Abstract
Background
Social support (SS) may contribute to the long-term recovery of critical illness survivors. This study focuses on survivors of acute respiratory distress syndrome (ARDS) to investigate the causal relationship between SS and health-related quality of life (HRQoL) and healthcare utilisation in critically ill patients.
Methods
A cohort study with 877 ARDS survivors in 61 intensive care units (ICUs) was conducted in Germany between 2014 and 2019. SS was measured using the F-SozU K-14 (Fragebogen zur sozialen Unterstützung) scale and HRQoL was assessed using the Physical and Mental Component Summaries of Short Form-12 at 3, 6, 12 and 24 months after ICU discharge. Healthcare utilisation was assessed after 12 and 24 months. To identify confounders and allow for causal inferences, a directed acyclic graph was developed.
Results
Adjusted regression models demonstrated significant positive impact of SS on mental HRQoL after 3 months onward (all β values >0.15, all p-values <0.05). This influence increases over time. In contrast, the influence of SS on physical HRQoL and healthcare utilisation remained inconclusive (only one significant association for physical HRQoL at 12 months: β=0.128, p<0.05, otherwise all p-values >0.05).
Conclusion
Results indicate SS plays an important and unique role in the long-term recovery of survivors of critical illness in terms of mental health. It appears that the more distal mechanism of SS unfolds progressively over time, perhaps as the immediate sequelae of critical illness after discharge subside. In contrast, SS does not appear to exert a substantial causal impact on physical HRQoL and healthcare utilisation.
Shareable abstract
This cohort study shows that social support (SS) plays a unique role in the mental recovery process of ARDS survivors (mental health-related quality of life (HRQoL)). In contrast, SS does not have an impact on physical HRQoL or healthcare utilisation. https://bit.ly/4gdTb80
Introduction
Survivors of the acute respiratory distress syndrome (ARDS) frequently describe enduring physical, cognitive and mental health impairments, commonly referred to as post-intensive care syndrome (PICS) [1–4]. Consequently, ARDS survivors often report significantly diminished health-related quality of life (HRQoL) and heightened healthcare utilisation in the years following discharge from an intensive care unit (ICU) (e.g. [5–7]). Given the profound and often persistent nature of PICS sequelae, identifying factors important for the long-term recovery of survivors of critical illness, such as ARDS, has become a central focus in current research [8–10].
In addition to medical considerations (e.g. general health status, medication or rehabilitation programmes), social factors, particularly social support (SS), may have a protective effect after hospital discharge and might contribute to the long-term recovery of ARDS survivors. SS may encompass instrumental support (e.g. providing transportation to medical appointments or aiding in physical therapy), emotional support (e.g. as protective factor for depression or post-traumatic stress disorder (PTSD) [11, 12]) or informational support (e.g. in evaluating treatment options) [13, 14]. The degree of feeling supported may depend on the specific situation, e.g. in the case of transportation to appointments, the type of transportation (public versus private) or whether a caregiver is present may have an impact.
However, research on determinants of post-ICU recovery in ARDS survivors primarily focuses on disease-related factors, such as comorbidities, pulmonary functioning or duration of hospitalisation. In contrast, SS is not commonly reported as a determinant of HRQoL following ARDS, as evidenced by Dodoo-Schittko et al.'s [15] review of 24 relevant articles. Similarly, in a review on the social determinants of recovery from critical illness by Jain et al. [13], only one of eight included studies looked at interpersonal level variables and found social isolation to be associated with an increased burden of disability among ICU survivors (albeit within 1 year and social isolation measured before the index ICU stay) [16]. Fittingly, in their recent consensus paper on the prediction and identification of long-term impairments after critical illness, Mikkelsen et al. [1] concluded that “social determinants of health could be key factors for post-ICU mental health problems; these have not been adequately researched, but should be” (p. 1674).
Furthermore, the existing body of research on the influence of SS on post-ICU recovery after critical illness is often limited to post-ICU time spans of 1 year or less or focuses on SS only during the ICU stay or the transition period from the ICU to home [13, 17–22]. Moreover, the reported outcomes are mostly limited to mental outcomes, whereas physical outcomes (including physical HRQoL) or behavioural indicators for recovery (e.g. healthcare utilisation) are rarely included. For instance, Milton et al. [18] demonstrated in a prospective cohort study that (lack of) perceived SS predicts psychological morbidity 3 months after ICU discharge. In a study on ARDS survivors, Deja et al. [17] found that perceived SS measured during the ICU stay was associated with a reduction in PTSD symptoms (on average 57±32 months) after discharge. Thus, even though some research exists, the possible protective role of SS for the long-term recovery process of ARDS survivors has not been thoroughly explored.
Against this background, the aim of this study was to investigate the possible long-term causal protective role of SS on HRQoL and healthcare utilisation in survivors of ARDS.
Materials and methods
The study was pre-registered at clinical clinicaltrials.gov (identifier: NCT02637011). It was approved by the Ethics Committee of the University of Regensburg (file number: 13-101-0262) and (if required) additionally by the Ethics Committees of the participating hospitals. Written informed consent to participate and consent to publish was obtained from 1225 patients (see figure 1). The study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement [23].
FIGURE 1.
Patient flow diagram. FUP: follow-up; FUP 5: no assessment of social support at 36 months due to study design; ICU: intensive care unit. Reproduced from [27] with permission.
Study design and sample
For the prospective, multicentre DACAPO cohort study (Surviving ARDS: the influence and quality of care and individual patient characteristics on health-related quality of life; funded by German Federal Ministry of Education and Research), 1900 adult patients with diagnosed ARDS were recruited between September 2014 and May 2016 at 61 ICUs in Germany [24–26]. Of the initially included patients, 877 ICU survivors (DACAPO cohort) received self-report questionnaires by mail at 3, 6, 12, 24 and 36 months after hospital discharge (follow-up period lasted until May 2019 [27]). For this study, we analysed data collected at 3 (n=456), 6 (n=389), 12 (n=396) and 24 (n=218) months after discharge from all participants who returned at least one valid questionnaire at any follow-up (figure 1). The methods of the DACAPO study have been reported in detail elsewhere [25, 26].
Measures
SS, subjective social status (SSS) and HRQoL were assessed at 3, 6, 12 and 24 months after ICU discharge. Healthcare utilisation was only assessed at 12 and 24 months, as the time frame covered was “the previous 12 months”. Age, sex and socioeconomic status (SES) were assessed at ICU admission.
Socioeconomic status
To measure objective social status, we evaluated SES based on education and occupational level before ICU admission [28]. Education is operationalised by an individual's highest academic qualification (e.g. certificate of secondary education or university diploma), while occupational level reflects their professional status (e.g. unskilled worker, qualified employee or civil servant). For each dimension, a minimum of one and a maximum of seven points are assigned (in-between points are possible). A final SES score ranging from 1 to 7 was calculated by aggregating the minimum of one and maximum of seven points assigned for each dimension [28].
Subjective social status
The MacArthur Scale [29–31] was used to assess SSS. Participants were asked to place themselves on the scales of a ladder (range 1 (lowest) to 10 (highest)) according to their perceived social position in society. The validated German version of the scale was used and participants were asked to rate themselves in the context of German society [30–32].
Social support
The F-SozU K-14 (Fragebogen zur sozialen Unterstützung) instrument [33] was used to assess SS. This measure assesses SS as perceived or anticipated support [33–35]. The 14 items were rated on a five-point Likert scale (ranging from “not agree” to “exactly right”), with higher values indicating higher levels of SS (range 1–5). Sample items are “I have no trouble in finding someone to take care of my flat when I am away” or “I know multiple people I enjoy spending time with”. For the final score, the mean value of all answered items is computed. The F-SozU K-14 has been normalised for the general German population [35].
Health-related quality of life
The Short Form-12 (SF-12) self-reported questionnaire was used to measure HRQoL [36–38]. Scores for the Physical and Mental Component Summaries of the SF-12 (PCS-12 and MCS-12) range from 0 to 100 (higher values indicate better HRQoL) and are normalised for the general German population, resulting in a standard score with a mean±sd of 50±10 [38, 39].
Healthcare utilisation
To assess healthcare utilisation, participants were asked to report their visits to any of the following 13 ambulatory medical specialties during the past 12 months: general practitioner, internist, gynaecologist, ophthalmologist, orthopaedist, otolaryngologist, neurologist/psychiatrist, psychotherapist, surgeon, dermatologist, radiologist, dentist/orthodontist and “other”. As a measure for healthcare utilisation, the total number of yearly visits (across all specialties) was computed [40, 41].
Statistical analysis
Although observational data methods can only establish associations, the objective of our study is causal [42–44]. To increase the likelihood of modelling causal associations in our data, we employed directed acyclic graphs (DAGs), which encode causal relationships and conceptualise confounding based on prior knowledge. The process of developing DAGs is formalised and analytical, facilitating the systematic identification and visual representation of causal structures between variables. It also helps to distinguish confounders, mediators and colliders [44–48].
To select covariates and model the proposed causal pathways, we used the model described by Wilson and Cleary [49–51] and the Andersen [52] behavioural model for health service utilisation [53, 54] as theoretical frameworks. The Wilson and Cleary [49–51] model is a conceptual model for HRQoL and integrates biological/physiological factors, psychological and functional aspects, as well as general health perception. The Andersen [52] behavioural model offers a theoretical framework to understand access to and utilisation of health services and to identify the factors that affect an individual's decision to use or not use available health services. The model predicts that a sequence of predisposing, enabling and need factors influence an individual's utilisation of health services.
The following covariates were selected for the DAG: age, sex, SES, SSS, living situation (living with partner (yes/no), marital status and number of persons living in household), employment situation, health insurance (private versus statutory), lifestyle (smoking, drinking and body mass index), activities of daily living (as a measure for disability) and comorbidity (mental and physical). Confounding variables were selected from among these covariates by using a DAG of proposed associations between variables to guide our analyses (figure 2). Specifically, SS (exposure) and HRQoL (outcome) are linked by a direct hypothesised causal pathway, visualising the central research hypothesis. Sex, age, SSS and SES are confounders with regard to the central research hypothesis. No mediators or colliders were identified for the association between SS (exposure) and HRQoL/healthcare utilisation (outcome) [46]. Therefore, the DAG-implied adjustment variables are age, sex, SSS and SES. Supplementary file 1 provides further information on the theoretical derivation of the DAG.
FIGURE 2.
A directed acyclic graph showing the association between social support and health-related quality of life (HRQoL)/HCU. Exposure: social support. Outcome: HRQoL/HCU. Dashed boxes: covariates. Grey boxes: unmeasured covariates. Living situation: composite measure of living with partner (yes/no), marital status (married, single, divorced, widow) and number of persons in household. ADL: activities of daily living; MCS-12: Mental Component Summary of Short Form-12; PCS-12: Physical Component Summary of Short Form-12; SSS: subjective social status; SES: socioeconomic status. #: Causal path health insurance only for healthcare utilisation (HCU).
For continuous variables, either the median and interquartile range (IQR) or mean±sd are reported. Counts and percentages are used for categorical variables. Effect sizes of bivariate associations between variables of interest were computed using Pearson's correlation. Regression models adjusted for age, sex, SSS and SES were then computed for each outcome and at each measurement point, namely linear regression for HRQoL and negative binomial regression for healthcare utilisation. Adjusted R2 and ΔR2 were computed to ascertain the explanatory power of each model [55, 56]. No predefined cut-offs for R2 were used. ΔR2 is reported in comparison to baseline a model with age, sex, SES and SSS as predictors. All analyses were performed using SPSS version 26.0 and R version R-4.3.2.
Results
Sample
Characteristics of the study participants (study sample, n=587, figure 1) are presented in table 1. Variables include sociodemographic and medical characteristics (see also [25] for details on the DACPO cohort profile). The mean±sd age of participants was 53.9±15.4 years (median: 55; IQR: 19) and about two thirds were male. More than half of the participants (56.9%) had at least an intermediate school level education and about three quarters of the sample were living with a partner.
TABLE 1.
Sociodemographic and medical characteristics of study participants (n=587)
| Variable | n | Result |
|---|---|---|
| Sex# male, n (%) | 587 | 397 (67.6) |
| Age, years (mean±sd) | 587 | 53.9±15.4 |
| Educational level | 493 | |
| No school leaving certificate, n (%) | 10 (2.0) | |
| No school leaving certificate yet, n (%) | 4 (0.8) | |
| Secondary school leaving certificate, n (%) | 194 (39.4) | |
| Intermediate school leaving certificate, n (%) | 170 (34.4) | |
| University entrance level, n (%) | 111 (22.5) | |
| Other | 4 (0.8) | |
| Education score (median, IQR)¶ | 521 | 3.6 (3.0–3.6) |
| Employment situation before onset of ARDS | 518 | |
| Full time, n (%) | 207 (40) | |
| Part time, n (%) | 41 (7.9) | |
| Irregular, n (%) | 12 (2.3) | |
| Unemployed/retired, n (%) | 258 (49.8) | |
| German nationality, n (%) | 568 | 541 (95.2) |
| Living with partner (yes), n (%) | 560 | 420 (75) |
| Marital status | 555 | |
| Married/civil partnership, n (%) | 349 (62.9) | |
| Single, n (%) | 142 (25.6) | |
| Divorced, n (%) | 43 (7.7) | |
| Widowed, n (%) | 21 (3.8) | |
| Number of persons in household (including patient) | 530 | |
| 1 | 90 (17.0) | |
| 2 | 281 (53.0) | |
| 3 | 83 (15.7) | |
| 4 | 59 (11.1) | |
| ≥5 | 17 (3.2) | |
| Health insurance | 541 | |
| Statutory, n (%) | 471 (87.1) | |
| Private, n (%) | 60 (11.1) | |
| Other, n (%) | 10 (1.8) | |
| SOFA score at admission (without GCS), median (IQR) | 514 | 8 (6–10) |
| SAPS-II score at admission (without GCS), median (IQR) | 527 | 38 (31–47) |
| Cause of ARDS | 556 | |
| Pulmonary, n (%) | 463 (83.3) | |
| Extrapulmonary, n (%) | 93 (16.7) | |
| Diagnosis of ARDS | 572 | |
| Diagnosis in participating ICU, n (%) | 357 (62.4) | |
| Diagnosis in other ICU (transferred after diagnosis to participating ICU), n (%) | 215 (37.6) | |
| Severity of ARDS+ | 574 | |
| Mild, n (%) | 65 (11.3) | |
| Moderate, n (%) | 268 (46.7) | |
| Severe, n (%) | 241 (42.0) | |
| Length of ICU stay until discharge (days), median (IQR) | 563 | 23 (14–36) |
| Mechanical ventilation at discharge from ICU, n (%) | 575 | 92 (16) |
#: Sex was assessed at intensive care unit (ICU) baseline (T0) at the clinic of admission, categories were binary: male or female. ¶: Educational score: derived from educational and professional level [28]. +: According to the Berlin definition [59]. ARDS: acute respiratory distress syndrome; GCS: Glasgow coma scale; IQR: interquartile range, SAPS-IIL Simplified Acute Physiology Score-II [57]; SOFA: sequential organ failure assessment [58].
Overview results
Descriptive results for SS, HRQoL and healthcare utilisation are presented in table 2. SS remains stable across all measurement points with a median of 4.3–4.4 (IQR 3.8–4.0 to 4.8–4.9, range 1–5) (figure 3). PCS-12 and MCS-12 scores show a slight increase from 3 to 12 months and remain stable afterwards, but remain below the population norm throughout [27], while PCS-12 scores are consistently lower than MCS-12 scores, which is in line with previous literature [10]. Healthcare utilisation remains stable from 12 to 24 months, but is markedly elevated compared to the German general population [41].
TABLE 2.
Main exposure and outcomes
| Variable | 3 months | 6 months | 12 months | 24 months |
|---|---|---|---|---|
| n | 456 | 389 | 396 | 218 |
| F-SozU K-14 # , median (IQR) | 4.4 (4.0–4.9) | 4.4 (3.9–4.9) | 4.4 (3.9–4.9) | 4.3 (3.8–4.8) |
| Valid (n) | 428 | 366 | 382 | 212 |
| PCS-12, median (IQR) | 36 (31–43) | 39 (31–49) | 42 (34–52) | 42 (35–54) |
| Valid (n) | 322 | 298 | 313 | 184 |
| MCS-12, median (IQR) | 44 (32–54) | 46 (33–55) | 47 (33–57) | 45 (30–55) |
| Valid (n) | 322 | 298 | 313 | 184 |
| HCU ¶ , median (IQR) | – | – | 15 (8–25), | 14 (8–21) |
| Valid (n) | – | – | 367 | 218 |
#: F-SozU K-14 (Fragebogen zur sozialen Unterstützung) social support score (range 1–5). ¶: Healthcare utilisation (HCU) (total number of out-patient medical specialists visited during past 12 months). –: no data for healthcare utilisation at 3 and 6 months, since time span of measure is 12 months. IQR: interquartile range; MCS-12: Mental Component Summary of Short Form-12; PCS-12: Physical Component Summary of Short Form-12.
FIGURE 3.
Histograms of perceived social support at 3, 6, 12 and 24 months follow-ups. Social support assessed with F-SozU K-14 (Fragebogen zur sozialen Unterstützung) (composite score of 14 items, range 1 to 5).
HRQoL
Bivariate correlations between SS and PCS-12/MCS-12 scores were statistically significant at all measurement points, except for PCS-12 scores at 3 and 6 months (with small to large effect sizes: r=0.183–0.6 [60]) (table 3). After adjusting for confounders, SS is statistically associated only with PCS-12 score at 12 months and MCS-12 score at 6, 12 and 24 months (β=0.128–0.540, all p-values <0.05). The direct impact of SS on MCS-12/PSC-12 scores at these measurement points is also reflected in the increase of explanatory power (ΔR2=0.014–0.232, p<0.05) when SS is added as a predictor into the baseline model with sex, SES and SSS as predictors. Interestingly, the absolute impact of SS on MSC-12 score consistently and substantially increases with time.
TABLE 3.
Summary of causal effect of social support (SS) on the Physical and Mental Component Summaries of Short Form-12 (PCS-12 and MCS-12) at 3, 6, 12 and 24 months follow-up after adjusting for age, sex, socioeconomic status (SES) and subjective social status (SSS)
| Follow-up | n | Outcome | r# | B ¶ | β¶ | Adjusted R2 | ΔR2+ |
|---|---|---|---|---|---|---|---|
| 3 months | 296 | PCS-12 | 0.071 | −0.360 | −0.028 | 0.123 | 0.001 |
| MCS-12 | 0.183** | 1.207 | 0.068 | 0.168 | 0.004 | ||
| 6 months | 256 | PCS-12 | 0.094 | −1.063 | −0.066 | 0.203 | 0.004 |
| MCS-12 | 0.278*** | 3.043 | 0.155* | 0.158 | 0.014* | ||
| 12 months | 265 | PCS-12 | 0.268*** | 1.955 | 0.128* | 0.25 | 0.015* |
| MCS-12 | 0.437*** | 5.287 | 0.289*** | 0.249 | 0.075*** | ||
| 24 months | 160 | PCS-12 | 0.231** | 0.821 | 0.053 | 0.211 | 0.002 |
| MCS-12 | 0.600*** | 10.398 | 0.540*** | 0.423 | 0.232*** |
*: p<0.05. **: p<0.01. ***: p<0.001. #: r bivariate correlation between SS and PCS-12/MCS-12. ¶: B and β for final regression model. +: ΔR2 in comparison to baseline model with age, sex, SES and SSS as predictors.
Healthcare utilisation
Negative binomial regression models were used to assess the causal influence of SS on healthcare utilisation, with adjustments made for age, sex, SES and SSS. The results indicate that there was no significant association between SS and healthcare utilisation at 12 and 24 months (all p-values >0.15).
Discussion
Key findings
Our findings indicate that SS is causally associated with MCS-12 score at 6, 12 and 24 months after ICU discharge, following adjustment for age, sex, SES and SSS. Moreover, this positive effect increases over time. In contrast, no causal associations between SS and PCS-12 score/healthcare utilisation were found, except for PCS-12 score at the 12-months follow-up.
Interpretation, in relation to literature
The present study indicates a long-term causal association between SS and MCS-12 score, which is consistent with previous research demonstrating the positive effects of SS on mental outcomes in ICU survivors (e.g. [17, 22, 61, 62]). A particularly interesting finding in our study is the increasing strength of this association over time. At 3 months, no association was found (after controlling for confounder), whereas from 6 months onwards, SS had a significant impact on mental HRQoL. It is possible that the initial months after being discharged from the ICU are marked by physical discomfort and medical interventions, such as pain management or rehabilitation programmes, that are not significantly impacted by SS. However, as recovery progresses following the initial trauma (cf. the “big hit” trajectory of recovery [63]), the more distal effects of SS may become more apparent.
In contrast to MCS-12 score, PCS-12 score and healthcare utilisation were not substantially affected by SS. This discrepancy may be due to the fact that the impact of SS on physical outcomes and healthcare utilisation is necessarily mediated and therefore more indirect (e.g. through instrumental SS), as opposed to the more direct effects of SS on mental outcomes (e.g. through emotional support). This is a possibility that future research should investigate. In this respect, SS is conceptually closer to mental HRQoL. For instance, social contact with other people is often considered an explicit aspect of mental HRQoL (in fact, one item of the PCS-12 scale specifically asks about social contact with other people). However, in contrast to our findings, Tilburgs et al. [18] found that instrumental and emotional SS have a buffering effect on the physical dimension of quality of life in a cross-sectional survey of former ICU patients. In a study on older adults with critical illness, Falvey et al. [16] found that social isolation before ICU hospitalisation was associated with greater disability and higher mortality in the year following critical illness. Therefore, SS may have an impact on physical outcomes, albeit to a considerably lower extent (see also [61]). Future research is needed to evaluate this question and possible underlying mechanisms, especially since most studies on the effect of SS on post-ICU measures tend to focus on mental health outcomes [17].
To the best of our knowledge, this is the first study to examine the possible link between SS and healthcare utilisation in ARDS survivors. The finding of no association between SS and healthcare utilisation 12 and 24 months after ICU discharge could be explained in several ways. For example, the German healthcare system offers universal coverage for all citizens, meaning that SS at home might not be as relevant as a resource in this respect, as people do not depend on it as much. Additionally, the German social system includes long-term care insurance, which provides patients with in-home care when necessary. Therefore, the peculiarities of the German healthcare system, as well as cross-cultural differences in SS in general [64], may limit the generalisability of our results to other countries.
It is noteworthy that SS was generally high in our sample, with a median of 4.3–4.4 (mean±sd of 4.17±0.72–4.29±0.77) on a five-point scale and a distribution skewed to the left (skewness −1.3–−1.9). This result suggests that SS remains stable and unaffected by ARDS, despite challenges posed by critical illness, such as job loss (and loss of contact with coworkers) or reduced social participation due to disability. In their standardisation study of the F-SozU K-14 instrument on a representative German sample, Fydrich et al. [33] reported similar results (mean±sdL 3.97±0.68; skewness: −0.69). This suggests that SS in our study is comparable to that of the general German population and may even be somewhat more pronounced. However, it is important to note that these findings could be partly explained by a bias towards responses from patients with higher levels of SS, as one form of instrumental SS could be assistance with survey completion.
Our findings expand existing research on the association between SS and HRQoL and healthcare utilisation in critically ill patients by providing a longer-term perspective in a population of survivors of ARDS. SS appears to have differential effects on mental and physical HRQoL at different points in the recovery process. This finding is consistent with the perspective that recovery is a process with changing patient support needs, depending on the stage and time. The “Timing it Right Framework” [21, 65, 66] supports this view. During the first months, the focus is on immediate physical medical care, such as rehabilitation programme, pain management and adapting to physical impairments. As patients adapt to their new circumstances and gain more functional independence, SS becomes more effective.
Strengths and limitations
The present study has a key strength in its substantial and diverse patient sample across various ICUs, including both university and nonuniversity hospitals, with multiple, discrete follow-up time points over 24 months. Characteristics of ARDS patients from the DACAPO cohort are similar to other large ARDS cohorts regarding age and sex distribution [25]. The analysis and causal interpretation of the present study were strengthened by establishing a theoretically deduced a priori DAG to guide confounder selection. Additionally, healthcare utilisation was incorporated as an objective behavioural measure to complement the subjective assessment of HRQoL.
Several limitations require consideration. Firstly, it is not possible to control for potential selective dropout/nonresponse between measurement points, which introduces the risk of self-selection bias. For example, it is possible that individuals with higher HRQoL or SS are more likely to respond. In our study, approximately one-third of the total number of ICU survivors did not respond to any follow-up and the follow-up sample at 3 months (n=456) was further reduced by approximately 55% compared to 24 months (n=218) (see figure 1). Dropout analyses between 3 and 24 months showed that there were no statistically significant differences for age, sex, SS, PCS-12 score and MCS-12 score between the study sample at 24 months (n=218) and the dropout sample (n=238; t-tests and chi-squared tests, all p>0.05). However, as the specific reasons for nonresponses remain unknown, the risk of nonrandom dropout limits the interpretation and generalisability of results. For instance, it is still possible that our sample was healthier than the average ARDS survivor, due to the possibility of highly morbid participants or survivors with psychopathology dropping out with a higher likelihood (self-selection bias). Furthermore, we cannot exclude the possibility that in some instances caregiver may have helped to complete the questionnaires to varying degrees, which may have introduced bias.
Secondly, the reciprocal nature of the relationship between SS and HRQoL/healthcare utilisation and subsequent measurement points cannot be modelled by a DAG and requires attention in future research. The interplay between SS and its outcomes is intricate. Support may influence recovery, but critical illness and comprised recovery may also affect patients’ access to SS [13, 16]. It is further possible that confounding factors not assessed in our study or addressed by our theory-driven DAG could influence the findings.
With regard to healthcare utilisation, recall bias is possible as patients self-reported their visits to medical specialties over the course of 1 year. This needs to be considered in the interpretation of the results. Future research might include additional and more objective measures for healthcare utilisation assessment, such as medical records.
Future research
To better understand the long-term benefits of SS for survivors of critical illness, it is necessary to conduct more longitudinal studies with multiple measurement points. This will allow for the monitoring of changes and the effects of SS among ARDS survivors over an extended period, possibly beyond the 24-month follow-up period used in the present study. Future studies of this type could offer greater understanding of the temporal dynamics of SS and its effects on long-term outcomes by analysing the evolution of SS after ICU discharge and its influence on patients’ well-being.
Additionally, future research could explore the feasibility of early detection of unmet social needs in ICU survivors. This could aid in identifying at-risk populations by implementing screening tools for SS and social isolation. Once vulnerable individuals are identified, interventions to enhance SS can be implemented. Interventional studies aimed at improving SS for ARDS survivors during the post-ICU recovery period could involve social worker-initiated care, peer support groups (digital or in-person) or caregiver training initiatives tailored to the needs of survivors and their social networks. The aim of these studies will be to assess the effectiveness of targeted support interventions in improving the (mental) well-being of ARDS survivors. This will be achieved by comparing the outcomes of participants who receive the SS intervention with those who receive standard care. Additionally, it would be valuable to investigate the cost-effectiveness and scalability of such interventions to inform healthcare policies and practices.
Conclusions
Recovering from critical illness and regaining independence is a multifaceted and complex process. Coping with the sequelae of recovery presents different needs and challenges at different points in time, often many years after hospital discharge. This study provides a comprehensive, long-term examination of how SS influences key outcomes during post-ICU recovery. We found a significant causal association between SS and mental HRQoL. This relationship strengthens over time as immediate medical concerns subside. In contrast, there is no conclusive evidence of a direct causal impact of SS on physical HRQoL or healthcare utilisation. Our results demonstrate that is important to evaluate SS beyond the “critical first year” after discharge while distinguishing between outcomes. Future research should explore whether identifying at-risk populations and implement SS interventions for survivors of critical illness could improve their well-being in the long-term recovery process.
Acknowledgement
The study was conducted in multiple areas in Germany. During the preparation of this work, the authors used ChatGPT and DeepL Write in order to improve the language and readability of their paper. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Footnotes
Provenance: Submitted article, peer reviewed.
Ethics statement: The study was approved by the Ethics Committee of the University of Regensburg (file number: 13-101-0262) and (if required) additionally by the Ethics Committees of the participating hospitals. Written informed consent to participate and consent to publish was obtained from 1225 patients. Written informed consent was given by the patients or their caregivers or legal guardians during the intensive care unit length of stay.
Author contributions: Conceptualisation: H. Szymczak, C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha and T. Bein. Data curation: S. Brandsetter, F. Dodoo-Schittko and M. Rohr. Formal analysis: H. Szymczak, C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko and M. Rohr. Funding acquisition: C. Apfelbacher and T. Bein. Investigation: C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha, T. Bein and M. Rohr. Methodology: H. Scymczak, C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha, T. Bein and M. Rohr. Project administration: C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha, T. Bein and M. Rohr. Resources and software: H. Szymczak, C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha and T. Bein. Supervision: C. Apfelbacher and S. Brandsetter. Validation: H. Szymczak, C. Apfelbacher, S. Brandsetter, F. Dodoo-Schittko, S. Blecha, T. Bein and M. Rohr. Visualisation: H. Szymczak. Writing – original draft: H. Szymczak and C. Apfelbacher. Writing – review and editing: S. Brandsetter, F. Dodoo-Schittko, S. Blecha and T. Bein.
Conflict of interest: All authors declare that the DACAPO study was funded by a research grant from the German Federal Ministry of Education and Research (01GY1340). No further conflicts of interest are declared.
Support statement: The DACAPO study was funded by a research grant from the German Federal Ministry of Education and Research (01GY1340). Grant holders were T. Bein (University Hospital Regensburg, principal investigator) and C. Apfelbacher (University of Regensburg, co-principal investigator). S. Brandsetter, F. Dodoo-Schittko, M. Rohr and S. Blecha were funded by this grant for parts of or the entire study period. All other authors received payments from the grant to support patient recruitment. The funding body had no role in the design of the study; in the collection, analysis or interpretation of the data; or in writing the manuscript. Funding information for this article has been deposited with the Crossref Funder Registry.
Supplementary material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
01113-2024.SUPPLEMENT
Data availability
The datasets used and analysed during the current study are available from C. Apfelbacher on reasonable request.
References
- 1.Mikkelsen ME, Still M, Anderson BJ, et al. Society of critical care medicine's international consensus conference on prediction and identification of long-term impairments after critical illness. Crit Care Med 2020; 48: 1670–1679. doi: 10.1097/CCM.0000000000004586 [DOI] [PubMed] [Google Scholar]
- 2.Needham DM, Davidson J, Cohen H, et al. Improving long-term outcomes after discharge from intensive care unit: report from a stakeholders’ conference. Crit Care Med 2012; 40: 502–509. doi: 10.1097/CCM.0b013e318232da75 [DOI] [PubMed] [Google Scholar]
- 3.Preiser J-C, Herridge M, Azoulay E, eds. Post-Intensive Care Syndrome. Cham, Springer, 2020. [Google Scholar]
- 4.Bein T, Weber-Carstens S, Apfelbacher C. Long-term outcome after the acute respiratory distress syndrome: different from general critical illness? Curr Opin Crit Care 2018; 24: 35–40. doi: 10.1097/MCC.0000000000000476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gerth AMJ, Hatch RA, Young JD, et al. Changes in health-related quality of life after discharge from an intensive care unit: a systematic review. Anaesthesia 2019; 74: 100–108. doi: 10.1111/anae.14444 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ruhl AP, Huang M, Colantuoni E, et al. Health care resource use and costs in long-term survivors of ARDS: a 5-year longitudinal cohort study. Crit Care Med 2017; 45: 196–204. doi: 10.1097/CCM.0000000000002088 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Dowdy DW, Eid MP, Dennison CR, et al. Quality of life after acute respiratory distress syndrome: a meta-analysis. Intensive Care Med 2006; 32: 1115–1124. doi: 10.1007/s00134-006-0217-3 [DOI] [PubMed] [Google Scholar]
- 8.Cagino LM, Seagly KS, McSparron JI. Survivorship after critical illness and post-intensive care syndrome. Clin Chest Med 2022; 43: 551–561. doi: 10.1016/j.ccm.2022.05.009 [DOI] [PubMed] [Google Scholar]
- 9.Voiriot G, Oualha M, Pierre A, et al. Chronic critical illness and post-intensive care syndrome: from pathophysiology to clinical challenges. Ann Intensive Care 2022; 12: 58. doi: 10.1186/s13613-022-01038-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fazzini B, Battaglini D, Carenzo L, et al. Physical and psychological impairment in survivors of acute respiratory distress syndrome: a systematic review and meta-analysis. Br J Anaesth 2022; 129: 801–814. doi: 10.1016/j.bja.2022.08.013 [DOI] [PubMed] [Google Scholar]
- 11.Gariépy G, Honkaniemi H, Quesnel-Vallée A. Social support and protection from depression: systematic review of current findings in Western countries. Br J Psychiatry 2016; 209: 284–293. doi: 10.1192/bjp.bp.115.169094 [DOI] [PubMed] [Google Scholar]
- 12.Wang Y, Chung MC, Wang N, et al. Social support and posttraumatic stress disorder: a meta-analysis of longitudinal studies. Clin Psychol Rev 2021; 85: 101998. doi: 10.1016/j.cpr.2021.101998 [DOI] [PubMed] [Google Scholar]
- 13.Jain S, Hauschildt K, Scheunemann LP. Social determinants of recovery. Curr Opin Crit Care 2022; 28: 557–565. doi: 10.1097/MCC.0000000000000982 [DOI] [PubMed] [Google Scholar]
- 14.Hope AA, McPeake J. Healthcare delivery and recovery after critical illness. Curr Opin Crit Care 2022; 28: 566–571. doi: 10.1097/MCC.0000000000000984 [DOI] [PubMed] [Google Scholar]
- 15.Dodoo-Schittko F, Brandstetter S, Blecha S, et al. Determinants of quality of life and return to work following acute respiratory distress syndrome. Dtsch Arztebl Int 2017; 114: 103–109. doi: 10.3238/arztebl.2017.0103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Falvey JR, Cohen AB, O'Leary JR, et al. Association of social isolation with disability burden and 1-year mortality among older adults with critical illness. JAMA Intern Med 2021; 181: 1433–1439. doi: 10.1001/jamainternmed.2021.5022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Deja M, Denke C, Weber-Carstens S, et al. Social support during intensive care unit stay might improve mental impairment and consequently health-related quality of life in survivors of severe acute respiratory distress syndrome. Crit Care 2006; 10: R147. doi: 10.1186/cc5070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Milton A, Schandl A, Soliman IW, et al. Development of an ICU discharge instrument predicting psychological morbidity: a multinational study. Intensive Care Med 2018; 44: 2038–2047. doi: 10.1007/s00134-018-5467-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Eaton TL, Scheunemann LP, Butcher BW, et al. The prevalence of spiritual and social support needs and their association with postintensive care syndrome symptoms among critical illness survivors seen in a post-ICU follow-up clinic. Crit Care Explor 2022; 4: e0676. doi: 10.1097/CCE.0000000000000676 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tilburgs B, Nijkamp MD, Bakker EC, et al. The influence of social support on patients’ quality of life after an intensive care unit discharge: a cross-sectional survey. Intensive Crit Care Nurs 2015; 31: 336–342. doi: 10.1016/j.iccn.2015.07.002 [DOI] [PubMed] [Google Scholar]
- 21.King J, O'Neill B, Ramsay P, et al. Identifying patients’ support needs following critical illness: a scoping review of the qualitative literature. Crit Care 2019; 23: 187. doi: 10.1186/s13054-019-2441-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Haines KJ, Beesley SJ, Hopkins RO, et al. Peer support in critical care: a systematic review. Crit Care Med 2018; 46: 1522–1531. doi: 10.1097/CCM.0000000000003293 [DOI] [PubMed] [Google Scholar]
- 23.Elm EV, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61: 344–349. doi: 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 24.Apfelbacher C, Brandstetter S, Blecha S, et al. Influence of quality of intensive care on quality of life/return to work in survivors of the acute respiratory distress syndrome: prospective observational patient cohort study (DACAPO). BMC Public Health 2020; 20: 861. doi: 10.1186/s12889-020-08943-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Dodoo-Schittko F, Brandstetter S, Brandl M, et al. German-wide prospective DACAPO cohort of survivors of the acute respiratory distress syndrome (ARDS): a cohort profile. BMJ Open 2018; 8: e019342. doi: 10.1136/bmjopen-2017-019342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Brandstetter S, Dodoo-Schittko F, Blecha S, et al. Influence of quality of care and individual patient characteristics on quality of life and return to work in survivors of the acute respiratory distress syndrome: protocol for a prospective, observational, multi-centre patient cohort study (DACAPO). BMC Health Serv Res 2015; 15: 563. doi: 10.1186/s12913-015-1232-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Szymczak H, Dodoo-Schittko F, Brandstetter S, et al. Trajectories of quality of life, return to work, psychopathology, and disability in survivors of the acute respiratory distress syndrome (ARDS): a three-year prospective cohort study (DACAPO). J Crit Care 2023; 78: 154356. doi: 10.1016/j.jcrc.2023.154356 [DOI] [PubMed] [Google Scholar]
- 28.Lampert T, Kroll L, Müters S, et al. Measurement of socioeconomic status in the German health interview and examination survey for adults (DEGS1). Bundesgesundheitsblatt 2013; 56: 631–636. doi: 10.1007/s00103-012-1663-4 [DOI] [PubMed] [Google Scholar]
- 29.Miyakawa M, Magnusson Hanson LL, Theorell T, et al. Subjective social status: its determinants and association with health in the Swedish working population (the SLOSH study). Eur J Public Health 2012; 22: 593–597. doi: 10.1093/eurpub/ckr064 [DOI] [PubMed] [Google Scholar]
- 30.Hoebel J, Müters S, Kuntz B, et al. Messung des subjektiven sozialen Status in der Gesundheitsforschung mit einer deutschen Version der MacArthur Scale. Bundesgesundheitsbl 2015; 58: 749–757. doi: 10.1007/s00103-015-2166-x [DOI] [PubMed] [Google Scholar]
- 31.Euteneuer F, Süssenbach P, Schäfer SJ, et al. Subjektiver sozialer Status. MacArthur-Skalen zur Erfassung des wahrgenommenen sozialen Status im sozialen Umfeld (SSS-U) und in Deutschland (SSS-D). Verhaltenstherapie 2015; 25: 229–232. doi: 10.1159/000371558 [DOI] [Google Scholar]
- 32.Hegar R. “Subjektiver sozialer Status”: Stellenwert für die Untersuchung und Verringerung von gesundheitlicher Ungleichheit. Prävention und Gesundheitsförderung 2010; 5: 389–400. doi: 10.1007/s11553-010-0261-2 [DOI] [Google Scholar]
- 33.Fydrich T, Sommer G, Brähler E. F-SozU: Fragebogen zur sozialen Unterstüzung (F-SozU). Göttingen, Hogrefe, 2007. [Google Scholar]
- 34.Brandl M, Brandstetter S, Dodoo-Schittko F, et al. Soziale Unterstützung ist bei Überlebenden eines akuten Lungenversagens signifikant mit der psychischen Lebensqualität assoziiert: Ergebnisse der prospektiven Patientenkohortenstudie DACAPO. Gesundheitswesen 2019; 81: 745. doi: 10.1055/s-0039-1694607 [DOI] [Google Scholar]
- 35.Fydrich T, Sommer G, Tydecks, S, et al. Fragebogen zur sozialen Unterstützung (F-SozU): Normierung der Kurzform (K-14). Z Med Psychol 2009; 18: 43–48. [Google Scholar]
- 36.Ware J, Kosinski M, Keller SD. A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996; 34: 220–233. doi: 10.1097/00005650-199603000-00003 [DOI] [PubMed] [Google Scholar]
- 37.Jenkinson C, Layte R, Jenkinson D, et al. A shorter form health survey: can the SF-12 replicate results from the SF-36 in longitudinal studies? J Public Health Med 1997; 19: 179–186. doi: 10.1093/oxfordjournals.pubmed.a024606 [DOI] [PubMed] [Google Scholar]
- 38.Bullinger M, Kirchberger I. Der SF-36 Fragebogen zum Gesundheitszustand: Handbuch für die deutschsprachige Fragebogenversion. Boston, Medical Outcomes Trust, 1995. [Google Scholar]
- 39.Morfeld M, Kirchberger I, Bullinger M. SF-36 Fragebogen zum Gesundheitszustand. 2nd Edn. Göttingen, Hogrefe, 2011. [Google Scholar]
- 40.Brandstetter S, Dodoo-Schittko F, Brandl M, et al. Ambulatory and stationary healthcare use in survivors of ARDS during the first year after discharge from ICU: findings from the DACAPO cohort. Ann Intensive Care 2019; 9: 70. doi: 10.1186/s13613-019-0544-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Szymczak H, Dodoo-Schittko F, Brandstetter S, et al. Health care utilization in ARDS survivors 2–3 years after discharge. Respir Care 2023; 68: 676–679. doi: 10.4187/respcare.10478 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hernán MA. The C-Word: scientific euphemisms do not improve causal inference from observational data. Am J Public Health 2018; 108: 616–619. doi: 10.2105/AJPH.2018.304337 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Hernán MA, Hernández-Díaz S, Werler MM, et al. Causal knowledge as a prerequisite for confounding evaluation: an application to birth defects epidemiology. Am J Epidemiol 2002; 155: 176–184. doi: 10.1093/aje/155.2.176 [DOI] [PubMed] [Google Scholar]
- 44.Tennant PWG, Murray EJ, Arnold KF, et al. Use of directed acyclic graphs (DAGs) to identify confounders in applied health research: review and recommendations. Int J Epidemiol 2021; 50: 620–632. doi: 10.1093/ije/dyaa213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Schipf S, Knüppel S, Hardt J, et al. Directed Acyclic Graphs (DAGs) – Die Anwendung kausaler Graphen in der Epidemiologie. Gesundheitswesen 2011; 73: 888–892. doi: 10.1055/s-0031-1291192 [DOI] [PubMed] [Google Scholar]
- 46.Tönnies T, Kahl S, Kuss O. Collider bias in observational studies. Dtsch Arztebl Int 2022; 119: 107–122. doi: 10.3238/arztebl.m2022.0076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.VanderWeele TJ, Robins JM. Directed acyclic graphs, sufficient causes, and the properties of conditioning on a common effect. Am J Epidemiol 2007; 166: 1096–1104. doi: 10.1093/aje/kwm179 [DOI] [PubMed] [Google Scholar]
- 48.Rohrer JM. Thinking clearly about correlations and causation: graphical causal models for observational data. Adv Meth Pract Psychol Sci 2018; 1: 27–42. doi: 10.1177/2515245917745629 [DOI] [Google Scholar]
- 49.Wilson IB, Cleary SM. Linking clinical variables with health-related quality of life. A conceptual model of patient outcomes. JAMA 1995; 273: 59–65. doi: 10.1001/jama.1995.03520250075037 [DOI] [PubMed] [Google Scholar]
- 50.Ferrans CE, Zerwic JJ, Wilbur JE, et al. Conceptual model of health-related quality of life. J Nurs Scholarsh 2005; 37: 336–342. doi: 10.1111/j.1547-5069.2005.00058.x [DOI] [PubMed] [Google Scholar]
- 51.Bakas T, McLennon SM, Carpenter JS, et al. Systematic review of health-related quality of life models. Health Qual Life Outcomes 2012; 10: 134. doi: 10.1186/1477-7525-10-134 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav 1995; 1: 1–10. doi: 10.2307/2137284 [DOI] [PubMed] [Google Scholar]
- 53.Babitsch B, Gohl D, Lengerke TV. Re-revisiting Andersen's behavioral model of health services use: a systematic review of studies from 1998–2011. GMS Psychosocial Medicine 2012; 9: Doc11. doi: 10.3205/psm000089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Andersen RM, Davidson PL. Improving access to care in America: individual and contextual indicators. In: Andersen RM, Rice TH, Kominski GF, eds. Changing the US Health Care System: Key issues in Health Services Policy and Management. 4th Edn. San Francisco, CA, Jossey-Bass, 2014; pp. 3–31. [Google Scholar]
- 55.Lewis-Beck C, Lewis-Beck M. Applied Regression: An Introduction. 2nd Edn. London, UK, Sage Publications, 2015. [Google Scholar]
- 56.Field A. Discovering Statistics Using IBM SPSS. 5th Edn. London, UK, Sage Publications, 2018. [Google Scholar]
- 57.Le Gall J-R, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270: 2957–2963. doi: 10.1001/jama.1993.03510240069035 [DOI] [PubMed] [Google Scholar]
- 58.Vincent J-L, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure, On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22: 707–710. doi: 10.1007/BF01709751 [DOI] [PubMed] [Google Scholar]
- 59.Force ADT, Ranieri VM, Rubenfeld GD, et al. Acute respiratory distress syndrome: the Berlin definition. JAMA 2012; 307: 2526–2533. doi: 10.1001/jama.2012.5669 [DOI] [PubMed] [Google Scholar]
- 60.Cohen J. A power primer. Psychol Bull 1992; 112: 155–159. doi: 10.1037/0033-2909.112.1.155 [DOI] [PubMed] [Google Scholar]
- 61.Langerud AK, Rustøen T, Småstuen MC, et al. Health-related quality of life in intensive care survivors: associations with social support, comorbidity, and pain interference. PLoS One 2018; 13: e0199656. doi: 10.1371/journal.pone.0199656 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Orwelius L, Fredriksson M, Bäckman C, et al. Social support: a significant factor for health-related quality of life after critical illness? Crit Care 2010; 14: P434. doi: 10.1186/cc8666 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63.Pfoh ER, Wozniak AW, Colantuoni E, et al. Physical declines occurring after hospital discharge in ARDS survivors: a 5-year longitudinal study. Intensive Care Med 2016; 42: 1557–1566. doi: 10.1007/s00134-016-4530-1 [DOI] [PubMed] [Google Scholar]
- 64.Taylor SE. Social support: a review. In: Friedman HS, ed. The Oxford Handbook of Health Psychology. Oxford, Oxford University Press, 2011; pp. 189–214. [Google Scholar]
- 65.Czerwonka AI, Herridge MS, Chan L, et al. Changing support needs of survivors of complex critical illness and their family caregivers across the care continuum: a qualitative pilot study of Towards RECOVER. J Crit Care 2015; 30: 242–249. doi: 10.1016/j.jcrc.2014.10.017 [DOI] [PubMed] [Google Scholar]
- 66.Cameron JI, Gignac MAM. “Timing it right”: a conceptual framework for addressing the support needs of family caregivers to stroke survivors from the hospital to the home. Patient Educ Couns 2008; 70: 305–314. doi: 10.1016/j.pec.2007.10.020 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material
01113-2024.SUPPLEMENT
Data Availability Statement
The datasets used and analysed during the current study are available from C. Apfelbacher on reasonable request.