ICU mortality has been decreasing over the past 2 decades (1–3). In contrast with other countries, the rate of increase in intensive care beds in the U.S. outpaces general hospital beds (4). Increasing capacity and decreasing mortality have created an evolving and diverse population of ICU survivors. Today’s survivors of critical illness are increasingly burdened by extensive physical and psychological comorbidities, often resulting in reduced quality of life (5,6).
Functional status deterioration (FS↓) occurs to some degree in at least 50% of critically ill survivors (7–10). In those greater than 80 years old, functional status limitations can be as high as 97% (11). Alarmingly, deficits in physical and cognitive functioning have been shown to persist for months following recovery from critical illness (12). The term post-intensive care syndrome (PICS) has been used to encompass the physical, emotional, and neuropsychologic sequelae from critical illness. In response to increasing recognition of PICS, international critical care societies have developed and endorsed guidelines and bundles targeted at improving the delivery of care, particularly as it relates to sedation management, delirium recognition and mitigation, and early mobility practices. Societies have also recognized the need for longitudinal care of the ICU survivor population with post-ICU clinics being established at a growing number of institutions (13,14).
The purpose of this study was to evaluate the change in FS↓ after critical illness over the past decade. We hypothesized that functional status decline among ICU admissions has changed over time, both overall and within specific disease processes. Using a national prospective cohort, we also sought to quantify the amount of FS↓ in different diseases. In our secondary outcome, we sought to characterize the magnitude of decline and temporally model this change to further stratify these temporal changes. Characterization of these trends could help determine high yield targets for future research, intervention, and resource allocation. Some of the results of this study have been previously reported in the form of an abstract (15).
MATERIALS AND METHODS
We used data from the Cerner Acute Physiology and Chronic Health Evaluation (APACHE) outcomes database (Cerner Corporation, Kansas City, MO) and included all patients requiring ICU admission at 236 participating hospitals in the United States (16). Of these hospitals, 90 reported functional status during this time period. These data were manually collected by trained clinical abstractors at all hospitals participating in this national quality improvement and benchmarking program (16–18). Abstractors underwent rigorous training and recurrent audits to ensure data accuracy and reliability. Information regarding data collection, data auditing, and validity are described in further detail elsewhere (17,18). Our study was approved by the Institutional Review Board at our institution (STUDY00001489).
All adult patients (age ≥ 18 yr) admitted to the ICU at participating centers between January 1, 2008, and December 31, 2016, with a documented functional status at admission and hospital discharge were included in the study.
We excluded patients who had a primary surgical admission as the majority of patients may have a higher propensity for recovery due to the nature of the procedure, while other procedures’ normal postoperative trajectory may include an expected functional status decline prior to discharge. Furthermore, surgical patients tend to experience less morbidity and mortality compared with nonsurgical critical care patients in other large ICU studies (2,3). Resultantly, we were concerned the differing phenotypes may create an inaccurate depiction of critical illness associated functional status decline and proceeded to focus on the nonsurgical patients for our study. We also excluded those who died during their hospitalization and focused on survivors from critical illness. As hospitals were not required to participate each year, we sought to minimize selection bias by excluding patients who had an admission diagnosis with less than 100 occurrences, were admitted to a hospital with less than 3 years of data, or admitted to a hospital with less than 100 total cases during the study time period (Fig. 1). Patient (Table E1, Supplemental Digital Content 1, http://links.lww.com/CCM/F689) and hospital (Table 1) characteristics are presented in table form.
TABLE 1. -
ICU Characteristics During Study Period
||No Change, n = 91,801
||Functional Status Decline, n = 38,116
|Admissions per ICU, per year
||3,983.3 ± 2,838
||3,878.4 ± 2,781
|Hospital bed size
||562 ± 335
||600 ± 343
| Large teaching hospital
| Small teaching hospital
| Nonteaching hospital
Descriptive statistics include n (%) or mean ± sd where indicated.
Definitions and Outcomes
Functional status was documented on ICU admission and hospital discharge based on data manually abstracted by trained individuals who would categorize the patient’s functional status as fully independent, partially dependent, or fully dependent. Functional status was obtained from the patient or surrogate to determine functional status prior to ICU admission (termed admission functional status). Functional status was also recorded just prior to discharge based on information provided by the healthcare team (i.e., provider, nursing, physical therapy, occupational therapy documentation). Categories were determined by an assessment of basic activities of daily living (ADLs). Definitions for each category included: fully independent—the patient was able to perform all ADLs without assistance, partially dependent—the patient needed assistance to perform ADLs, and fully dependent—the patient cannot perform ADLs and must be cared for by another person. We categorized the primary admission diagnoses into organ system-disease based groups to facilitate data modeling (for classification details, see Table E2, Supplemental Digital Content 2, http://links.lww.com/CCM/F690)
Our primary outcome was the secular changes of FS↓ among ICU survivors. In our primary outcome, we categorized FS↓ as a dichotomous variable—decline or no decline. A decline was defined as a decline from the admission functional status category (status before critical illness) to the hospital discharge functional status category.
We hypothesized FS↓ may be increasing over time; however, FS↓ is a spectrum. In our secondary outcome, we sought to characterize the magnitude of the decline and temporally model this change to further stratify these temporal changes. The outcome was ordered as, no change, fully independent to partially dependent, partially dependent to fully dependent, and fully independent to fully dependent, in order of progressing severity. We modeled our outcome in this fashion to address concerns that classification at the margins favors partial dependence, whereas being completely independent or dependent in ADLs are very specific definitions. Given this issue, we separated independent to partially dependent and partially dependent to fully dependent into their own categories with the latter category as more severe than the former.
We collected patient-level variables including: patient demographics (age, sex, race, and insurance status), illness severity (APACHE III score, based on worst parameters in the first 24 hr), shock index (heart rate/systolic blood pressure), and admission characteristics (vital signs, organ system-disease based category, year of admission, code status, and length of stay [LOS]).
For descriptive purposes, parametric continuous variables are expressed as mean and sd, and nonparametric continuous variables are expressed as the median and interquartile range (IQR). Categorical variables are expressed as n (%). Imputation was not performed as missingness of data was less than 5% (n = 1,179 [0.5%]) (19).
We fit quasi-Poisson models to assess the prevalence of FS↓ from admission to hospital discharge. Data were aggregated to the hospital-level and covariates, including age and APACHE III score, were averaged by hospital and year. All variables were scaled and centered prior to model fitting. For the overall model, we used an offset of the log number of admissions in each hospital and year. In contrast, we used the log number of patients in each hospital, year, and disease-system category as offsets to model our outcomes within each disease-system category. The estimate for the exponentiated centered year term is reported with a 95% CI to determine if the prevalence rate ratio (IRR) per year of FS↓ is changing over time.
We used proportional odds logistic regression (a form of ordinal logistic regression) to evaluate the change in the magnitude of FS↓ over time. The outcome was ordered as, no change, fully independent to partially dependent, partially dependent to fully dependent, and fully independent to fully dependent, in order of progressing severity. The effect of year was modeled overall and within each epidemiological group (disease-system category) at the patient level via an interaction term adjusting for age and APACHE III score. To evaluate the odds of FS↓, within each disease process, we used proportional odds logistic regression, adjusted for age, APACHE III, and year. Sum-to-zero parameterization was implemented to compare the odds for each disease group against the overall odds of FS↓. All analyses were performed using R Version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria), Stata MP, Version 15 (StataCorp, College Station, TX), or Microsoft Excel (Microsoft, Redmond, WA).
A total of 129,917 patients were included in the final analysis with FS↓ occurring in 38,116 patients (29.3%). The overall patient population had a median age of 63 years old (IQR, 50–75 yr old). They were more likely to be male (54.0%) and white (80.9%). Table 1 shows characteristics for patients who had functional status decline versus those who did not. Patients with FS↓ after admission were older (68 vs 60 yr old) and had a higher proportion of patients that were white (83.6% vs 79.2%). Other variables associated with FS↓ included: higher APACHE III score (54.4 vs 43.6), increased ICU LOS (2.9 vs 1.7), and increased hospital LOS (9.1 vs 4.7).
Prevalence of FS↓
The prevalence of FS↓ increased overall throughout the period of observation (IRR, 1.15; 95% CI, 1.13–1.17; p < 0.001) (Fig. 2). This trend was present across all disease categories: neurologic (IRR, 1.09; 95% CI, 1.05–1.13; p < 0.001), cardiac (IRR, 1.12; 95% CI, 1.07–1.17; p < 0.001), pulmonary (IRR, 1.11; 95% CI, 1.07–1.16; p < 0.001), gastrointestinal (IRR, 1.15; 95% CI, 1.10–1.20; p < 0.001), renal (IRR, 1.08; 95% CI, 1.03–1.13; p = 0.003), nonsurgical trauma (IRR, 1.13; 95% CI, 1.09–1.17; p < 0.001), toxicology (IRR, 1.15; 95% CI, 1.09–1.21; p< 0.001), vascular (IRR, 1.12; 95% CI, 1.06–1.18; p < 0.001), septic infection (IRR, 1.10; 95% CI, 1.06–1.15; p < 0.001), endocrine/metabolic (IRR, 1.08; 95% CI, 1.02–1.14; p = 0.008), and hematologic/oncologic (IRR, 1.10; 95% CI, 1.04–1.17; p = 0.003).
Magnitude of FS↓
The magnitude of FS↓ increased overall (odds ratio [OR], 1.04; 95% CI, 1.04–1.04; p < 0.001) and in most categories. The only category showing a decrease in the magnitude of FS↓ overtime was nonsurgical trauma (OR, 0.97; 95% CI, 0.97–0.97; p < 0.001) (Fig. 3). However, despite the decreasing magnitude of FS↓ in nonsurgical trauma, many admission diagnoses in this category remain in the top quartile of higher risk for FS.
Diagnosis Variation of FS↓ and Magnitude
Using the average hospital admission as the reference group, FS↓ was highest in patients admitted with head and polytrauma (OR, 4.47; 95% CI, 4.46–4.48; p < 0.001), nonsurgical intracranial hemorrhage (OR, 3.39; 95% CI, 3.37–3.40; p< 0.001), chest and spine trauma (OR, 3.38; 95% CI, 3.38–3.39; p < 0.001), and spine trauma (OR, 3.19; 95% CI, 3.18–3.19; p< 0.001). Lowest odds of FS↓ was seen in diabetic ketoacidosis (OR, 0.27; 95% CI, 0.27–0.27; p < 0.001) and asthma (OR, 0.35; 95% CI, 0.35–0.36; p < 0.001) (Fig. 4). The top quartile of diseases with the highest odds for FS↓ included nonsurgical trauma, neurologic, pulmonary, and gastrointestinal diseases.
We sought to identify temporal changes and disease variation in FS↓ during hospitalization among critical illness survivors. We found that, after controlling for age and severity of illness, the prevalence of FS↓ was increasing, both overall and throughout all system-disease groups for the duration of the study. Additionally, we found that in addition to prevalence, the magnitude of FS↓ was also increasing overall and in each category (aside from nonsurgical trauma) over time. The magnitude of FS↓ varied in relation to the admission diagnosis. The top quartile of admission diagnoses with the highest odds of FS↓ included the following system-disease categories: nonsurgical trauma, neurologic, pulmonary, and gastrointestinal. All these high-risk system-disease groups that make up the aforementioned top quartile have an increasing prevalence of FS↓ and apart from nonsurgical trauma, also have an increasing magnitude of FS↓.
Functional status before, during, and after admission is critical to patient outcomes. Baseline functional status has been associated with short- and long-term mortality for many subsets of hospitalized patients which include critically ill, immunocompromised, severe pneumonia, and lung transplant patients (20–26). Functional status is comparable, if not superior to common variables (i.e., age and illness severity) used to predict patient outcomes (27–29). In fact, premorbid FS was shown to be superior to antibiotic timing for predicting mortality in patients with pneumonia (30). These associations have led to growing interest in assessing frailty, which Fried et al (31) have advocated not be used synonymously with functional status. Both are risk factors for poor outcomes and will be key to improving risk-stratification in research and clinical care (22,32). Beyond predicting absolute endpoints (i.e., mortality) after admission, preoperative FS has been shown to affect not only long-term FS but also the trajectory of functional status change (33). Other important outcomes linked to baseline functional status include: quality of life, readmission, rate of admission to the ICU, and decisions to withhold or withdraw care (34,35).
As ICU admissions are generally not elective, baseline functional status serves mainly in risk assessment without targets for intervention; however, in-hospital functional status can be used for both. A 2016 study of ICU survivors by Rydingsward et al (36) found decreased functional status on discharge, irrespective of admission functional status, was associated with mortality. Additionally, they found patients with improving functional status prior to discharge, versus no improvement, had a 64% lower 90-day mortality. While our current study is limited to functional status decline during hospitalization, the duration this decline is rarely short and may persist without return to baseline (9,37). A 2015 study showed only 10% of patients had improvement in their functional status at 3 months (9). Furthermore, in relation to other health-related quality of life measures, functional status is the least likely to return to population norms after critical illness at 1 year (37). The increasing prevalence of FS↓ may be in part to increasing survivorship as mortality in the ICU has decreased over prior decades (2). New interventions, such as prone positioning (38) and neuromuscular blockade (39) for acute respiratory distress syndrome have shown survival benefit; however, the increasing prevalence of FS↓ may be the cost exchanged for these advances.
Increasing survivorship from critical illness has led to new challenges for both patients and providers (40,41). ICU patients that survive to discharge have higher mortality after critical illness, especially in the first 3 months, when matched to comparable general hospital patients (42). When compared to the general population, ICU survivors have decreased quality of life, due in part, to FS↓ which can persist up 12 years later (5). In contrast, patients undergoing planned surgery and requiring ICU admission may be an exception, as one study found a majority of patients can return to their baseline functional status and quality of life after ICU admission (43). The ICU involves caring for a remarkably large variety of illness and severity. Identifying the patients with the highest risk and magnitude for FS↓ can assist in resource allocation and quality improvement projects. We identified admission diagnoses associated with the highest odds of FS↓ (Fig. 4) while also showing that most epidemiological categories have increasing rates of FS↓ (Fig. 3). Models were attempted to identify each diagnosis, however, given the large number of admission diagnoses, we could only fit models to attempt temporal change within the larger epidemiological categories.
Patient-centered decision-making hinges on being able to prognosticate outcomes, including level of function. Unfortunately, our ability to predict functional status outcomes is worse than our ability to predict mortality (44). This highlights the need to identify high-risk patient groups. Currently, several tools exist and recommendations have been laid out to accurately assess a patient’s risk for FS↓ (7). This includes using patient and surrogate information to identify a patient’s ability to perform ADLs and independent mobility (45). In the modern age with large amounts of data, we need to use these databases to help determine critical care needs and services (46). Future studies using validated functional scales are needed to further refine risk factors and tailor interventions to prevent deterioration.
Interventions for prevention of FS↓ are limited, and future research is essential (41,47). Currently, efforts aimed at improving early mobility and delirium prevention have shown the most promise in mitigating deterioration in functional status. A study in 2009 showed early physical therapy and interruption of sedation were associated with regaining function after critical illness (OR, 2.7) (48). Implementation of Awakening and Breathing Coordination, Delirium monitoring/management, and Early exercise/mobility bundles has led to improvements in mobilizing patients sooner in ICUs, but this has not translated to improved rates of discharge to independent functional status or return home (49). More recently, implementation specifically of early mobility, added to an abbreviated bundle of awakening, breathing, and delirium was associated with reduced mechanical ventilation, LOS, and cost (50). Delirium is known to be associated with long-term cognitive decline and long-term disability in ADLs in ICU survivors (12). Multifaceted interventions to reduce delirium have been found to increase the rate of patients being able to return home (51). Daily cognitive screening and delirium prevention are key in improving patient outcomes. Taken in composite, these studies suggest that continued efforts targeted at the prevention of ICU delirium may be integral to preventing further FS↓. Efforts aimed at improving the identification of patients at high risk for FS↓ will be an important area of further study.
This study highlights the significant temporal and disease variation effect on FS↓ in the critically ill. Limited research has been able to capture functional status on a national level. There are several limitations to this study. First, the APACHE database was not prospectively built to answer this question. Second, there are limited hospitals that report consistently to the database, which may introduce selection bias. To mitigate this, we eliminated hospitals that had less than 3 years of data included during the study period. Furthermore, we were only able to use hospitals that reported functional status. Table E3 (Supplemental Digital Content 3, http://links.lww.com/CCM/F691) compares hospitals that reported functional status to those that did not report functional status. Third, there were only three functional categories and a validated tool for functional status designation was not used, limiting the replicability of the study. Also, the source from which the ADL data was from is not included in the database (i.e., patient vs surrogate). The categories also limited our ability to discern signals that existed within the partially dependent category. We hypothesized that a subset of patients within the partially dependent category may have had a small, but clinically significant, functional status decline that was not captured due to limitation in our categorization. In our ordinal categorization, we attempted to mitigate the dilutional effect (while maintaining clinically significant changes) by separating the stepwise decline to and from partial dependence (independent to partially dependent; partially dependent to fully dependent) into separate categories instead of labeling them as the same. Another important limitation includes the generalizability. Postoperative patients, which were excluded from this study, account for greater than 20% of critical illness survivors. These findings cannot and should not be extrapolated to these patients. Analysis of the primary surgical patients from this cohort, and preliminary analysis supports our concern for large heterogeneity and thus separation of these patient populations (52). Although we were able to determine that most system-disease cohorts had increasing FS↓ and magnitude of FS↓, we were not able to parse this out by specific admission diagnoses due to issues with model convergence. Furthermore, model convergence also limited our ability to include other covariates. Subsequent analyses to identify independent predictors of FS↓ are ongoing are not included in this article. This study highlights the increasing prevalence of FS↓ in a large multicenter cohort while also identifying the variation that exists across different admission diagnosis categories.
FS↓ following nonsurgical admission to the ICU occurred in almost one-third of patients in our study. Overall prevalence and magnitude of FS↓ increased over the study period and there was significant variation among admission diagnoses associated with varying levels of decline. We believe our study provides important information that can be used in beginning to identify patients at high risk of functional status decline. Improving the identification of these patients and targeting appropriate interventions to mitigate this decline will be important directions for future studies in this area.
The authors would like to thank Cerner Corporation and Laura Freeseman-Freeman for the use of APACHE Outcomes data for research purposes.
1. Hutchings A, Durand MA, Grieve R, et al. Evaluation of modernisation of adult critical care services in England: Time series and cost effectiveness analysis. BMJ. 2009; 339:b4353
2. Zimmerman JE, Kramer AA, Knaus WA. Changes in hospital mortality for United States intensive care unit admissions from 1988 to 2012. Crit Care. 2013; 17:R81
3. Lilly CM, Swami S, Liu X, et al. Five-year trends of critical care practice and outcomes. Chest. 2017; 152:723–735
4. Wallace DJ, Angus DC, Seymour CW, et al. Critical care bed growth in the United States. A comparison of regional and national trends. Am J Respir Crit Care Med. 2015; 191:410–416
5. Flaatten H, Kvåle R. Survival and quality of life 12 years after ICU. A comparison with the general Norwegian population. Intensive Care Med. 2001; 27:1005–1011
6. Herridge MS, Tansey CM, Matté A, et al.; Canadian Critical Care Trials Group. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med. 2011; 364:1293–1304
7. Brummel NE, Balas MC, Morandi A, et al. Understanding and reducing disability in older adults following critical illness. Crit Care Med. 2015; 43:1265–1275
8. Ferrante LE, Pisani MA, Murphy TE, et al. Factors associated with functional recovery among older intensive care unit survivors. Am J Respir Crit Care Med. 2016; 194:299–307
9. Oliveira R, Rosa R, Ascoli A, et al. Factors influencing decline of physical functional status among ICU survivors: A prospective cohort study. Intensive Care Med Exp. 2015; 3:A361
10. Schandl A, Bottai M, Holdar U, et al. Early prediction of new-onset physical disability after intensive care unit stay: A preliminary instrument. Crit Care. 2014; 18:455
11. de Rooij SE, Govers AC, Korevaar JC, et al. Cognitive, functional, and quality-of-life outcomes of patients aged 80 and older who survived at least 1 year after planned or unplanned surgery or medical intensive care treatment. J Am Geriatr Soc. 2008; 56:816–822
12. Pandharipande PP, Girard TD, Jackson JC, et al.; BRAIN-ICU Study Investigators. Long-term cognitive impairment after critical illness. N Engl J Med. 2013; 369:1306–1316
13. Haines KJ, Sevin CM, Hibbert E, et al. Key mechanisms by which post-ICU activities can improve in-ICU care: Results of the international THRIVE collaboratives. Intensive Care Med. 2019; 45:939–947
14. Prescott HC, Angus DC. Enhancing recovery from sepsis: A review. JAMA. 2018; 319:62–75
15. Ingraham NE, Vakayil V, Coughlan A, et al. Dysfunction junction: 8-year trends in deteriorating intensive care functional status. Am J Respir Crit Care Med. 2019; 199
16. Cerner Corporation: APACHE Outcomes. 2019. Available at: https://apacheoutcomes.cernerworks.com/criticaloutcomes-home/
. Accessed December 6, 2019
17. Knaus WA, Wagner DP, Draper EA, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest. 1991; 100:1619–1636
18. Zimmerman JE, Kramer AA, McNair DS, et al. Acute Physiology and Chronic Health Evaluation (APACHE) IV: Hospital mortality assessment for today’s critically ill patients. Crit Care Med. 2006; 34:1297–1310
19. Jakobsen JC, Gluud C, Wetterslev J, et al. When and how should multiple imputation be used for handling missing data in randomised clinical trials - a practical guide with flowcharts. BMC Med Res Methodol. 2017; 17:162
20. Sacanella E, Pérez-Castejón JM, Nicolás JM, et al. Mortality in healthy elderly patients after ICU admission. Intensive Care Med. 2009; 35:550–555
21. Krinsley JS, Wasser T, Kang G, et al. Pre-admission functional status impacts the performance of the APACHE IV model of mortality prediction in critically ill patients. Crit Care. 2017; 21:110
22. Muscedere J, Waters B, Varambally A, et al. The impact of frailty on intensive care unit outcomes: A systematic review and meta-analysis. Intensive Care Med. 2017; 43:1105–1122
23. Casalino E, Mendoza-Sassi G, Wolff M, et al. Predictors of short- and long-term survival in HIV-infected patients admitted to the ICU. Chest. 1998; 113:421–429
24. Sligl WI, Eurich DT, Marrie TJ, et al. Only severely limited, premorbid functional status is associated with short- and long-term mortality in patients with pneumonia who are critically ill: A prospective observational study. Chest. 2011; 139:88–94
25. Banga A, Sahoo D, Lane CR, et al. Characteristics and outcomes of patients with lung transplantation requiring admission to the medical ICU. Chest. 2014; 146:590–599
26. Ingraham NE, Tignanelli CJ, Menk J, et al. Pre- and peri-operative factors associated with chronic critical illness in liver transplant recipients. Surg Infect (Larchmt). 2020; 21:246–254
27. Spicher JE, White DP. Outcome and function following prolonged mechanical ventilation. Arch Intern Med. 1987; 147:421–425
28. Mayer-Oakes SA, Oye RK, Leake B. Predictors of mortality in older patients following medical intensive care: The importance of functional status. J Am Geriatr Soc. 1991; 39:862–868
29. Tabak OF, Clark RA, Lepre D, et al. Nosocomial central venous catheter-related infections in intensive care units
. J La State Med Soc. 1996; 148:399–402
30. Marrie TJ, Wu L. Factors influencing in-hospital mortality in community-acquired pneumonia: A prospective study of patients not initially admitted to the ICU. Chest. 2005; 127:1260–1270
31. Fried LP, Tangen CM, Walston J, et al.; Cardiovascular Health Study Collaborative Research Group. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001; 56:M146–M156
32. McDermid RC, Stelfox HT, Bagshaw SM. Frailty in the critically ill: A novel concept. Crit Care. 2011; 15:301
33. Stabenau HF, Becher RD, Gahbauer EA, et al. Functional trajectories before and after major surgery in older adults. Ann Surg. 2018; 268:911–917
34. Cohen RI, Eichorn A, Silver A. Admission decisions to a medical intensive care unit are based on functional status rather than severity of illness. A single center experience. Minerva Anestesiol. 2012; 78:1226–1233
35. Brown SM, Wilson E, Presson AP, et al.; with the National Institutes of Health NHLBI ARDS Network. Predictors of 6-month health utility outcomes in survivors of acute respiratory distress syndrome. Thorax. 2017; 72:311–317
36. Rydingsward JE, Horkan CM, Mogensen KM, et al. Functional status in ICU survivors and out of hospital outcomes: A cohort study. Crit Care Med. 2016; 44:869–879
37. Gerth AMJ, Hatch RA, Young JD, et al. Changes in health-related quality of life after discharge from an intensive care unit: A systematic review. Anaesthesia. 2019; 74:100–108
38. Guérin C, Reignier J, Richard JC, et al.; PROSEVA Study Group. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013; 368:2159–2168
39. Papazian L, Forel JM, Gacouin A, et al.; ACURASYS Study Investigators. Neuromuscular blockers in early acute respiratory distress syndrome. N Engl J Med. 2010; 363:1107–1116
40. Angus DC, Kelley MA, Schmitz RJ, et al.; Committee on Manpower for Pulmonary and Critical Care Societies (COMPACCS). Caring for the critically ill patient. Current and projected workforce requirements for care of the critically ill and patients with pulmonary disease: Can we meet the requirements of an aging population?. JAMA. 2000; 284:2762–2770
41. Iwashyna TJ, Netzer G. The burdens of survivorship
: An approach to thinking about long-term outcomes after critical illness. Semin Respir Crit Care Med. 2012; 33:327–338
42. Wunsch H, Guerra C, Barnato AE, et al. Three-year outcomes for Medicare beneficiaries who survive intensive care. JAMA. 2010; 303:849–856
43. Magnette C, De Saint Hubert M, Swine C, et al. Functional status and medium-term prognosis of very elderly patients after an ICU stay: A prospective observational study. Minerva Anestesiol. 2015; 81:743–751
44. Detsky ME, Harhay MO, Bayard DF, et al. Discriminative accuracy of physician and nurse predictions for survival and functional outcomes 6 months after an ICU admission. JAMA. 2017; 317:2187–2195
45. Katz S, Ford AB, Moskowitz RW, et al. Studies of illness in the aged. The index of ADL: A standardized measure of biological and psychosocial function. JAMA. 1963; 185:914–919
46. Wunsch H, Angus DC, Harrison DA, et al. Variation in critical care services across North America and Western Europe. Crit Care Med. 2008; 36:2787–2793, e1–e9
47. Doiron KA, Hoffmann TC, Beller EM. Early intervention (mobilization or active exercise) for critically ill adults in the intensive care unit. Cochrane Database Syst Rev. 2018; 3:CD010754
48. Schweickert WD, Kress JP. Implementing early mobilization interventions in mechanically ventilated patients in the ICU. Chest. 2011; 140:1612–1617
49. Balas MC, Vasilevskis EE, Olsen KM, et al. Effectiveness and safety of the awakening and breathing coordination, delirium monitoring/management, and early exercise/mobility bundle. Crit Care Med. 2014; 42:1024–1036
50. Hsieh SJ, Otusanya O, Gershengorn HB, et al. Staged implementation of awakening and breathing, coordination, delirium monitoring and management, and early mobilization bundle improves patient outcomes and reduces hospital costs. Crit Care Med. 2019; 47:885–893
51. Skrobik Y, Ahern S, Leblanc M, et al. Protocolized intensive care unit management of analgesia, sedation, and delirium improves analgesia and subsyndromal delirium rates. Anesth Analg. 2010; 111:451–463
52. Vakayil V, Ingraham NE, Robbins AJ, et al. Epidemiological trends of surgical admissions to the intensive care unit in the United States. J Trauma Acute Care Surg. 2020 May 6. [online ahead of print]