Optimizing recovery following critical illness: A systematic review of home-based interventions

Abstract

Around 5 million Americans are treated in an intensive care unit (ICU) annually. Upon discharge, it is not uncommon for ICU survivors to experience psychological, physical, or cognitive symptoms related to their ICU stay. Home-based interventions have been touted as a potential treatment modality for post-ICU sequelae. However, limited evidence exists regarding the effectiveness of home-based interventions for patients in the post-ICU recovery period. As such, the purpose of this review was to aggregate and summarize the findings of studies focused on post-ICU rehabilitation, following critical illness, delivered in the home setting. A literature search was performed in MEDLINE, CINAHL, EMBASE, APA PsycINFO, and Google Scholar. Studies were included if they: used a RCT or quasi-experimental study design; included participants aged ≥18 years discharged home from an ICU; examined the effectiveness of a home-based, post-ICU intervention; were published in English after the year 2010; and were peer-reviewed. Nine studies met inclusion criteria. Sample sizes ranged from 21 to 386, with most participants receiving mechanical ventilation. Target outcomes included: physical function, psychological well-being, cognitive function, quality of life, and healthcare utilization. Interventions included face-to-face, web-based, telephone, or self-directed activities. Findings of included studies were mixed or inconclusive. Limitations of this review include: inclusion of only adult ICU survivors, exclusion of Post-Intensive Care Syndrome as a search term, and search restricted to pre-pandemic studies. Findings suggest a need for more rigorous research to develop and test home-based interventions.

Background

Over 5 million people in the United States require care in an intensive care unit (ICU) each year.^1,2 Despite the complexity of care needed for those requiring an ICU stay, expectations for survival are realistic, given recent advances in therapeutics and care processes.³ Care options for patients following hospitalization include long-term care, long-term acute care,^4,5 acute rehabilitation, and discharge to home.⁶ Patients discharged to home often have a higher level of function during their hospital stay⁷ or require fewer long-term institutional care resources following ICU treatment. However, patients discharged to home may still require support services from an interdisciplinary team to address potential psychological, physical, or cognitive impairments that may arise during the post-ICU recovery period.⁸

The physical, psychological, and cognitive symptoms, which are prolonged and persistent following an ICU stay, are collectively known as Post-Intensive Care Syndrome.^9,10 The impact of Post-Intensive Care Syndrome can be devastating for patients, their families, and the community. Follow-up services to assure a favorable recovery trajectory are vital to achieving pre-illness function, ability to return to work,^11,12 and heighten quality of life.^13,14 Despite becoming more common, post-ICU follow-up and care are structurally diverse, inconsistently available, and offer mixed evidence for effectiveness in alleviating Post-Intensive Care Syndrome symptoms.^15,16

Home health care is a group of health care services delivered in the home for illness or injury.^8,17 These services include support for activities of daily living, monitoring for serious illness progression, rehabilitation services, and resources for end-of-life care. Home health care is the most rapidly growing option for post-acute care service; over 40% of Medicare beneficiaries are prescribed home health care services from a home health care agency.^18,19 Previous research studies have found home health care services to be safe, effective, and efficacious for a number of medical conditions and post-hospitalization needs (i.e., cardiovascular events, orthopedic surgery, chronic wound care, and chronic obstructive pulmonary disease).^20,21 Similarly, home care services are an option for patients discharged home following a critical illness or injury that required an ICU stay.²² With more healthcare providers offering telehealth delivery options to their patients, as opposed to in-person facility visitation, home health care services have only grown in popularity—as home health care offers expanded care options, greater access to providers, and possibility of quarantine for those with infectious disease concerns (e.g., SARS-CoV-2).

Interventions delivered in the home (i.e., home-based interventions) decrease risk of travel to post-ICU care sites for patients who are physically vulnerable, who have limited social support, or whose caregivers are ill-prepared for the caregiving of family members.²³ Home-based interventions can be particularly attractive to patients who live at a distance, are returning to work, or who come from rural or underserved areas.¹⁹ However, given the unique nature of post-ICU sequelae, limited evidence of the effectiveness of home-based interventions for patients in the recovery period following an ICU stay is available. As such, the purpose of this literature review was to aggregate and summarize the findings of randomized controlled trials (RCTs) and quasi-experimental design studies examining the effects of home-based interventions on post-ICU health outcomes in adults following discharge.

Methods

This review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Four databases (MEDLINE, Cumulative Index to Nursing and Allied Health Literature [CINAHL], Excerpta Medica Database [EMBASE], and APA PsycINFO) were searched in February 2020 using a combination of the following search terms: “intensive care,” “ICU,” “critical care,” “rehab*,” “intervention,” “treat*,” “home based,” “home-based,” “home care,” “post,” and “discharge.” Databases were accessed through the EBSCOhost database system. Boolean connectors “AND” and “OR” were used to link search terms. The Boolean connector “NOT” was used to exclude references relating to populations, settings, or interventions deemed not applicable for this review (i.e., “NICU,” “neonatal intensive care unit,” “newborn,” “neonate,” “child*,” “pediatric*,” and “paediatric*”). Utilization of the web-based software Covidence accounted for, and removed, duplicate publications. Search terms were created to capture results in both American and British English.

The initial search yielded 1,314 studies. A brief scoping search of Google Scholar (i.e., search of first 100 studies) yielded an additional eight studies. After 14 duplicates were removed from the combined searches, 1,308 articles remained. Citation tracking was used to identify references not found through the aforementioned searches. During the study selection process, a PRISMA flow diagram was completed (see Figure 1.). Studies were included (1) if they used a RCT or quasi-experimental study design; (2) if participants were adults aged ≥18 years who were discharged home from an ICU; (3) if they examined the effectiveness of a home-based, post-ICU intervention; (4) if they were available in English and published no earlier than the year 2010; and (5) if they were peer-reviewed. Studies were excluded if they (1) enrolled pediatric patients; (2) used a solely hospital-based or ICU-based intervention; (3) focused on patients discharged or transferred to a skilled nursing facility, rehabilitation center, or hospital step-down unit; (4) included patients treated for an accident or non-critical illness; or (5) focused on interventions for caregivers of post-ICU patients. A coding guide detailing desired study characteristics was established a priori. After identification, all selected studies were coded based on desired characteristics and underwent a brief quality assessment.

Figure 1.

PRISMA flowchart

Screening and Selection

Two authors (AH, JT) independently evaluated applicable citations generated through the search strategy at the title, abstract, and full-text level. Disagreements were resolved by discussion, and when no consensus was reached, a third party was selected to arbitrate (JC). Independent screening and review of articles utilized the web-based software platform Covidence. A total of 1,308 articles were identified for screening at the title and abstract level. After screening, 1,270 articles were excluded due to irrelevant titles or abstracts. Full-text review was initiated for 38 articles, of which 29 were further excluded, yielding a total of 9 studies for final analysis (see Supplementary Material, Table 1).

Quality

Overall quality and risk of bias for included studies were independently assessed using The Joanna Briggs Institute Critical Appraisal Checklist for Randomized Controlled Trials. Articles were rated across five domains: 1) selection bias, 2) blinding, 3) study design, 4) methodological quality, and 5) statistical analysis propriety (see Table 1). Eight studies were RCTs, with two^24,25 self-identifying as pilots. All studies had experimental and control groups that were similar at baseline.

Table 1.

Sample Eligibility Criteria for Included Studies.

Author	Purpose	Design	Country	Setting	Participant Characteristics (Exclusion)
Battle et al31	Investigate impact of a 6-week, supervised exercise program on cardiopulmonary fitness, balance, muscle strength, anxiety, and depression following ICU discharge.	Single-center, parallel RCT.	United Kingdom	Mixed	Inclusion—medical/surgical ICU patient for >48 hours. Exclusion—enrolled in formal pulmonary rehabilitation program,
Cox et al24	Evaluate the feasibility and acceptability of a self-directed mindfulness app, compared to a telephone-based mindfulness or web-based critical illness education program, on psychological distress among ICU survivors.	Pilot RCT.	United States	Home	Inclusion—ICU management for ≥24 hours, Exclusion—pre-existing/current cognitive impairment, treatment of severe mental illness within 6 months of ICU admission, hospitalized within 3 months of ICU admission, active substance abuse, expected survival ≤6 months, ICU length of stay ≥30 days, not discharged to home, required complex medical care upon discharge, poor English fluency, lack of reliable phone or internet access.
Elliott et al30	Test the effect of an individualized 8-week, home-based physical rehabilitation on health-related quality of life and physical function in survivors of a critical illness.	Multi-center RCT.	Australia	Home	Inclusion—ICU length of stay >48 hours, received mechanical ventilation >24 hours, discharged home, lived inside a commutable area, cognitively able to participate. Exclusion-neurological/spinal or skeletal dysfunction, receiving palliative care, organized rehabilitation,
Jackson et al25	Evaluate the feasibility of a structured 12-week, in-home rehabilitation program in a cohort of ICU survivors with a cognitive and/or physical impairment.	Single-site, feasibility, pilot RCT.	United States	Home	Inclusion—English-speaking, patients of a medical or surgical ICU, enrolled in BRAIN-ICU study. Exclusion—cumulative ICU time >5 days in past 30 days, severe cognitive or neurodegenerative disease, ICU admission post cardiopulmonary resuscitation with suspected anoxic injury, active substance abuse or psychotic disorder, recent suicidal gesture necessitating hospitalization within 6 months, actively blind or deaf, unable to speak English, moribund or requiring palliative care, lived outside a commutable area, was a prisoner or homeless, experienced respiratory failure or shock >72 hours prior to admission, had cardiac bypass within 3 months.
Jensen et al28	Test the effectiveness of a post-ICU recovery program compared to usual care during the first year after an ICU discharge.	Pragmatic, non-blinded, multicenter, parallel-group RCT.	Denmark	Mixed	Inclusion—Danish-speaking, received mechanical ventilation >48 hours Exclusion—not oriented in personal data according to verbal response in Glasgow Coma Score, had detected delirium in Confusion Assessment Methods for the ICU at randomization, enrolled in other follow-up study.
Jónasdóttir et al26	Describe a structured, 3-month nurse-led follow-up of patients after ICU discharge and follow-up effects on health status.	Prospective, quasi-experimental, with a non-equivalent control group.	United Kingdom, Iceland	Mixed	Inclusion—acute or elective ICU admission, length of stay >72 hours. Exclusion—unlikely to survive general ward stay, unlikely to be alert or able to communicate after discharge, had dementia, active drug and/or alcohol user.
McDowell et al27	Determine the effect of a personalized exercise program on physical function in patients discharged to home following a critical illness.	Multicenter, prospective phase 2, RCT.	Ireland	Mixed	Inclusion—mechanical ventilation >96 hours, being discharged home, medically fit to participate, not participating in another rehabilitation program.
Shelly et al29	Determine the effect of a 4-week, individualized home-based exercise on health-related quality of life post-ICU discharge.	Block RCT.	India	Mixed	Inclusion—Mechanically ventilated for 24 hours, ICU LOS ≥ 48 hours. Exclusion—had musculoskeletal trauma, neurological deficit, psychological/psychiatric disorder, malignancies, organophosphorus poisoning, substance abuse, ICU admission in <6 months.
Vitacca et al32	Determine if a 6-month home physiotherapy program can improve outcomes in survivors of a critical illness.	Randomized, prospective controlled trial.	Italy	Home	Inclusion—were discharged post-acute CC patients with an episode of acute respiratory failure in <3 months, had an ICU length of stay >20 consecutive days with difficulty weaning from mechanical ventilation, completed a step-down unit rehabilitation program, had a caregiver living with them. Exclusion—≥18 years of age, had cardiovascular instability, had amyotrophic lateral sclerosis, remained critically ill, had electrodiagnostic evidence of myopathy/polyneuropathy, had organ failure, required, hemodialysis, had a life expectance >6 months, lived outside a commutable area, were transferred to another hospital or nursing home, refused participation, or were enrolled in another trial.

Note. RCT = randomized control trial; ICU = intensive care unit; CC = critical care; MV = mechanical ventilation; LOS = length of stay.

All studies had the following inclusion critiera: Age ≥18, admitted to ICU.

While eight studies self-identified as an RCT, one study²⁶ self-identified as a non-equivalent control group design. Due to the nature of the intervention, participant blinding was not feasible or did not occur in four of the studies.^24,26–28 Attempts to ensure blinding of participants in the remaining five articles is unclear. All studies utilized appropriate measurement selection, study design, and statistical analysis of outcomes. Follow-up, albeit at varying time points, was also completed at least once across all nine studies.

Results

Sample

Universally, samples were drawn from adults ≥18 years of age who received mechanical ventilation in an ICU, were discharged to home, and lived within a commutable distance. Diagnostic criteria for inclusion in this sample were not specific but reflected the case mix of the ICU setting. For instance, participants who were cared for in a medical-surgical ICU would be included. None of the settings were specifically specialty ICUs (e.g., cardiothoracic or trauma). Duration of mechanical ventilation as an inclusion criterion varied from a minimum of 24^24,29,30 to 96²⁷ hours and an ICU length of stay ranging from 24 hours³¹ to 20 consecutive days.³² Two studies excluded patients who had a hospital admission or an ICU stay within a specified time period.^25,29,32 Exclusion criteria for these studies varied widely, but most excluded patients with pre-illness neurocognitive or psychiatric conditions such as dementia, substance abuse, suicidal ideation, or profound communication disorders.^{24–26,29,32} Studies that targeted physical recovery (n = 3)^25,29,32 excluded patients with neurological or skeletal dysfunction. Several studies (n = 3)28,31,32 excluded patients who were participating in other studies or who were already receiving rehabilitation services. Sample sizes ranged from 21 to 386. Table 1 includes further information about eligibility criteria for included studies. Studies were conducted in a variety of geographic areas, including: the US (n= 2),^24,25 Australia (n=1),³⁰ the United Kingdom (n=2),^26,31 the European Union (n=4),^26–28,32 and India (n=1).²⁹ Intervention settings were divided into home-based exclusively (n=4)^24,25,29,30 or combination home and healthcare/community facility (n=5).^{26–28,31,32} There are no studies that indicate a threshold of “exposure” to ICU therapies, such as mechanical ventilation or severity of illness, that would influence a potential for recovery and ability to participate in home-based rehabilitation.

Design of the Studies

Eight studies were prospective RCTs, while one²⁶ self-identified as having a non-equivalent control group design. One study was a single site, pilot RCT designed to demonstrate feasibility.²⁴ Three studies were multi-center.^26,27,30 Seven of the studies used usual care as the control arm.^{25,26,28–32} Most studies represented multiple disciplines, including: nursing, medicine, physical therapy, occupational therapy, and psychology. Duration of interventions ranged from 4 weeks (1 month) to 40 weeks (10 months)—with most intervention sessions occurring either weekly or biweekly. The timing of follow-up measures was inconsistent and varied based on the study. Studies used varying follow-up time points, which ranged from 1 week³⁰ post-ICU discharge to 1 year.^26,28,31 A majority of studies followed up with participants at 3 months^24–26,28 or 6 months.^26,27,31,32 Two studies indicated that they performed baseline measures at hospital discharge.^26,32

Studies utilized a wide range of measures for physical function, psychological well-being, and cognitive function. Five studies targeted improved physical function,^27,29–32 while 3 targeted psychological well-being and quality of life.^24,26,28 One study targeted both cognitive and physical function.²⁵ Measures common to these studies include the Six-Minute Walk Test (6 MWT),³³ the EuroQoL-5 Dimension (EQ-5D),³⁴ and the 36-item Short Form Health Survey (SF-36).³⁵ See Table 2 for a full list of measures utilized in each respective study. See Figure 2 for a breakdown of the measurement strategies utilized.

Table 2.

Characteristics of the Home-based Interventions in Included Studies.

		Intervention
Author	Control	Duration	Interval	Components	Type/Target?	Outcome Measures
Battle et al31	UC	6 weeks	Twice weekly; one weekly session at home independently	Cardiopulmonary, balance and strengthening exercises Instructions for completing	Physical function	6MWT, BBS, Jamar Dynamometer, HADS.
Cox et al24	—	1 month	Weekly	Mindfulness mobile app Or Telephone-based mindfulness Or Web-based critical illness education.	Psychological wellbeing	Feasibility, CSQ, SSU, PHQ-9, GAD-7, PTSS, Euro-QOL VAS, PHQ-15 CAMS-R, Brief COPE
Elliott et al30	UC	8 weeks	Weekly PT home visits weeks 1, 3, 5 Phone visits weeks 2,4,5,7	Strength training, endurance exercises	Physical function	SF-36, 6MWT
Jackson et al25	UC	12 weeks In person at home alternating with tele-visit; supplemental phone call;	Weekly	Physical exercise, cognitive training, functional rehab	Cognitive and physical function	TOWER tests, FAQ, TUG, MMSE, DEX, ABCS, Katz ADL
Jensen et al28	UC	10 months First visit 1–3 months post hospital discharge; follow up at 5 months and 10 months post hospital discharge	2–11 months	3 consultations with nurse—individualized illness narratives and completion of “reflection” sheets for second and third visits	Psychological wellbeing	SF-36, SOC-13, HADS, Use of health care system, HTQ-IV
Jónasdóttir et al26	UC	3 months	Contact at 1 week, appointment 3 months post-discharge	(1) Booklet delivered at ICU discharge; (2) ward visits; (3) phone contact 1 week after hospital discharge to home at and (4) an appointment 3 months after discharge from the ICU—could include revisiting ICU, or could take place in home or other setting	Quality of life	SF-36, 1 year post hospital discharge
McDowell et al27	—	6 weeks	Weekly 2 supervised in hospital gym and 2 unsupervised	Aerobic exercise, hand strengthening, dexterity	Physical Function	RMI, Hydraulic hand dynamometer, 9-HPT, ISWT, SF-36, FLP, EuroQol-5D-5L, MRC
Shelly et al29	UC	4 weeks	Daily × 5 days per week at home	Physical and breathing exercises, 30–40 minutes duration Booklet describing exercises	Physical Function	SF-36
Vitacca et al32	UC	6 months	Twice daily exercise; 60–90 minutes sessions; 7 days per week at home	Pulmonary rehab; Included—bronchial hygiene, physical activity, advanced physical activity; Physiologic monitoring for feedback and safety during exercise periods monitored by family caregiver Therapist home visit at end of 1 month and as needed.	Physical function	MIP/MEP, FEV, ABG, RR BADL MRC-SS, EQ-5D, Survival, Readmission, ED visits, Adherence to protocol, Decannulation Safety, Patient satisfaction, Costs

Note. UC = usual care; ED = emergency department; BBS = Berg Balance Scale; 6MWT = 6 minutes Walk Test; PHQ-9 = Patient Health Questionnaire-9; GAD-7 = General Anxiety Disorder-7; PTSS = Posttraumatic Stress Survey; Euro-QOL VAS = European Quality of Life Visual Analog Scale; PHQ-15 = Patient Health Questionnaire-15; CAMS-R = Cognitive and Affective Mindfulness Scale-Revised; Brief COPE = Brief Coping Orientation to Problems Experienced Inventory; SF-36 = 36-Item Short Form Survey; TUG = Timed Up and Go Test; MMSE = Mini-Mental State Examination; SOC-13 = Validation of Sense of Coherence-13 Scale; HADS = Hospital Anxiety and Depression Scale; HTQ-IV = Harvard Trauma Questionnaire-Part IV; RMI = Rivermead Mobility Index; ISWT = Incremental Shuttle Walk Test; FLP = Functional Limitations Profile; EuroQol-5D-5L = European Quality of Life-5 Dimension-5-Levels; MRC = Medical Research Council Scale for Muscle Strength; MIP/MEP = maximal inspiratory pressure/maximal expiratory pressure; FEV = forced expiratory volume; ABG = air blood gases; RR = respiratory rate; BADL = Basic Activities of Daily Living; MRC-SS = Medical Research Council Sum Score; EQ-5D = European Quality of Life-5 Dimension; SOC = Validation of Sense of Coherence; HRQOL = Health-Related Quality of Life; SSU = System Usability Scale; DEX = Dysexecutive Questionnaire; ABCS = Activities Balance and Confidence Scale; FAQ = Functional Activities Questionnaire; Katz ADL = Katz Activities of Daily Living Scale; 9-HPT = 9 Hole Peg Test; CSQ = Client Satisfaction Questionnaire.

Figure 2. Domains of outcome measures.

6MWT = Six Minute Walk Test; 9-HPT = Nine Hole Peg Test; ABCS = Activities Balance and Confidence Scale; ABG = arterial blood gases; BBS = Berg Balance Scale; Brief COPE = Brief Coping Orientation to Problems Experienced Inventory; CAMS-R = Cognitive and Affective Mindfulness Scale-Revised; CSQ = Client Satisfaction Questionnaire; CIRS = Cumulative Illness Rating Scale; DEX = Dysexecutive Questionnaire; EuroQol-5D-5L = European Quality of Life-5 Dimension-5-Levels; Euro-QOL VAS = European Quality of Life Visual Analogue Scale; FAQ = Functional Activities Questionnaire; FEV = forced expiratory volume; FLP = Functional Limitations Profile; GAD-7 = General Anxiety Disorder-7; EuroQol-5D-5L = European Quality of Life-5 Dimension-5-Levels; HTQ-IV = Harvard Trauma Questionnaire-Part IV; HADS = Hospital Anxiety and Depression Scale; ISWT = Incremental Shuttle Walk Test; Katz ADL = Katz Activities of Daily Living Scale; MIP/MEP = maximal inspiratory pressure/maximal expiratory pressure; MRC = Medical Research Council Scale for Muscle Strength; MRC-SS = Medical Research Council Sum Score; MMSE = Mini–Mental State Examination; PHQ-9 = Patient Health Questionnaire-9; PHQ-15 = Patient Health Questionnaire-15; PTSS = Posttraumatic Stress Scale; RR = respiratory rate; RMI = Rivermead Mobility Index; SF-36 = 36-Item Short Form Survey; SOC-13 = Validation of Sense of Coherence-13 Scale; SUS = System Usability Scale; TUG = Timed Up and Go

The interventions in these studies range from single component to multi-level and multi-context (home+visits). Three studies employed nurse-led interventions,^26,28,30 while the rest were led by other disciplines such as medicine,^24,32 psychology,²⁵ or rehabilitation science.^27,29,31 Several studies utilized telephone^24,30-or web-based interventions,²⁴ while others relied on face-to-face^{26–28,30–32} or self-directed activities.^26–29,32 Given the complexity of the interventions in these studies, it is important to assure consistency of protocol application using a plan for intervention fidelity. Careful attention to intervention fidelity allows greater certainty that significant findings were actually due to the intervention. Documentation of these procedures allows for replication of the intervention as it was designed—thus facilitating translation of effective interventions from research settings to clinical practice. Future investigators can become aware of, and control, the impact of nonspecific treatment effects and any unintended processes on the intervention. Intervention fidelity was assured in only one study; McDowell et al²⁷ addressed training updates for interventionists and weekly team discussions about deploying the intervention. Jackson et al,25 conducted weekly intervention fidelity meetings with the principal investigators to assure adherence to study protocols. Although several studies documented adherence to the intervention protocol by study staff^,24,26,31 only 2 studies provided evidence of participant adherence to unsupervised activities at home.^30,32 No other studies documented a plan or described any strategies to assure intervention fidelity.

Main Findings of the Studies

Several studies (n = 3)^26,30,31 showed no significant differences in physical function at the conclusion of the study. In the study by McDowell et al,²⁷ the intervention group showed improvements in physical function and self-efficacy at the first measurement. These changes were not sustained at 6 months. However, a simple exercise intervention by Shelly et al²⁹ demonstrated significant differences in the physical and mental health components of the SF-36 at 1 month follow-up. Pulmonary function improved for intervention groups in the study by Vitacca et al.³² The study by Jackson et al²⁵ showed statistically significant changes in cognitive function at 3 months, while the intervention group, in the study by Cox et al²⁴ showed clinically significant changes in psychological and physical symptoms at 3 months. The intervention group in the study by Jensen et al²⁸ showed no difference in quality of life or psychological symptoms upon study completion.

Discussion

Our review targeted studies that tested interventions focused on rehabilitation following critical illness that were delivered exclusively, or primarily, in the home setting. The outcomes of studies that met review criteria were mixed and likely a result of the infancy of the science. It is imperative to note, however, that impairments experienced by ICU survivors are heterogeneous and the course of recovery is unpredictable.^36,37 The lack of significant or sustained improvements in physical, psychological, and cognitive function may be due to the heterogeneity of study design, measurement, timing, duration, or intervention targets. In addition, ICU survivors are a heterogeneous population, ranging in pre-illness and chronic illness burden, age, access to resources, and functional differences at hospital discharge, which make linear recovery improbable. Our findings, therefore, need cautious interpretation due to the influence of these heterogeneous elements.

This review is timely given the impact of the SARS-CoV-2 pandemic. With the emergence of SARS-CoV-2, colloquially known as COVID-19, there has been an increased demand for home-based, and post-ICU, care via telehealth services. In the U.S., from 2010 to 2017, the percentage of hospitals connecting healthcare providers to patients through various telehealth technologies rose from 35% to 76%.³⁸ This transition from traditional, in-person facility visitation to telehealth delivery has only continued to rise with the persistence of SARS-CoV-2.³⁹ In this review, none of the selected studies delivered an intervention via telehealth technology. While home-based interventions have the potential to provide benefit for post-ICU patients, they are not without risk to interventionists, healthcare providers, and patients themselves. Prior research has cited several household-related hazards that make providing home health care a challenge.⁴⁰ With this in mind, it is important to note that not one of the studies included in this review sought to determine differences in safety, feasibility, or efficacy of in-person versus telehealth home-based intervention delivery.

As a result of acute critical illness and hospitalization, nearly all ICU survivors experience one or more new post-ICU problems that include impairment of physical, psychological, or cognitive function.⁴¹ Recognition and management of Post-Intensive Care Syndrome, or post-ICU symptoms, may be difficult for specialists responsible for these patients during the recovery period.⁴² Despite the prevalence of psychological or cognitive dysfunction, most of the interventions identified in this review targeted physical recovery. This may be due to the make-up of the study team. For instance, when led by physical therapists, interventions tended to prioritize physical or pulmonary function improvement. Although ICU-acquired weakness is common, conducting physical exercise interventions can be difficult, as direct observation and coaching by physical therapists are not possible. Participant adherence to prescribed exercise programs may vary.

Despite these issues, patients who are discharged home following an ICU stay have better mobility outcomes at 1 year compared to those discharged to a care facility.⁴³ Home-based recovery is a model of care that makes use of an interdisciplinary team of professionals, decreases cost, increases patient satisfaction, and improves health outcomes.³⁶ However, home-based care for ICU survivors is bound by local and federal regulations. Home-based care services also vary in quality and levels of service. Reimbursement for home care services, either through private insurance or Medicare, relies on strict inclusion criteria and is often short-term rather than long-term. The lack of consistency in inclusion criteria makes these studies difficult to replicate. For instance, durations of mechanical ventilation and length of ICU or hospital stay varied. The justification of 24 hours versus 48 hours for either duration of mechanical ventilation and length of stay are not described either in these studies or in the extant literature within the ICU community.

Despite the inconsistency and variability of study findings, home-based interventions are a vital and desired care option for many adult ICU patients pending discharge to home. The value of home visits is that healthcare providers can review contextual factors that may not be possible with other modes of delivery during rehabilitation. Further, the home environment is familiar to the patient and can be conducive to full deployment of interventions. For patients with post-ICU cognitive impairment, post-ICU care facilities can be unfamiliar, and resultantly increase the risk of manifesting confusion or agitation.⁴⁴ Additionally, travel to post-ICU care facilities can be a challenge for patients experiencing functional impairment, who fear exposure to infectious disease (e.g., SARS-CoV-2), who live outside a commutable distance, or who have caregivers with schedules that conflict with facility hours of operation.⁴⁴

We recommend conducting additional studies of home-based interventions to improve recovery for ICU survivors. These might include testing interdisciplinary strategies that would target a range of outcomes, such as: ICU-acquired weakness and mobility, psychological well-being, and cognitive function. These studies should include strategies to sustain results beyond the post-ICU recovery period.

Limitations

Several limitations are noted in this review. First, our search was limited to adult ICU survivors, as we recognize the difference in recovery trajectories and configuration of home care services⁴⁵ for those in the pediatric population. Second, we did not examine Post-Intensive Care Syndrome as a search term for this review, as we wanted to capture articles focused on post-ICU sequelae and not Post-Intensive Care Syndrome exclusively, thus decreasing the inclusion of studies that might have directly aimed to resolve Post-Intensive Care Syndrome in ICU survivors. Third, this search took place shortly before the onset of the SARS-CoV-2 pandemic, resultantly causing potentially applicable articles to be missed if they related to, or focused solely on, SARS-CoV-2. Fourth, none of these studies captured the unique problems of ethnic or cultural minorities; most included studies were conducted in Western countries. Finally, we chose to review only RCTs or quasi-experimental studies conducted in the past 10 years, in order to capture studies with recent advances in therapeutics and care processes. This design restriction may have eliminated potentially older, yet still innovative, interventions from our analysis.

Conclusion and Future Directions

In conclusion, surviving critical illness is becoming more commonplace—and recovery at home is becoming a more desired post-ICU care modality. Findings of this review suggest a need for more rigorous research geared toward the development, evaluation, and implementation of home-based interventions for adult patients discharged to home from an ICU. Additional research efforts should be made to examine home-based intervention delivery methods (e.g., telehealth or in-person) for feasibility and efficacy. For those post-ICU patients recovering from a critical illness, home-based interventions have the potential to alleviate persistent physical, psychological, and cognitive symptoms originating from their ICU stay. Home-based interventions also have the ability to decrease risk of hospital readmission, provide more rapid follow-up care, cut costs, and increase patient and healthcare provider satisfaction.⁴⁴ Patients discharged home from the ICU have a number of unique challenges and recovery hurdles. However, home-based interventions have the potential to mitigate some of these challenges and further support ICU survivors in their return to pre-ICU function.

Table 3.

Sample Size and Results of Reviewed Studies.

Author	Sample Size	Results
Battle et al31	n = 60	No differences in 6MWT at any time point (p > .05). Treatment group with significantly improved anxiety levels and balance in the treatment group at 12 months (p = .006 and p = .040).
Cox et al24	n = 80	Feasibility, acceptability, and usability exceeded targets. Clinically significant changes for group in depression, anxiety, post-traumatic stress, and physical symptoms.
Elliott et al30	n = 195	Both groups improved physical function and 6MWT distance at 8 and 26 weeks. However, no significant group effects or group by time interactions for physical function, 6MWT and SF-36 summary scores.
Jackson et al25	n = 21	At 3 months, intervention group significantly improved executive function and functional status.
Jensen et al28	n = 386	No difference in HRQOL, self-reported SOC, anxiety, depression, or utilization of health care services.
Jónasdóttir et al26	n = 148	No difference in patient health status for any time point to 12 months.
McDowell et al27	n = 60	No difference between groups in the primary outcome (SF-36). The intervention group improved significantly in SF-36 physical, ISWT, functional limitations profile self-efficacy and readiness to exercise. Improvements were not sustained at 6 months except for readiness to exercise. Other outcomes were not significantly different.
Shelly et al29	n = 35	Differences in physical and mental components of SF-36 statistically significant at 4 weeks.
Vitacca et al32	n = 48	For cardiorespiratory patients only: MIP/MEP, QoL, FEV, dyspnea and respiratory rate improved significantly for intervention group. Decannulation of treatment group was significantly different for treatment group. Majority of treatment group was satisfied. Cost was 459 euro/month.

Note. QoL = Quality of Life; 6MWT = 6 Minutes Walk Test; SF-36 = 36-Item Short Form Survey; HRQOL = Health-Related Quality of Life; SOC = Validation of Sense of Coherence; ISWT = Incremental Shuttle Walk Test; MIP/MEP = maximal inspiratory pressure/maximal expiratory pressure; FEV = forced expiratory volume.