Psychological Support Based on Positive Suggestions (PSBPS) on Mental Health Morbidity and Cognitive Function

Background:

Psychological morbidity is a major health problem among survivors of critical illness, with no effective intervention available to mitigate its burden. Advances in intensive care have increased survival to hospital discharge, but resolution of critical illness does not end a patient's struggles. New or worsening physical, cognitive, or mental health impairments may persist for years after an ICU stay and are collectively termed post-intensive care syndrome (PICS). These impairments are associated with rehospitalization, increased health care costs, reduced quality of life, and inability to return to work. Although physical rehabilitation, delirium prevention, and cognitive outcomes have received substantial attention, prevention of mental health morbidity remains largely unaddressed.

More than one-third of critical illness survivors report symptoms of anxiety, depression, or post-traumatic stress disorder (PTSD) 6 to 12 months after discharge. Rates are even higher among survivors of acute respiratory failure, with up to two-thirds experiencing impairment in at least one psychiatric domain, often in all three, for years. No evidence-based intervention is available to prevent or effectively treat this morbidity. Most recognized risk factors, including female sex, alcohol misuse, unemployment, and premorbid psychopathology, are not modifiable. Among modifiable factors, frightening and delusional memories from the ICU appear to be especially important. Anxiety, depression, and PTSD can be conceptualized partly as disorders of memory content in which the emotional representation of a traumatic event persists beyond its adaptive purpose. Once established, these emotionally charged memories are difficult to treat.

Psychological support may therefore need to begin early in critical illness, before traumatic memories become consolidated. Conventional psychological evaluation and treatment are difficult during an ICU stay, and little is known about psychological dysfunction in this setting. During critical illness, thinking may become literal and unusually sensitive to both direct and implied meanings. Patients may accept suggestions without critical appraisal and use communicated messages to interpret their circumstances.

To address this heightened suggestibility, Dr. Varga and colleagues in Hungary adapted therapeutic suggestion to the ICU setting as psychological support based on positive suggestions (PSBPS). Therapeutic suggestion has been used for more than a century in anesthesia, surgery, palliative care, radiology, chronic pain, somatization disorders, obstetrics, and pediatrics. Critically ill patients may experience panic, loss of control, lack of privacy, painful procedures, and dependence on complex technology. They may perceive staff performing painful interventions as adversaries and may develop learned helplessness because they cannot control basic physiological needs. An effective ICU intervention should therefore include the patient as a valued member of the care team who can assist in recovery and participate in rehabilitation.

Investigators developed specific suggestions that can be used verbatim with mechanically ventilated patients. PSBPS has been associated with earlier ventilator weaning and reduced sedative and analgesic requirements. Subsequent analyses suggested that early positive reframing was the most important component. This is consistent with research showing that fear conditioning may be altered during the initial memory-consolidation window. Mitigating information introduced after memory formation and recall, during the reconsolidation window, may also modify the emotional experience of the memory. Thus, early positive reframing may weaken or prevent traumatic emotional memories by shaping their consolidation and reconsolidation while memories remain labile. For critically ill patients, this approach requires psychological support as soon as feasible after admission and in parallel with medical care.

PSBPS can be delivered regardless of a patient's level of participation and may begin during sedation. Semantic processing can continue during altered states of consciousness, and some patients later recall statements made while they were comatose. These preliminary findings support further evaluation of PSBPS as an early preventive intervention for psychological morbidity after critical illness.

Methods:

Design. This is a two-arm, single-center, prospective, randomized controlled trial comparing PSBPS with no PSBPS in critically ill adults. Trained ICU doulas will administer PSBPS to eligible patients.

Eligible patients will be randomized at the earliest feasible time after meeting inclusion criteria. Randomization will be 1:1 to the intervention or non-intervention arm and stratified by ICU location. Both groups will receive standard medical care at the discretion of the ICU team. The intervention group will receive PSBPS as soon as feasible and daily thereafter. ICU doulas will also liaise with the patient's family, clinical team, and nursing staff.

After ICU discharge and again at 6 months by telephone, participants in both groups will complete questionnaires assessing anxiety, depression, acute stress, and cognitive function. Questionnaire results will not be entered into the medical record. The intervention will occur once daily, and all participants assigned to PSBPS will receive the same intervention. Session duration will depend on the clinical situation. In prior experience, bedside sessions sometimes lasted more than 20 minutes, whereas procedures, tests, or other clinical needs occasionally required shorter sessions.

Retention strategy. Questionnaire length and frequency have been minimized to reduce burden. Additional retention strategies include emphasizing the study's potential benefits, engaging participants as partners with the research team, and using the organized and persistent follow-up processes of the anesthesia clinical research unit. All available telephone numbers will be collected to facilitate follow-up. Participants will receive financial compensation after completing the 6-month questionnaires.

Sample Size and Power

A 2-point difference in the Hospital Anxiety and Depression Scale anxiety subscale (HADS-A) was considered clinically meaningful based on prior studies. In our preliminary study, the mean baseline HADS-A score after ICU discharge was 7.7 (SD 3.9). Assuming a pooled standard deviation of 4.0, 64 participants per arm would provide 80% power to detect a 2-point difference using a two-sided two-sample t-test with α=0.05. To account for anticipated attrition, enrollment was planned until 182 participants (91 per arm) had been randomized and completed the baseline post-ICU discharge HADS-A assessment.

This sample size was expected to provide adequate power to evaluate the prespecified psychological and cognitive outcome measures assessed after ICU discharge and at 6-month follow-up, including symptoms of anxiety, depression, post-traumatic stress, cognitive function, and health-related quality of life. The study was also intended to provide data to inform the design of a future multicenter trial.

Analysis Plan

Demographic characteristics and in-hospital clinical characteristics abstracted from the electronic health record were summarized using descriptive statistics (means, medians, standard deviations, interquartile ranges, counts, and percentages, as appropriate). These variables were used to characterize the study population and, where appropriate, as covariates in adjusted analyses; they were not prespecified efficacy outcome measures.

The effect of the intervention on the prespecified psychological and cognitive outcome measures was estimated using analysis of covariance (ANCOVA) with adjustment for relevant prerandomization characteristics identified in earlier studies. A nonparametric Wilcoxon rank-sum test was planned when appropriate. Interactions with relevant variables, including sex and race/ethnicity, were explored to assess whether intervention effects differed across key subgroups, although the study was not powered for these comparisons.

Missing questionnaire data were anticipated primarily because of inability to complete assessments, death, or loss to follow-up. The primary analysis followed the intent-to-treat principle, with complete-case analyses supplemented by sensitivity analyses using multiple imputation and worst-case imputation where appropriate. Statistical significance was defined as a two-sided p<0.05.