An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Christopher E. Cox, MD, MPH, Catherine L. Hough, MD, MSci, Shannon S. Carson, MD, Douglas B. White, MD, MAS, Jeremy M. Kahn, MD, MSci, Maren K. Olsen, PhD, Derek M. Jones, MDiv, Tamara J. Somers, PhD, Sarah A. Kelleher, PhD, and Laura S. Porter, PhD.
Author Information and AffiliationsMany survivors of critical illness and their family members experience significant psychological distress after hospital discharge.
To determine if assigning participants to a postdischarge coping-skills training (CST) program reduces patient and family psychological distress compared with assigning participants to an educational program.
A multicenter randomized clinical trial conducted between December 2013 and February 2016.
Five academic and community hospitals.
Adult patients (≥18 years) who received mechanical ventilation >48 hours were randomized individually after discharge; 1 family member per patient was also offered a chance to participate. Ten patient and family member stakeholders participated as advisors in the planning and conduct of this study.
Participants assigned to CST were offered six 30-minute weekly telephone calls with a psychologist. CST was designed to teach coping strategies for managing distress related to critical illness. CST participants also received access to a complementary study website. An education program comparator included 2 telephone calls to review printed and web-based information about critical illness.
The primary outcome was the Hospital Anxiety and Depression Scale (HADS) summary score assessed among patients 3 months postrandomization. Secondary outcomes included HADS depression and anxiety subscale scores and posttraumatic stress disorder symptoms measured using the Impact of Events Scale-Revised scale assessed at 3 and 6 months.
Among 175 patients randomized to CST (n = 86) or education (n = 89), 131 (75%) completed the 6-month follow-up. Of the patients, 44 (25%) did not complete the study, generally because of death or medical illness. At 3 months, there were no significant differences between the CST and education groups in HADS summary score improvement (difference 1.3 [95% CI, −0.9 to 3.4], P = .24). Secondary outcomes for patients and family members were similar by treatment group at 3 and 6 months. A priori exploratory secondary analyses revealed that among patients with high baseline distress (HADS summary score >14), those randomized to CST had improved HADS summary scores (difference −4.6, [95% CI, −8.6 to −0.6], P = .02) compared with education participants at 6 months.
Nearly 40% more participants than planned to offset higher-than-expected attrition observed after consent but before randomization. Additionally, the generalizability of our results may be limited because the cohort was drawn largely from academic institutions.
Among critical illness survivors, CST did not reduce symptoms of psychological distress compared with an education program. However, among patients with high baseline distress, CST may have improved symptoms of distress at 6 months. Future efforts to reduce psychological distress among critical illness survivors will require greater patient engagement and should specifically target high-risk populations.
Does a postdischarge coping-skills training (CST) intervention delivered by telephone and online reduce symptoms of depression, anxiety, and posttraumatic stress disorder compared with a similarly delivered critical illness education program?
In this randomized clinical trial that included 175 patients and 86 family members, CST did not reduce psychological distress compared with education overall. Exploratory analyses suggest that symptoms of psychological distress may have been reduced among patients with high levels of baseline distress in the CST group, but not the education group.
CST did not reduce symptoms of psychological distress among a general population of intensive care unit survivors. The CST intervention requires further study among patients with high baseline distress and should include features that enhance participant engagement to aid in adherence.
Acute respiratory failure requiring mechanical ventilation in an intensive care unit (ICU) affects more than a million persons in the United States alone each year.1 In addition to the physical and cognitive disability associated with critical illness that has been the focus of much research,2,3 ICU survivors and their family members suffer from symptoms of depression, anxiety, and posttraumatic stress disorder (PTSD) that may persist for more than 1 year at 10-fold higher rates than in the general population.4-10
However, few effective postdischarge interventions address ICU survivors' psychological distress. More broadly, patients commonly experience barriers accessing traditional postdischarge psychological interventions because of acquired disability, distance from specialized centers, and financial distress.11,12
Coping is an active response for managing the demands of stressful events such as the experience of critical illness.13,14 While the use of adaptive coping skills (eg, relaxation, positive reframing, problem solving) is associated with increased self-efficacy and decreased psychological distress, ICU patients and their family members report infrequent use of such strategies.15,16 CST interventions based on cognitive-behavioral principles have been effective in helping patients with cancer and other medical conditions manage a variety of symptoms and types of psychological distress.17,18 To address ICU survivors' distress as well as the inherent difficulty of accessing personalized care during critical illness recovery, we developed a telephone-based CST program. In a pilot evaluation, ICU patients and their family members acquired adaptive coping skills and experienced improvement in symptoms of psychological distress.19
To further evaluate the effects of the CST intervention augmented with web content, we conducted a multicenter randomized trial testing the hypothesis that a postdischarge CST program compared with a standardized education program would result in fewer symptoms of depression, anxiety, and PTSD.
The institutional review boards of the participating sites (Duke, University of Washington, University of North Carolina at Chapel Hill, University of Pittsburgh) approved the study protocol (Online Appendix). A 10-member panel of layperson research stakeholders from diverse backgrounds participated through the entirety of the project to ensure a patient-centered approach was applied to the conception of the key research question, the development of the protocol, the general approach to inclusion and exclusions, and the choice of outcomes (eMethods, pp. 2-3). An independent data safety monitoring board approved the protocol and reviewed performance, data quality, and safety at 6-month intervals.
We attempted to enroll patient–informal caregiver dyads for participation in this trial. Between December 2013 and April 2015, clinical research coordinators identified all patients meeting inclusion criteria through daily electronic health record–based screening of consecutive medical and surgical ICU patients. The coordinator obtained permission from the ICU attending physician to approach the patient and/or his or her legal representative for written informed consent at the time of transfer from the ICU to the ward.
Patient inclusion criteria were age ≥18, receipt of mechanical ventilation for >48 consecutive hours, and successful extubation before hospital discharge. For each patient, the unit of subsequent randomization, we attempted to enroll for intervention participation 1 family member aged ≥18 years who expected to provide significant postdischarge assistance.20 Patients without an available family member were still eligible. Patient exclusions included preexisting or current cognitive impairment (defined by chart review for evidence of dementia or 3 or more errors on a brief 6-item in-person cognitive screen), treatment for severe mental illness during the 6 months preceding admission, residence at a location other than home immediately before admission, poor English fluency, ICU attending expectation of survival < 3 months, inability to complete study procedures as determined by study staff, and failure to return home within 3 months after discharge. Exclusions for family members included history of cognitive impairment and poor English fluency determined by need for a translator to participate in medical decision making.
Randomization to CST or education program groups was performed by the electronic data system after completion of the first postdischarge interview, which was completed within 2 weeks of patient arrival home from either the hospital or a postacute care facility. A password- protected computerized system allocated patients at a 1:1 treatment group ratio with blocks of 4, stratifying by 3 factors to ensure balance: severity of baseline patient psychological distress (Hospital Anxiety and Depression Scale [HADS] summary score [≤14 vs >14]), days of mechanical ventilation (≤7 vs >7), and study site. Printed material was mailed to each participant that included instructions for accessing a password-protected, treatment group–specific website hosting videos and other content. The interventions (described in detail on eMethods pp. 4-6) began 1 week after randomization.
One of 2 clinical psychologists (T.J.S., S.A.K.) was randomly assigned to each patient. The CST intervention was delivered in 6 weekly telephone sessions, each lasting about 30 minutes, that addressed (1) relaxation exercises, (2) progressive muscle relaxation, (3) pleasant activities and the activity–rest cycle, (4) communication, (5) cognitive restructuring and pleasant imagery, and (6) review and planning for sustainability.19 Psychologists taught each skill by providing a description and rationale for its use, leading participants through practice with feedback in the context of any self-reported ongoing stressors, helping participants plan how to apply the skill in real life, and highlighting relevant web-based content. In addition to the patient participants learning the skills themselves, family members were encouraged to coach the patients in applying the skills and using the web content on a day-to-day basis.
As a comparator and control condition, we designed an education program to address poor comprehension of critical illness, a common finding that is associated with distress and that has been responsive to past interventions.21 This program consisted of 6 informational videos with accompanying web-based content. Study staff with content expertise in critical illness conducted two 30-minute calls with participants during the 6-week intervention period to review materials and answer related questions.
Trained coordinators blinded to treatment group collected clinical data. Participants completed study surveys by telephone with treatment group–blinded Duke coordinators or by a password-protected electronic patient-reported outcomes system at 3 time periods: within 2 weeks after arrival home from either the hospital or a postacute care facility (preintervention), 3 months postrandomization (immediate postintervention), and 6 months postrandomization.22 Interviews were delayed for at least 4 weeks for family members of patients known to have died during follow-up. Study staff used scheduled phone calls, postcards, letters, and emails to prompt completion of intervention sessions and surveys.
Participants were compensated $10 for each of 3 interviews completed. Follow-up was completed in February 2016.
The primary outcome measure was the Hospital Anxiety and Depression Scale (HADS) summary score assessed at 3 months after randomization.23 HADS summary scores can range from 0 (best) to 42 (worst). The minimal clinically important difference in HADS score is 1.5 to 2 units.24,25
Secondary outcomes included the 7-item HADS anxiety and depression subscale scores, which can range from 0 to 21. A HADS subscale score >7 represents significant distress. Other secondary outcomes included the Impact of Events Scale-Revised (IES-R) PTSD symptom score, a 22-item scale with scores ranging from 0 (best) to 88 (worst)26; quality of life as measured with the EuroQOL-5 visual analog scale (range 0 [worst] to 100 [best])27; global mental and physical health status assessed with the PROMIS Global Short Form (range of each 4 [worst] to 20 [best])28; adaptive coping behaviors (active coping, use of emotional support, positive reframing, planning, and acceptance) measured by the Brief Coping Orientation to Problems Experienced (COPE) (range 10 [worst] to 40 [best])29; and self-efficacy using a 4-item scale adapted from Lorig and colleagues (item range 1 [worst] to 10 [best]).30 At each scheduled survey, participants could report up to 3 stressors in their own words, rating each stressor's severity on a scale of 0 (not at all severe) to 10 (extremely severe), as well as record the number of weeks spent at home (ie, not hospitalized) throughout follow-up, an outcome suggested by the stakeholder group.
Based on previous studies,19,31 190 (95 in each treatment group) patients were needed to detect a HADS summary score difference of 4 units at 3 months postrandomization with 90% power and a type I error rate of 5%, assuming a baseline SD of 8.7, a correlation between baseline and 3 months of 0.4, and a 3-month dropout rate of 10%.32
We conducted the primary analysis of the HADS summary score and the secondary survey outcomes using a linear mixed-effects model with an unstructured covariance (using PROC MIXED in SAS [computer program]. Version 9.4. Cary, NC). Model parameters included a common intercept (which constrains the baseline means to be equal), stratification variables (baseline HADS summary score stratum [centered], duration of mechanical ventilation stratum [centered], and site [centered]), indicator variables for months 3 and 6, and a CST indicator interacted with months 3 and 6. We used contrast statements to estimate the difference between the groups from baseline to each of the follow-ups. We tested the primary hypothesis by examining the 3-month estimated group difference in mean HADS summary score improvement, 95% CI, and P value.
We originally chose the stratification variables of baseline HADS summary score (≤14 or >14) and ventilator days (≤7 or >7) because of their suspected association with the outcome variables.
Therefore, we also examined the intervention group effect on primary and secondary outcomes within each of these strata using a general linear model with unstructured covariance. Model parameters included separate intercepts for the low and high levels of each strata; indicator variables for months 3 and 6; indicators for the CST intervention interacted with months 3 and 6; indicators of the strata interacted with months 3 and 6; and the 3-way interaction of the strata indicator variable, CST intervention, and months 3 and 6.
We used all available data, including data from patients who were lost to follow-up or died during the study, in all analyses. We conducted primary and secondary survey outcome analyses with full likelihood methods, which are valid under a missing data framework in which the missing values may depend on intervention group, site, initial HADS summary score stratum, duration of ventilation stratum, and any observed outcome values.33,34 We also conducted a sensitivity analysis in which additional predictors of missing values were explored and included in a multiple imputation procedure.35 We examined all baseline patient demographic and clinical characteristics presented in Tables 1 and 2 for potential inclusion in the imputation model. We examined these variables descriptively and via chi-square and t tests by dropout yes/no and intervention complier yes/no. We included in the imputation model variables with P < .20, which included randomization group, site, marital status, medical ICU, education, discharged home, caregiver enrolled, length of ICU stay, HADS summary, length of mechanical ventilation, and financial distress. Additionally, we included all baseline variables for all outcomes in the imputation model.
Baseline Characteristics of Patients and Their Family Members.
Model Estimated Primary and Secondary Outcomes, Patients With Baseline HADS >14 (n = 60).
We generated 10 multiple imputations for the missing 3- and 6-month outcomes via the MCMC option in PROC MI. We assessed model convergence via EM iterations to convergence as well as ACF and time series plots of the mean week 56 weight. We drew imputations every 100 iterations. We analyzed the 10 imputed data sets with a general linear model including a common intercept, stratification variables (baseline HADS summary score stratum [centered], duration of mechanical ventilation stratum [centered], and site [centered]), indicator variables for months 3 and 6, and indicators for the CST intervention interacted with each of the follow-up time point indicator variables. We combined the 10 sets of model estimates and standard errors using PROC MIANALZE.
Estimated mean differences, 95% confidence intervals, and P values are reported for all comparisons, with P < .05 considered statistically significant. We identified primary and secondary outcomes a priori, and we made no adjustments for multiple comparisons. We used SAS version 9.4 for analyses. Further details, including the statistical code, are provided in the eMethods (pp. 7-14).
Of 411 potentially eligible patients, 277 proved informed consent, and 175 were randomized to CST (n = 86) or education (n = 89) (Figure 1). Of these, 86 (49%) had an eligible family member who was enrolled. Treatment groups appeared balanced in baseline characteristics (Table 1, eTable 1). Patients on average were middle-aged and severely ill, with an expected hospital mortality of about 50% based on the mean Acute Physiologic Assessment and Chronic Health Evaluation (APACHE II) score. Family members were primarily spouses/partners (n = 41, 48%) or parents (n = 12, 14%). Approximately a third of patients had been treated for a psychiatric condition in the 3 months preceding admission.
Flow of Patients and Family Members.
Patients experienced relatively long hospitalizations (mean ~25 days) (eTable 2). Of the patients, 52 (30%) were discharged to a postacute care facility before eventually returning home. Readmission occurred among 43 (25%) of patients during follow-up.
A total of 54 (63%) patients in the CST group and 58 (65%) in the education group were adherent to intervention procedures, defined as completion of 1 or more calls (CST group) or self-reported viewing of intervention material (education group) (Figure 2). Missed CST sessions were mostly related to medical illness (n = 24, 42%), loss to contact (n = 9, 16%), or death (n = 5, 9%). Only 5 (6%) CST patients who were physically and cognitively able completed no sessions.
Post-randomization interviews were completed by 136 (81%) and 131 (80%) of patients alive at 3 and 6 months, respectively. Similarly, 69 (80%) and 66 (77%) interviews were completed by family members at 3 and 6 months, respectively. The reasons for patient dropout at 6 months included loss to contact (n = 18, 10%), serious illness (n = 14, 8%), and death (n = 12, 7%) (eFigures 2, 3).
Results of the primary analysis revealed no significant difference in mean HADS summary score between the CST and education patient groups at either 3 months (16.6 and 15.3; between-group difference in change 1.3 [95% CI, −0.9 to 3.4], P = .24) or 6 months (15.6 and 15.9; between-group difference in change −0.3 [95% CI, −2.7 to 2.0], P = .78; see Table 2; eTable 3 demonstrates similar results from the multiple imputation analysis). Similarly, no significant between-group differences were present in HADS anxiety and depression subscales. IES-R scores were similar for CST patients and education patients at 3 months (31.0 vs 27.9; between-group difference in change 3.1 [95% CI, −1.9 to 8.1], P = .22) and 6 months (29.4 vs 25.8; between-group difference in change 3.6 [95% CI, −2.7 to 10.0], P = .26). Additionally, there were no statistically significant differences in physical or mental health status, quality of life, self- efficacy, and coping (Table 2) or days at home during follow-up (eTable 4). We observed no significant between-group differences among family members for primary and secondary outcomes at 3 and 6 months (eTable 5). Patients' self-reported stressors persisted largely unchanged throughout follow-up in severity and type, with physical disability and financial distress mentioned frequently (eTable 6, eFigure 4).
There were significant group-based differences in outcomes by the prespecified stratification variables in exploratory analyses. Among patients with high distress at baseline (HADS summary score >14; n = 60, 34%), CST recipients experienced significant improvement in 6-month, though not 3-month, outcomes including HADS summary scores (15.5 vs 20.2, between-group difference −4.6 [95% CI, −8.6 to −0.6], P = .02), HADS depression scores (6.6 vs 9.6, between-group difference −3.0 [95% CI, –{value missing} to −0.7], P = .01), global mental health (12.0 vs 9.7, between-group difference 2.3 [95% CI, 0.05-4.1], P = .01), and quality of life (59.5 vs 48.3, between-group difference 11.2 [95% CI, 0.02-22.4], P = .0496; Table 3). Proposed mechanistic metrics were improved as well, including self-efficacy (5.9 vs 4.5; between-group difference = 1.4 [95% CI, 0.2-2.5], P = .02).
Model Estimated Primary and Secondary Outcomes, Patients Ventilated >7 Days (n = 47).
In contrast, among those with >7 days of ventilation (n = 47, 27%), education program patients had improvements at 3 months in mean HADS summary score (12.5 vs 8.5, between-group difference 4.0 [95% CI, 0.05-8.1] P = .047), HADS depression score (6.6 vs 4.0, between-group difference 2.6 [95% CI, 0.2-5.1], P = .03), global mental health (12.3 vs 14.5, between-group difference −2.2 [95% CI, −4.2 to −0.2], P = .04), and quality of life (62.1 vs 77.9, between-group difference −15.8 [95% CI, −28.5 to −3.1], P = .02; Table 4). However, these differences were not sustained at 6 months.
Outcomes Analyzed by Days of Mechanical Ventilation Stratification Variable: Ventilation >7 Days.
In a large multicenter clinical trial of ICU survivors and their family members, we found that a telephone- and web-based CST program initiated after discharge did not reduce symptoms of depression, anxiety, and PTSD at 3 or 6 months among either patients or family members compared with an education program control. A priori exploratory subgroup analyses, however, revealed that among patients with high levels of baseline distress, CST reduced symptoms of psychological distress and improved quality of life at 6 months compared with the education control at a magnitude of clinical importance.
There are a number of potential explanations for lack of an overall effect among a general population of ICU survivors. First, the intervention's impact was likely attenuated by the decision to include a broadly representative sample of patients without consideration of their degree of functional debilitation or distress at the time of randomization in order to increase the generalizability of the results. For example, patients were allowed up to 3 months to return home after hospital discharge to allow participation by those discharged to postacute care facilities, a subgroup known to have high levels of distress.36 However, these patients were more likely to drop out of the study due to unresolved illness. Also, similar to recent approaches among ICU survivor–targeted interventions,37 we opted not to exclude patients with low baseline distress because we hypothesized that the intervention might prevent an increase in distress over time as the stresses of chronic critical illness and caregiving burden accumulated.38 Yet participants' baseline distress on average was less than observed in a previous pilot,19 thus limiting the intervention's potential for effect. Second, patients' illnesses were severe, persistent, and disruptive throughout follow-up, leading to a lower-than-planned intervention dose despite 6620 contact attempts (mean 27.6 [SD 12.1] per participant) by study staff (eTable 9).
Other interventions for ICU survivors' distress have been evaluated in both inpatient (eg, ICU diaries,39 in-ICU psychologist intervention40) and outpatient (eg, self-help rehabilitation manuals,41 ICU survivor clinics,42 case management programs37) settings, though these have demonstrated little effect on depression and anxiety symptoms.43,44 But despite CST's similar lack of a primary effect, this study provides valuable insights that can improve future efforts to alleviate psychological distress among those who experience a serious illness or who may be unable to access traditional face-to-face therapy.
Our findings suggest that future approaches to addressing distress among ICU survivors should possibly focus on improving the identification of ICU survivors with high levels of distress. However, illness severity and delirium limit the accuracy of in-hospital distress screening among patients, while family members' distress is not clearly associated with patient demographics or clinical characteristics.38 In addition to developing prediction models for functional and psychological phenotyping,45 screening for distress soon after discharge using web- or SMS text-based surveys may better capture patients with high levels of distress.16,22
Novel strategies are also required to increase participant retention, a chronic and complex problem encountered in all long-term follow-up studies involving critically ill patients. As expected, nearly two-thirds of missed intervention sessions were among those whose illness persisted or worsened.
Surprisingly, however, adherence was lower among participants with shorter durations of life support, perhaps because they recovered sufficiently to resume bandwidth-limiting duties at work and home (eTable 10). Past work has shown that acute critical illness may make a greater global impact in comparison with that related to a worsening of a chronic condition.46 While addressing all of these factors will be challenging, improving intervention usability could significantly increase retention through better engagement with patients during the dynamic processes of recovery and social reintegration.47
Study limitations are notable. We needed to enroll nearly 40% more participants than planned to offset higher-than-expected attrition observed after consent but before randomization. While the effect on the observed results is unknown, our ability to characterize those who dropped out will inform future efforts to conduct more efficient ICU-based trials. Because the cohort was drawn largely from academic institutions, our results may not reflect the experience of those treated in other settings. However, the cohort was geographically diverse and the investigators included a group of patient and family member research stakeholders to ensure a patient- and family-centered perspective.20 Last, recent research has highlighted novel adaptive coping styles in the context of ICU survivorship that, while not included in our intervention, could better personalize future iterations.16,48
Among a general population of ICU survivors and their family members, a telephone- and web-based coping-skills training program did not reduce symptoms of depression, anxiety, and PTSD compared with an education program. However, exploratory analyses suggest that among patients with high baseline levels of distress, CST might have reduced symptoms of psychological distress in comparison with control at 6 months. Future efforts to reduce psychological distress in this population will require greater patient engagement and likely target patients with elevated levels of baseline distress.
Thanks to the patients and family members who participated, the patient and family research stakeholders (Emily Bass, Ron Bass, Lisa Boyd, Lynelle Fox, Joyce Odom (deceased), Darius Odom, Frieda Waisner, Harold Waisner, Rhonda Zack, and Mark Zack), the clinical research coordinator team (Khalida Arif, Summer Choudhury, Colin Johnston, Joyce Lanier, Mary Key, Melissa McCullough, Stella Ogake, Megan Potter, Wen Regan, Anne-Marie Shields, Anna Ungar, and Brenda Walton), the Data Coordinating Center team at the University of North Carolina- Chapel Hill (Roger Akers, Brian Cass, Mattias Jonsson, and Maria Tobin), the Data Safety Monitoring Board (Todd Rice, Tim Ahles, and Vernon Chinchilli), Verena Knowles, and Francis J. Keefe.
Study design and concept: C.E.C., L.S.P., C.L.H., S.S.C., J.M.K., D.B.W., M.K.O.
Writing and review: C.E.C., M.K.O., D.M.J., J.M.K., L.S.P., C.L.H., S.S.C., D.B.W., T.J.S., S.A.K.
Analyses: M.K.O., C.E.C., D.M.J.
Obtaining data: D.M.J., C.E.C., C.L.H., S.S.C., D.B.W., T.J.S., S.A.K.
Acquisition of funding: C.E.C.
Final approval: C.E.C., M.K.O., D.M.J., L.S.P., C.L.H., S.S.C., D.B.W., J.M.K., T.J.S., S.A.K.
Accountable for all aspects of work: C.E.C.
Grant support: This research was made possible by contract number 195 from the Patient-Centered Outcomes Research Institute (PCORI). The content is solely the responsibility of the authors and does not necessarily represent the official views of PCORI. Dr Olsen was also supported by the Center of Innovation for Health Services Research in Primary Care (CIN 13-410) at the Durham VA Medical Center.
Research reported in this report was funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (# 195). Further information available at: https://www.pcori.org/research-results/2012/can-coping-skills-training-help-patients-who-have-received-intensive-hospital
Online Appendix (PDF, 2.8M)
eMethods (PDF, 1.2M)
eResults (PDF, 1.8M)
eTable 4. Weeks not at home analysis (PDF, 110K)
eTable 6. Self-rated stressors, patients (PDF, 84K)
eTable 9. Study staff contact attempts (PDF, 70K)
eTable 13. Outcomes by HADS ≤14, patients (PDF, 101K)
eFigure 1. Timeline of study events for patients (PDF, 216K)
eReferences for Supplement (PDF, 68K)
Cox CE, Hough CL, Carson SS, et al. (2018). Can Coping-Skills Training Help Patients Who Have Received Intensive Hospital Care to Cope with Depression and Anxiety? Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/9.2018.CER.195
The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee.
This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/
Your browsing activity is empty.
Activity recording is turned off.
See more...
