The two aims of this study will be executed sequentially. Methods for each aim are shown in Figure 1 and described below.
2.1. Development of the CPOT-Fam
2.1.1. Working Group
To adapt the CPOT for use by families, we will convene a virtual working group with multiple stakeholders including patient partners (BS, SL), an ICU physician (HTS), an ICU nurse (VO), trainee researcher (AS), research staff (RBM, KK, KP, SM), and an experienced investigator (KF). The working group will meet monthly to: 1) determine the informational needs of family caregivers of critically ill adults regarding pain; 2) identify the steps required to adapt the CPOT to use by family caregivers and; 3) develop an accompanying pain education module to teach family caregivers to use the CPOT-Fam. After each working group meeting, minutes will be circulated by email to group members. The research lead (AS) will incorporate the feedback from group discussion to adapt the CPOT-Fam and build the educational module. Any re-drafted contents will be circulated back to the group by email. This process will occur iteratively until a consensus is reached on the content of the CPOT-Fam and educational module.
2.1.2. Adapting the CPOT for family use
The CPOT is designed for use by ICU clinicians and contains complex medical terminology. To adapt the CPOT for use by family caregivers, the working group will discuss simplification of the CPOT to create the CPOT-Fam. The content areas and scoring of the CPOT-Fam will be identical to the CPOT, but the scoring descriptions will be modified to better suit (i.e., improve readability, clarity, and conciseness) family caregiver use.
2.1.3. Educational module content
The educational module will be developed as a multi-part educational package which will include: 1) information on ICU pain; 2) information on how and when to use the CPOT-Fam and; 3) sample cases that family caregivers can score using the CPOT-Fam. A draft outline of the educational module containing further detail is presented in Table 1. The educational module will be written at a sixth-grade reading level and will be assessed using the Patient Education Materials Assessment Tool for print materials (PEMAT-P)27 to ensure that individuals with varying educational backgrounds and health literacy can understand the sample cases and utilize the information presented.
2.1.3.1. Sample Cases
The sample cases will be based on four dimensions used to assess pain in the original CPOT: 1) facial expression; 2) body movements; 3) compliance with ventilator (if intubated) or vocalization (if not intubated); and 4) muscle tension.26 Each sample case will represent a scoring option for dimensions 1-4, generating a unique combination adding up to a total CPOT score. For example, a sample case of a non-ventilated patient could reflect a total CPOT score of 3 and be comprised of individual dimensional scores such as: a score of 1 on vocalization, 2 on facial expression, 0 on body movements, and 0 on muscle tension. As such, there will be 81 unique combinations (4 dimensions across 9 scoring options = 81 combinations) of scores available for intubated patients and 81 for non-intubated patients for a total of 162 combinations. Of these, we anticipate some score combinations to be clinically unlikely (e.g., a score of 2 on vocalization, 0 on facial expression, 0 on body movements, and 0 on muscle tension) and suitable for removal, leaving approximately n=120 sample cases. Clinically unlikely scoring combinations will be identified through discussion with research team nurses and excluded. Family caregivers will be able to score pain for any given sample case by using the newly adapted CPOT-Fam. A sample case is attached in Supplementary Item 1.
2.1.4. CPOT-Fam Pre-Clinical Testing
After the CPOT-Fam and educational module are finalized, we will conduct pre-clinical testing with members of the general public and family caregivers of critically ill adults. Study participants will use the educational module and score the provided sample cases using the CPOT-Fam. This will be executed using a cross-sectional study design.
2.1.4.1. Study setting, participants, and recruitment
A group of participants for pre-clinical testing of the CPOT-Fam will be recruited from social media (e.g., Twitter channels), Bethecure.ca (a resource providing members of the general public information on participating in research and available opportunities), and four adult medical-surgical ICUs, and one cardiovascular surgical ICU in Calgary, Canada. Participants will be eligible for inclusion in this study if they are 18 years of age or older, able to communicate in English (understand, read, speak), and able to provide informed consent. To minimize risk of COVID-19 exposure to study participants, study materials will be provided to participants of this phase digitally through email. All participants will provide verbal consent prior to participating.
2.1.4.2. Procedures
After recruitment, participants will be asked to complete the educational module before using the CPOT-Fam to score four sample cases. After this, participants will be asked to complete a short demographic questionnaire administered using Qualtrics (Provo, UT, USA) providing information such as their age, sex, highest level of education completed, language spoken at home, and their relationship with the patient in ICU (if applicable) (Supplementary Item 2). Finally, participants will be asked to complete a short Qualtrics-based follow-up survey on the CPOT-Fam (Table 2). A research team member will facilitate participants’ completion of the educational module, sample cases, the demographic survey, and the follow-up survey by being present to answer questions by telephone or email. All data will be entered into a secure Research Electronic Data Capture (REDCap) database hosted at the University of Calgary, which provides an interface for validated data capture, audit trails for tracking data manipulation and export procedures, and seamless data downloads to common statistical software.28,29 For each sample case, participant CPOT-Fam scores will be compared with the reference standard CPOT score for a given sample case.
2.1.4.3. Sample size and power considerations
For pre-clinical testing of the CPOT-Fam, we aim to recruit participants according to an estimate of 0.85 for the intraclass correlation coefficient (ICC) between participant-assigned CPOT-Fam scores and reference scores (i.e., researcher-assigned scores). To estimate a 95% confidence interval with a width of 0.099(± 0.050), we will require a total of n=120 sets of scores from n=30 subjects (n=15 family caregivers of ICU patients, n=15 members of the general public) based on each subject scoring 4 sample cases.
2.1.4.4. Statistical Analysis
We will present the proportion of completion of the CPOT-Fam by family caregivers as a number and percentage. We will use the ICC to compare the participants’ CPOT-Fam scores with the reference CPOT-Fam scores. We will interpret the results according to the following accepted recommendations: an ICC ≤0.20 indicates slight agreement, 0.21-0.40 indicates fair agreement, 0.41-0.60 indicates moderate agreement, 0.61-0.80 indicates substantial agreement, and ≥0.81 indicates close to perfect agreement.30
2.2. Pilot testing the CPOT-Fam
After pre-clinical testing of the CPOT-Fam, the tool will be revised and pilot tested with family caregivers in adult ICU settings. These family caregivers will have a chance to learn to use the CPOT-Fam (with help from the educational module and sample cases), and practice using the CPOT-Fam to provide a proxy pain assessment when they visit their loved one in the ICU. Baseline measures for 1) family caregivers’ anxiety, 2) family caregivers’ sense of self-efficacy to care for the patient, and 3) family caregivers’ satisfaction with patient care will be collected. We will then compare family caregiver CPOT-Fam scores and nurse CPOT scores.
2.2.1. Study setting, participants, and recruitment
Consecutive critically ill adults and their family caregivers will be recruited from the four adult medical-surgical ICUs and one cardiovascular surgical ICU in Calgary, Canada. Patients will be eligible for inclusion in this study if they are 18 years of age or older, are being assessed for pain by nursing staff with the CPOT (i.e., are not able to communicate), and have a Richmond Agitation-Sedation Scale (RASS)31 score of ≥ -4. Family caregivers of patients will be eligible for inclusion in this study if they are 18 years of age or older, are able to communicate in English (understand, read, speak), and are able to provide written informed consent.
ICU nurses will inform a research team member if a patient and family caregiver(s) meet the study criteria. A trained member of the research team will approach family caregivers of critically ill adults, provide them with information about the study, and, if the family caregiver agrees to participate, obtain written informed consent. If the family caregiver consents to participate in the study, the patient will be enrolled through surrogate consent as patients who meet the inclusion criteria will be non-communicative.
2.2.2. Procedure and data collection
After recruitment, family caregivers will be asked to complete a demographic questionnaire providing information such as their age, sex, highest level of education completed, language spoken at home, and relationship with the patient (Supplementary Item 2). Family caregivers will be asked to complete the State-Trait Anxiety Inventory (STAI-Y),32,33 the Revised Scale for Caregiving Self-Efficacy (CSES-R),34 and the Critical Care Family Satisfaction Survey (CCFSS)35,36to provide baseline measures of anxiety, self-efficacy, and satisfaction with care in the ICU. 32–35, 37 Following these assessments, a trained member of the research team will present the educational module to family caregivers via the web or in paper-based format, depending on the participant’s preference.
After receiving the educational module, the family caregiver will be asked to provide a proxy assessment of the patient’s pain by scoring a time-stamped CPOT-Fam during each of their visit(s) with the patient until the patient is discharged from ICU, up to a maximum of three visits. Family members will complete the CPOT-Fam only if the patients remain unable to communicate. At the same time, the family caregiver will be asked to provide a subjective assessment of the patient’s pain using a short questionnaire, which will include an open-ended question and a visual analog scale (shown in Table 3). Family caregivers will be asked to note whether a patient’s pain was assessed at rest or during a procedure to provide context for their score. All data will be entered into a secure REDCap database by the researcher.
Following the last CPOT-Fam assessment (i.e., after the patient is discharged from the ICU, or after three assessments, whichever comes first), the family caregiver will be asked to complete a short follow-up survey (Table 2) to describe their experience with using the CPOT-Fam and their view of their role in assessing pain in their critically ill loved one. A research team member will be present to facilitate the completion of study tasks.
A trained research team member will coordinate with ICU nurses to provide their CPOT scores at the same time as the family caregivers provide their CPOT-Fam scores for a given patient on a given day. Assessments by family caregivers and ICU nurses will be conducted independently from one another. If this is not possible, the closest time stamped CPOT score(s) by an ICU nurse will be pulled from eCritical, a population-based provincial critical care clinical information system which captures demographic, clinical, and outcomes data for all admitted ICU patients, including notes on sedation and pain.38 Time-stamped CPOT scores by ICU nurses will be obtained, and any notes on sedation and pain will be reviewed to establish context for each CPOT score.
2.2.3. Sample size and power considerations
As this study is one of the first to adapt the CPOT to family caregivers and accompany it with an educational module,24 we aim to enroll a sample of 30 family caregivers based on projected timelines and enrollment rates in previous studies.39 With a sample size of 30 participants, we will be able to estimate an acceptability measure of 50% (i.e., 15 out of 30) to within a 95% confidence interval (CI) of ± 18%.
2.2.4. Statistical Analysis
We will calculate descriptive statistics (median (interquartile range); number (percentage)) for all study variables as appropriate. We will measure acceptability of the CPOT-Fam and educational intervention by calculating the proportion (with 95% CI) of eligible family caregivers approached who consented to participate in the study, and feasibility by calculating the proportion (with 95% CI) of consented family caregivers who complete their participation in the study.
We will calculate intraclass correlation coefficients to measure agreement between CPOT-Fam scores by family caregivers and CPOT scores by ICU nurses. The data will be stratified based on time between CPOT-Fam scores by family caregivers and CPOT scores by nurses to determine whether agreement differs with the time that scores are obtained (i.e., closer together vs. further apart). A similar approach will be used to assess agreement between CPOT-Fam scores and subjective assessments by family caregivers.