This study has been reported in accordance with the CONSORT statement [18, 19] that offers guidance for the transparent reporting of randomized controlled studies.
This was a pilot study because we were the first group to study the feasibility, adherence, patient satisfaction and effectiveness of an online multimedia psychoeducational intervention aimed at promoting patients’ resilience in the context of contagious diseases, such as the novel coronavirus disease. We chose to use a cluster randomized parallel-controlled trial with hospital ward as the unit of randomization (rather than individual patient). The primary reason for selection of the cluster randomized design was to protect against the ‘contamination’ that could occur in individually randomized trials [18, 20]. The risk of contamination was minimized by the fact that the hospitalized patients in the intervention and control wards (clusters) were not in contact with each other. Ethical approval was obtained from the local Ethics Committee of Shiraz University of Medical Sciences (IR.SUMS.REC.1399.011). The study was also registered in the Iranian Registry of Clinical Trials (IRCT20201001048893N1).
Participants and settings
The sample included 50 hospitalized patients with COVID-19 recruited from 4 hospital wards at 2 hospitals (two wards in each hospital) in Shiraz, Iran.
As suggested by Campbell et al. , the primary eligibility criterion in a cluster trial is often all clusters in a defined geographical area. Accordingly, the present study incorporated the two hospitals that admitted patients with COVID-19 in Shiraz in April 2020. Patients were included in the study, if they met the following criteria: age over 18 years, diagnosed with mild (nonpneumonia/ mild pneumonia) or severe (dyspnea, respiratory frequency> 30/min, blood oxygen saturation≤ 93%) COVID-19 , willingness to take part in the study, being literate, having been hospitalized during the past 48 hours, having internet access and having the ability to work with the media. The following exclusion criteria were applied: diagnosed with critical (respiratory failure, septic shock, and/ or multiple organ dysfunction or failure) COVID-19 , having a previous experience of quarantine, being unwilling or unable to continue contributing to the study, having a history of psychiatric disorders or taking psychiatric medications, and death or transfer to the ICU.
Patient recruitment was conducted between 1-30 April 2020. Two psychologists (not involved in the study) at the study hospitals informed the patients about the purpose of the study and asked them whether they were willing to be contacted by the study staff to undergo screening. The study coordinator contacted the interested patients by telephone and screened the eligible patients until the target number of 50 patients was reached (11-14 patients in each hospital ward). The eligible participants were informed about the voluntary nature of their participation, and online informed consent was obtained from them. Afterwards, online questionnaires were sent to them via WhatsApp (within 48 hours of admission to the ward). The patients who did not return the questionnaires were telephoned and encouraged to complete and return the questionnaires. It should be noted that no more than one follow-up attempt was made for each patient. The data were collected anonymously without name lists. Immediately at the end of the second week, the online questionnaires were reapplied and the post-treatment scores were obtained (Figure 1).
Information about the ‘intervention’ and ‘control’ wards could affect the types of patients recruited (recruitment bias). . Therefore, randomization of wards (clusters) was performed after recruitment of the patients. This was done by an independent observer not involved in this study using a coin toss (allocation concealment). In order to minimize imbalance across the intervention and control groups, a pair-matched randomization of clusters (wards) was used . Accordingly, to ensure comparability of the intervention and control groups, it was attempted to identify the pair of wards in the same hospital, so we randomly allocated one ward in each hospital to the intervention group and the other to the control group. Both hospitals were under the supervision of the Vice Chancellor for Medical Care of Shiraz University of Medical Sciences and provided the same services to all patients with COVID-19.
Online multimedia psychoeducational intervention condition
All eligible patients in the wards allocated to the intervention condition received online multimedia psychoeducational interventions during 2 weeks. Psychoeducational interventions consisted of 14 daily modules that were based on cognitive–behavioural techniques, stress management techniques, mindfulness-based stress reduction and positive psychotherapy. The patients were asked to complete 1 module per day, which was designed to be 60 minutes in total. Each module consisted of videos, audio files, educational texts, and one or two exercises related to the module content that were designed by a team of psychologists and psychiatric nurses supervised by the first author. WhatsApp was used to deliver daily multimedia psychoeducational contents (videos, podcasts, and educational texts) to the patients between 9 AM and 9 PM with approximately two-hour intervals.
Cognitive–behavioral modules were designed to teach patients how to recognize and mitigate their cognitive biases, especially in relation to the disease and the likelihood of adverse events due to the disease . In addition, various types of relaxation techniques including progressive muscle relaxation, imagination exercises, and diaphragmatic breathing were taught to the patients via video clips and audio files. Mindfulness techniques were also incorporated to help patients recognize their negative thoughts and emotions about the disease and reduce the intensity and impact of those thoughts and emotions on their stress levels . In this technique, the patients were trained to allow their negative emotions to be there without attempting to alter them or push them away patients were encouraged every day to practice these techniques and provide feedback on the techniques that worked best for them and adapted to their conditions. The patients were informed that they were not required to do all the techniques every day. Instead, they were advised to choose the most effective technique for themselves and practice it daily. In order to increase positive emotions and optimism in patients, positive psychotherapy exercises such as “Positive Reminiscence”, “Hope, Optimism, and Posttraumatic Growth”, “Gratitude Text” and “Finding Meaning”, were designed. During the “Positive Reminiscence Exercise”, the patients were encouraged to think about events in the past that evoked positive emotions, visualize the events in detail, and focus on the pleasant feelings arising during the exercise . During “Hope, Optimism, and Posttraumatic Growth” exercise, patients were encouraged to think about the times when important things were lost, but other opportunities transpired. During “Gratitude Text” exercise, the patients were encouraged to write and send a gratitude letter to someone he/she had never properly thanked using WhatsApp .
All eligible and interested patients in the wards allocated to the control condition were offered the opportunity to receive telephone-based counseling from the psychological team if needed. After the second assessment (T2), the patients in the control condition were offered the multimedia psychoeducational interventions.
The patients were blinded to the patient group assignments and did not know what the other interventions were. Besides, evaluators and analyzer of the outcomes were not informed about the patients’ treatment assignments.
Data were collected using online questionnaires and forms. Socio-demographic and clinical assessment form developed by the researchers was used to assess the patient' sociodemographic characteristics (age, gender, marital status and educational level) and clinical features (dyspnoea, fever, cough, tiredness, anorexia, nausea, diarrhea and hemoptysis). The outcome measures were as follows:
Primary clinical outcome
Resilience was considered as the primary clinical outcome because it was assumed that resilience acts as a protective factor against stress in patients . Resilience was evaluated by the Connor-Davidson resilience scale (CD-RISC) . This scale consists of 25 items rated on a 5-point Likert scale, ranging from 0 (not true at all) to 4 (true nearly all the time). The CD-RISC score could range from 0 to 100, with higher scores reflecting greater resilience. Internal consistency (Cronbach’s alpha) for the full scale is 0.89 . The scale demonstrated good convergent validity, and factor analysis yielded five factors . The Persian version also showed high internal consistency (Cronbach’s alpha=0.89) and sufficient validity [29, 30].
Secondary clinical outcome
Perceived stress was estimated employing the Perceived Stress Scale (PSS). The PSS was designed to measure the degree to which situations in one's life were appraised as stressful . It was a self-report 14-item questionnaire rated on a 5-point Likert scale, ranging from 0 (never) to 4 (very often). Thus, the total score of the scale could range from 0 to 56, with higher scores indicating higher levels of perceived stress. This measure exhibited sufficient reliability (Cronbach’s alpha =0.84-0.86) and validity . The Persian version also showed excellent internal consistency (Cronbach’s alpha=0.90) and convergent validity .
Feasibility, adherence, and satisfaction with the online multimedia psychoeducational interventions
The feasibility was assessed using the percentage of eligible patients who were enrolled and retained in the study. We defined the study feasible if 70% of patients were adherent to the study . The number of modules and exercises that patients completed was used as the definition of adherence to the intervention. Full adherence was defined as completing all daily modules and providing feedbacks on daily exercises.
In order to measure the level of satisfaction and to gather the necessary feedback on the online multimedia psychoeducational interventions, the client satisfaction questionnaire adapted to internet-based interventions (CSQ-I) was used t consists of 8 items answered on a four-point Likert scale ranging from 1 (does not apply to me) to 4 (does totally apply to me). Hence, the total score of the scale varied from 8-32. The scale demonstrated excellent internal consistency (McDonald omega = 0.93-0.95) as well as (convergent and discriminant) validity . The Persian version of CSQ-I also demonstrates an excellent internal consistency in the present study (Cronbach’s alpha =0.92). In the present sample, the construct validity of the Persian version of the CSQ-I was confirmed by significant correlations of the CSQ-I score and changes in the scores of resilience (r=0.41, P=0.03) and perceived stress (r=0.54, P=0.004) between T1 and T2.
Compliance test for normal distribution was applied using Kolmogorov–Smirnov test. Levene's test was used to examine the heterogeneity of the variances. Chi-square test was performed to compare the groups concerning demographic and clinical variables. Since the assumptions of analysis of covariance (ANCOVA) were not established , independent samples t-tests with adjustment for clustering effect were carried out to evaluate the differences between the two groups with regard to the dependent variables (perceived stress and resilience). Because of the hierarchical structure of the data (with hospital wards as the unit of randomization and patients as the unit of analysis), an adjustment for clustering was needed . To this end, the required Intra-cluster Correlation Coefficients (ICCs) were first calculated by the formula derived by Donner and Klar based on an analysis of variance . Then, the variance inflation factor (VIF) known as the ‘design effect’ was calculated from the ICC. In order to adjust for clustering effect, test statistics based on the t-tests were divided by the square root of the design effect . Between group effect size for the mean differences of groups with unequal sample sizes within a pre-post-control design (dppc2) was calculated according to Morris’s recommendations .
Descriptive statistics, such as means, standard deviations (SDs), frequencies and percentages were used to assess the feasibility, adherence, and satisfaction with the online multimedia psychoeducational interventions. Pearson’s correlation coefficients were calculated between the CSQ-I score and changes in resilience and perceived stress scores between T1 and T2. A p value <0.05 was considered to be statistically significant. The analyses were conducted with SPSS® for Windows® version 22.0 (SPSS Inc, Chicago, IL, USA).
Based on the results of a previous study , assuming a two-tailed test, α = 0.05, 20% attrition, mean difference=12, standard deviations (S1=12.6, S2=9.2), and using the MedCalc software, totally 32 patients were needed to ensure 80% power to detect a significant difference between the intervention and control groups. The calculated sample size was multiplied by the design effect to estimate the effective sample size in the present study (32×1.57=50). The design effect represents the factor by which the sample size must be increased when a cluster design is used in order to provide the same power as a study with individual allocation and analysis . The design effect for calculating an effective sample size was calculated using the following formula: VIF=1+(m-1) ICC, where m was the mean number of individuals per cluster (12.5) and ICC was considered as 0.05 .