Journal Logo

Feature Article

Mental Health Interventions to Improve Psychological Outcomes in Informal Caregivers of Critically Ill Patients: A Systematic Review and Meta-Analysis*

Cherak, Stephana J. MSc, PhD(c)1–4; Rosgen, Brianna K. BHSc1–4; Amarbayan, Mungunzul BSc, MPP2; Wollny, Krista MN, PhD(c)1,3,5; Doig, Christopher J. MD, MSc1–3; Patten, Scott B. MD, PhD1,3,4,6; Stelfox, Henry T. MD, PhD1–3; Fiest, Kirsten M. PhD1–4,6

Author Information
doi: 10.1097/CCM.0000000000005011


Critically ill patients admitted to ICUs frequently rely on informal caregivers (e.g., family, friends) to act as surrogate decision-makers and emotional supporters (1,2). Caregivers of critically ill patients have long-lasting negative psychologic sequalae following an ICU stay including anxiety, depression, and post-traumatic stress disorder (PTSD) (3,4).

Mental health interventions improve psychologic outcomes in informal caregivers of critically ill patients (5). It is unknown whether these interventions (e.g., to improve knowledge about mental health [6] or to prevent mental disorders [7,8]) affect informal caregivers by reducing negative (e.g., anxiety) or increasing positive (e.g., humanity) psychologic outcomes.

We performed a systematic review and meta-analysis of interventional trials conducted in adult ICU, neonatal ICU, and PICU to determine how mental health interventions affect negative and positive psychologic outcomes in informal caregivers of critically ill patients. We hypothesized that mental health interventions improve psychologic outcomes in informal caregivers of critically ill patients.


This systematic review was conducted and reported to the Preferred Reporting Items for Systematic Review and Meta-Analyses reporting guideline (9) (Supplemental Table 1,, registered on PROSPERO (CRD42019133700) prior to data extraction.

Outcomes Measures

The main outcome was informal caregiver psychologic outcomes categorized: 1) negative psychologic outcomes (anxiety [state only], depression, PTSD, distress [discomforting emotional state in response to stressor or demand], and burden) as defined by the American Psychologic Association (10) or others (11,12) and 2) positive psychologic outcomes (courage, humanity, justice, transcendence, temperance, and wisdom and knowledge) as defined in the Character Strengths and Virtues book (that provides a classification system and theoretical framework for six classes of positive virtues made up of 24 character strengths) (13–15). Both symptoms (i.e., self-report) and diagnoses included. Operational definitions provided in Supplemental Table 2 (

Populations, Interventions, Comparators, Settings, and Study Designs

We included quantitative and qualitative studies reporting trials targeted at informal caregivers of critically ill patients, any age. We excluded interventions for healthcare professionals. We included studies where the intervention was performed outside ICU (e.g., recovery psychology follow-up clinics). We defined: 1) a caregiver as any informal (i.e., nonclinical) person who regularly provides patient support and is in some way implicated in patient care or directly affected by patient health (e.g., family, friend) (16) and 2) critically ill patients as any persons currently or previously admitted to ICU (16).

Data Sources and Searches

Literature searches were conducted in MEDLINE (Supplemental Table 3,, Embase, PsycINFO, CINAHL, and Cochrane CENTRAL Register of Controlled Trials. Search strategies were developed with a Medical Librarian (D.L.L.), revised after preliminary results. Controlled vocabulary terms and text words relating to caregivers and critical care, interventions, and psychology were used. We searched Cochrane Database of Systematic Reviews to identify related review articles; reference lists were screened to identify studies missed in search. All databases were searched from inceptions to October 31, 2019. Reference lists of included papers were reviewed to identify potentially missed studies. No language or date limits were applied.

Study Selection

After a subset of the team achieved 100% agreement on pilot-test of 50 random citations, titles, and abstracts were reviewed independently in duplicate (S.J.C., B.K.R.). Full abstracts of five foreign language studies (three French, two Italian) were translated by a fluent study volunteer blinded to the review objective (M.C.). Any study selected by either reviewer progressed. Full-texts were reviewed independently in duplicate (S.J.C., B.K.R.); articles selected by both reviewers were included. No foreign language abstract was progressed to full-text review. Disagreements resolved by a third reviewer (M.A.). References managed in Endnote X9 (Clarivate Analytics, Philadelphia, PA).

Data Extraction and Quality Assessment

Two reviewers (S.J.C., B.K.R.) extracted data independently in duplicate using a data collection sheet developed and piloted by the review team (Supplemental Methods, Discrepancies resolved by a third reviewer (M.A., K.K.). Intervention aim categorized as caregiver experience, role, or support, based on a literature review of informal caregivers of critically ill patients by Finucane et al (17). Intervention stage was categorized to Medical Research Council and National Institutes of Health Research (18). Operational definitions provided in Supplemental Table 2 ( We used the Adapted Cochrane Risk of Bias Assessment Tool for assessing risk of bias (19). For grading quality of evidence, we used BMJ Best Practice Grading of Recommendations Assessment, Development and Evaluation (GRADE) of Evidence Tool for quantitative studies (20) and GRADE-Confidence in the Evidence from Reviews of Qualitative research for qualitative studies (21).

Data Synthesis and Analysis

We used random-effects models with inverse variance weighting to pool outcome data from each randomized controlled trial (RCT). Pooled data are reported as ratio of means (RoM [experimental group to control group]) (22). Studies were excluded from meta-analysis if additional data was not provided and we were not able to estimate mean and sd. Two-sided p value of less than 0.05 was considered statistically significant. For each outcome, we estimated the I2 (23). Magnitude of heterogeneity was quantified using τ2 (24). The I2 was viewed as a proportion of variability due to τ2; low (25–49%), moderate (50–74%), or high (≥ 75%) (25). Estimates of heterogeneity were used to guide detailed meta-analytic stratified analyses. We performed subgroup meta-analyses according to intervention type (i.e., caregiver experience, role, or support) and patient population (i.e., adult, neonatal, or pediatric). If heterogeneity diminished upon stratification, this was interpreted as evidence that the stratification variable was a source of heterogeneity. Outcomes not amenable to meta-analysis are presented by intervention type and statistically significant effects. We synthesized results reported from included qualitative studies using thematic synthesis for reviews on health research (26). We developed discrete themes that represented the findings reported in primary studies and considered these themes to generate new interpretive constructs, explanations, or hypotheses (27). We then integrated our qualitative and quantitative findings by using qualitative results to interrogate quantitative results, aiding interpretation of heterogeneity as a rich source of information and to identify research gaps (28). Data analysis was performed using STATA-IC16 (StataCorp LLC, College Station, TX) and NVivo12 (QSR International Pty Ltd., version 12, Melbourne, Australia). Additional details on data synthesis and analysis in the Supplemental Methods (


Results of the Search

We screened 11,201 unique abstracts and reviewed 460 full-text articles; 358 full-text articles were excluded, most common reasons being the study did not include an intervention (n = 251/358) or was not relevant to our research question (n = 58/358) (Fig. 1). Hand searching identified 17 studies, of which all were excluded (zero interventions). Frequency of disagreements and agreements between reviewers for the full-text screen was calculated using Cohen’s Kappa statistic (κ = 0.89).

Figure 1.
Figure 1.:
Study flow diagram. PTSD = post-traumatic stress disorder.

Included Studies

Trial characteristics are in Supplemental Table 4 (, overall results by intervention type in Table 1. Classification of trials according to intervention aim is shown in Supplemental Figure 2 ( Among 102 studies, 63 were RCTs, 16 quasi-experimental trials, and 23 uncontrolled trials. Eight studies were qualitative, of which six studies performed thematic analysis and two studies used grounded theory. Sixty-four trials were single-centered, 38 multicentered. Fifty-eight studies were in neonatal ICUs (NICUs), 38 in adult ICUs, and two in PICUs (Supplemental Fig. 1, Several studies focused on specific clinical circumstances (Supplemental Table 5, and long-term follow-up of interventions (Supplemental Table 6,

TABLE 1. - Effect of Mental Health Interventions for Informal Caregiver Psychologic Outcomes
Intervention Typea,b Negative Psychologic Outcomes Positive Psychologic Outcomes
Anxiety, n = 33 (52%)f Depression, n = 34 (54%)f Post-Traumatic Stress Disorder, n = 21 (33%)f Distress, n = 22 (35%)f Burden, n = 12 (19%)f Courage, n = 7 (11%)f Humanity, n = 4 (6%)f Justice, n = 0 (0%)f Temperance, n = 2 (3%)f Transcendence, n = 14 (22%)f Wisdom and Knowledge, n = 2 (3%)f
Caregiver experiencec 18 20 13 15 6 5 2 0 0 7 0
n = 30 (48%)f ↑4 ↔14 ↓0 ↑5 ↔15 ↓0 ↑3 ↔9 ↓10 ↑3 ↔11 ↓1 ↑1 ↔5 ↓0 ↑1 ↔4 ↓0 ↑0 ↔2 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑3 ↔2 ↓2 ↑0 ↔0 ↓0
 Preventive psychology ↑2 ↔3 ↓0 ↑2 ↔3 ↓0 ↑1 ↔1 ↓0 ↑0 ↔3 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0
 Recovery psychology ↑1 ↔2 ↓0 ↑0 ↔4 ↓0 ↑1 ↔4 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0 ↑0 ↔2 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0
 Skin-to-skin contact ↑0 ↔2 ↓0 ↑0 ↔3 ↓0 ↑0 ↔1 ↓ ↑0 ↔3 ↓1 ↑0 ↔1 ↓0 ↑0 ↔2 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0
Caregiver roled 4 3 2 2 1 1 0 0 0 2 0
n = 6 (10%)f ↑0 ↔4 ↓0 ↑0 ↔3 ↓0 ↑0 ↔2 ↓0 ↑0 ↔2 ↓0 ↑0 ↔1 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑1 ↔1 ↓0 ↑0 ↔0 ↓0
 Individual patient care ↑0 ↔0 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0
 Patient advocation ↑0 ↔1 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0 ↑0 ↔1 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0
Caregiver supporte 11 11 6 5 5 1 2 0 2 5 2
n = 20 (32%)f ↑4 ↔7 ↓0 ↑5 ↔5 ↓1 ↑1 ↔5 ↓0 ↑0 ↔5 ↓0 ↑0 ↔4 ↓0 ↑0 ↔1 ↓0 ↑0 ↔2 ↓0 ↑0 ↔0 ↓0 ↑2 ↔0 ↓0 ↑2 ↔3 ↓0 ↑1 ↔1 ↓0
 Caregiver liaison ↑0 ↔3 ↓0 ↑1 ↔1 ↓0 ↑0 ↔1 ↓0 ↑0 ↔1 ↓0 ↑0 ↔1 ↓0 ↑1 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0
 Psychoeducation ↑1 ↔4 ↓0 ↑1 ↔4 ↓0 ↑1 ↔3 ↓1 ↑2 ↔2 ↓0 ↑1 ↔2 ↓0 ↑0 ↔1 ↓0 ↑0 ↔1 ↓0 ↑0 ↔0 ↓0 ↑0 ↔0 ↓0 ↑1 ↔1 ↓0 ↑0 ↔0 ↓0
↑ = significant improvement, ↓ = significant detriment, ↔ = no effect.
aAs reported by Finucane et al (17).
bSelected interventions represented by at least three included studies.
cInterventions aimed at caregiver distress, difficulty or deteriorating relationships, balancing need to relieve suffering with desire to communicate, or helplessness versus control. Preventive psychology (i.e., trauma-preventive psychology, psychologic communication strategy, creating opportunities for parent engagement). Recovery psychology (i.e., coping skills training, mindfulness training, brief-cognitive behavioral therapy, recovery programme). Skin-to-skin contact (i.e., skin-to-skin, kangaroo care).
dInterventions aimed at caregiver detection and prevention of patient delirium, monitoring patient symptoms, or advocating for the patient. Individual patient care (i.e., newborn individualized developmental care and assessment program). Patient advocation (i.e., family-staff communication training, patient communication training).
eInterventions aimed at caregiver knowledge on delirium, advice on how to respond to the patient, caregiver support systems, or patient outcomes. Caregiver liaison (i.e., liaison nurse, communication facilitator, transitional consultation program). Psychoeducation (i.e., emotional support meetings, stress management interviews, psychoeducational tool, self-help manual, patient psychoeducational support).
fn, number of studies that reported each objective or conclusion; %, n as a proportion of total randomized controlled trials (n = 63).

Meta-analysis included 56 RCTs; 34 RCTs (61%) in NICUs or PICUs. Caregiver outcomes were reported within (i.e., short-term, n = 39) or after (i.e., long-term, n = 17) 3 months postdischarge. Additional descriptions in the Supplemental Results ( Pooled estimates from meta-analyses in Supplemental Table 7 (, summary of findings in Supplemental Table 8 ( Summary of findings from qualitative studies in Supplemental Table 9 (

Negative Psychologic Outcomes

Five negative psychologic outcomes from 56 RCTs were amenable to meta-analysis (Fig. 2A). Meta-analysis of 33 trials (n = 16 [49%] in adult ICUs) demonstrated interventions to significantly reduce anxiety within 3 months postdischarge (RoM, 0.92; 95% CI, 0.87–0.97; I2 = 97.06%; n = 3,478 caregivers). No significant effect was observed after 3 months (RoM, 0.96; 95% CI, 0.91–1.02; I2 = 90.17%; n = 2,845 caregivers) (Fig. 3A). Seven studies demonstrated significant reductions in anxiety; four targeted caregiver experience (29–32) and three targeted caregiver support (33–35). Three were multisite (30,31,34), the largest a cluster-crossover RCT investigating flexible family visitation policies across 36 adult ICUs in Brazil (31). Two studies noted significantly positive long-term effects; Family Participation in Patient Care (35) and Family Nurture Interventions (32) improved anxiety after 3 months postdischarge.

Figure 2.
Figure 2.:
Ratio of means in psychological outcomes in the randomized trials amenable to meta-analysis. Ratio of means is the difference between the experimental group to the control group. Horizontal lines denote 95% confidence intervals. Diamonds represent point estimates for informal caregiver (A) negative and (B) positive psychological outcomes grouped by follow-up. PTSD = post-traumatic stress disorder.
Figure 3.
Figure 3.:
Ratio of means in psychological outcomes in the randomized trials amenable to meta-analysis. Ratio of means is the difference between the experimental group to the control group. Horizontal lines denote 95% confidence intervals. Diamonds represent point estimates for informal caregiver (A) anxiety and (B) depression grouped by follow-up. Parentheses following study reference denote intervention type: caregiver experience (E); caregiver role (R); caregiver support (S). Asterisk indicates adult patient population. PTSD = post-traumatic stress disorder.

Mental health interventions (n = 32 total; n = 15 [47%] in adult ICUs) significantly decreased depression within 3 months of ICU discharge (RoM, 0.83; 95% CI, 0.69–0.99; I2 = 97.09%; n = 3,650 caregivers). No effect on depression after 3 months was observed (RoM, 0.99; 95% CI, 0.83–1.20; I2 = 98.44%; n = 2,870 caregivers) (Fig. 3B). Six RCTs observed significant improvement, four evaluation trials within 3 months discharge (30,31,33,34) and two single-site studies targeting caregiver experience (32) and caregiver support (36) at 4 and 6 months postdischarge, respectively. All six RCTs but the piloting and feasibility trial by Samuel et al (36), a Pediatric Intensive Care Follow-Up Clinic, reported significant improvements in anxiety. Conversely, increased depression after 3 months postdischarge was reported by four trials (37–40). Kentish-Barnes et al (39) found in relatives of patients who died in ICU, condolence letters failed to alleviate psychologic distress and worsened depression and symptoms of PTSD. In addition to increased depression, the multisite evaluation RCT by Bohart et al (38) reported increased anxiety and decreased courage at 12 months postdischarge for their 3-month Recovery Programme.

Outcomes for informal caregiver PTSD were inconsistent and meta-analysis of 21 RCTs (n = 14 [67%] in adult ICUs) demonstrated no significant effect of interventions on PTSD symptoms at any time (Supplemental Fig. 3a, Four studies on caregiver experience interventions reported significantly decreased PTSD symptoms within 3 months discharge from adult mixed ICU (30,41), adult medical-surgical ICU (42), or NICU (43), while two studies reported significantly increased PTSD symptoms in adult medical ICU (44) and NICU (45). Studies with longer-term follow-up reported conflicting results (36,39,46,47). Two of 21 trials tested ICU Family Diaries to reduce risk of PTSD symptoms; the piloting and feasibility trial by Jones et al (41) informed the evaluation trial across four adult medical-surgical ICUs by Nielsen et al (42). Among studies reporting on symptoms of PTSD, only the evaluation RCT by Lautrette et al (30) that assessed end-of-life conferences and bereavement leaflets across 22 adult mixed ICUs observed a consistent reduction in anxiety, depression, and PTSD at 3 months postdischarge.

Meta-analysis of 22 trials (n = 7 [32%] in adult ICUs) identified no significant effect of on psychologic distress at any time (Supplemental Fig. 3b, Two studies reported significantly improved caregiver distress within 3 months discharge (34,48) while two reported increased distress (49,50); all studies conducted in NICUs. No study reported significant improvement after 3 months discharge.

Interventions pooled across five RCTs (n = 3 [60%] in adult ICUs) significantly increased caregiver burden [feeling need to do more] within 3 months of ICU discharge (RoM, 1.08; 95% CI, 1.05–1.12; I2 = 6.96%; n = 819 caregivers) (Fig. 4A). There was no significant effect after 3 months across four trials that reported long-term outcomes (RoM, 0.81; 95% CI, 0.81–1.06; I2 = 0.00%; n = 127 caregivers). The single study that reported significantly increased burden was the Creating Opportunities for Parent Engagement evaluation trial by Melnyk et al (34) in two NICUs.

Figure 4.
Figure 4.:
Ratio of means in psychological outcomes in the randomized trials amenable to meta-analysis. Ratio of means is the difference between the experimental group to the control group. Horizontal lines denote 95% confidence intervals. Diamonds represent point estimates for informal caregiver burden (A) humanity (B), and transcendence (C) grouped by follow-up. Parentheses following study reference denote intervention type: caregiver experience (E); caregiver role (R); caregiver support (S). Asterisk indicates adult patient population. PTSD = post-traumatic stress disorder.

Positive Psychologic Outcomes

Three positive psychologic outcomes from 22 RCTs were amenable to meta-analysis (Fig. 2B). These included courage (e.g., coping, resilience) (n = 5 RCTs [n = 3; 60% adult ICUs], Supplemental Fig. 4,, humanity (e.g., kindness, social intelligence) (n = 3 RCTs [n = 0; 0% adult ICUs], Fig. 4B), and transcendence (e.g., gratitude, satisfaction) (n = 14 RCTs [n = 7; 50% adult ICUs], Fig. 4C). Humanity (RoM, 1.11; 95% CI, 1.07–1.15; I2 = 0.01%; n = 338 caregivers) and transcendence (RoM, 1.04; 95% CI, 1.02–1.06; I2 = 74.39%; n = 2,617 caregivers) were significantly increased within 3 months postdischarge. There were no significant differences in humanity and transcendence after 3 months postdischarge or courage at any time. Findings on positive psychologic outcomes detailed in the Supplemental Results (

Stratified Analyses

Estimates grouped by intervention type and patient population were inconsistent; no outcome was consistently significantly improved across all interventions (Supplemental Table 10, Significant positive effects of mental health interventions on caregiver depression and PTSD, as well as a significant negative effect on caregiver burden, were not consistently statistically significant due to smaller number of studies in the subgroup analyses. Stratified analyses for quasi-experimental and uncontrolled studies not amenable to meta-analysis in Supplemental Figure 5 ( Additional details in the Supplemental Results (

Risk of Bias and Certainty of Evidence

Risk of bias scores in Supplemental Table 10 ( Certainty of evidence assessments summarized in Supplemental Tables 11–13 ( Detailed results in the Supplemental Results (


Our systematic review of mental health interventions targeting informal caregivers of critically ill patients included 102 interventional trials, including 56 RCTs. Mental health interventions improved short-term anxiety, depression, humanity, and transcendence while increasing burden. No significant long-term effects after 3 months were observed. These findings suggest mental health interventions improve short-term but not long-term psychologic outcomes.

Three types of interventions had greatest supporting evidence for positive effects on short-term psychologic outcomes regardless of patient population. Specifically, 1) preventive psychology (i.e., to prevent negative psychologic sequelae; e.g., Trauma-Preventive Psychology [43]), 2) recovery psychology (i.e., to emphasize potential for recovery; e.g., Mindfulness Training [29]), and 3) psychoeducation (i.e., to facilitate knowledge translation for enhanced coping, treatment efficacy and adherence; e.g., Patient-Partner Psychoeducational Support [51]). Effective short-term interventions involved the critical care team (commonly nurses and/or liaisons) to enhance caregiver experience (i.e., regardless of need), as well as specialists (e.g., psychologist, psychiatrist) to provide targeted support for highly complex needs. Greater effect of mental health interventions on long-term outcomes might require including multiple levels of support (i.e., universal and targeted) to capture various levels of caregiver need while providing longer time set of supports.

We found significant heterogeneity our results. Some predisposing factors to poor psychologic outcomes among caregivers of the critically ill are more prominent in NICU and PICU populations, such as younger patient age and lower education, while other factors are predominantly related to adult ICU populations (e.g., being the patient’s spouse) (52). Environmental factors related to the ICU setting that vary based on patient population are also known to predispose informal caregivers to poor psychologic outcomes (53).

Multiple factors may explain discordance between short- and long-term outcomes. Few studies included process evaluations to assess implementation (i.e., fidelity, dose, or reach) and outcome assessments (i.e., use of wrong tool). It is possible interventions were not reliably delivered as intended or consistently adhered to as required for mental health interventions to have full effect (4). PTSD symptoms have been reported to develop further from the time in ICU as informal caregivers process their reality of critical illness (54). Few studies designed interventions based on theories of long-term psychologic change (e.g., Post-Traumatic Growth Inventory [55]). Mental health trajectories in informal caregivers may require different outcomes targeted by different interventions at different time points. What works in immediate aftermath of critical illness may not be helpful months later.

We evaluated diverse psychologic outcomes. In addition to negative consequences, we also explored positive adaptations (56,57). Humanity (i.e., love, kindness) and transcendence (i.e., gratitude, hope) are positive psychologic adaptations to evolving demands of stressful experiences encapsulated within psychologic resilience (58,59). Although highly resilient individuals proactively cultivate positive adaptations (60,61), it is unknown whether positive adaptations are byproducts of intervention or whether they improve coping behaviors. Mental health in informal caregivers of critically ill patients should be considered holistically as the aggregate of negative and positive psychologic outcomes.

We found that some mental health interventions increased caregiver burden. Many authors noted that what constitutes appropriate mental healthcare with adequate follow-up is unclear. Individuals cope differently with critical illness such that any intervention that changes coping trajectories has the potential to both help and harm (62). Furthermore, clinicians are uncertain how to provide mental health interventions to offer support rather than to reduce negative psychologic outcomes (39). We suggest that those who provide mental health interventions should carefully weigh possible adverse effects against potential benefits (12). There is no magic bullet to improve informal caregiver mental health.

Our results need to be interpreted within context of limitations. Primarily, heterogeneity limited pooled statistical analyses. Effectiveness of interventions should be generalized with great caution to age groups other than that in which they have been studied. Second, included studies used different assessment tools. It remains uncertain which tools are most appropriate for informal caregivers (63) and theoretical frameworks to aid tool selection are scarce (64). Third, given diverse and multicomponent nature of interventions, it was not possible to determine specific components of interventions associated with more favorable caregiver psychologic outcomes in particular patient populations. Fourth, no universal definition for positive caregiver psychology added complexity to study selection. However, our broad inclusion criteria resulted in a comprehensive summary of literature assessing this concept that is understudied in critical care. Finally, perspectives from non-Western countries were under-represented. Differences in cultural values such as family responsibilities and support networks may influence caregiving experiences (65).


In this systematic review and meta-analysis, mental health interventions improved short-term anxiety, depression, humanity, and transcendence among informal caregivers of critically ill patients while increasing burden. There were no significant effects on long-term psychologic outcomes. Clinicians should consider short-term prescriptions of mental health interventions for informal caregivers of critically ill patients with capacity to manage interventions.


We would like to acknowledge Dr. Diane L. Lorenzetti (University of Calgary) for the development of the search strategies and Mr. Michal S. Cherak (University of Calgary) for assistance with article translation. We would also like to acknowledge Dr. Karla D. Krewulak (University of Calgary) for resolving discrepancies in screening and providing valuable feedback on the article.


1. Tate JA, Devito Dabbs A, Hoffman LA, et al.: Anxiety and agitation in mechanically ventilated patients. Qual Health Res 2012; 22:157–173
2. Jackson JC, Mitchell N, Hopkins RO: Cognitive functioning, mental health, and quality of life in ICU survivors: An overview. Crit Care Clin 2009; 25:615–628, x
3. Wendler D, Rid A: Systematic review: The effect on surrogates of making treatment decisions for others. Ann Intern Med 2011; 154:336–346
4. Johnson CC, Suchyta MR, Darowski ES, et al.: Psychological sequelae in family caregivers of critically iii intensive care unit patients. A systematic review. Ann Am Thorac Soc 2019; 16:894–909
5. Corry M, While A, Neenan K, et al.: A systematic review of systematic reviews on interventions for caregivers of people with chronic conditions. J Adv Nurs 2015; 71:718–734
6. Jorm AF, Korten AE, Jacomb PA, et al.: “Mental health literacy”: A survey of the public’s ability to recognise mental disorders and their beliefs about the effectiveness of treatment. Med J Aust 1997; 166:182–186
7. Min JA, Lee CU, Lee C: Mental health promotion and illness prevention: A challenge for psychiatrists. Psychiatry Investig 2013; 10:307–316
8. Haggerty RJ, Mrazek PJ: Can we prevent mental illness? Bull N Y Acad Med 1994; 71:300–306
9. Moher D, Liberati A, Tetzlaff J, et al.; PRISMA Group: Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. J Clin Epidemiol 2009; 62:1006–1012
10. VandenBos GRAPA Dictionary of Psychology. 2007Washington, DC, American Psychological Association,
11. Ridner SH: Psychological distress: Concept analysis. J Adv Nurs 2004; 45:536–545
12. Kim KW, Lim HC, Park JH, et al.: Developing a basic scale for workers’ psychological burden from the perspective of occupational safety and health. Saf Health Work 2018; 9:224–231
13. Peterson C, Seligman MEPCharacter Strengrhs and Virtures: A Handbook and Classification. 2004Oxford, United Kingdom, Oxford University Press,
14. Park N, Peterson C: Does it matter where we live? The urban psychology of character strengths. Am Psychol 2010; 65:535–547
15. Park N, Peterson C: Moral competence and character strengths among adolescents: The development and validation of the Values in Action Inventory of Strengths for Youth. J Adolesc 2006; 29:891–909
16. Fiest KM, McIntosh CJ, Demiantschuk D, et al.: Translating evidence to patient care through caregivers: A systematic review of caregiver-mediated interventions. BMC Med 2018; 16:105
17. Finucane AM, Lugton J, Kennedy C, et al.: The experiences of caregivers of patients with delirium, and their role in its management in palliative care settings: An integrative literature review. Psychooncology 2017; 26:291–300
18. Moore GF, Audrey S, Barker M, et al.: Process evaluation of complex interventions: Medical Research Council guidance. BMJ 2015; 350:h1258
19. Sterne JAC, Savović J, Page MJ, et al.: RoB 2: A revised tool for assessing risk of bias in randomised trials. BMJ 2019; 366:l4898
20. Guyatt GH, Oxman AD, Vist GE, et al.; GRADE Working Group: GRADE: An emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008; 336:924–926
21. Lewin S, Bohren M, Rashidian A, et al.: Applying GRADE-CERQual to qualitative evidence synthesis findings-paper 2: How to make an overall CERQual assessment of confidence and create a summary of qualitative findings table. Implement Sci 2018; 13:10
22. Friedrich JO, Adhikari NK, Beyene J: The ratio of means method as an alternative to mean differences for analyzing continuous outcome variables in meta-analysis: A simulation study. BMC Med Res Methodol 2008; 8:32
23. Higgins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21:1539–1558
24. Higgins JP: Commentary: Heterogeneity in meta-analysis should be expected and appropriately quantified. Int J Epidemiol 2008; 37:1158–1160
25. Higgins JP, Thompson SG, Deeks JJ, et al.: Measuring inconsistency in meta-analyses. BMJ 2003; 327:557–560
26. Thomas J, Harden A: Methods for the thematic synthesis of qualitative research in systematic reviews. BMC Med Res Methodol 2008; 8:45
27. O’Cathain A, Goode J, Drabble SJ, et al.: Getting added value from using qualitative research with randomized controlled trials: A qualitative interview study. Trials 2014; 15:215
28. Petticrew M, Rehfuess E, Noyes J, et al.: Synthesizing evidence on complex interventions: How meta-analytical, qualitative, and mixed-method approaches can contribute. J Clin Epidemiol 2013; 66:1230–1243
29. Cox CE, Porter LS, Buck PJ, et al.: Development and preliminary evaluation of a telephone-based mindfulness training intervention for survivors of critical illness. Ann Am Thorac Soc 2014; 11:173–181
30. Lautrette A, Darmon M, Megarbane B, et al.: A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med 2007; 356:469–478
31. Rosa RG, Falavigna M, da Silva DB, et al.: Effect of flexible family visitation on delirium among patients in the intensive care unit: The ICU visits randomized clinical trial. JAMA 2019; 322:216–228
32. Welch MGH, Halperin MS, Austin J, et al.: Depression and anxiety symptoms of mothers of preterm infants are decreased at 4 months corrected age with family nurture intervention in the NICU. Arch Women’s Mental Health 2016; 19:51–61
33. Cobiella CW, Mabe PA, Forehand RL: A comparison of two stress-reduction treatments for mothers of neonates hospitalized in a neonatal intensive care unit. Child Health Care 1990; 19:93–100
34. Melnyk BM, Feinstein NF, Alpert-Gillis L, et al.: Reducing premature infants’ length of stay and improving parents’ mental health outcomes with the creating opportunities for parent empowerment (COPE) neonatal intensive care unit program: A randomized, controlled trial. Pediatrics 2006; 118:e1414–e1427
35. Rodriguez Martinez MdC, Morilla FR, Roncero del Pino A, et al.: [Family involvement in the critically ill patient basic care]. Enferm Intensiva 2003; 14:96–108
36. Samuel VM, Colville GA, Goodwin S, et al.: The value of screening parents for their risk of developing psychological symptoms after PICU: A feasibility study evaluating a pediatric intensive care follow-up clinic. Pediatr Crit Care Med 2015; 16:808–813
37. Affleck G, Tennen H, Rowe J, et al.: Effects of formal support on mothers’ adaptation to the hospital-to-home transition of high-risk infants: The benefits and costs of helping. Child Dev 1989; 60:488–501
38. Bohart S, Egerod I, Bestle MH, et al.: Reprint of Recovery programme for ICU survivors has no effect on relatives’ quality of life: Secondary analysis of the RAPIT-study. Intensive Crit Nursing 2019; 50:111–117
39. Kentish-Barnes N, Chevret S, Champigneulle B, et al.; Famirea Study Group: Effect of a condolence letter on grief symptoms among relatives of patients who died in the ICU: A randomized clinical trial. Intensive Care Med 2017; 43:473–484
40. Rosenbaum JL, Smith JR, Yan Y, et al.: Impact of a neonatal-bereavement-support DVD on parental grief: A randomized controlled trial. Death Stud 2015; 39:191–200
41. Jones C, Bäckman C, Griffiths RD: Intensive care diaries and relatives’ symptoms of posttraumatic stress disorder after critical illness: A pilot study. Am J Crit Care 2012; 21:172–176
42. Nielsen AH, Angel S, Egerod I, et al.: The effect of family-authored diaries on posttraumatic stress disorder in intensive care unit patients and their relatives: A randomised controlled trial (DRIP-study). Aust Crit Care 2020; 33:123–129
43. Jotzo M, Poets CF: Helping parents cope with the trauma of premature birth: An evaluation of a trauma-preventive psychological intervention. Pediatrics 2005; 115:915–919
44. Carson SS, Cox CE, Wallenstein S, et al.: Effect of palliative care-led meetings for families of patients with chronic critical illness: A randomized clinical trial. JAMA 2016; 316:51–62
45. Bernard RS, Williams SE, Storfer-Isser A, et al.: Brief cognitive-behavioral intervention for maternal depression and trauma in the neonatal intensive care unit: A pilot study. J Trauma Stress 2011; 24:230–234
46. Cox CE, White DB, Hough CL, et al.: Effects of a personalized web-based decision aid for surrogate decision makers of patients with prolonged mechanical ventilation: A randomized clinical trial. Ann Intern Med 2019; 170:285–297
47. Cox CE, Hough CL, Carson SS, et al.: Effects of a telephone- and web-based coping skills training program compared with an education program for survivors of critical illness and their family members. A randomized clinical trial. Am J Respir Crit Care Med 2018; 197:66–78
48. Lee S-Y, Aycock DM, Moloney MF: Bright light therapy and maternal sleep and well-being: A pilot randomized clinical trial. Sleep 2010; 33(Suppl 1):A225
49. Roberts KL, Paynter C, McEwan B: A comparison of kangaroo mother care and conventional cuddling care. Neonantal Netw 2000; 19:31–35
50. van der Pal SM, Manguire CM, le Cessie S, et al.: Parental experiences during the first period at the neonatal unit after two developmental care interventions. Acta Paediatr 2007; 96:1611–1616
51. Ågren S, Eriksson A, Fredrikson M, et al.: The health promoting conversations intervention for families with a critically ill relative: A pilot study. Intensive Crit Care Nurs 2019; 50:103–110
52. Wang S, Allen D, Kheir YN, et al.: Aging and post-intensive care syndrome: A critical need for geriatric psychiatry. Am J Geriatr Psychiatry 2018; 26:212–221
53. Schmidt M, Azoulay E: Having a loved one in the ICU: The forgotten family. Curr Opin Crit Care 2012; 18:540–547
54. Wong ML, Cavanaugh CE, Macleamy JB, et al.: Posttraumatic growth and adverse long-term effects of parental cancer in children. Fam Syst Health 2009; 27:53–63
55. Tedeschi RG, Calhoun LG: The posttraumatic growth inventory: Measuring the positive legacy of trauma. J Trauma Stress 1996; 9:455–471
56. Loiselle KA, Devine KA, Reed-Knight B, et al.: Posttraumatic growth associated with a relative’s serious illness. Fam Syst Health 2011; 29:64–72
57. Tugade MM, Fredrickson BL: Resilient individuals use positive emotions to bounce back from negative emotional experiences. J Pers Soc Psychol 2004; 86:320–333
58. Block J, Kremen AM: IQ and ego-resiliency: Conceptual and empirical connections and separateness. J Pers Soc Psychol 1996; 70:349–361
59. Lazarus RS: From psychological stress to the emotions: A history of changing outlooks. Annu Rev Psychol 1993; 44:1–21
60. Kumpfer KL, Bluth B: Parent/child transactional processes predictive of resilience or vulnerability to “substance abuse disorders.” Subst Use Misuse 2004; 39:671–698
61. Kumpfer KL, Summerhays JF: Prevention approaches to enhance resilience among high-risk youth: Comments on the papers of Dishion & Connell and Greenberg. Ann N Y Acad Sci 2006; 1094:151–163
62. Kagee A: Concerns about the effectiveness of critical incident stress debriefing in ameliorating stress reactions. Crit Care 2002; 6:88
63. Mosquera I, Vergara I, Larrañaga I, et al.: Measuring the impact of informal elderly caregiving: A systematic review of tools. Qual Life Res 2016; 25:1059–1092
64. Gérain P, Zech E: Informal caregiver burnout? Development of a theoretical framework to understand the impact of caregiving. Front Psychol 2019; 10:1748
65. Knight BG, Sayegh P: Cultural values and caregiving: The updated sociocultural stress and coping model. J Gerontol B Psychol Sci Soc Sci 2010; 65B:5–13

critical care; informal caregivers; intensive care unit; psychology; systematic review

Supplemental Digital Content

Copyright © 2021 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the Society of Critical Care Medicine and Wolters Kluwer Health, Inc.