A series of qualitative interviews will be conducted among a purposive sample of clinicians throughout BC who deliver care to homeless clients with problematic substance use. Recruitment will be conducted by email advertisement distributed by infectious control leaders in each health authority, and sampling will be done until theoretical saturation is reached (estimated to be 20 participants). Maximum-variation sampling will be employed, which allows exploration of both typical and unusual concepts across a broad range of settings . Critical social theory and decision-making theory will be used to create a semi-structured interview guide to interview participants about their current practice for assessing CTC, i.e., the conceptual elements they consider when assessing capacity, the methods they use to assess these elements, and the threshold they use to make a final decision about capacity. Interviews will be conducted in a private office at the participant’s place of work.
A series of qualitative interviews will also be conducted with a convenience sample of adults throughout BC who are homeless and who self-report as abusing substances. Recruitment will be conducted through advertisement at community-based organizations and through word of mouth. Sampling will be done until theoretical saturation is reached (estimated to be 25 participants). The purpose of these interviews is to explore the experience of providing consent or obtaining medical care from the perspective of these individuals. Power inequities will also be explored. Interviews will be conducted in a private room where the community organization meets. A $20 cash honorarium will be provided to participants.
All interviews will be audiotaped and transcribed verbatim without personal identifiers. Concepts and themes will be identified in the data, and an interpretive description analysis , using both inductive (allowing concepts and themes to emerge from the data) and deductive (using concepts and themes that are described in the literature) approaches, will be conducted to determine how assessment of CTC can be translated into clinical practice.
A comprehensive review of the literature will be conducted to examine the existing instruments for determining CTC. In addition, theoretical concepts related to CTC will be reviewed. The proposed tool will be constructed using theoretical concepts, legal concepts, and knowledge gained from qualitative interviews. Reliable, validated questions from existing instruments will be added to the list of possible questions.
The proposed questions will be provided to a panel of experts to establish item-objective congruency by rating each question (item) using the Osterlind method . The panel of experts will include four doctors and four nurses from public health, one practitioner who delivers clinical care to the target population, one lawyer, one ethicist, an addictions specialist, and a psychiatrist. Experts will be encouraged to suggest other questions which may have been omitted. A modified Delphi process will be used to achieve group consensus about the relevancy of each question. Questions with discordant ratings will be further discussed in person until consensus is reached.
Once a near-final version of the instrument is created, pilot testing  will be conducted with five individuals from the target population. During this process, participants will be asked to respond to the questions and asked why they chose the response they chose and what the question meant to them. At this time, the number of response options will be assessed for appropriateness.
Three hundred clients who are more than 18 years old, and who speak and read English, self-report as being homeless, and self-report abuse of substances will be recruited for a validation study through a recruitment poster placed in a downtown Vancouver community-based organization. A short questionnaire on demographics, substance use, and history of mental illness will be administered. Participants will be presented with a simulated consent for a hypothetical medical scenario. They will then be assessed by a psychiatrist to capture a clinical assessment of capacity. Next, participants will be administered the MacCAT for treatment (MacCat-T) tool followed by the administration of the new assessment tool, conducted by a research nurse blinded to the clinical and MacCAT-T result. A second researcher will observe the interview and score it independently and a Kappa statistic  will be calculated using SPSS 14 to determine inter-rater reliability. A Kappa score of >0.8 will be considered acceptable .
A combination of classical test theory and advanced psychometric methods will be used for the psychometric analysis. An item analysis will be conducted by examining the corrected item-total correlation to identify items that discriminate poorly [15, 16]. Removal of items with corrected item-total correlation of <0.20 will be considered if the content is not considered clinically important. We aim to create an instrument that has the fewest items while maintaining good psychometric properties. The minimum number of items required to provide a reliability of at least 0.80 will be calculated using the Spearman Brown prophecy formula .
Guided exploratory factor analysis will be conducted on the final selection of items to confirm the assumptions for a Rasch model . If multi-dimensionality is confirmed multidimensional Rasch analysis will be conducted. If unidimensionality is confirmed an unstandardized Cronbach Alpha will be calculated and a global summary score will be created. The global score and the MacCAT-T score will be compared using a Pearson’s correlation test. A cutoff value for the new instrument will be created by conducting a receiver operating characteristic analysis using a combination of the clinical assessment of capacity and the four domain scores from the MacCAT-T as the gold standard. If the clinical assessment and all four MacCAT-T domain scores indicate capacity, then the gold standard assessment will be recorded as “has capacity”. If any of these five elements indicates a lack of capacity, then the gold standard assessment will be recorded as “does not have capacity”. Results will be stratified by alcohol use only, drug use only, and mental status.