % and 36 % vs 25 % respectively), less likely to undergo mechanical ventilation (66 % vs 75 %). However, frail patients were just as likely as fit patients to receive other life-sustaining treatments and had similar time to death, ICU readmission rates, and durations of ICU and hospital stay.	ICU mortality: 22 % In-hospital mortality: 35 %
Le Maguet et al., Intensive Care Med, 2014 Multicenter Prospective cohort	196	≥ 65	41 %	CFS> 4 Adapted Physical Frailty Phenotype> 2	CFS-defined frailty: independently associated with increased in-hospital and 6 month-mortality Physical phenotype FI-defined frailty was associated with increased ICU mortality. Limitation of treatment orders: significantly increased for CFS-frail patients. No significant difference in ICU or hospital length of stay. CFS and adapted physical frailty phenotype were well correlated.	ICU mortality: 21 % In-hospital mortality:33 %
Kizilarslanoglu, M. C et al., Aging Clinical and Experimental Research, 2017 Single-center Prospective cohort	122	> 60	30.3 % Pre-frail 2.13 % Frail	FI> 0.4	ICU mortality: higher in frail patients compared to pre-fail and fit subjects (69.2 % v. 56.8 % v. 40.7 %, respectively). Overall survival: lower in frail group (23 v. 31 v. 140 days, respectively). Long-term mortality over 3 and 6 months: higher in frail patients (80.8 % and 84.6 % respectively). FI is an independent predictor of ICU mortality. (HR 39.019, 95 % CI 1.235–1232.537)	ICU mortality: 51.6 % In-hospital mortality: 56.6 %
Zeng, A., et al., The Journals of Gerontology, 2015 Single-center (geriatric ICU), prospective cohort	155	> 65	60 %	FI 52 > 0.22	FI was significantly higher in patients who died within 30 days (FI = 0.41 ± 0.11) than in survivors (FI = 0.22 ± 0.14). All subjects that scored a FI> 0.46 died within 90 days. The FI is an accurate predictor of 30-day mortality (AUC = 0.89).	30-day mortality: 45 % 300-day mortality: 38.7 %
Bruno et al., J Crit Care, 2020 Multicenter prospective cohort	415	> 80	51 %	CFS> 4	ICU mortality (24.5 % v. 14.9 %) and 30-day-mortality (43.4 % v. 23.9 %) were significantly higher in the frail patients’ group. ADL was significantly lower in the frail population (4 v 6 activities) CFS was an independent early predictor of 30-day-mortality in a multivariate logistic regression model. (OR 1.23)	ICU mortality:19.8 %, 30-day mortality: 34 %
Dres et al., Annals of Intensive Care, 2021 Multicenter prospective cohort	1199	> 70	24 %	CFS> 4	Day 90 mortality correlated with increases in age and CFS in COVID-19. 71 % mortality for CFS> 5 (HR=2.83) and 61 % for CFS= 4 (HR=2.24) on day 90.	28 mortality: 41 % day 60 mortality: 45 % day 90 mortality: 46 % day
Jung, C et al. Critical Care, 2021 Multicenter prospective cohort	1346	≥ 70	20,7 %	CFS> 4	Frailty: independently linked to lower survival in COVID-19. 30 days-survival: 66 % fit, 53 vulnerable (CFS=4), 41 % frail patients. 90-day survival: decreased with increasing CFS. (59 % v. 47 % v. 33 %) 30 days-mortality: 41 % in patients aged 70 or older 90 days-mortality: 48 % Frailty was associated with less use of mechanical ventilation and an increased incidence of treatment limitation orders.	30-day survival: 59 % 90-day survival: 52 %

ADL: Activities of Daily Living, AE: Adverse Events, CFS: Clinical Frailty Score, CGA: Comprehensive Geriatric Assessment., FI: Frailty Index, LOS: Length of Stay, QoL: Quality of Life

2.2.1. Interventions

When an older patient is admitted to the ICU, every possible appropriate treatment is to be offered. However, as stated during triage, this does not necessarily mean maximal treatment and should conform with the prior decision-making process (Philippart et al., 2013). Older patients commonly receive a lower level of treatment intensity and earlier decisions to limit life sustain measures than their young counterparts, either in the form of withholding or withdrawing care (Azoulay et al., 2009). Even though the timing of this decision remains arbitrary, in patients that do not show improvement in their clinical status, the therapeutic intensity level needs to be weighed against the patient’s chances of long-term survival. Apart from withdrawing or withholding care, not all therapeutic interventions seem to favor the elderly’s chances of survival. Dialysis does not seem to impact the death rates of critically ill patients with acute kidney injury aged 70 + years (Teles et al., 2019). Even in the context of the COVID-19 pandemic, the role of corticosteroids in the elderly remained controversial. A secondary analysis of the COVIP study found a higher 30-day-mortality in critically ill COVID-19 patients aged 70 years or older who received steroids as part of their treatment (Jung et al., 2021a). Similarly, RECOVERY did not find a beneficial effect of corticosteroid or tocilizumab use in patients of the same age (Group, 2021). However, the use of mechanical ventilation, vasoactive therapy, or renal replacement therapy does not seem to significantly differ between the frail and non-frail groups (Xia et al., 2021) since no recommendation for the differentiated treatment of elderly patients exists at the moment.

2.2.2. Infections and sepsis

The incidence of sepsis increases with advanced age. In addition, atypical or subtle clinical manifestations that delay diagnosis and age-related specificities, e.g., indwelling catheters, polypharmacy, institutionalization, reduced renal function, and poor nutritional status, contribute to the challenging management and poor outcomes following sepsis (Angus et al., 2001). Data from an international, observational, point-prevalence study (EPIC II) evaluating the effect of age on patterns of infection and outcomes in ICU populations from 75 countries highlight a large number of elderly ICU patients with infection, with individuals aged > 85 y/o accounting for a substantial proportion of the ICU population with infection (4.7 %) (Dimopoulos et al., 2013). Several significant differences occur between elderly and younger individuals as well as in-between elderly age subgroups (Dimopoulos et al., 2013). Those aged > 85 y/o had fewer bloodstream infections than those < 75 y/o, fewer central nervous system infections than those < 65, and more abdominal infections than those < 45 y/o. Although the microbiological diagnosis was less frequently established in participants > 85 y/o, gram-negative microorganisms were more regularly isolated, commonly exhibiting a multi-drug resistance phenotype (Poulakou et al., 2019). The latter is attributed to multiple antibiotic consumption, frequent hospitalizations in this age group, various comorbidities, and residence in long-term care facilities. Besides, as older the population grows, so does the number of patients likely to receive chemotherapeutic regimens for cancer or immunosuppressive drugs for nonmalignant diseases. Moreover, aging per se leads to variable physiological alterations in the immune system, coherently described as immunosenescence, that render older patients more vulnerable to fungal colonization and infections, as reflected in the increased incidence of C. glabrata fungemia (Flevari et al., 2013) and invasive aspergillosis (Matthaiou et al., 2018). Regarding the former, high mortality and increased rates of resistance to fluconazole are notable, calling for the use of echinocandins since they display a safer safety profile in the setting of an elderly patient with various comorbidities and co-medications. Specifically, they lack the risk of nephrotoxicity and have fewer drug interactions than azoles (Flevari et al., 2013). Regarding the latter, invasive aspergillosis' challenging diagnosis calls for a high index of suspicion. Even though no difference in survival between elderly and non-elderly has been noted, a decision not to administer antifungal agents in more critically ill patients, as reflected by higher SOFA scores, may have an impact on the survival of older patients calling for early and active search for evidence of mycological infection (Matthaiou et al., 2018). In general, ICU and in-hospital mortality were significantly higher in patients > 85 y/o (Dimopoulos et al., 2013). However, data as to whether age represents an independent risk factor for worse outcomes in younger decades remains heterogeneous (Poulakou et al., 2019). In a secondary analysis of the EU-VAP database (Koulenti et al., 2009), old age was not identified as a risk factor for Ventilator-associated Pneumonia (VAP) (Blot et al., 2014). On the other hand, VAP is associated with a significantly higher mortality rate in older age groups (Blot et al., 2014). The authors state that such a disparity could be attributed to survival bias since VAP is a complication that primarily affects patients who survived the initial phase of critical illness, meaning that numerous other patients died before the onset of VAP (Blot et al., 2014). All in all, age-related particularities should be considered while drawing a therapeutic plan. Hepatic and renal impairment, sarcopenia, and gastrointestinal pathology pose important implications in treatment decisions, requiring continuous vigilance for adverse events. Finally, site-specific pathogen identification is pivotal to the selection of the best available antibiotic regimen (Poulakou et al., 2019).

2.2.3. COVID-19 infection

Like most respiratory viruses that ravage the elderly seasonally every year, SARS-COV-2 seemed predestined to affect older individuals disproportionately. Advanced age has been identified as the most significant risk factor for poor outcomes in critically ill patients (Grasselli et al., 2020a). Frail patients are likely to be more susceptible to COVID-19 disease, with a prevalence of frailty as high as 51.1 % (Hewitt et al., 2020). Almost two years into the pandemic, people over 65 years old account for almost three-quarters of all deaths by COVID-19 in the US and two-thirds of ICU admissions, despite representing a mere 15 % of the total population ($author1$ et al., 39]</id><AuthGrp><Author><au>European Centre for Disease Prevention and Control</au></Author></AuthGrp>). Pertinent metanalyses have established the increased risk of infection, ICU admission, and death in these populations with overwhelming results/odd ratios (Pijls et al., 2021). Although the elderly may have been overrepresented in the initial statistics concerning ICU mortality, reports from the second surge of the pandemic showed both a short- and long-term survival decrease, despite advances in treatment options and disease management (Jung et al., 2021a). Therefore, one cannot help but wonder if power managing this group of patients in the ICU is the appropriate therapeutic direction. The American Geriatric Society published a COVID-19 position statement regarding resource allocation strategies and age-related considerations in COVID-19 to guide healthcare systems in developing emergency rationing strategies (Farrell et al., 2020). Their position statement declares that age should not drive resource allocation, nor should it be used as a means for exclusion from therapeutic interventions. "Life years saved" or "long-term predicted life expectancy" are criteria that, by definition, disadvantage older individuals (Farrell et al., 2020).

The impact of frailty on COVID-19 outcomes remains controversial, and data in global literature are conflicting (Dadras et al., 2022). Cardiovascular diseases, respiratory diseases, nervous system diseases, renal conditions, and malignancies are associated with higher mortality from COVID-19 in elderly populations. Among specific diseases, peripheral artery diseases, hypertension, coronary heart diseases, heart failure, hypertension, COPD, asthma, bronchitis, dementia, cognitive impairment, diabetes, and both low (<18.5 kg m²) and high (>35 kg m²) body mass index increased mortality (Dadras et al., 2022). An overall mortality rate of 80.5 % has been reported in a Turkish cohort, whereas mortality reached 55 % in an Italian cohort (Grasselli et al., 2020b), 66.8 % in a French cohort (Dres et al., 2021), and 72 % in a German study (Karagiannidis et al., 2020). A 30-day mortality rate of 41 % in patients aged 70 or older is reported in the COVIP study, which increased to 48 % at 90 days and was found to be as high as 67 % in frail individuals (Jung et al., 2021b). A recent meta-analysis identified no increased mortality risk among frail patients after adjusting for age and other covariates (Subramaniam et al., 2022). However, comparisons are not drawn easily between different COVID-19-related studies since various age cutoffs have been employed: 60 years, 65 years, 70 years, 75 years, 80 years, and 85 years (Jung et al., 2021b, Demiselle et al., 2021, Falandry et al., 2021). Due to the sheer pathophysiology of COVID-19, complications and rapid deterioration may arise acutely, especially in frail older individuals with reduced physiologic reserve capacity; hence, advanced directives are to be in place to accommodate patient wishes should they become incapacitated. As stated before, the use of antiviral and immunomodulatory therapies in this population remains controversial due to conflicting data on efficiency and effectiveness in such age extremes (Jung et al., 2021a).

2.2.4. Surgery and trauma

In general, postoperative patients after elective surgery show better long-term outcomes than those following acute or emergency admissions. Evidently, patients undergoing planned surgery are most probably younger and less frail; thus, 1-month mortality was lower than acutely admitted patients (Karakus et al., 2017). In a recent study of critically ill patients over 80 years that were admitted to the ICU for elective surgery, 72 % were discharged home (Bagshaw et al., 2009), while long-term prognosis was much more favorable for scheduled than urgent admissions (Guidet et al., 2018a). Patients with a frailty score > 5 according to a multidimensional frailty model (composed of the CCI, dementia or dementia risk, nutritional status, and independent daily activities) showed increased postoperative mortality and longer length of stay. The latter constructed model predicted outcomes more accurately than the commonly employed score of the American Society of Anesthesiology (Zampieri and Colombari, 2014).

2.3. After the ICU

Age, comorbidities, disease severity, and organizational factors, including discharge time and step-down facility availability, play a role in-hospital mortality following ICU discharge (Lin et al., 2009). The elderly suffer from sequelae due to prolonged sedation, ventilation, and immobilization, which may complicate mortality (Karakus et al., 2017). Even though survival remains the cornerstone for clinical outcome evaluation in critically ill patients admitted to the ICU, the concept of patient-important (or patient-centered) outcomes, including long-term quality of life (QoL), functional, cognitive, and mental health outcomes, has been recently introduced (Pallanch et al., 2022). Following ICU discharge, older patients commonly suffer from exacerbated comorbidities or secondary complications requiring early physical rehabilitation (Guidet et al., 2018a). Elderly patients who survive the ICU are also expected to have worse QoL (swallowing disorders, neuromuscular dysfunction, delirium) compared to younger survivors. To a lesser extent, they may experience worse QoL compared to age-matched community controls or even their own QoL up to 1 month before ICU (Ariyo et al., 2021). Older patients frequently lose functional autonomy and may develop neurocognitive disorders, which represent an additional burden also for relatives and caregivers (Vallet et al., 2019). Pre-admission frailty has been well identified as an independent factor driving post-ICU morbidity (Ferrante et al., 2018). Similarly, among COVID-19 patients, frailty was associated with a need for a higher level of post-discharge care (Research et al., 2021) and also a higher mortality risk two weeks post-discharge among survivors (Hewitt et al., 2020, Kow et al., 2021).

This cluster of multidimensional impairments was defined as "Post-Intensive Care Syndrome" only as recently as 2010 and affects not only patients but also caregivers and society (Morgan, 2021). Even though its prevalence remains unclear, PICS is believed to impact approximately 25–55 % of ICU survivors (Ohtake et al., 2018). The quality of life in elderly ICU survivors is often significantly lower than that of their young counterparts (Guidet et al., 2018a). In fact, it is worse among patients requiring mechanical ventilation for more than seven days (Herridge et al., 2016) or those exhibiting neurocognitive problems (Pandharipande et al., 2013). A decline in physical functioning has also been reported in older ICU survivors at 3,6 and 12 months following discharge compared to age and gender-matched controls, while in 64 % of patients, at least one domain in the ADL score decreased at 6–21 months, especially in more baseline dependent individuals (Guidet et al., 2017, Heyland et al., 2015). Similarly, in a UK-based study, the SF36 HRQoL questionnaire showed no improvement from 6 to 12 months (Guidet et al., 2017), while the reported quality of life was lower for the elderly than younger patients following ICU discharge (Kaarlola et al., 2006).

Attempts to evaluate and possibly quantify patients' feelings and values in a holistic and comprehensive approach have been put forward to address recent evidence suggesting that patients discharged from the ICU often consider survival less critical than their subsequent quality of life (Pallanch et al., 2022). Although most people would prefer to be in moderate pain than to die, when considering the potential need for permanent breathing support, feeding tubes, or incontinence, the general feeling was the same or worse than death (Rubin et al., 2016). Finally, the caregiver represents the "second victim" in this sequela of events (Cameron et al., 2016). Post-Intensive Care Syndrome Family (PICS-Family) describes a parallel concept of how outcomes affect family members of ICU survivors that includes depression, grief, anxiety, or post-traumatic stress disorder (Cameron et al., 2016), illustrating the need for broader management and increased awareness.

2.4. Issues to be resolved

The COVID-19 pandemic and the vast need for ICU care that came with it exposed the otherwise pre-existing issue of limited resource availability and the tough decisions that need to be made on a daily basis. A clear protocol to allow personnel training and preparation should be set in place to optimize clinical decision-making and ensure the best outcomes by identifying optimal practice scenarios for each patient population in every step of ICU-related care. Without a set of guidelines or proper training, decisions would risk being left to a lottery or on a first come, first served basis, thus, compromising outcomes (Vincent and Creteur, 2022, Maves et al., 2020).

Setting things in an individual context, discussions with surrogate decision-makers, and subsequent arrangements are to occur well before the need for an intensive level of care to ensure an appropriate treatment plan in an ethical climate. However, communication, especially in times of pandemic, can be problematic. Success in conveying details about critical situations over a phone call or teleconferencing and evoking comprehension while respecting the sensibilities involved in distant communication is, by definition, a difficult task. Establishing rapport and attaining attention in order to assist in end-of-life decisions, thus, requires a greater amount of time and effort from the side of the physician. So as to overcome all practical concerns and deal with conflicting demands on patient care in the short time framework of the ICU requiring immediate decisions, advanced care planning is highly recommended. Optimally, all elderly patients, regardless of general condition, must draft a living will if they do not have an advanced directive in place. Patients and families opting for a "Do Not Intubate" (DNI) code status or hospice care are encouraged to hold conversations with their primary care providers in advance. In all cases, the main question to be addressed by the surrogate decision maker is not what they themselves feel about ICU admission but what they know about the patient's wishes. Interestingly though, in many cases, intensivists' decisions seem to take priority over any advance directive, underlining the fact that disease severity will ultimately drive the dynamic decision process (Smirdec et al., 2020). One should also consider that the ICU stay is not but one phase in some patients' long hospitalization course, which may include a protracted "before the ICU" disease and an "after the ICU" extended and demanding rehabilitation period. The possibility of recovering a meaningful life following discharge is possibly the most crucial parameter in admitting an elderly patient to the ICU.

Health economic analysis of quality-adjusted life years (QALYs) of elderly ICU survivors may prove helpful for resource allocation and policy establishment purposes. However, no such analyses and no respective explicit data are currently available. In the effort to be formed, the analysts will have to grapple with the controversial notion that everyone is entitled to a 'normal' span of health or "fair innings' (Williams, 1997). As a strategy to improve the care of older patients and provide everyone with a fighting chance, the "ICU trial" has been suggested (Chang et al., 2021). This method would allow time to collect all patient relevant information, e.g., medical history, baseline functional status, and advance directives, and closely evaluate their progress under complete treatment before deciding on therapy continuation or withdrawal by providing initial full organ support with continuous re-evaluation in the next 48–72 h (Leblanc et al., 2017). However, withholding and withdrawing care is subject to regional differences and variable ethical and cultural frameworks that could permit less or more societal tolerance (Vincent and Creteur, 2022, Guidet et al., 2018b). ICU bed availability, religious issues, and growth in the domestic product directly correlate with life-sustaining measures availability and limitation orders (Guidet et al., 2018b). It is important to note that life-sustaining therapy limitation does not equal end-of-life decisions. The ICU and 30-day mortality in patients withholding or withdrawing care were 29 vs. 59 % and 82 vs. 93 %, respectively, even though results may vary among studies (Guidet et al., 2018b, Lobo et al., 2017). Disproportional care should be avoided so as not to prove futile. As stated in the Hippocratic oath, physicians should refuse to treat "those overmastered by the disease…[.]. in such cases medicine is powerless" (Jones) HtbW (1959)). In some cases, early end-of-life decisions may be unavoidable. Developing consensus recommendations as to who, when, and how will decide to "accompany" a patient to die with dignity is far from easy due to variable global attitudes (Vincent and Creteur, 2022).

Following ICU discharge, a lack of universal protocol is observed. Poor step-down care planning makes the elderly extremely vulnerable due to their complicated physiology and need for intensive monitoring (van Sluisveld et al., 2017). Studies have failed to identify discharge premises as risk factors for poor outcomes; hence no impact of discharge protocol on hospital mortality has ever been described (van Sluisveld et al., 2017). Despite the fact that this population calls for a multidimensional approach in the rehabilitation period, only two-thirds of the physicians in the VIP1 study recognized the need for a geriatric consultation when deciding on discharge for an elderly individual (Flaatten et al., 2017). Nonetheless, geriatric intervention can benefit patients' functional status after discharge from a medical or surgical unit to a similar extent as physical therapy (Rosa et al., 2019). System and technological innovations to complement care for the elderly post-ICU could prove beneficial in the long run.

Unfortunately, heterogenicity in outcome assessment tools and follow-up duration does not allow for precise results regarding post-ICU outcome comparison in the elderly. Of note, no studies have explored how changes in management within or following ICU could affect the incidence of frailty-associated outcomes. Determining the proper intervention for the right patient population and defining the circumstances under which it should take place are the key steps in avoiding under- or overuse of stringent ICU resources in patient-centered quality care.

3. Conclusion

Elderly patients currently represent a vast expanding population in the ICU. Multiple comorbidities and immune dysregulation drive variable responses to hospitalization and standard of care in the elderly, causing uncertainty and a lack of confidence in establishing solid and universal recommendations to guide clinical management in the volatile ICU environment. Albeit reviews as such can provide the initial context for a discussion, they cannot account for regional variations in clinical practice or disparities in patient populations and treatment option availability. Recognizing the presence or absence of potential benefits of ICU care for the elderly goes beyond standardized assessments and involves critical insights into patient pre-admission status and specific disease characteristics. Tailoring safe treatment plans to match patient's wishes, and personalized needs will guide critical care for the elderly from this time forward ( Fig. 2).

Fig. 2.

Comprehensive ICU management of the elderly.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declarations of interest

none.