Evaluation of the CPR video decision aid with patients with end stage renal disease

A prospective quasi-experimental design included pre/posttest measures of knowledge and confidence in decision-making, and posttest only measure of uncertainty about the decision (decisional conflict). Research Ethics Board approval was obtained prior to commencing the study (BEH 13–200) and all participants provided written informed consent. The study was also approved by the local patient and family advisory council and health authority partners.

Participants were recruited between late summer and mid-winter 2015 in an urban inpatient and outpatient hemodialysis center. At these centers, 260 patients receive hemodialysis 6 days per week. A convenience sample of eight physicians who cared for patients with ESRD receiving dialysis was recruited from the hemodialysis centers. Eligible physicians were staff physicians (nephrologists) or residents on their nephrology rotation. Recruited physicians and members from the renal health care team (social worker, clinical nurse specialist, and unit manager) helped identify eligible patients in the hemodialysis program and/or family members and introduced the study. The research nurse then determined the final eligibility of patients and family members who expressed desire to participate.

Eligible patients were over 55 years of age, had stage 5 renal failure, were dependent on dialysis, and could communicate in English. Patients were invited to identify the adult ‘family’ member who knew them best, including partners, significant others, and/or close friends. The family member had to be 18 years or older, speak and understand English, have capacity to make healthcare decisions, have accompanied the patient to dialysis at least once, and have provided assistance to the patient without pay. Ideally, we wanted to recruit dyads consisting of a patient and their family member with whom they shared healthcare decisions. However, willing patients who could consent were included even without the participation of a family member. Patients without decisional capacity were invited to assent to participate if their family member was a participant.

The CPR-VDA is a seven-minute video designed for participants to independently view on a portable screen. The video presents information about CPR and the alternative option (comfort care) as well as information about the patient experience and important health outcomes. A CPR decision worksheet, which included a values clarification exercise, tailored the generic patient decision aid format to the CPR decision, and was completed with the study nurse. A plain language script for the CPR-VDA was a significant adaptation from an earlier paper-based CPR information tool [17], and was informed by a rapid systematic review process of CPR outcomes data. In collaboration with the research team, a cinematographer produced the final video CPR-VDA that is publicly available at http://​vimeo.​com/​48147363. The study nurse provided non-directive support to help participants complete the questionnaires thus preparing them for the discussion with the physician, consistent with the role of a decision coach in a shared decision framework [18].

The study nurse conducted interviews during a scheduled dialysis treatment. The interview commenced with the patient and/or family member once dialysis was initiated and the patient remained physically stable (Table 1). After obtaining consent, a brief chart review was conducted to check documented CPR orders and pertinent health history. Interviews occurred in a semi private place with patients, family, and staff in close proximity.

The first part of the questionnaires included; demographics with a self-reported frailty index and health literacy test, pre-knowledge test about CPR and pre-intervention self-efficacy questions. Following this series of questions, the patient and/or family member viewed the seven-minute CPR-VDA video on an iPad screen. The interview continued with acceptability questions regarding the CPR-VDA video, post-knowledge test questions about CPR, post-decision self-efficacy questions and series of questions to assess any decision conflict related to the CPR decision. The study nurse asked patients their CPR preference after viewing the CPR-VDA and completing the values clarification worksheet.

Acceptability was assessed using eight validated questions about use, amount of information, the length, the clarity, balance in presentation, willingness to recommend to others and overall suitability for decision-making [19, 20]. Knowledge about CPR was tested using nine questions developed by the research team. Self-Efficacy is the participants’ self-confidence or belief in their abilities in decision-making, and was measured using the Decision Self-Efficacy Scale [21]. The scale gave a total score out of 100, and a higher score indicates higher self-efficacy for decision-making. Used to evaluate a decision aid for women with osteoporosis, the reliability of the scale was 0.92 and it correlated with decisional conflict subscales of feeling informed (r = 0.47) and supported (r = 0.45) [21]. Decision conflict was also assessed, which occurs when a patient is uncertain about what course of action is best for them [22]. The perception of uncertainty is related to modifiable factors such as feeling uninformed, being unclear about personal values for the options, or feeling unsupported in decision-making [22]. Decision conflict was measured using the low literacy version of the Decision Conflict Scale for patient/family [22] where scores lower than 25/100 were associated with decision implementation and scores over 37.5/100 were associated with delayed decisions. The scale has a reliability of 0.78 and has been used in many studies of patient decision aids [16, 22]. Decisional conflict was also measured using the SURE test [23] in the values clarification worksheet. The reliability of this scale is 0.65 [23]. Measurement of relevancy of the CPR decision for the patient at this stage of their healthcare journey was asked before and after viewing the video. After participant(s) completed the pre and posttest questionnaires, they were then invited to complete the paper based values clarification worksheet.

Once all patient/family questionnaires were completed, the physician caring for the patient discussed CPR with the patient and/or family member to assess whether or not they wanted CPR in the event of a cardiac or respiratory arrest. The research nurse observed this discussion and completed an Observing Patient Involvement (OPTION) tool that assessed physician and patient involvement in shared decision-making. The OPTION instrument was developed to evaluate shared decision-making communication and the reliability of this scale was 0.66 in a study evaluating physician encounters in primary care settings [24]. Responses on a five-point scale ranged from ‘the behavior is not observed’ to ‘observed and executed to a high standard’. The total summed score range from zero to 48 with higher scores indicating greater competency in shared decision-making [24]. Physicians completed a final questionnaire to report on the communication with patient/family. During the interview time, the study nurse used an observation matrix and field notes. The matrix captured observations such as time of day, distractions, and local environment, before and after the CPR-VDA intervention. The data collection strategy is summarized in Table 1.

Baseline characteristics and outcomes were reported using proportions for categorical variables, and means and standard deviations (SD) for continuous variables. For each outcome of interest, analysis was conducted at the participant level using descriptive statistics and when appropriate, a paired sample t-test for comparing means (before/after). Data management and statistical analyses were conduct using IBM SPSS Statistics (Version 23).

Of those invited to participate, 8/8 (100%) physicians accepted, 49/53 (92%) patients accepted, and 8/9 (89%) family members accepted (Table 2). Of the five people who declined to participate, four were not interested and one patient was unable to complete the interview due to declining health status. There were seven patient/family member dyads and one family member who participated without the patient. Physicians were mostly experienced clinical nephrology staff and some had either palliative care experience or training about goals of care communication. Fewer than half of patients were female, half were married, and the average age was 67 years. Most patients lived in their own home in an urban setting where they received dialysis. Still, one third of the patients were from a rural area and travelled into the city for treatment. Patients had relatively high health literacy score despite having high school or less education. Over half of patients considered themselves vulnerable to severely frail and most had prior communication about CPR, which varied in formality.

All patient and family member participants viewed the CPR-VDA and completed the values clarification worksheet during the interview time. Each had a follow-up discussion with their physician, although the discussion did not always occur on the same day as the initial interview. After viewing the CPR-VDA and completing the worksheet, participants were clear about the necessary decision, knew the options, could articulate their values, and could discuss these in varying degrees of detail. They were also able to identify their support person(s), decision-making needs and make a plan for next steps. During the video viewing, challenges included poor lighting, disruptive noise, physical discomfort (i.e., vascular access in their arm / needle in arm) positioning for treatment, fatigue, thirst, and confusion over wording. Multiple interruptions in the busy environment affected some people’s ability to attend to detail during the interview / intervention process resulting in 6% (3/49) of patients needing to review parts of the video. While working through the values clarification worksheet, participants revealed emotional struggles surrounding their current health state and concerns about the future.

Participants were asked how relevant the decision about CPR was for them before and after the intervention, and while the average rating was that it was relevant to them (2/5), the mean score increased from 2.1 to 2.3 (p = 0.01) after the intervention. Ninety-eight percent (56/57) of patient and family member participants rated the CPR-VDA as good to excellent. Seventy-seven percent (44/57) stated it contained the right amount of information, 75% (43/57) thought the information in the video was completely balanced, and 93% (53/57) found the information presented about CPR to be clear. The CPR-VDA was helpful in making decisions about CPR for 89% (51/57) of participants and almost everyone (98%) would recommend the video to other people who are considering CPR (Table 3).

Participant knowledge about CPR increased from a mean score of 4.8/9 (standard deviation [SD] = 1.65) before viewing the video to 7.5/9 (SD = 1.40) (p = 0.000) after viewing the video. Decisional self-efficacy improved slightly from 84% pre intervention (SD 17.04, range 20–100) to 86% after the intervention (SD 14.13, range 39–100) (p = 0.005) for patient participants; however, family members’ scores remained high in both periods (Table 4). Decisional conflict scores were relatively low overall (scores could range from 0 [no decisional conflict] to 100 [extremely high decisional conflict]); they were higher amongst patients (mean score of 13.57, SD = 18.34, range 0–70) but very low amongst family members (mean score of 1.25, SD = 3.54, range 0–10). Decisional conflict was also measured for patients using the clinical SURE test on the values clarification worksheet: 14 (28%) patients reported experiencing decisional conflict while 36 (72%) reported no decisional conflict.

After the intervention, a physician discussed the CPR decision with the patient and/or family member with variable quality of patient involvement as assessed using OPTION (M = 25.66 SD 7.41, range 9–47, maximum score possible 48). These discussions were held during the patient’s dialysis treatment with others (i.e., family, nurse, pharmacist, social work, clinical coordinator, as well as other patients and their families) present in 31/49 (62%) patient conversations. Only the physician and participants were directly involved in the CPR conversation. This is usual practice as the dialysis unit is mostly an open observation unit. In the exit survey, physicians reported the CPR decision as “very relevant” to their patients (M = 3.60, SD = 0.53, range = 2–4), were “very satisfied” with the discussion they had after the intervention (M = 3.18, SD = 0.79, range = 1–4), and reported that the overall discussion was “easy” to have with their patient (M = 0.80, SD = 0.80, range = 0–3) (Table 4).