Youth represent a substantial portion of smokeless tobacco (ST) users in Canada compared to the general population. Highest prevalence of use is typically seen in males, in current smokers, and in the Western provinces. ST use has also been associated with youth who participate in sports teams. Furthermore, ST has been marketed to youth through the use of flavours and sweeteners to make ST more attractive and appealing. The goal of this study was to determine the prevalence of ST use among Canadian youth and examine factors associated with its use. The study used self-report data from 29,007 grade 9-12 youth who participated in the 2010-2011 Canadian Youth Smoking Survey (YSS). Logistic regression analyses were used to determine factors associated with ST ever and current use among Canadian youth as well as among the sub-population of youth smokers. A mediation analysis was also used to understand if (a) sports team participation or (b) physical activity mediate an association between attempting to quit smoking and ST use among youth smokers. In 2010-2011, 5.3% of Canadian youth had ever used ST and 1.9% were current ST users. Odds of ST use were highest among males, grade 12 students, youth with more than $100 of weekly spending money and current smoking youth. This study was the first to identify associations between both physical activity and sports team participation and ST use among grade 9-12 Canadian youth. Continued monitoring of ST use among youth is recommended. Additionally, further research is needed to explore beyond individual-level factors and understand broader influences of ST use among youth.

Background. Inuit experience some of the highest rates of tobacco use and of tobacco-related diseases in Canada. Communication strategies, such as health warnings on tobacco products, are seen as a necessary means of informing the public of tobacco-related health risk and motivating smokers to want to quit smoking. However, there is little evidence to suggest how such strategies might be working among Inuit nor is there evidence to suggest how best to communicate tobacco-related health risk to and promote smoking cessation among Inuit smokers. Objectives. (1) To systematically examine the effects of textual message frame (i.e., loss- vs. gain-framed), graphic type (i.e., gruesome vs. personal suffering), and narrative style (i.e., testimonial vs. didactic) on measures of message acceptance (i.e., personal relevance and perceived credibility), affective response, and potential message effectiveness. (2) To examine fear as a potential mediator of the relation between textual message frame and measures of potential message effectiveness, as well as of the relation between graphic type and measures of potential message effectiveness. (3) To examine the potential impact of the message spokesperson (i.e., Caucasian, middle-aged male/female vs. Inuit middle-aged male/female vs. Inuit Elder male/female) on measures of message acceptance and potential message effectiveness. Experimental design. A repeated measures (i.e., within-subject) 2 x 2 x 2 factorial design was used to examine the effects of textual message frame, graphic type and narrative style. A separate ranking task assessed the potential impact of the message spokesperson. Methods. Eligible participants (Inuit, aged 18 years of age or older, having smoked at least one cigarette in the previous 30 days and smoked over 100 cigarettes in their lifetime) were recruited in October 2012 from two communities in Nunavut (Iqaluit and Rankin Inlet). Participants completed a survey, an experimental procedure (i.e., a health warning rating task) and a health warning ranking task on a hand-held electronic device with a trained research assistant. With data from the health warning rating task, a series of multinomial regression models using the Generalized Estimating Equation (GEE) method were fitted to examine the effects of three message characteristics on each of the outcome measures, controlling for known covariates. Outcome measures were categorized into 3-levels: (1) extremely, (2) somewhat, and (3) not really. The “not really” category was used as the comparison category for multinomial regression models. Multinomial regression was also used to examine the potential mediating effects of fear as it related to each of the measures of potential message effectiveness. With data from the health warning ranking task, frequencies of participant choices as related to the message spokesperson were examined. 129 participants were included in the analyses. Results. Participants were, on average, 37.3 years of age (STD = 12.7) and smoked 13.0 cigarettes per day (STD = 8.9). Just over half were female (56.6%) and most had less than a high school education (72.7%). Messages with gruesome images were more likely than those with images of personal suffering to be rated as extremely relevant (OR = 2.23, CI: 1.56-3.20), credible (OR = 2.46, CI: 1.67-3.62), emotionally arousing (OR = 3.40, CI: 2.27-5.08), and potentially effective (OR = 2.56, CI: 1.69-3.86). Loss-framed messages were more likely than gain-framed messages to be rated as extremely emotionally arousing (OR = 1.71, CI: 1.23-2.37), but no more likely to be rated as extremely relevant (OR = 1.03, CI: 0.61-1.74), credible (OR = 1.06, CI: 0.81-1.39), or potentially effective (OR = 1.24, CI: 0.98-1.58). Testimonial messages were no more likely than didactic messages to be rated as extremely relevant (OR = 0.90, CI: 0.60-1.35), credible (OR = 0.97, CI: 0.70-1.34), emotionally arousing (OR = 1.22, CI: 0.90-1.67), or potentially effective (OR = 1.08, CI: 0.85-1.37). Fear appeared to partially mediate the relation between textual message frame and all three indicators of potential message effectiveness suggesting loss-framed messages elicited greater feelings of fear, thereby enhancing the potential effectiveness of the message. There was also some evidence that fear partially mediated the relation between graphic type and some indicators of potential message effectiveness suggesting messages with gruesome images elicited greater feelings of fear, thereby enhancing the potential effectiveness of the messages. Finally, greater proportions of participants indicated health warnings with an Inuit Elder were most personally relevant (44.2%) and most credible (35.9%) compared to health warnings with middle-aged Inuit or Caucasian spokespersons. However, participants’ choice of which health warning was potentially most effective was split relatively evenly between all options. Conclusions. Findings from this study suggest health warnings accompanied by gruesome images are potentially more effective at communicating tobacco-related health risk and motivating cessation among Inuit compared to those with images of personal suffering. This provides some initial evidence that current communication strategies that use gruesome imagery, like some tobacco product health warnings in Canada, may be effective among Inuit populations. However, when a spokesperson is used in a communication campaign, Inuit Elders tend to be preferred. Together these findings suggest that an integrated communication strategy that includes complementary, targeted materials working synergistically alongside population-level approaches (like tobacco product warning labels) may work best among Inuit.

Introduction: Bandura’s self-efficacy theory postulates that confidence is a stronger determinant of behaviour than one’s actual skills or abilities. The Day and Night time Driving Comfort Scales (D-DCS and N-DCS) are new tools to measure perceived confidence in older drivers. Developed inductively with older drivers, the DCSs have good structural properties and preliminary psychometric support, including test-retest reliability. Purpose: This thesis builds on previous work by further examining the construct validity of the DCSs via (1) temporal associations (Study 1) and (2) cross-sectional relationships (Study 2). Study 1 prospectively examined the relationship between DCS scores and self-reported driving behaviour. Study 2, meanwhile, examined cross-sectional associations between DCS scores and selected sample characteristics, self-reported driving behaviour and driving problems, perceived abilities and seven objectively measured driving-related abilities (visual acuity, contrast sensitivity, disability glare, brake reaction time, lower body mobility, executive skills and visual attention). Methods: For Study 1, a convenience sample of 82 older drivers (aged 63 to 93) were assessed at baseline and follow-up (five to 17 months later). Study tools comprised the DCSs, multi-item ratings of driving frequency, avoidance and perceived abilities and a background questionnaire. Telephone interviews were conducted with a subgroup of 45 drivers to examine reasons for changes in driving comfort. For Study 2, cross-sectional relationships with the DCSs were examined using baseline data from 65 drivers (aged 63 to 93). A subgroup of 42 participants completed objective assessments of driving-related abilities assessed via ETDRS charts, Pelli-Robson charts, Brightness Acuity Tester, brake reaction time apparatus, the Rapid Paced Walk, the Trail Making Tests (Parts A and B) and the UFOV subtest 2. Results: Prospectively, lower baseline N-DCS scores (p<.001) and decreased N-DCS scores (p<.001) were significant predictors of lower self-reported situational driving frequency (R2 = 34%) and greater situational avoidance (R2 = 51%) at follow-up. While DCS scores did not change appreciably for those who continued driving, N-DCS scores were significantly lower at follow-up for seven individuals who stopped driving (p<.05). Cross-sectionally, lower DCS scores were significantly associated with lower self-reported situational driving frequency, higher situational avoidance and lower ratings of perceived abilities (p<.001). Poorer left eye acuity scores were significantly associated with lower D-DCS (p<.05) and N-DCS (p<.05), while slower performance on Part A of the Trail Making Test was significantly related to lower D-DCS scores (p<.05). Participants with a discrepancy between their perceived and actual abilities had significantly higher D-DCS, situational driving frequency and lower situational avoidance (p<.05). Conclusions: Findings are consistent with Bandura’s self-efficacy theory and Rudman’s model of driving self-regulation and, thus, provide further support for the construct validity of the DCSs. Further studies are needed with larger, more diverse samples, including those with diagnosed impairments, to establish benchmarks for driving comfort in healthy drivers and various clinical populations (such as those with stroke, Parkinson’s or visual conditions). Prospective studies should also involve longer follow-up periods, examination of actual driving behaviour and barriers to self-restriction, and attempt to pinpoint whether there is a critical level of discomfort at which voluntary cessation is likely to occur.

Objective: The primary aim of this study was to better understand the role of depressive and negative symptoms in patients with schizophrenia. As such, two specific research questions guide this analysis: (1) What factors are associated with depressive and negative symptoms at Time 1 across four major psychiatric diagnoses (patients with schizophrenia, mood disorder, both schizophrenia and mood disorders, and patients whose primary diagnosis is neither schizophrenia nor mood disorder)? (2) To what extent do depressive and negative symptoms improve over time among individuals with schizophrenia? More specifically, what variables predict an improvement in these symptoms? Methods: The study involved analysis of secondary data from 3269 in-patients from 15 psychiatric facilities in the Province of Ontario, Canada. Patients were assessed using the Resident Assessment Instrument – Mental Health (RAI-MH). Bivariate analyses were performed examining demographic, clinical, social, and other factors as independent variables and depressive and negative symptom scores among each of the four diagnostic groups: schizophrenia, mood disorder, both schizophrenia and mood disorder, and neither schizophrenia nor mood disorder. Logistic regression of depressive and negative symptoms, as dependent variables, were performed on demographic, psychiatric, clinical, social, and other variables, as the independent variables. Results: Variables associated with depressive and negative symptoms did not necessarily predict an improvement of depressive and negative symptoms over time. Findings from logistic regression models showed that statistically significant predictors of improvement in depressive and negative symptoms included the following variables: (1) not having a diagnosis of schizophrenia; (2) insight into one’s condition; (3) fewer number of recent psychiatric admissions (over the last two year period); and (5) being administered both atypical and typical antipsychotic medications. Conclusions: Depressive and negative symptoms are prevalent in schizophrenia and are associated with demographic, psychiatric, and social variables. Depressive and negative symptoms do not share the same pattern across diagnoses, suggesting that these symptoms represent a unique profile within each diagnostic group. Moreover, both atypical and typical antipsychotic medications, in combination, were shown to be more effective at treating depressive and negative symptoms than either typical or atypical medications alone.

Considerable socioeconomic disparities have been identified for smoking and cessation: lower socioeconomic status (SES) groups have higher rates of tobacco use, are less likely to successfully quit, and may also be less likely to intend or attempt to quit. However, results are inconsistent for some quitting-related outcomes, and little is known about how socioeconomic disparities may vary across countries and over time. This study examined the extent to which SES was associated with smoking cessation and related constructs among representative samples of smokers in Canada, the US, the UK, and Australia, using data from the first five waves (2002-2006) of the ITC Four Country Survey (35 532 observations from 16 458 respondents). Generalized estimating equations modeling was used to examine whether education and income were related to intentions to quit (any, and within the next six months), incidence of quit attempts, smoking abstinence (for at least one, six and 12 months), and reduction in daily cigarette consumption by at least half. Potential differences in the associations over time and across countries were also considered. In addition, logistic regression modeling examined associations between education and income, reasons for quitting, and use of cessation assistance, using a cross-sectional sample of the most recent survey wave. Respondents with higher education were more likely to intend to quit, have made a quit attempt, and be abstinent for at least one and six months, and those with higher income were more likely to intend to quit and be abstinent for at least one month. Associations were stable throughout the time period under study. Country differences were observed in quit intentions: UK and US respondents were less likely to intend to quit than Australians and Canadians. Also, UK respondents were least likely to attempt to quit overall, but those that did attempt were more likely to be abstinent for at least one and six months. Socioeconomic and between-country differences were also identified in the cross-sectional analyses of use and access to cessation assistance and reasons for quitting. The results suggest that socioeconomic disparities exist at multiple stages in the path to smoking cessation.

Background: Planning has long been assumed to be an important element of any successful quit smoking strategy. However, recent research findings indicate that unplanned, or spontaneous quit attempts, may lead to more successful and longer cessation periods than planned attempts. This calls into question continued advice to plan and the validity of planning based behaviour change theories. Objectives: To: 1) assess the prevalence of planning; 2) identify the attributes of planners; 3) examine the association of planning and intentions to quit; 4) examine the association of planning and use of quit aids; and 5) examine the impact of planning on smoking cessation outcomes. Methods: Data from the longitudinal Ontario Tobacco Survey (OTS) were examined. Analyses consisted of 418 smokers who made a planned or unplanned quit attempt between their waves 4 and 5 survey response. Descriptive, bivariate, and multivariate analyses were conducted using specialized survey analysis procedures to account for the complex sampling and design features of the OTS. Multiple Imputation (MI) was used to fill in missing data to reduce bias. Results: The prevalence of planning among Ontario smokers who made a quit attempt in the past six months was 70%. Regression models indicated that women, smokers who felt that it would be “very hard to quit”, and those with intentions to quit in the next 6 months had higher odds of making a planned quit attempt. Individuals who made a planned quit attempt had increased odds for using a quit aid compared to those making unplanned attempts. Univariate regression models indicated that planned attempts did not result in higher odds ratios for being abstinent for at least one week and at least one month. Conclusions: The present results suggest that a majority of smokers who made a quit attempt did so with a plan. Planned quit attempts had higher odds use of any quit aid compared with unplanned attempts. Planning was not found to increase the odds of being abstinent compared to unplanned attempts. Although more research is needed, the current study contributes to the dialogue of planning and smoking cessation.

Background: The rehabilitation of older persons is often complicated by increased frailty and medical complexity – these in turn present challenges for the development of health information systems. Objective investigation and comparison of the effectiveness of geriatric rehabilitation services requires information systems that are comprehensive, reliable, valid, and sensitive to clinically relevant changes in older persons. The Functional Independence Measure is widely used in rehabilitation settings – in Canada this is used as the central component of the National Rehabilitation Reporting System of the Canadian Institute of Health Information. An alternative system has been developed by the interRAI consortium. We conducted a literature review to compare the development and measurement properties of these two systems and performed a direct empirical comparison of the operating characteristics and validity of the FIM motor and the ADL items on the PAC in a sample of older adults receiving rehabilitation. Methods: For the first objective english language literature published between 1983 (initial development of the FIM) and 2008 was searched using Medline and CINAHL databases, and the reference lists of retrieved articles. Additionally, attention was paid to the ability of the two systems to address issues particularly relevant to older rehabilitation clients, such as medical complexity, comorbidity, and responsiveness to small but clinically meaningful improvements. For the second objective we used Rasch analysis and responsiveness statistics to investigate and compare the instruments dimensionality, item difficulty, item fit, differential item function, number of response options and ability to detect clinically relevant change. Results: The majority of FIM articles studied inpatient rehabilitation settings; while the majority of interRAI/MDS articles focused on nursing home settings. There is evidence supporting the reliability of both instruments. There were few articles that investigated the construct validity of the interRAI/MDS. The analysis showed that the FIM may be slightly more responsive than the PAC, especially in the MSK patients. However, both scales had similar limitations with regards the large ceiling effect and many unnecessary response options. Conclusions: Additional psychometric research is needed on both the FIM and MDS, especially with regard to their use in different settings and ability to discriminate between subjects with functional higher ability.

Background and Purpose: Quality measurement is an essential, yet, complex component of mental health services that is often limited by a lack of clinically meaningful data across service providers. Understanding how services are organized, delivered, and effective is vital for ensuring and improving health care quality. In quality measurement of mental healthcare, structural indicators are common with fewer process and outcome indicators available. Using data from the RAI - Mental Health (RAI-MH), a comprehensive assessment system mandated for use in Ontario, this dissertation aims to define a set of mental health quality indicators (MHQIs), effectiveness quality indicators (EQIs), and risk adjustment strategy that can be used to evaluate and compare quality at the facility- and regional-levels. Methodology: The MHQIs were developed using a retrospective analysis of two data sets: A pilot sample of 1,056 RAI-MH admission and discharge assessments collected from 7 inpatient mental health units in Ontario and a sample of 30,046 RAI-MH admission and discharge assessments collected from 70 Ontario hospitals as part of the Canadian Institute for Health Information Ontario Mental Health Reporting System. The MHQIs were chosen based on clinically meaningful domains identified by mental health and quality stakeholders, MHQI rates that were consistently above 5% or below 95% among hospitals, and appropriate variation in rates among hospitals in both sets of data. For each MHQI domain, regression modeling using generalized estimating equations was employed to choose risk adjustment variables and logistic or linear regression was used to perform risk adjustment to compare MHQI and EQI rates among hospitals and regions. Results: A set of 27 MHQIs was defined measuring improvement and incidence/failure to improve in the following domains: depressive/psychosis/pain symptoms, cognitive/physical/social functioning, aggressive/ disruptive/violent behaviours, and control procedures. Also, 13 EQIs were defined to identify the magnitude of change in MHQI domains per 7 days between assessments. Regression models using generalized estimating equations identified between 1 and 8 risk adjustment covariates for each MHQI. Risk adjustment using logistic and linear regression resulted in over 50% of hospitals and LHINs changing in rank based on MHQI and EQI scores. Conclusion: This dissertation has developed an evidence-based set of MHQIs and EQIs based on a clinically rich set of data. Since the data is available provincially, the MHQIs and EQIs can be used for hospital based, regional, and public reports on quality of inpatient mental health services. The MHQIs/EQIs can be linked to care planning and funding using the RAI-MH to promote quality improvement and accountability for recipients, providers, managers, governors, and funders of mental health services. Opportunities are also available to extend the use of the MHQIs to community mental health, so that system level evaluations of quality can be developed.

Introduction: The high prevalence of obesity in Aboriginal Canadian youth is a major public health concern. Although little is known about the diets of children on-reserve, it is hypothesized that high intake of low nutrient dense foods has a negative impact on overall diet quality. Objectives: To describe the dietary quality of First Nation students using a Canadian adaptation of the Healthy Eating Index (HEI-C) and determine the relationships between HEI-C and BMI, intake of non-food group (“other”) foods, game consumption, frequency of eating outside the home and remoteness (latitude, °N). Methods: Between 2004-2009, from a 24 h dietary recall data were obtained using a validated web-based food behaviour questionnaire, from students in grades 6,7 & 8 from the First Nations communities of Georgina Island, Christian Island, Fort Albany, Attawapiskat, Peawanuck, Moose Factory, Kashechewan and Ouje-Bougoumou (Quebec). HEI-C was assessed as good (81-100), needs improvement (51-80) or poor (0-50). Results: Mean community HEI-C scores ranged from 57.38-70.04, with differences by community (p=0.027) and season (p=0.007); more northerly communities seemed to have lower HEI-C scores and fall seemed to have higher HEI-C scores compared to winter and spring. A non-significant negative correlation between BMI and HEI-C was observed (r=-0.107, NS). As percent energy from ‘other foods’ increased, the HEI-C score tended to decrease. Game consumption appeared to contribute to higher HEI-C scores. Eating outside the home did not seem to affect HEI-C. Conclusions: Poor diet quality in remote on-reserve youth populations is a concern. Lack of consistent access to healthy foods may have a negative impact on diet quality. Programs that help improve the provision of healthy foods, decrease the intake of “other foods” and that emphasize game may help to improve diet quality.

Cigarette packaging is the most prominent form of tobacco promotion in Canada. Tobacco companies are increasingly selling cigarettes in innovative packaging, including the use of slim and super-slim “lipstick” sizes that are primarily marketed towards females. Australia is currently the only country that regulates the shape and size of cigarette packaging. The current study examined the relative importance of five cigarette packaging attributes—pack shape (e.g., “slims”) , brand, plain packaging, warning label size, and price—on perceptions of product taste, harm, and interest in trying, among young women in Canada. A discrete choice experiment was conducted online with smoking (n=211) and non-smoking (n=292) females, aged 16 to 24, recruited from a commercial sample. Respondents were shown 8 choice sets, each containing four packs displaying different combinations of the attributes: pack structure (slim, lipstick, booklet, standard); brand ( ‘Vogue’, ‘du Maurier’); branding (branded, plain); warning label size (50%, 75%); and price ($8.45, $10.45). For each choice set, respondents chose the brand that they: 1) would rather try, 2) would taste better, 3) would be less harmful, or “none”. For each outcome, the attributes’ impact on choices was analyzed using a multinomial logit model, and the relative importance (RI) of each attribute was calculated. The results showed that pack structure significantly influenced interest in trying (RI = 16%) and perceptions of taste (RI = 8%), whereas perceptions of harm were driven by pack structure (RI = 46%). Branding was the most important contributor to trial intent decisions (RI = 39%) and perceptions of taste (RI = 48%). Interest in trying among females significantly increased for booklet (p < 0.0001) packs compared to the traditional design. As well, females were significantly more interested in trying branded packs, female oriented ‘Vogue’ brand, and a 75% warning label size (p < 0.0001, for all). In terms of taste related perceptions, females believed that slim (p=0.02) and booklet packs (p=0.006) were significantly better tasting than traditional designs. Similarly, branded packs (p < 0.0001), ‘Vogue’ brand (p < 0.0001), 75% warning (p < 0.0001), and higher priced packs (p=0.04) significantly increased perceptions of taste among females. Among young females, booklet (p=0.03), lipstick (p < 0.0001) and slim (p < 0.0001) pack sizes were perceived as significantly less harmful compared to traditional designs. As well, women believed branded packs, ‘Vogue’ brand, and more expensive brands would be significantly less harmful (p < 0.0001, for all). Given that the discrete choice design did not include all pack profiles that could be generated with attribute-level combinations of branding, brand, and warning labels, and in particular, due to the absence of “branded Vogue packs with smaller warnings”, the findings on warning label size should be interpreted with caution. Overall, the findings suggest that “plain” packaging and prohibiting variations in pack shape and size may decrease interest in trying and reduce false perceptions of reduced product harm among young females.