The Hidden Burden After Critical Illnesses: A Systematic Review of Psychological Outcomes and Their Associated Factors Among ICU Survivors (2021–2025)

Introduction

The global burden of critical illness continues to increase due to aging populations and rising rates of chronic disease.^1,2 Millions of individuals require intensive care each year for conditions such as sepsis, acute respiratory failure, and major trauma, often requiring mechanical ventilation and sedation.^3–6 Improvements in critical care have enhanced survival,⁷ but a growing number of ICU survivors experience post-intensive care syndrome (PICS), affecting more than half of patients and involving physical, cognitive, and psychological impairments such as depression, anxiety, and post-traumatic stress disorder (PTSD).^8,9

Psychological outcomes are among the most frequently reported components of PICS, including depression, anxiety, and PTSD.^10–12 These symptoms are often linked to traumatic ICU experiences such as life-threatening events, invasive procedures, prolonged sedation, sleep disruption, and reduced communication or social interaction.^13–15 They may persist after hospital discharge, negatively affecting quality of life, functional recovery, and healthcare utilization.^{10,12,16–20} Despite growing recognition, existing evidence remains fragmented and inconsistent. Many studies focus on single outcomes, while variation in study design, follow-up timing, and measurement tools limits comparability and generalizability.^21–24 Moreover, the role of factors such as pre-existing mental health, ICU-related exposures, and length of stay remains unclear or inconsistently reported.^25–27

Previous systematic reviews have typically focused on either broader post-intensive care syndrome (PICS) outcomes or single psychological conditions, without integrating multiple psychological outcomes and their associated factors within a unified analytical framework. To address these gaps, this review defines psychological outcomes as clinically relevant symptoms of depression, anxiety, post-traumatic stress disorder (PTSD), and related emotional distress, and associated factors as demographic, clinical, functional, and psychosocial variables influencing these outcomes. Recent evidence (2021–2025) reflects evolving ICU practices and post-pandemic care contexts that may influence psychological outcomes; however, a focused synthesis of this evidence remains limited. Therefore, this review aims to synthesize recent evidence on psychological outcomes among ICU survivors and to identify associated factors, with the goal of supporting early identification of at-risk individuals and informing targeted strategies to improve long-term recovery.

Objective

To synthesize current evidence on psychological outcomes, measured using validated instruments, among ICU survivors, and to identify associated factors, including ICU-related exposures.

Materials and Methods

Identify Relevant Studies

This review was conducted in accordance with the PRISMA 2020 statement.²⁸ The literature search was performed independently by two reviewers, in consultation with an experienced university librarian, to ensure methodological rigor and comprehensiveness. A systematic search was conducted across six electronic databases: PubMed, MEDLINE via EBSCO, ScienceDirect, Scopus, Nursing & Allied Health Premium (ProQuest), and the Cochrane Library. All databases were searched from January 2021 to June 2025, and the final search was completed on 18 June 2025. This timeframe was selected to capture recent evidence reflecting contemporary ICU practices, including post–COVID-19 developments in critical care and survivorship. Eligible studies examined psychological outcomes among adult ICU survivors. Studies assessing outcomes during ICU stay were included only when findings were explicitly linked to post-ICU survivorship trajectories to ensure conceptual consistency (Table 1). The search strategy was developed using a combination of keywords and controlled vocabulary (eg., MeSH terms in PubMed), and was tailored to each database. Boolean operators (AND, OR) were applied to combine search terms. Truncation and field tags were used where appropriate to enhance sensitivity and specificity. An example of the full search strategy for PubMed is provided in Supplementary Table S1, including detailed search strings, field specifications, and filters. The core search structure included: (Critical Illness OR Intensive Care Units OR Critical Care OR ICU Survivors) AND (Risk Factors OR Predictors) AND (Depression OR Anxiety OR Stress Disorders OR Post-Traumatic Stress Disorder OR Mental Health OR Psychological Distress). Search limits included studies published in English and involving adult populations. The strategy was adapted for each database by modifying subject headings, field tags, and syntax according to indexing requirements. All records were imported into EndNote X7 for organization and duplicate removal. Full details are shown in Supplementary Table S1.

Table 1 Inclusion and Exclusion Criteria

Study Selection

A two-stage screening process identified eligible studies. Titles and abstracts were first screened for relevance, followed by full-text assessment based on predefined PICOS criteria (Population, Exposure, Comparison, Outcomes, Study design). Two reviewers independently conducted screening, with disagreements resolved through discussion or consultation with a third reviewer. The study selection process is presented in a PRISMA flow diagram (Figure 1).

Figure 1 PRISMA flow diagram of study selection. The diagram illustrates the process of study identification, screening, eligibility assessment, and inclusion.

Quality Assessment

The methodological quality of the included studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal tools,²⁹ with checklists selected according to the study design. Two reviewers independently conducted the appraisal, and any disagreements were resolved through discussion. The purpose of the quality assessment was to evaluate the methodological rigor of each study and identify potential sources of bias related to study design, conduct, and analysis. Rather than relying solely on numerical scores, appraisal findings were used to inform the interpretation and synthesis of results, particularly when comparing studies with varying levels of methodological quality.

Data Extraction

A structured extraction form (Supplementary Table S2) ensured consistency and completeness across studies. Data extraction was performed independently by two reviewers, and the extracted data were cross-checked for accuracy, with discrepancies resolved through discussion. Key data included bibliographic details (citation, year, country, setting), population characteristics (sample size, age, sex, diagnosis), and study variables (aim, design, and outcomes). Extracted outcomes covered psychological measures, predictive factors, and mental health results. Authors’ conclusions, limitations, and recommendations were also documented.

Data Synthesis

Data were synthesized using a convergent integrated approach based on JBI methodology for mixed-method reviews.³⁰ Quantitative findings were transformed into qualitative themes to enable integrated analysis. Specifically, extracted data were coded, grouped based on conceptual similarity, and iteratively refined through constant comparison across studies. This process resulted in the development of overarching themes that reflect patterns in factors associated with psychological outcomes among ICU survivors.

Results

Search results

Following PRISMA guidelines, records were identified from six electronic databases, yielding a total of 4,548 articles. After duplicate removal and title and abstract screening, a limited number of full-text articles were assessed for eligibility. Thirteen studies met the inclusion criteria and were included in the final synthesis.^{16,20,25,31–40} The detailed selection process is illustrated in Figure 1.

Assessment of Methodological Quality

The methodological quality of the included studies was evaluated using the JBI critical appraisal checklists.²⁹ Overall, the included studies demonstrated moderate to high methodological quality. However, variations were noted in areas such as sample representativeness, control of confounding factors, and consistency of outcome measurement. These differences were taken into account when interpreting and comparing findings across studies. Quality scores ranged from 72.7% to 100% (mean = 87.4%); however, these scores are reported descriptively and should be interpreted with caution. Detailed appraisal results for each study are presented in Supplementary Table S2.

Description of Included Studies

Thirteen studies were included (Table 2; see Supplementary Table S2 for details). Most were published in 2025 (n = 5, 38.5%) and 2024 (n = 4, 30.8%), with the remainder spanning 2021–2023. Studies were conducted across 10 countries, most commonly Germany (n = 3), Brazil (n = 2), and Taiwan (n = 2). Research settings included university hospitals, databases, and general hospitals (each n = 3, 23.1%). Prospective cohort studies were most common (n = 6, 46.2%), followed by retrospective cohort (n = 4, 30.8%), cross-sectional (n = 2, 15.4%), and observational (secondary longitudinal cohort analysis) (n = 1, 7.7%).

Table 2 The Characteristics of the Included Studies

Populations included general ICU patients (n = 6, 46.1%), those with severe sepsis/septic shock (n = 3, 23.1%), and trauma, acute respiratory distress syndrome (ARDS), or liver-transplant recipients (each n = 1). Sample sizes ranged from 30 to >2,000, with most enrolling 101–500 participants (n = 5, 38.5%). Older adults (60–80 years) represented the majority (n = 9, 69.2%). Psychological outcomes were assessed using both validated psychiatric instruments (eg., HADS, PTSD Checklist) general indicators of psychological distress based on non-standardized or self-reported measures, which may partially explain variability across findings. The Hospital Anxiety and Depression Scale (HADS) was most frequently used (n = 5, 26.3%), followed by European Quality of Life-5 Dimensions-5 Levels (EQ-5D-5L) and the PTSD Checklist–Civilian Version (each n = 2, 10.5%). Other instruments (eg., Beck Inventories, Impact of Event Scale-Revised (IES-R), Post-traumatic Symptom Scale −10 (PTSS-10), Visual Analog Scale for Anxiety (VASA), International Statistical Classification of Diseases and Related Health Problems, 10^th (ICD-10) criteria appeared in single studies, while two studies did not specify the tool. Timing of outcome measurement varied across studies, with assessments conducted during hospitalization (ICU stay) (n = 3, 23.07%), 3–12 months after ICU discharge (n = 7, 53.83%), and more than 12 months after discharge (n = 1, 7.69%), while two studies did not report the assessment timeframe (n = 2, 15.38%). This variation suggests that psychological outcomes may change over time, highlighting the importance of considering follow-up duration when interpreting findings.

Psychological and Mental Health Outcomes Among ICU Survivors

Psychological outcomes refer to emotional and cognitive responses experienced by ICU survivors, such as anxiety, depression, and PTSD. In contrast, mental health outcomes encompass these psychological manifestations as well as clinically diagnosed mental disorders. This distinction is important because not all psychological symptoms meet the diagnostic criteria for a mental disorder, yet they may still affect recovery and quality of life. Table 3 presents a summary of psychological and mental health outcomes reported in the 13 included studies. Depression was examined in nine studies (69.21%), making it the most frequently reported outcome. Anxiety and PTSD were each investigated in eight studies (61.52%). Mental health was reported in two studies (15.38%), and stress was addressed in one study (7.69%). Overall, psychological outcomes showed a consistent pattern across studies, with depression, anxiety, and PTSD being the most frequently reported conditions; however, substantial variability was observed across studies, which may be explained by differences in study design, timing of outcome assessment (eg., during ICU stay versus post-discharge follow-up), and the use of heterogeneous measurement tools (validated instruments versus non-standardized measures).

Table 3 Psychological Outcomes Among ICU Survivors

Factors Related to Psychological Outcomes Among ICU Survivors

Figure 2 and Table 4 summarize factors influencing psychological outcomes among ICU survivors, with detailed results provided in Supplementary Table S2. Themes were developed using a convergent integrated approach, allowing findings from diverse study designs to be grouped into conceptually meaningful categories. From 13 included studies, seven major themes emerged: (1) personal characteristics (including demographic and clinical factors such as biomarker, age, gender, education, comorbidities, injury type, treatment, and illness severity [APACHE II score]); (2) length of stay; (3) distress-related factors (physical and emotional); (4) pre-existing mental health or substance use; (5) physical function; (6) cognitive function; and (7) protective factors.

Table 4 Factors Related to Psychological Outcomes Among ICU Survivors

Figure 2 Factors related to psychological outcomes among ICU survivors. This figure summarizes key psychological outcomes, including depression, anxiety, stress, post-traumatic stress disorder (PTSD), and overall mental health, along with associated factors such as personal characteristics, length of stay, distress-related factors, pre-existing mental health or substance use, physical function, cognitive function, and protective factors.

Personal Characteristic

Biomarker

One study³¹ found that elevated serum S100B levels were significantly associated with an increased risk of PTSD symptoms (p = 0.046), suggesting that S100B may serve as a potential biomarker for identifying individuals who are more vulnerable to developing PTSD after traumatic events.

Age

Seven studies^{16,20,25,31,33,35,38} examined the effect of age on psychological outcomes in ICU survivors. For instance, Calsavara et al (2021) found that age was significantly associated with PTSD (p =0.014) and depressive symptoms (p =0.021), with older age associated with lower odds of these outcomes.³¹ Nicholson et al (2025) reported a positive correlation between age and anxiety,³³ while Teixeira et al (2021) observed higher anxiety among patients younger than 65 years (p =0.009).²⁰

Gender

Four studies^16,36,38,40 examined gender differences in psychological outcomes among ICU patients. Chang et al (2025) found that females had higher odds of depression, anxiety, and PTSD than males (aOR = 1.33; 95% CI: 1.27–1.40).¹⁶ Similarly, Schmidt et al (2024) reported that female sepsis survivors were more likely to experience worsening post-traumatic stress symptoms up to two years after discharge (OR = 2.45; 95% CI: 1.11–5.41).³⁶ Overall, evidence indicates that women are more vulnerable to post-ICU psychological distress.

Education

Two studies^20,38 examined the role of education in psychological outcomes among ICU survivors. Shih and Pai (2023) found that higher education predicted greater anxiety during ICU admission (β =0.24, p <0.05), possibly due to heightened awareness of illness severity.³⁸ In contrast, Teixeira et al (2021) reported that higher education reduced the risk of post-discharge mental health outcomes (PR = 0.74; 95% CI: 0.55–0.99; p =0.04).²⁰

Comorbidities

Two studies^16,34 found that comorbidities were significantly associated with adverse psychological outcomes in ICU survivors. Chang et al (2025) reported higher risks of depression, anxiety, and PTSD among ICU survivors (aHR = 1.17), with specific conditions such as heart failure, stroke, and COPD linked to greater readmission and mortality.¹⁶ Similarly, Oh et al (2022) showed that a high comorbidity burden (Charlson Comorbidity Index; CCI ≥6) increased the risk of new-onset depression (OR = 1.67; 95% CI: 1.21–2.29), which was associated with higher two-year mortality (HR = 1.49; 95% CI: 1.18–1.88).³⁴

Penetrating Injury

Nicholson et al (2025) identified penetrating injury as a significant predictor of psychological distress among ICU survivors. In a multicenter cohort of 290 trauma patients in the United Kingdom (median age 55; 26.2% female), 63% reported anxiety or depression one year post-discharge. Penetrating injury independently increased this risk (OR = 10.10; 95% CI: 1.90–44.4), along with younger age, prior psychological morbidity, pain, and functional limitations.³³

Treatment

One study^16,37 examined treatment-related predictors of psychological outcomes in ICU survivors. Chang et al (2025) found that patients receiving mechanical ventilation were more likely to develop depression, anxiety, or PTSD than those not ventilated.¹⁶

Apache II Score

Shih and Pai (2023) examined the predictive role of the APACHE II score in ICU survivors and found it to be a significant predictor of anxiety (β =0.22, p <0.05). Higher illness severity, reflected by elevated APACHE II scores, was associated with greater psychological distress among critically ill patients.³⁸

Length of Stay

Two studies^16,25 identified length of stay as a predictor of adverse psychological outcomes, including both hospital and ICU length of stay. Chang et al (2025) examined total hospital length of stay (including both ICU and ward stay) and found that patients hospitalized for 7–13 days (aOR = 1.30; 95% CI: 1.22–1.39) or ≥14 days (aOR = 1.67; 95% CI: 1.55–1.80) had higher odds of depression, anxiety, and PTSD.¹⁶ Similarly, Paul et al (2024) reported that extended ICU stays were associated with poorer mental health at 3- and 6-month follow-ups (p =0.038).²⁵

Distress-Related Factors

Physical

Four studies^32,33,35,39 linked pain, discomfort, injury severity, and physical restraints to adverse psychological outcomes. Sun et al (2024) found that greater postoperative pain (OR = 1.06, p =0.007) and distress (OR = 1.89, p <0.001) increased PTSD risk (OR = 2.21, p <0.001).³⁹ Nicholson et al (2025) reported that persistent pain one year post-ICU was associated with anxiety, depression, and functional impairment (OR = 1.61; 95% CI: 1.10–2.30).³³ Francken et al (2025) noted that physical restraints elevated depressive symptoms (p =0.001) and reduced quality of life.³²

Emotional

One study³⁸ found that a patient’s emotional responses to their illness were significantly associated with higher levels of stress (β=0.30, p <0.01) and anxiety (β=0.48, p <0.001), indicating that emotional responses could serve as a critical predictor of psychological outcomes among critically ill patients.

Pre-Existing Mental Health/Substance Use

Five studies^{20,33–35,37} examined the impact of pre-existing mental health and substance use conditions on ICU survivors’ outcomes. Oh et al (2022) found that pre-existing anxiety (OR = 2.35; 95% CI: 1.93–2.86), substance use disorder (OR = 2.18; 95% CI: 1.48–3.20), and PTSD (OR = 15.52; 95% CI: 2.57–93.51) were significantly linked to a higher prevalence of depression.³⁴ Additionally, Sell et al (2025) reported that pre-existing sleep disorders were associated with increased risk of anxiety, depression, acute stress disorder, and PTSD.³⁷ Although sleep disorders are not strictly classified as mental health, they may reflect underlying psychological vulnerability and are therefore considered within this category.

Physical Function

Two studies^20,33 linked reduced physical function to adverse psychological outcomes in ICU survivors. Nicholson et al (2025) found that difficulty performing usual activities (EQ-5D-5L domain) was associated with anxiety and depression (OR = 1.40; 95% CI: 1.02–2.29; p =0.04).³³ Likewise, Teixeira et al (2021) reported that decreased functional status at six months post-discharge was associated with anxiety, depression, and PTSD.²⁰

Cognitive Function

One study by Calsavara et al (2021) examined cognitive function as a risk factor for psychological outcomes among ICU survivors.³¹ In a longitudinal cohort of 33 sepsis patients, higher Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) scores were significantly associated with increased PTSD risk, suggesting that cognitive impairment may heighten vulnerability to post-ICU psychological distress.

Protective Factors

Two studies identified protective factors against post-ICU psychological distress, including resilience and social support.^39,40 Sun et al (2024) reported that higher resilience was associated with a reduced risk of PTSD one year after liver transplantation (OR = 2.21; p <0.001).³⁹ Yoshino et al (2024) found that greater pre-ICU social support was associated with fewer depressive symptoms after discharge (β = –0.018; p =0.002).⁴⁰

Across themes, psychological outcomes appear to be influenced by the interaction between pre-existing vulnerability, illness severity, ICU-related exposures, and post-discharge recovery factors, rather than by single isolated variables.

Discussion

This review synthesized current evidence on psychological outcomes among ICU survivors and identified associated factors. Depression, anxiety, PTSD, and other psychological symptoms were commonly reported and were influenced by a range of personal, clinical, psychosocial, and protective factors. Overall, the findings indicate that psychological outcomes among ICU survivors are driven by interconnected mechanisms rather than isolated determinants, reflecting the dynamic interaction between biological vulnerability, illness severity, ICU-related exposures, and psychosocial recovery environments. Pre-ICU vulnerability factors, including demographic and health-related characteristics, may shape baseline susceptibility to psychological distress. ICU-related exposures, such as illness severity, mechanical ventilation, and prolonged hospitalization, may further intensify this risk. In contrast, post-ICU recovery resources, including physical and cognitive functioning as well as psychosocial support, may help mitigate adverse psychological outcomes.

Serum S100B was associated with increased PTSD risk,³¹ likely reflecting inflammatory brain injury that disrupts neural functioning and contributes to psychiatric symptoms [41]. Similar findings in craniocerebral trauma suggest its potential role as a predictive biomarker [42], indicating that biological pathways may contribute to post-ICU psychological outcomes and may inform future risk stratification approaches.

Seven studies linked age with psychological outcomes.^{16,20,25,31,33,35,38} Younger individuals demonstrated higher PTSD, anxiety, and depression rates,^20,31,33 whereas older adults adapted more effectively to illness-related limitations [43, 44]. Younger age has similarly predicted post-ICU distress [45] and PTSD in COVID-19 survivors [46]. This pattern may be linked to differences in coping strategies, life-stage expectations, or social roles. However, variation across studies indicates that age-related effects are not uniform and may depend on contextual factors such as recovery environment and follow-up duration.

Gender differences were consistent: females had higher odds of depression, anxiety, and PTSD,¹⁶ and greater worsening of PTSD over two years post-sepsis.³⁶ This aligns with population data suggesting higher psychological vulnerability among women, especially those with lower socioeconomic or educational status [47, 48], and prior ICU research linking female sex to PTSD [48] and depression.²⁷ These findings suggest the potential value of considering gender in post-ICU psychological assessment and support strategies.

Education showed mixed associations. Higher education predicted greater anxiety during ICU admission,³⁸ possibly due to heightened awareness,^{38, 49} yet also appeared protective against later psychological disorders,²⁰ consistent with reports linking health literacy and coping with lower risk [50–52]. This pattern suggests that education may influence both immediate emotional responses and longer-term coping capacity, highlighting the importance of tailored communication and support across different stages of recovery.

Comorbidities also influenced long-term psychological outcomes. Despite an unexpected inverse association between higher CCI and psychiatric diagnoses,¹⁶ heart failure, stroke, and COPD were linked to worse mental health and increased mortality,¹⁶ and CCI ≥6 predicted new depression and higher two-year mortality.³⁴ The association between comorbid conditions and adverse psychological outcomes may reflect the cumulative effects of physical illness, functional limitations, and ongoing healthcare needs. These findings suggest that systematic screening and integrated physical–mental health follow-up may support early detection and management of depression, psychological distress, and other post-ICU mental health outcomes.^{16, 53}

Among trauma patients, penetrating injury was associated with an increased risk of anxiety or depression one-year post-ICU,³³ likely reflecting the combined effects of physical injury and traumatic experience [54]. Similarly, treatment-related exposures, particularly mechanical ventilation, were associated with psychological outcomes, including depression and PTSD.^16,37 Longer ventilation duration was associated with greater depression [55], consistent with prior reports demonstrating higher PTSD, anxiety, and psychotropic use among ventilated patients [56, 57]. These findings suggest that clinical practices such as minimizing unnecessary sedation, optimizing communication, and structured post-ICU follow-up may help support psychological recovery.

Illness severity and ICU-related factors, including higher APACHE II scores and longer lengths of stay, were associated with increased psychological distress. Higher APACHE II scores predicted anxiety and distress,³⁸ consistent with prior work associating severity with anxiety [58]. APACHE II may help identify high-risk survivors. Longer ICU stays were consistently associated with worse psychological outcomes,^16,25 with symptoms peaking around four months post-discharge [55]. Prolonged hospitalization similarly contributed to post-ICU morbidity [56]. These factors may represent the intensity of physiological stress and exposure to invasive interventions, which together contribute to vulnerability during recovery. This suggests the potential value of early identification of high-risk patients and targeted follow-up care.

Pain, discomfort, injury severity, and physical restraints were linked to PTSD, anxiety, and depression across studies.^32,33,35,39 Pain increased PTSD risk,³⁹ persistent pain correlated with anxiety and depression,³³ and restraints worsened depressive symptoms and reduced quality of life.³² Additional studies also linked pain and restraints with PTSD, anxiety, and agitation [45, 59], supporting the importance of optimized pain management and restraint minimization.

Negative emotional responses during critical illness increased stress and anxiety,³⁸ while peritraumatic distress mediated later mental health problems [60]. Stress-management and psychological interventions may improve outcomes. Pre-existing psychiatric or substance use disorders increased risk of PTSD, anxiety, and depression,^{20,33–35,37} consistent with earlier findings in sepsis and ARDS survivors [61–63]. Early assessment and continued psychological support may be beneficial. Reduced physical function correlated with anxiety, depression, and PTSD,^20,33 consistent with COVID-19 findings linking physical and mental impairment [64]. Rehabilitation may support both domains [65].

Cognitive impairment was both a predictor and consequence of psychological morbidity,³¹ and has been associated with poorer quality of life and persistent problems [66, 67]. Integrated cognitive–psychological rehabilitation may be beneficial in addressing these interrelated outcomes. In contrast, protective factors including resilience, self-acceptance, and social support, were associated with better psychological outcomes [^{39, 40}, 69–71], highlighting the role of psychosocial resources in promoting recovery and adaptation.

These findings are broadly consistent with evidence from larger cohort studies and previous systematic reviews, which similarly identify depression, anxiety, and PTSD as common outcomes among ICU survivors. However, variations in reported prevalence and associated factors across studies may be attributed to differences in population characteristics, timing of assessment, and measurement approaches. Overall, the findings of this review suggest that psychological outcomes among ICU survivors are shaped by a complex interplay of biological, clinical, and psychosocial factors. This highlights the importance of comprehensive, individualized approaches to post-ICU care that address both physical recovery and psychological well-being. Future research should prioritize longitudinal designs and more standardized assessment methods to further clarify these relationships and support the development of targeted interventions.

Study Limitations

This review has several limitations. The included studies varied in design, sample size, measurement tools, and follow-up duration, limiting comparability. Many relied on self-reported psychological outcomes, introducing recall bias. Most were observational, preventing causal inference, and key confounders—such as pre-existing mental illness, treatment variation, and post-discharge care—were inconsistently controlled. Because most studies were conducted in high-income countries, generalizability to resource-limited settings is limited. Finally, the synthesis was constrained by variable study quality and possible publication bias. Future research should include more diverse populations, standardized assessments, longitudinal or interventional designs, and improved reporting.

Conclusion

Although survival after critical illness has improved, psychological morbidity among ICU survivors remains common. This review of 13 studies identified several factors associated with adverse psychological outcomes. These findings highlight the importance of translating risk factors into clinical practice to support early risk stratification and the development of targeted, multidisciplinary interventions aimed at improving long-term psychological recovery and quality of life. However, given the relatively small number and heterogeneity of included studies, these implications should be interpreted with caution. Further research is needed to strengthen the evidence base and to better inform targeted interventions in this population.