Abstract
Background
Although comorbid medical diseases are important determinants of outcome among the critically ill, the role of psychiatric comorbidity is not well defined. The objective of this study was to determine the occurrence of psychiatric comorbidity and its effect on the outcome of patients admitted to adult intensive care units (ICU) in Queensland.
Methods
Admissions among adults to 12 ICUs in Queensland during 2015–2021 were included and clinical and outcome information was obtained through linkages between the ANZICS Adult Patient Database, the state-wide Queensland Hospital Admitted Patient Data Collection, and death registry.
Results
A total of 89,123 admissions were included among 74,513 individuals. Overall, 7,178 (8.1%) admissions had psychiatric co-morbidity with 6,270 (7.0%) having one major psychiatric diagnosis and 908 (1%) having two or more. Individual diagnoses of mood, psychotic, anxiety, or affective disorders were present in 1,801 (2.0%), 874 (1.0%), 3,241 (3.6%) and 354 (0.4%) admissions respectively. Significant differences were observed among the main groups (mood, affective, anxiety, psychotic, or multiple disorders) and those without psychiatric comorbidity with respect to main diagnosis, Acute Physiology and Chronic Health Evaluation (APACHE II) score, sex, age, and medical comorbidity. Crude 30-day case-fatality rates were significantly lower (5.1%) compared to the general ICU population (10.1%) (p < 0.001). After controlling for confounding variables in the logistic regression model, patients with psychiatric comorbidity were at lower odds of death.
Conclusions
Psychiatric comorbidity is common among ICU presentations and is associated with a lower risk of death. This association is likely to be more complex than being a simple protective factor, and future research needs to further delineate how psychiatric comorbidity informs outcomes of specific ICU presentations.
Keywords: Psychiatry, Intensive care, Critical illness
Background
Although illness severity is a major determinant of outcome following admission to intensive care units (ICU), there has been increasing recognition of the role comorbid illnesses have on the development of persistent critical illness and overall survival [1, 2]. Technological improvements with transplants, cancer care, as well as advancing age and older populations have all increased the prevalence of comorbidities such as diabetes, cancer, and cardiovascular and renal diseases among patients admitted to ICUs in many countries [3–5]. Several studies have shown that overall comorbidity as well as a range of specific medical conditions increase the risk for death among patients admitted to ICUs, and that comorbidity is a major determinant of persistent critical illness outcome [6–9]. While the influences of comorbidities on critical illness have focused on medical illnesses, psychiatric conditions are also an important consideration. Meta-analyses have demonstrated that people living with mental illness die almost 10–20 years earlier than their peers with 14.3% of deaths attributable to mental disorders annually [10]. Mental illnesses can contribute to ICU admission [11, 12], and influence treatment approaches (such as medications, sedation, and participation in rehabilitation) impacting their length of stay [13–15].
Despite the high prevalence of psychiatric comorbidities among ICU patients [16], there is a limited body of evidence examining the relationship between psychiatric comorbidity and the determinants and outcomes of critical illness. Some studies have explored specific decision points such as admission [14], or overall outcomes of specific psychiatric presentations such as overdose [12]. More studies have explored the impact a specific psychiatric diagnosis has on the trajectory of ICU subpopulations [13, 17–24]. However, studies have shown conflicting results on the role of psychiatric comorbidity on survival, which may in part reflect small sample sizes and/or studies being conducted in single centres or highly selected cohorts. The objective of this study was therefore to identify the occurrence, clinical characteristics, and case-fatality associated with major psychiatric comorbidities among a large multicentred cohort of patients admitted to Australian ICUs.
Methods
This multicentre, retrospective observational study examined the association between psychiatric comorbidities and ICU outcomes and is described below following STROBE and RECORD guidelines.
Population
Patients aged 18.0 years or older, who were admitted to 12 participating publicly funded ICUs in Queensland between January 1st 2015 and December 31st 2021 were included.
Participating ICUs included five tertiary, three metropolitan and four regional ICUs. Together, they encompassed all state-wide referral centres for neurosurgical, transplant, burns, cardiothoracic and trauma admissions.
Ethical approval
Ethical approval was granted by the Metro South Hospital and Health Service Human Research Ethics Committee with an individual waiver of consent granted (HREC/2022/QMS/82024).
Data collection
Patients eligible for the study were identified through data linkage of local, state-wide, and national electronic health records. At each participating centre, routinely collected clinical information relating to patient demographics, physiology and investigations was retrieved from the eCritical MetaVision™ (iMDsoft, Boston, MA, USA) clinical information system. Further routinely collected patient information, such as illness severity based on the Acute Physiology and Chronic Health Evaluation (APACHE) II score, and details of ICU management were collected from the Australian New Zealand Intensive Care Society (ANZICS) Centre for Outcome and Resource Evaluation (CORE) Adult Patient Database (APD).
Once demographic and ANZICS-CORE-APD data had been established for all eligible patients, further data linkages were made with the state-wide Queensland Hospital Admitted Patient Data Collection (QHAPDC) and death registry. Linkage was performed by the Statistical Analysis and Linkage Unit of the Statistical Service Branch, Queensland Health using confidential unique patient identifiers. Once confidential identifiers were removed, the authors had full access to the linked datasets.
Diagnoses were grouped according to organ system based on APACHE III codes to reflect the admission to ICU. Sepsis was coded according to the site of infection, using International Classification of Diseases 10th Revision Australian Modification (ICD-10AM) codes where not specified. Further general groupings of infection, trauma, neurological, cardiovascular surgery, pregnancy, overdose, toxicity, and environment were based on all available codes.
Psychiatric and medical comorbidities
Details of comorbidities were established from routinely documented diagnostic information, using the ICD-10AM). Psychiatric diagnoses were categorised as Psychotic Disorders (F20-F29), Mood Disorders (F32-39), Affective Disorders (F30, F31), and Anxiety Disorders (F40-48). Medical comorbidities were classified as per the Charlson Comorbidity Index [25]. The presence of alcohol (F10, G62.1, I42.6, K29.20, K 29.21, K70.1-4, K70.5), cannabis (F12), and tobacco use disorders (F17) were similarly established using ICD-10AM codes.
Statistical analysis
Data were analysed using Stata 18.0 (StataCorp, College Station, Texas, USA). Missing data were not replaced. Prior to analysis of continuous variables, histograms were created to examine the underlying distribution. Skewed variables were reported as medians with interquartile ranges (IQR) and compared using the k-equality of medians test. Grouped categorical data were compared using the Chi2 test. A multivariable logistic regression model was developed to examine factors associated with 30-day all cause case-fatality. Variables included in the model were psychiatric diagnosis, age, sex, surgical category, main diagnostic category, ventilation status, Charlson score, APACHE II score, and presence of tobacco, alcohol, and cannabis use disorders. The model was limited to first ICU admissions among those who had more than one. The main effects model was presented without variable elimination. P-values less than 0.05 were considered significant.
Results
Overall, 89,779 admissions were identified among the 12 participating ICUs for which 89,494 (99.7%) were successfully linked to hospital admissions and death data. Of these a further 371 (0.4%) were excluded based on enrolment criteria and/or data adequacy leaving 89,123 admission episodes among 74,513 individuals for analysis. Of these many patients had multiple admissions; with 10,671 having two, 2,425 having three, 774 having four, and 740 having five or more admissions during the study.
Overall, psychiatric comorbidity was identified in 7,178 (8.1%) of ICU admission episodes, with 6,270 (7.0%) having one major psychiatric diagnosis and 908 (1%) having two or more. Among those with single psychiatric diagnoses, 874 (1.0%) had psychotic, 1,801 (2.0%) mood, 3,241 (3.6%) anxiety, and 354 (0.4%) affective disorders.
There were several differences in clinical features observed between these groups, as shown in Table 1. Subjects with psychotic disorders were more likely to be male, had more infectious diagnoses, and highest proportion of tobacco and alcohol use disorders. Subjects with mood disorders had the highest use of alcohol and the lowest proportion of infections. Patients with anxiety disorders were characterised by the highest proportion of elective surgical patients with more comorbid medical illnesses and cardiopulmonary diagnoses. Affective disorders were associated with fewer comorbid medical diseases and had the shortest length of hospital stay.
Table 1.
Clinical features associated with major psychiatric co-morbidity among patients admitted to intensive care units in Queensland
| Variable | Multiple (n=908) | Psychotic (n=874) | Mood (n=1801) | Anxiety (n=3241) | Affective (n=354) | None (n=81945) | P value |
|---|---|---|---|---|---|---|---|
| Median Age (IQR), years | 44.8 (34.1-58.3) | 47.0 (37.1-56.8) | 52.5 (39.2-64.0) | 54.8 (40.5-66.1) | 50.2 (37.2-60.5) | 61.7 (48.3-71.7) | <0.001 |
| Male | 478 (52.6%) | 583 (66.7%) | 987 (54.8%) | 1559 (48.1%) | 162 (45.8%) | 50658 (61.8%) | <0.001 |
| Surgical group | <0.001 | ||||||
| Non-surgical | 709 (78.1%) | 661 (75.6%) | 1425 (79.1%) | 1805 (55.7%) | 285 (80.5%) | 37380 (45.6%) | |
| Elective | 99 (10.9%) | 86 (9.8%) | 186 (10.3%) | 953 (29.4%) | 36 (10.2%) | 32591 (39.8%) | |
| Non-elective | 100 (11.0%) | 127 (14.5%) | 190 (10.6%) | 483 (14.9%) | 33 (9.3%) | 11974 (14.6%) | |
| Median APACHE II score (IQR) | 17 (12-22) | 16 (11-21) | 17 (12-23) | 15 (11-20) | 16 (12-22) | 15 (11-21) | <0.001 |
| Median Charlson score (IQR) | 1 (0-2) | 1 (0-2) | 0 (0-2) | 1 (0-2) | 0 (0-2) | 1 (0-2) | <0.001 |
| Charlson category | <0.001 | ||||||
| 0 | 410 (35.2%) | 341 (39.0%) | 943 (52.4%) | 1109 (34.2%) | 198 (55.9%) | 27610 (33.7%) | |
| 1-2 | 320 (35.2%) | 362 (41.4%) | 512 (28.4%) | 1355 (41.8%) | 101 (28.5%) | 34766 (42.4%) | |
| 3-4 | 137 (15.1%) | 130 (14.9%) | 235 (13.1%) | 509 (15.7%) | 36 (10.2%) | 12691 (15.5%) | |
| 5+ | 41 (4.5%) | 41 (4.7%) | 111 (6.2%) | 268 (8.3%) | 19 (5.4%) | 6878 (8.4%) | |
| Diagnosis | <0.001 | ||||||
| Cardiovascular | 134 (14.8%) | 96 (11.0%) | 202 (11.2%) | 867 (26.8%) | 39 (11.0%) | 25410 (31.0%) | |
| Pulmonary | 115 (12.7%) | 177 (20.3%) | 143 (7.9%) | 735 (22.7%) | 47 (13,3%) | 10667 (13.0%) | |
| Musculoskeletal | 85 (9.4%) | 106 (12.1%) | 176 (9.8%) | 294 (9.1%) | 28 (7.9%) | 5885 (7.2%) | |
| Gastrointestinal | 64 (1.1%) | 79 (9.0%) | 156 (8.7%) | 347 (10.7%) | 27 (7.6%) | 11700 (14.3%) | |
| Neurologic | 153 (16.9%) | 155 (17.7%) | 235 (13.1%) | 413 (12.7%) | 53 (15.0%) | 14826 (18.1%) | |
| Metabolic | 302 (33.3%) | 195 (22.3%) | 778 (43.2%) | 338 (10.4%) | 144 (40.7%) | 5271 (6.4%) | |
| Genitourinary | 20 (2.2%) | 28 (3.2%) | 49 (2.7%) | 101 (3.1%) | 2 (0.5%) | 4171 (5.1%) | |
| Sepsis no source | 30 (3.3%) | 37 (4.2%) | 52 (2.9%) | 117 (3.6%) | 14 (4.0%) | 3301 (4.0%) | |
| Othera | |||||||
| Trauma | 134 (14.8%) | 125 (14.3%) | 238 (13.2%) | 338 (19.4%) | 42 (11.9%) | 8177 (10.0%) | <0.001 |
| Infection | 137 (15.1%) | 197 (22.5%) | 233 (12.9%) | 565 (17.4%) | 49 (13.8%) | 13850 (16.9%) | <0.001 |
| Cardiovascular surgery | 43 (4.7%) | 34 (3.9%) | 79 (4.4%) | 495 (15.3%) | 13 (3.7%) | 16163 (19.7%) | <0.001 |
| Ethanol | 171 (18.8%) | 124 (14.2%) | 358 (19.9%) | 355 (11.0%) | 51 (14.4%) | 6477 (7.9%) | <0.001 |
| Tobacco | 188 (20.7%) | 281 (32.2%) | 324 (18.0%) | 567 (17.5%) | 69 (19.5%) | 9131 (11.1%) | <0.001 |
| Cannabis | 33 (3.6%) | 68 (7.8%) | 49 (2.7%) | 89 (2.8%) | 15 (4.2%) | 711 (0.9%) | <0.001 |
| Hospital | <0.001 | ||||||
| Metro | 157 (17.3%) | 147 (16.8%) | 344 (19.1%) | 359 (11.1%) | 63 (17.8%) | 7217 (8.8%) | |
| Rural/Regional | 99 (10.9%) | 134 (15.3%) | 246 (13.7%) | 350 (10.8%) | 41 (11.6%) | 9430 (11.5%) | |
| Tertiary | 652 (71.8%) | 593 (67.9%) | 1211 (67.2%) | 2532 (78.1%) | 250 (70.6%) | 65298 (79.7%) | |
| Median ICU length of stay (IQR) | 2.5 (1.1-6.0) | 2.6 (1.3-5.7) | 2.0 (1.1-4.9) | 2.1 (1.0-5.3) | 2.1 (1.2-4.7) | 1.5 (0.9-3.2) | <0.001 |
| Median hospital length of stay (IQR) | 14 (6-31) | 13 (7-23) | 11 (5-29) | 15 (9-28) | 9 (5-21) | 10 (6-17) | <0.001 |
APACHE Acute physiology and chronic health evaluation, ICU Intensive care unit, IQR Interquartile range
Othera includes 759 other subjects but detailed data not shown due to low cell numbers with some conditions
The case fatality associated with ICU admission associated with the presence of psychiatric comorbidity is displayed in Table 2.
Table 2.
Case fatality associated with ICU admission among patients with psychiatric comorbidity
| Variable | None (n=81,945) | Multiple (n=908) | Psychotic (n=874) | Mood (n=1801) | Anxiety (n=3241) | Affective (n=354) | p-value |
|---|---|---|---|---|---|---|---|
| Survive | 61,586 (75.2%) | 698 (76.9%) | 708 (81.1%) | 1,379 (76.6%) | 2,566 (79.2%) | 292 (82.5%) | <0.001 |
| ICU death | 5,661 (6.9%) | 27 (3.0%) | 23 (2.6%) | 103 (3.2%) | 11 (3.1%) | ||
| Hospital death | 2,051 (2.5%) | 20 (2.2%) | 15 (1.7%) | 58 (3.2%) | 79 (2.4%) | 9 (2.5%) | |
| Post-hospital death | 12,647 (15.4%) | 163 (18.0%) | 128 (14.7%) | 49 (2.7%) 315 (17.5%) | 493 (15.2%) | 42 (11.9%) | |
| 30-day | 8,293 (10.1%) | 39 (4.3%) | 36 (4.1%) | 87 (4.8%) | 185 (5.7%) | 17 (4.8%) | <0.001 |
| 90-day | 9,898 (12.1%) | 62 (6.8%) | 54 (6.2%) | 146 (8.1%) | 263 (8.1%) | 22 (6.2%) | <0.001 |
| 365-day | 13,360 (16.3%) | 119 (13.1%) | 87 (10.0%) | 254 (14.1%) | 425 (13.1%) | 31 (8.8%) | <0.001 |
ICU Intensive care unit
While the ICU case fatality was similar among each of the psychiatric comorbidities, they were consistently lower than for patients without psychiatric disorders. Post-ICU in-hospital case fatality was similar across all groups. The post-hospital case fatality was highest for multiple and mood psychiatric comorbidity. While the 30-day case-fatality was significantly lower among episodes associated with psychiatric comorbidity as compared to those without (364/7,178; 5.1% versus 8,293/81,945; 10.1%; RR 0.50 95% CI, 0.45-0.55; p<0.001), the relative risks for 90-(547/7,178; 0.76% versus 9,898/81,945; 12.1%; RR 0.63; 95% CI 0.58-0.69; p<0.001) and 365 day (916/7,178; 12.8% versus 13,360/81,945; 16.3%; RR 0.78; 95% CI 0.74-0.83; p<0.001) case fatality were of decreasing magnitude.
A multivariable logistic regression model was developed to examine the effect of psychiatric comorbidity on 30-day all cause case-fatality and summarised in Table 3. After controlling for diagnostic, severity of illness, medical comorbidity, and substance use, psychiatric comorbidities were found to be associated with reduced risk for 30-day all cause case-fatality. Adjusted risks for death associated with psychiatric comorbidities at 30-, 90-and 365-days post ICU admission are shown in Table 4.
Table 3.
Multivariable modelling of risk factors for 30-day all cause case fatality
| Variable | Odds Ratio | 95% CI | P-value |
|---|---|---|---|
| No psych comorbidity (reference) | 1 | - | - |
| Multiple | 0.37 | 0.24-0.56 | <0.001 |
| Psychotic | 0.37 | 0.24-0.57 | <0.001 |
| Mood | 0.36 | 0.27-0.48 | <0.001 |
| Anxiety | 0.58 | 0.47-0.71 | <0.001 |
| Affective | 0.47 | 0.25-0.90 | 0.011 |
| Male | 0.99 | 0.93-1.05 | 0.7 |
| Age (per year) | 1.01 | 1.01-10.1 | <0.001 |
| Ventilated | 1.37 | 1.29-1.46 | <0.001 |
| Non-surgical (reference) | 1 | - | - |
| Elective Surgical | 0.11 | 0.09-0.12 | <0.001 |
| Unscheduled surgical | 0.62 | 0.57-0.68 | <0.001 |
| Diagnosis Cardiovascular (reference) | 1 | - | - |
| Pulmonary | 0.67 | 0.61-0.75 | <0.001 |
| Musculoskeletal | 0.56 | 0.49-0.65 | <0.001 |
| Gastrointestinal | 0.78 | 0.70-0.86 | <0.001 |
| Hematology | 0.89 | 0.60-1.32 | 0.6 |
| Neurologic | 1.40 | 1.28-1.52 | <0.001 |
| Metabolic | 0.16 | 0.13-0.20 | <0.001 |
| Genitourinary | 0.20 | 0.17-0.24 | <0.001 |
| Pregnancy | 0.08 | 0.02-0.35 | <0.001 |
| Sepsis no source | 0.61 | 0.530.69 | <0.001 |
| Tobacco Use | 0.58 | 0.52-0.65 | <0.001 |
| Cannabis Use | 0.46 | 0.30-0.70 | <0.001 |
| Alcohol Use | 0.90 | 0.82-1.00 | 0.053 |
| Charlson Comorbidity Index | 1.16 | 1.14-1.18 | <0.001 |
| APACHE II Score | 1.16 | 1.15-1.16 | <0.001 |
APACHE Acute physiology and chronic health evaluation, CI Confidence interval
Table 4.
Adjusted risk for 30-, 90-, and 365-day all cause case fatality associated with psychiatric comorbidities
| Variable | 30-days Adjusted Odds Ratio (95% confidence interval) | 90-days Adjusted Odds Ratio (95% confidence interval) | 365-days Adjusted Odds Ratio (95% confidence interval) |
|---|---|---|---|
| No psych comorbidity (reference) | 1 | 1 | 1 |
| Multiple | 0.37 (0.24-0.56) | 0.53 (0.37-0.75) | 0.83 (0.64-1.10) |
| Psychotic | 0.37 (0.24-0.57) | 0.48 (0.34-0.69) | 0.67 (0.51-0.88) |
| Mood | 0.36 (0.27-0.48) | 0.57 (0.45-0.73) | 0.79 (0.66-0.96) |
| Anxiety | 0.58 (0.47-0.71) | 0.74 (0.63-0.88) | 0.86 (0.74-0.99) |
| Affective | 0.47 (0.25-0.90) | 0.50 (0.28-0.89) | 0.46 (0.28-0.77) |
Discussion
Our study found that patients admitted to ICU with pre-existing psychiatric diagnoses had lower mortality up to one-year post-presentation compared to the general ICU population. This supports the findings of Gacouin et al. who had found a reduced mortality rate among patients with psychiatric morbidity up to 1 year post ICU [20]. However, this low mortality rate overall is likely to consist of a range of heterogeneous subpopulations. Gacouin et al. reported 72% of their population to be ICU admissions for intentional self-harm [20], which is known to be associated with much lower mortality than other ICU presentations [26].
Such presentations must therefore not be conflated with other ICU presentations, all of which can present with a pre-existing mental illness. For example, Large et. al. found that patients admitted with major trauma and comorbid depression had a lower mortality than their peers, but also had longer ICU and hospital stays, and were more likely to undergo surgery [13]. Conversely, Fond et al. found that patients managed for COVID-19 with comorbid bipolar disorder had an increased mortality compared to their peers [17]. Another important complication to consider in the context of comorbid mental illness is delirium [27]. Further distinction must be made between patients with a history of mental illness, and those with ongoing symptoms at the time of ICU presentation.
Unfortunately, the data presented in this study is unable to differentiate between these specific subpopulations. As such, our findings do not inherently contradict the body of literature which has described how specific mental illnesses can complicate both the inpatient and post-discharge recovery processes and the relationship between mental illness, critical illness and recovery is likely to be nuanced and complex. Each mental illness diagnostic group in our study had slightly different characteristics. For example, those with anxiety disorders had the lowest APACHE II scores, and yet the longest ICU length of stay. This is partially consistent with the findings of May et al, who found both increased length of stay and illness severity [28]. As supported in other settings, those with psychotic disorders have high rates of tobacco and cannabis use [29], and those with mood disorders had the highest rate of ethanol use [30]. However, other associations, such as patients with psychotic disorders having a higher rate of infectious disorder presentations, or bipolar disorder being associated with a shorter length of ICU stay would benefit from further exploration.
Future studies should further explore and delineate how certain psychiatric comorbidities interface with different ICU presentations, and how this might influence the best approach to their ICU care as well as impact functional and other outcomes beyond mortality [31]. Differentiating the association between mental illness and ICU presentation can be challenging. For example, medication overdose can be recreational, unintentional, or with suicidal intent [32]. The same could be said for many trauma presentations, such as falls from a height or motor vehicle collisions. While the circumstances can be clear in some situations, such as when the person speaks to someone or leaves a note before the event, at other times it can remain unclear for the entirety of the ICU admission and beyond. This is even more complicated in the setting of a person who presents with both a serious medical condition and a relapse of their mental illness, such as a person presenting with sepsis and a manic episode. A manic episode can cause misadventure and personal neglect that precipitates the onset of sepsis, but it is equally possible for the illness processes of sepsis (such as pain, distress, and sleep disturbance) as well as its management (such as medications) to precipitate a manic episode.
While our study benefits from a large sample size and a case-mix reflective of a population at large, some limitations merit discussion. First. although the data available in the study databases was extensive and well standardized, it was not specifically collected for the study. We did not include psychiatric medication administrations. In addition, the presence of a psychiatric diagnosis is best made by a psychiatrist in a prospective manner. However, we do not have data to validate the accuracy for use of ICD10-AM codes in our hospitals in comparison to this “gold standard”. Studies conducted elsewhere in Australia have indicated that ICD10-AM codes may underestimate true rates of psychiatric diagnoses as established by chart review [33]. Second, the relationship between the psychiatric comorbidities and the index presentation could not be delineated. This study was unable to differentiate presentations where psychiatric comorbidities contributed to the presentation and those where this was incidental. This includes during the hospital admission, ICU admission, or both. Additionally, discharge codes may represent either acute new conditions or a pre-existing exacerbated diagnoses and by limiting to discharge codes only associated with the present admission we could not establish chronicity. Furthermore, we did not make any attempt to determine the severity of the psychiatric diagnoses. Finally, it is likely that there is a highly complex interaction between the variables evaluated in this study that we were unable to elucidate. We observed that not only psychiatric diagnoses were associated with lower risk for death but also substance use disorders. While potentially these could have some protective effect, it seems likely that they are interacting in some way with other confounding variables in the association with outcome. Further research that delves deeper into the interactions between acuity and severity of psychiatric comorbidity, underlying medical illnesses, severity and diagnosis of critical illnesses, and socioeconomic determinants is needed.
Conclusions
Psychiatric comorbidities are common among patients admitted to ICUs and are associated with improved survival for at least one-year post admission. It remains unclear as to why patients with psychiatric comorbidities suffer lower mortality related to critical illness although there are many variables that may confound this observation. Future research should further explore the nuance between psychiatric comorbidities and outcomes for various ICU presentations and whether psychiatric interventions may improve outcomes for critically ill populations at large.
Research in context
Evidence before this study
Pre-existing mental illness is a common comorbidity in the ICU population. While some studies have shown psychiatric comorbidity being associated with lower mortality, others have shown mental illness complicating treatment and outcomes of various critical illnesses, including increased mortality.
Added value of this study
This study provides a large sample of patients admitted to 12 ICUs with psychiatric comorbidity and shows decreased mortality at all time points regardless of underlying psychiatric diagnosis.
Implications of all the available evidence
At a population level, comorbid mental illness is not associated with increased short- or long-term mortality among ICU patients. However, this relationship is likely to be nuanced, and future studies need to further explore how specific mental illnesses interact with specific ICU presentations.
Acknowledgements
The authors acknowledge the Statistical Analysis and Linkage Unit of the Statistical Services Branch (SSB), Queensland Health for linking the data sets used in this project. The authors thank the ANZICS CORE management committee and the clinicians, data collectors and researchers at the following contributing sites: Caboolture Hospital, Cairns Hospital, Gold Coast University Hospital, Logan Hospital, Mackay Base Hospital, Princess Alexandra Hospital, Redcliffe Hospital, Rockhampton Hospital, Royal Brisbane and Women’s Hospital, Sunshine Coast University Hospital, The Prince Charles Hospital, and The Townsville Hospital. The Queensland Critical Care Research Network Group includes James McCullough, Kerina J Denny, Mandy Tallott, Andrea Marshall, David Moore, Sunil Sane, Aashish Kumar, Lynette Morrison, Pam Dipplesman, Ahmad Nasser, David Stewart, Vikram Shah, Kyle White, Adam Suliman, Jason Meyer, Ra’eesa Doola, Rod Hurford, Meg Harward, James Walsham, Neeraj Bhadange, Wayne Stevens, Hannah McCabe, Vijo Kuruvilla, Adam Visser, Judy Smith, Kevin B. Laupland, Felicity Edwards, Tess Evans, Jayesh Dhanani, Pierre Clement, Nermin Karamujic, Kiran Shekar, Jayshree Lavana, George Cornmell, Alexis Tabah, Stuart Baker, Hamish Pollock, Kylie Jacobs, Mahesh Ramanan, Prashanti Marella, Jatinder Grewal, Patrick Young, Julia Affleck, Peter Garrett, Paula Lister, Vikram Masurkar, Lauren Murray, Jane Brailsford, Janine Garrett, Anamika Ganju, Langa Lutshaba, Cameron Anderson, Antony G. Attokaran, Jaco Poggenpoel, Josephine Reoch, Stephen Luke, Anni Paasilahti, Jennifer Taylor, Christopher Smart, Siva Senthuran, Stephen Whebell, Sananta Dash, Philippa McIlroy, Sebastiaan Blank, Ben Nash, Michelle Gatton, Zephanie Tyack, Sam Keogh
Clinical trial number
Clinical trial number: not applicable.
Authors’ contributions
The study conception and design (KW, MR, AT, KL); data acquisition (DF, KW, FE, SB, SS, MR, AGA, AK, JM, KS, PM, AT, SL, KBL); analysis (KL); interpretation of data (DF, KW, FE, SB, SS, MR, AGA, AK, JM, KS, PM, AT, SL, KBL); article drafting (DF, KL), article revision for important intellectual content (DF, KW, FE, SB, SS, MR, AGA, AK, JM, KS, PM, AT, SL, KBL); final approval of the version submitted for publication (DF, KW, FE, SB, SS, MR, AGA, AK, JM, KS, PM, AT, SL, KBL); agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved (KW, KL).
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors
Data availability
Data cannot be shared publicly due to institutional ethics, privacy, and confidentiality regulations. Data released for research under Sect. 280 of the Public Health Act 2005 requires an application to the Director-General of Queensland Health (PHA@health.qld.gov.au).
Declarations
Ethics approval and consent to participate
This study was approved by the Metro South Hospital and Health Service Human Research Ethics Committee (HREC/2022/QMS/82024) with an individual waiver of consent granted. Individual informed consent was not required.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Kevin B Laupland, Email: Kevin.laupland@qut.edu.au.
on behalf of the Queensland Critical Care Research Network (QCCRN):
Kyle White, Felicity Edwards, Stuart Baker, Siva Senthuran, Mahesh Ramanan, Antony G Attokaran, Aashish Kumar, James McCullough, Kiran Shekar, Philippa McIlroy, Alexis Tabah, Stephen Luke, Peter Garrett, Kevin B Laupland, Kerina J Denny, Mandy Tallott, Andrea Marshall, David Moore, Sunil Sane, Lynette Morrison, Pam Dipplesman, Ahmad Nasser, David Stewart, Vikram Shah, Adam Suliman, Jason Meyer, Ra’eesa Doola, Rod Hurford, Meg Harward, James Walsham, Neeraj Bhadange, Wayne Stevens, Hannah McCabe, Vijo Kuruvilla, Adam Visser, Judy Smith, Tess Evans, Jayesh Dhanani, Pierre Clement, Nermin Karamujic, Jayshree Lavana, George Cornmell, Hamish Pollock, Kylie Jacobs, Prashanti Marella, Jatinder Grewal, Patrick Young, Julia Affleck, Paula Lister, Vikram Masurkar, Lauren Murray, Jane Brailsford, Janine Garrett, Anamika Ganju, Langa Lutshaba, Cameron Anderson, Jaco Poggenpoel, Josephine Reoch, Anni Paasilahti, Jennifer Taylor, Christopher Smart, Stephen Whebell, Sananta Dash, Sebastiaan Blank, Ben Nash, Michelle Gatton, Zephanie Tyack, and Sam Keogh
References
- 1.Esper AM, Martin GS. The impact of cormorbid conditions on critical illness. Crit Care Med. 2011;39(12):2728–35. [DOI] [PubMed] [Google Scholar]
- 2.Zador Z, Landry A, Cusimano MD, Geifman N. Multimorbidity states associated with higher mortality rates in organ dysfunction and sepsis: a data-driven analysis in critical care. Crit Care. 2019;23:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380(9836):37–43. [DOI] [PubMed] [Google Scholar]
- 4.Harrison C, Henderson J, Miller G, Britt H. The prevalence of complex multimorbidity in Australia. Aust N Z J Public Health. 2016;40(3):239–44. [DOI] [PubMed] [Google Scholar]
- 5.Prados-Torres A, Calderón-Larrañaga A, Hancco-Saavedra J, Poblador-Plou B, van den Akker M. Multimorbidity patterns: a systematic review. J Clin Epidemiol. 2014;67(3):254–66. [DOI] [PubMed] [Google Scholar]
- 6.Simpson A, Puxty K, McLoone P, Quasim T, Sloan B, Morrison DS. Comorbidity and survival after admission to the intensive care unit: a population-based study of 41,230 patients. J Intensive Care Soc. 2021;22(2):143–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ongel EA, Karakurt Z, Salturk C, Takir HB, Burunsuzoglu B, Kargin F, Ekinci GH, Mocin O, Gungor G, Adiguzel N. How do COPD comorbidities affect ICU outcomes? Int J Chronic Obstr Pulm Dis 2014:1187–96. [DOI] [PMC free article] [PubMed]
- 8.Rezoagli E, McNicholas BA, Madotto F, Pham T, Bellani G, Laffey JG, LUNG SAFE Investigators tETG. Presence of comorbidities alters management and worsens outcome of patients with acute respiratory distress syndrome: insights from the LUNG SAFE study. Ann Intensiv Care. 2022;12(1):42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jung C, Flaatten H, Fjølner J, Bruno RR, Wernly B, Artigas A, Bollen Pinto B, Schefold JC, Wolff G, Kelm M. The impact of frailty on survival in elderly intensive care patients with COVID-19: the COVIP study. Crit Care. 2021;25:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Walker ER, McGee RE, Druss BG. Mortality in Mental disorders and Global Disease Burden implications: a systematic review and Meta-analysis. JAMA Psychiatry. 2015;72(4):334–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Nohl A, Ohmann T, Kamp O, Waydhas C, Schildhauer TA, Dudda M, Hamsen U. Major trauma due to suicide attempt: increased workload but not mortality. Eur J Trauma Emerg Surgery: Official Publication Eur Trauma Soc. 2022;48(1):519–23. [DOI] [PubMed] [Google Scholar]
- 12.Athavale V, Green C, Lim KZ, Wong C, Tiruvoipati R. Characteristics and outcomes of patients with drug overdose requiring admission to Intensive Care Unit. Australasian Psychiatry: Bull Royal Australian New Z Coll Psychiatrists. 2017;25(5):489–93. [DOI] [PubMed] [Google Scholar]
- 13.Large J, Naumann DN, Fellows J, Connor C, Ahmed Z. Clinical outcomes following major trauma for patients with a diagnosis of depression: a large UK database analysis. Trauma Surg Acute care open. 2021;6(1): e000819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.King DA, Hussein E, Shochet GE, Bar-Lavie YP. Admission Rate of Patients With Major Psychiatric Disorders to the Intensive Care Unit. American J Crit. 2020;29(6):480–3. [DOI] [PubMed] [Google Scholar]
- 15.Casey E, Hughes JW, Waechter D, Josephson R, Rosneck J. Depression predicts failure to complete phase-II cardiac rehabilitation. J Behav Med. 2008;31(5):421–31. [DOI] [PubMed] [Google Scholar]
- 16.Pilowsky JK, Elliott R, Roche MA. Pre-existing mental health disorders in patients admitted to the intensive care unit: a systematic review and meta‐analysis of prevalence. J Adv Nurs. 2021;77(5):2214–27. [DOI] [PubMed] [Google Scholar]
- 17.Fond G, Pauly V, Leone M, Orleans V, Garosi A, Lancon C, Auquier P, Baumstarck K, Llorca PM, Boyer L. Mortality among inpatients with bipolar disorders and COVID-19: a propensity score matching analysis in a national French cohort study. Psychol Med. 2023;53(5):1979–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lakbar I, Leone M, Pauly V, Orleans V, Srougbo KJ, Diao S, Llorca PM, Solmi M, Correll CU, Fernandes S, et al. Association of severe mental illness and septic shock case fatality rate in patients admitted to the intensive care unit: a national population-based cohort study. PLoS Med. 2023;20(3): e1004202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kao LT, Liu SP, Lin HC, Lee HC, Tsai MC, Chung SD. Poor clinical outcomes among pneumonia patients with depressive disorder. PLoS ONE. 2014;9(12): e116436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gacouin A, Maamar A, Fillatre P, Sylvestre E, Dolan M, Le Tulzo Y, Tadié JM. Patients with preexisting psychiatric disorders admitted to ICU: a descriptive and retrospective cohort study. Ann Intensive Care. 2017;7(1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shen HN, Lu CL, Yang HH. Increased risks of acute organ dysfunction and mortality in intensive care unit patients with schizophrenia: a nationwide population-based study. Psychosom Med. 2011;73(7):620–6. [DOI] [PubMed] [Google Scholar]
- 22.Obed D, Schroeter A, Gruber L, Salim M, Krezdorn N, Vogt PM. Outcomes following burn injury in intensive care patients with major psychiatric disorders. Burns: J Int Soc Burn Injuries. 2023;49(4):830–7. [DOI] [PubMed] [Google Scholar]
- 23.Barel N, Bdolach-Abraham T, Levin P, Einav S. Psychiatric patients’ intensive care admission characteristics, weaning from mechanical ventilation and sedative drug use: a single center retrospective case-control study. J Crit Care. 2023;77:154331. [DOI] [PubMed] [Google Scholar]
- 24.Rosgen BK, Moss SJ, Soo A, Stelfox HT, Patten SB, Fiest KM. Healthcare utilization and mortality outcomes in patients with pre-existing psychiatric disorders after intensive care unit discharge: a population-based retrospective cohort study. J Crit Care. 2021;66:67–74. [DOI] [PubMed] [Google Scholar]
- 25.Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–9. [DOI] [PubMed] [Google Scholar]
- 26.Fernando SM, Reardon PM, Ball IM, Van Katwyk S, Thavorn K, Tanuseputro P, Rosenberg E, Kyeremanteng K. Outcomes and costs of patients admitted to the intensive care unit due to accidental or intentional poisoning. J Intensive Care Med. 2020;35(4):386–93. [DOI] [PubMed] [Google Scholar]
- 27.Aung Thein MZ, Pereira JV, Nitchingham A, Caplan GA. A call to action for delirium research: Meta-analysis and regression of delirium associated mortality. BMC Geriatr. 2020;20:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.May AD, Parker AM, Caldwell ES, Hough CL, Jutte JE, Gonzalez MS, Needham DM, Hosey MM. Provider-documented anxiety in the ICU: prevalence, risk factors, and associated patient outcomes. J Intensive Care Med. 2021;36(12):1424–30. [DOI] [PubMed] [Google Scholar]
- 29.Khokhar JY, Dwiel LL, Henricks AM, Doucette WT, Green AI. The link between schizophrenia and substance use disorder: a unifying hypothesis. Schizophr Res. 2018;194:78–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Puddephatt JA, Irizar P, Jones A, Gage SH, Goodwin L. Associations of common mental disorder with alcohol use in the adult general population: a systematic review and meta-analysis. Addiction. 2022;117(6):1543–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Flaws D, Patterson S, Bagshaw T, Boon K, Kenardy J, Sellers D, Tronstad O. Caring for critically ill patients with a mental illness: a discursive paper providing an overview and case exploration of the delivery of intensive care to people with psychiatric comorbidity. Nurs Open. 2023;10(11):7106–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Harriss L, Hawton K, Zahl D. Value of measuring suicidal intent in the assessment of people attending hospital following self-poisoning or self-injury. Br J Psychiatry. 2005;186(1):60–6. [DOI] [PubMed] [Google Scholar]
- 33.Nguyen TQ, Simpson PM, Braaf SC, Cameron PA, Judson R, Gabbe BJ. Level of agreement between medical record and ICD-10-AM coding of mental health, alcohol and drug conditions in trauma patients. Health Inf Manag. 2019;48(3):127–34. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data cannot be shared publicly due to institutional ethics, privacy, and confidentiality regulations. Data released for research under Sect. 280 of the Public Health Act 2005 requires an application to the Director-General of Queensland Health (PHA@health.qld.gov.au).