Patients Preferences for Enrolment Into Critical Care Trials

Patients’ Preferences For Enrolment into Critical Care Trials

On-Line Supplement

Damon C. Scales MD, PhD1,2, Orla M. Smith, RN, MN3,4, Ruxandra Pinto, PhD2,

Kali A. Barrett, MD2, Jan O. Friedrich, MD, DPhil1,4, Neil M. Lazar, MD1,5,

Deborah J. Cook, MD, MSc6,7, and Niall D. Ferguson, MD, MSc1,5

for the Canadian Critical Care Trials Group

1 Interdepartmental Division of Critical Care Medicine, University of Toronto, ON, Canada

2 Department of Critical Care Medicine, Sunnybrook Health Science Centre, Toronto, ON, Canada

3 Faculty of Nursing, University of Toronto, ON, Canada

4 Critical Care and Medicine Departments, and Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, ON, Canada

5 Department of Medicine, Division of Respirology, University Health Network, Toronto, ON, Canada

6 Departments of Medicine and Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, ON, Canada

7 Department of Critical Care Medicine, St. Joseph’s Hospital, Hamilton, ON, Canada

Herein we elaborate on the full methods of our study, including the screening of patients, assessment of capacity, conduct of the interview, and analyses of results. Following the description of the methods, we present the interview script used for the study.

Methods

Development and Pre-testing of the Study Instrument

We first conducted a literature review and consulted with local experts (physicians, nurses, an ethicist, a lawyer, and a past patient) to develop the construct domains. The major consent frameworks identified were (1) consent by a substitute decision-maker (SDM) prior to enrolment; (2) study enrolment followed by delayed consent by a SDM when one becomes available; (3) study enrolment followed by delayed consent by the patient when they become capable; and (4) study enrolment with no formal consent obtained. This process also identified 3 important study permutations that might influence preference of consent framework: risk (very low versus high), treatment type (new versus currently available), and availability of substitute decision-maker (yes versus no).

We next developed a structured interview script to evaluate preference and acceptability of these different consent frameworks across study permutations. We performed sensibility testing using a 10-item questionnaire administered to 4 critical care physicians, 4 critical care nurses, and 4 critical care research coordinators. Respondents assessed the instrument for simplicity (1 question), clarity (1 question), subjectivity (1 question), validity (3 questions), necessity (1 question), bias (1 question), and reliability (1 question) and rated items using 7-point Likert scales (1=poor, 7=excellent). The results from this sensibility testing (Figure A1) were used to improve and refine the interview script. We then pilot-tested the instrument using 13 patients and 2 interviewers to obtain qualitative feedback that was used to further improve the final interview script. All interviewers in this study underwent centralized training under the direction of a senior research coordinator (OMS) to ensure consistency in approach across sites and interviewers.

Figure A1: Results of Sensibility Testing

Figure A1 Legend: Median responses obtained on Likert scales during sensibility testing. Likert 1 = poor; Likert 4 = neutral; Likert 7 = excellent.

Patients

All patients from 5 Ontario teaching hospitals who survived to ICU discharge were screened for study eligibility. Screening occurred between August 2005 and August 2006 for 12 months at one hospital, 8 months at 2 hospitals, and 4 months at 2 hospitals. Patients were considered potentially eligible if they demonstrated capacity to provide consent for the interview, but were excluded if they were not fluent in spoken English, were age ≤ 16 years, died or were discharged from hospital prior to demonstrating capacity, or were previously enrolled in this study.

Capacity Screening

We modified the Aid to Capacity Evaluation (ACE)[2], a validated tool for assessing capacity for medical decision-making, to systematically evaluate patients for capacity to provide research consent. We changed all references to ‘treatment’ to refer to ‘research’. Accordingly, if the potential subject was able to communicate and speak English, the interviewer would ask the questions outlined in Box A1. After ICU discharge, patients were screened for capacity at the following intervals: every 72 hours during the first week, every week during the next 4 weeks, and every 2 weeks thereafter until hospital discharge. Once a patient was deemed capable, we sought consent for the interview and no further capacity screening occurred.

Structured Interview

We asked participants to consider a scenario where they were critically ill and incapable of providing consent for a randomized clinical trial. The participants were informed that the study under consideration in the scenario had been carefully reviewed and approved by a research ethics board, and was a randomized, placebo-controlled trial with a very low risk profile (<1/1000 of serious complications); a SDM was available. We asked participants to first select which one framework for obtaining consent they would prefer for the scenario (Box A in main manuscript), and then asked them to rate the acceptability of each consent framework using a 7-point Likert scale (1 = highly unacceptable to 7 = highly acceptable). We repeated this process for the following 3 permutations of the baseline scenario: (1) increased risk (<1/20 instead of <1/1000); (2) treatment type (testing 2 currently available treatments instead of a new treatment); (3) and availability of SDM (unavailable instead of available).

Interviewers evaluated patients’ comprehension of the structured interview by documenting whether they could correctly identify the major change between each scenario permutation and the baseline scenario. Interviewers also documented their subjective assessments of global participant comprehension (definitely inadequate, probably inadequate, probably adequate, or definitely adequate).

Patient Characteristics

We collected the following data on all patients discharged alive from participating ICUs during the study period: age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II score within 24 hours of ICU admission[3], ICU admission diagnosis, need for mechanical ventilation, duration of ICU admission, and history of the patient or SDM ever being approached for research consent during the recent ICU admission. From enrolled patients, we also documented birthplace (Canada versus other), ethnic origin (according to Statistics Canada categories)[4], highest education level attained, and the relationship with their self-identified SDM.

Outcomes

The primary outcome was the frequency distribution of participants’ preferred consent framework for the baseline research scenario. Secondary outcomes included the frequency distribution of the preferred consent framework for each permutation of the baseline scenario; proportion of participants finding each consent framework to be ‘acceptable’ to ‘highly acceptable’ (Likert scores = 6 or 7) and ‘unacceptable’ to ‘highly unacceptable’ (Likert scores = 1 or 2) in each scenario; and proportion of participants who changed their baseline preference with each permutation of the study scenario.

Analysis

We report mean and standard deviation for normally distributed continuous variables and median and interquartile range (IQR) for non-normally-distributed continuous data. We used the student t test or Mann Whitney U test, where appropriate, to test for differences in continuous variables between enrolled and non-enrolled patients. Dichotomous variables are reported as proportions. Bowker’s test of symmetry was used to compare proportions of patients selecting each consent framework across different study scenarios[1]. We also examined whether preference for ‘consent by SDM prior to enrolment’ in the baseline scenario versus other consent frameworks was associated with participant characteristics (age, female sex, APACHE II score, surgical diagnosis, mechanical ventilation during ICU admission, patient or SDM approached for research consent during ICU stay, spouse as SDM versus other, education level high school or higher, and born in Canada versus elsewhere) using the Chi-square test or Fisher exact test for dichotomous variables and the student t test for continuous variables. For all comparisons, a p-value of <0.05 was considered to be statistically significant.

Ethics

This study was approved by the Research Ethics Boards of all participating hospitals. Informed consent was obtained from all participants prior to the interview.

Role of the Funding Source

This study was funded by grants from the Physicians’ Services Incorporated Foundation and the Canadian Intensive Care Foundation. None of the funding agencies had any role in the design and conduct of the study; collection, management, analysis, the interpretation of the data; or the preparation, review, and approval of the manuscript.

Supplementary Results

Excluding the minority of patients whose comprehension was deemed to be inadequate did not change our results. For example, when the 8 (3.6%) patients with ‘probably inadequate’ or ‘definitely inadequate’ comprehension were excluded, the proportions of respondents favouring the SDM framework in each scenario were as follows: (1) baseline scenario – 77%; (2) scenario comparing 2 accepted treatments - 82%; (3) scenario involving increased risk of harm – 79%. When no SDM was available, 77% favoured enrolment followed by delayed consent to patient.

Characteristics Associated with Choice of SDM Framework

We tested whether patient characteristics were associated with choosing ‘consent by SDM prior to enrolment’ as the preferred framework for the baseline scenario. None of the following characteristics were associated with preference of the SDM framework: female sex, surgical diagnosis, mechanical ventilation during ICU admission, patient or SDM approached for research consent during ICU stay, spouse as identified SDM versus other, education level high school or higher, born in Canada versus elsewhere, APACHE II score.

Table A1: Characteristics associated with choice of SDM Framework

Characteristics / SDM Framework Preferred
(n=180) / Other Framework Preferred
(n=60) / p-value
Age / 51 / 49 / 0.38
Female Sex / 41% / 48% / 0.35
Mean APACHE II Score / 13.6 / 13.8 / 0.91
Surgical admission diagnosis / 48% / 57% / 0.22
Need for mechanical ventilation / 58% / 68% / 0.18
Duration of ICU admission (median) / 2 / 2 / 0.11
Approached for research consent / 4% / 7% / 0.16
Education greater than high school / 80% / 77% / 0.72
Born in Canada / 61% / 66% / 0.51
Spouse as SDM / 52% / 54% / 0.80

Box A1: Modified Aid to Capacity Assessment

Box A2: Interview Script

[Interviewer reads the following script to patient during interview]

Preamble

Before any new treatment or medication can be used in the routine treatment of patients, it must first be carefully studied through research. Many patients in the ICU are very sick, and therefore there are many people whose lives could potentially be saved through these new treatments. This is why research in the Intensive Care Unit (ICU) is important. Most of the time patients decide for themselves if they want to participate in a research study after hearing about its purpose, and its possible risks and benefits. This process of informed consent usually can’t happen in the ICU because almost all ICU patients are too sick to make this decision. In the ICU, several other potential ways for getting permission to include patients in research studies have been used; here is a card that lists four of them. As someone who has recently been a patient in the ICU, we want to ask you how you feel about them - let me go over them with you…

[Interviewer hands patient card listing the following four different frameworks].

1.Substitute Decision Maker. A substitute decision maker (usually the closest family member) is asked to decide whether or not the patient would want to be included in the research study. For example, in your case – who would be your substitute decision maker? [Interviewer may prompt patient to determine substitute decision maker (spouse, partner, child, parent, sibling, power of attorney, etc.); this information is inserted into subsequent scenarios]

2.Delayed to Substitute. A patient who meets all the criteria to participate in a research study is automatically enrolled in the study. At a later point in time, the substitute decision maker (usually the closest family member) is asked to decide whether or not the patient would want to continue to be included in the research study.

3.Delayed to Patient. A patient who meets all the criteria to participate in a research study is automatically enrolled in the study. As soon as the patient is able to make this decision, he/she is asked to decide whether or not they want to continue to be included in the research study.

4.No Formal Consent. A patient who meets all the criteria to participate in a research study is automatically enrolled in the study. No formal consent is obtained.

Do you have any questions about these ways of obtaining permission?

It is important that you understand that before any research study begins (no matter what way is used to obtain permission) the study has already been reviewed and approved by a research ethics board. A research ethics board is an independent and impartial group of individuals made up of physicians, nurses, ethicists, and members of the general public. They have the responsibility of carefully evaluating research studies to make sure that any risks to a patient’s well being are minimized and that the patient’s rights are protected.

I am going to present some hypothetical scenarios to you. In each scenario I want you to imagine that you are the patient and back in the ICU. I will then have you pick the one way for obtaining research permission out of those listed that you would most prefer we use in that situation. I will also ask you to rate how acceptable each of the ways would be in that situation. There are no ‘right’ or ‘wrong’ answers.

Do you have any questions before we begin?

Scenario 1- Baseline:

I want you to imagine that you are the patient. You are very sick and have been admitted to the intensive care unit. You are in a coma, on life-support, and unable to communicate. Your doctors and the team working on a research project have determined that you are eligible to be enrolled in a study. This study has been carefully evaluated and approved by a research ethics board. As I mentioned earlier, a research ethics board is a group of people including doctors, nurses, and members of the general public who evaluate studies to make sure that patients’ rights are protected and that the potential risks in studies are minimized.

•Your [insert relationship of SDM to patient] is in the waiting room.

•There is a very low risk (less than 1 in 1000) of serious complications related to the study treatment.

•When you are enrolled in the study, there is a 50/50 chance that you will receive either the study treatment or a placebo based on random chance (e.g. toss of a coin).

In this scenario, imagining yourself as the patient, which of the four ways for obtaining permission would you most prefer we use in your case? [Interviewer directs patient to refer to card –answer is recorded on response sheet]

Now please rate each of these ways using this scoring sheet [Interviewer hands response sheet to patient]. You are to assign a score from 1 to 7 for each way – as indicated, 1 is the lowest score (highly unacceptable), while 7 is the highest score (highly acceptable)

Scenario 2- Different Risk:

Here is the 2nd scenario.

This time the treatment being studied has a higher risk of serious complications (less than 1 in 20 compared with less than 1 in 1000).

As before:

•Your [insert relationship of SDM to patient] is in the waiting room.

•There is a risk (less than 1 in 20) of serious complications related to the study treatment.

•When you are enrolled in the study there is a 50/50 chance that you will receive either the study treatment or a placebo based on random chance (e.g. toss of a coin).

To review – can you tell me what the major difference is in this scenario compared with the original situation? [Interviewer records patient’s response] Do you think this is an important difference? [record yes/no]

So, in this scenario, imagining yourself as the patient, which of the four ways for obtaining permission would you most prefer we use in your case? [Interviewer directs patient to refer to card –answer is recorded on response sheet]

Scenario 3- Two accepted strategies:

Here is the 3rd scenario.

This time the research study does not involve a new treatment. Instead, two different treatments that are commonly used are being compared (for example, evaluating two different medications that are used to treat the same condition).