Introduction : Les statines ont prouvé leur efficacité dans le traitement des dyslipidémies. Cependant, ces molécules sont associées à des effets secondaires d’ordre musculaire. Puisque ces effets peuvent avoir des conséquences graves sur la vie des patients en plus d’être possiblement à l’origine de la non-observance d’une proportion importante des patients recevant une statine, un outil pharmacogénomique qui permettrait d’identifier a priori les patients susceptibles de développer des effets secondaires musculaires induits par une statine (ESMIS) serait très utile. L’objectif de la présente étude était donc de déterminer la valeur monétaire d’un tel type d’outil étant donné que cet aspect représenterait une composante importante pour sa commercialisation et son implantation dans la pratique médicale courante. Méthode : Une première simulation fut effectuée à l’aide de la méthode de Markov, mais celle-ci ne permettait pas de tenir compte de tous les éléments désirés. C’est pourquoi la méthode de simulation d'évènements discrets fut utilisée pour étudier une population de 100 000 patients hypothétiques nouvellement initiés sur une statine. Cette population virtuelle a été dupliquée pour obtenir deux cohortes de patients identiques. Une cohorte recevait le test et un traitement approprié alors que l'autre cohorte recevait le traitement standard actuel—i.e., une statine. Le modèle de simulation a permis de faire évoluer les deux cohortes sur une période de 15 ans en tenant compte du risque de maladies cardio-vasculaires (MCV) fatal ou non-fatal, d'ESMIS et de mortalité provenant d’une autre cause que d’une MCV. Les conséquences encourues (MCV, ESMIS, mortalité) par ces deux populations et les coûts associés furent ensuite comparés. Finalement, l’expérience fut répétée à 25 reprises pour évaluer la stabilité des résultats et diverses analyses de sensibilité ont été effectuées. Résultats : La différence moyenne des coûts en traitement des MCV et des ESMIS, en perte de capital humain et en médicament était de 28,89 $ entre les deux cohortes pour la durée totale de l’expérimentation (15 ans). Les coûts étant plus élevés chez celle qui n’était pas soumise au test. Toutefois, l’écart-type à la moyenne était considérable (416,22 $) remettant en question la validité de l’estimation monétaire du test pharmacogénomique. De plus, cette valeur était fortement influencée par la proportion de patients prédisposés aux ESMIS, par l’efficacité et le coût des agents hypolipidémiants alternatifs ainsi que par les coûts des traitements des ESMIS et de la valeur attribuée à un mois de vie supplémentaire. Conclusion : Ces résultats suggèrent qu’un test de prédisposition génétique aux ESMIS aurait une valeur d’environ 30 $ chez des patients s’apprêtant à commencer un traitement à base de statine. Toutefois, l’incertitude entourant la valeur obtenue est très importante et plusieurs variables dont les données réelles ne sont pas disponibles dans la littérature ont une influence importante sur la valeur. La valeur réelle de cet outil génétique ne pourra donc être déterminée seulement lorsque le modèle sera mis à jour avec des données plus précises sur la prévalence des ESMIS et leur impact sur l’observance au traitement puis analysé avec un plus grand nombre de patients. / Introduction: Statins have proven their efficacy in the treatment of dyslipidemias. However, these molecules are associated with muscular side effects. Since these side effects may have adverse consequences on patients’ daily life and have an important role in the discontinuation of statin therapy in a large proportion of patients, it would be useful to develop a pharmacogenomic test that identifies a priori the individuals who are likely to develop statin-related muscular side effects (SRMSE). The objective of the present study was to determine of the monetary value of such a type of test considering that this aspect would represent an important component of its marketing and implementation into medical practice. Method: The first simulation was carried out using the method of Markov, but this one did not allow consider all the desired elements. This is why the discrete events simulation method have been used to study a population of 100 000 hypothetical patients newly initiated on a statin. This virtual population was duplicated to have two identical cohorts of patients. The first one was administered the test and a suitable treatment while the second received the current standard treatment—that is, a statin. The model allowed the two cohorts to evolve over a period of 15 years taking into account the risks of fatal and non fatal cardiovascular diseases (CVD), SRMSE and mortality from other causes than CVD. The consequences (CVD, SRMSE, death) incurred in these two populations and the associated costs were then compared. Finally, the process was repeated 25 times to assess the stability of the results and various sensitivity analyses were carried out. Results: The mean difference of CVD and SRMSE treatments, lost of human capital and drugs costs between the two cohorts was of 28.89 $, these costs being higher in the cohort who was not administered the test. However, the standard deviation with the average was considerable (416.22 $) calling in question the validity of the monetary estimate of the test pharmacogenomic.This difference varied a lot as a function of the proportion of patients being predisposed to SRMSE, the efficacy and the costs of the alternative treatments, the SRMSE cost, and the value assigned to one additional month of life. Conclusion: The results suggest that a test of genetic predisposition to SRMSE would have a value around 30 $ in patients who start a statin treatment. However, uncertainty surrounding the value obtained is very important and several variables for which the real data are not available in the literature have an important influence on the value. The real value of this genetic tool could thus be given only when the model is updated with more precise data on the prevalence of the ESMIS and their impact on the observance at the treatment and then analyzed with a higher number of patients.

Comparisons of raw span (i.e., number of staff who report directly to a manager) within and across organizations can misrepresent managerial capacity to support staff because managers may not allocate the same amount of time to staff contact. The purpose was to examine the influence of alternative measures of managerial span on nurse satisfaction with manager’s supervision and on multidisciplinary teamwork. The alternative measures were (a) raw span as a measure of reporting structure and (b) time in staff contact as a measure of closeness of contact by the manager. The main effects of the alternative measures, leadership, hours of operation, and other covariates on outcomes were examined. The interaction effects of the alternative measures with leadership and hours of operation were investigated. The study framework was based on Open System Theory and the boundary spanning functions of managers. A descriptive, correlational design was used to collect survey and administrative data from employees, managers, and organizations. Managerial time allocation data were collected through self-logging and validated through observation. Acute care hospitals were selected through purposive sampling. For supervision satisfaction, the final sample size was 31 first-line managers and 558 nurses. For teamwork, the final sample size was 30 first-line managers and 754 staff. The Leadership Practices Inventory, the Satisfaction with my Supervisor Scale, and the Relational Coordination Scale were used. Hierarchical linear modeling was the main type of analysis conducted. Raw span interacted with leadership and hours of operation to explain supervision satisfaction. Teamwork was explained by leadership, clinical support roles, hours of operation, total areas, and non-direct reports, but not by raw span or time in staff contact. Large acute care hospitals can improve satisfaction with supervision and teamwork by modifying first-line management positions.

The purpose of this study was to explore the relationship between the formalization of inter-organizational partnerships and collaboration among staff members working together across primary care, mental health and addition organizations to provide services to adults with complex mental health and addiction needs. Phase I of the study provided an environmental scan of existing partnerships among Family Health Teams (FHTs) and Community Health Centres (CHCs), and the Mental Health and/or Addiction (MHA) organizations they partner with, in the province of Ontario (Canada). Phase II explored the relationship between formalization and a) administrative collaboration and b) and service delivery collaboration. The hypotheses proposed that staff members who are part of formalized partnerships would report higher levels of collaboration. Phase III explored how formal and informal partnerships and collaboration are experienced by the administrative and service provider staff members who work across FHTs, CHCs and MHAs organizations. Using a mixed methods approach, data were collected using electronic surveys and telephone interviews. The results of Phase I indicated that FHTs and CHCs in Ontario have between 1-3 partnerships with MHA organizations. Most are informal partnerships, have existed for less than 5 years, and most staff members (partners) interact on a monthly basis. The quantitative results of Phase II showed no significant relationship between formalization and either form of collaboration. The qualitative findings from Phase III provide two key contributions. First, the results of the interviews may help explain why collaboration was not higher in formalized partnerships, as demonstrated by the range of advantages and disadvantages experienced by administrators and service providers in both formal and informal partnerships. Second, the findings illuminate factors related to the process of creating and/or formalizing partnerships, suggesting that there may be other factors that mediate or have a direct impact on the relationship between formalization and collaboration. By bringing together the study findings, the study addresses a gap in the literature by proposing a pathway through which formalization may be associated with collaboration. The results of the study provide opportunities for future research to help improve the quality and accessibility of services to adults with complex mental health and addiction needs.

The objective of this study was to explore inpatients’ opinions on their hospital paper medical records after they had incidental access to them. One hundred inpatients in the C.T. department at St. Michael's Hospital were surveyed: 65 patients who read their records and 35 who did not. Overall, 75.4% of readers found their records easy to understand, and most found their records correct, complete and did not find anything unexpected or distressing. Seventy-nine percent of all respondents would trust the hospital, approximately half would trust Google Health or Microsoft Healthvault and 5.6% would trust Facebook to provide online medical records. Being female, under 60 years and having a higher education predicted readership. Younger patients were also more likely to think that accessing their records would help decrease errors. Patients with higher education were more likely to find their records useful and trusted the hospital to provide online medical records.

The objective of this study was to explore inpatients’ opinions on their hospital paper medical records after they had incidental access to them. One hundred inpatients in the C.T. department at St. Michael's Hospital were surveyed: 65 patients who read their records and 35 who did not. Overall, 75.4% of readers found their records easy to understand, and most found their records correct, complete and did not find anything unexpected or distressing. Seventy-nine percent of all respondents would trust the hospital, approximately half would trust Google Health or Microsoft Healthvault and 5.6% would trust Facebook to provide online medical records. Being female, under 60 years and having a higher education predicted readership. Younger patients were also more likely to think that accessing their records would help decrease errors. Patients with higher education were more likely to find their records useful and trusted the hospital to provide online medical records.

Patient safety experts identify changes in culture as critical to creating safer care (Flin, 2007; Leape, 1994; Reason, 1997; Vincent, Taylor-Adams & Stanhope, 1998). Yet there is limited understanding of how to best study, evaluate and make changes to patient safety culture. The literature on organizational culture, safety sciences and health services research suggests varying perspectives on studying culture and an evolving approach to creating tools to measure culture change. This thesis reports two projects. The first project used the Manchester Patient Safety Culture Assessment Tool, the Modified Stanford Instrument, and qualitative interviews to examine whether safety culture profiles varied by research method and instrument used to assess culture. Comparative assessment of the results suggests that while the quantitative measurement tools provide a high level organizational summary of safety issues, the qualitative interviews provide a more fine-grained understanding of the contextual and local features of the culture. The second research project used a multiple case study design to understand what hospitals have learned from trying to improve patient safety culture. Interviews in three organizations were used to determine how these organizations shifted their cultures. Although each organization had different experiences and used varying methods, they all created culture change through the simultaneous implementation of practice, policies and strategic framing of patient safety culture concepts in their everyday work. The third research paper examined how leaders measured changes in patient safety culture. Both leaders and front line workers look to both process measures (e.g., talking about safety and encouraging patient safety activities) together with outcome measures (e.g., adverse events, infection rates, and culture survey results) to evaluate their success in culture change. Overall this dissertation deepens our knowledge of how methods influence our assessment of patient safety culture and how leaders influence culture change. Future research needs to assess in more detail the roles of leaders and middle managers to understand how these individuals are able to reconcile the practice environment challenges while continuing to create a culture of patient safety.

Comparisons of raw span (i.e., number of staff who report directly to a manager) within and across organizations can misrepresent managerial capacity to support staff because managers may not allocate the same amount of time to staff contact. The purpose was to examine the influence of alternative measures of managerial span on nurse satisfaction with manager’s supervision and on multidisciplinary teamwork. The alternative measures were (a) raw span as a measure of reporting structure and (b) time in staff contact as a measure of closeness of contact by the manager. The main effects of the alternative measures, leadership, hours of operation, and other covariates on outcomes were examined. The interaction effects of the alternative measures with leadership and hours of operation were investigated. The study framework was based on Open System Theory and the boundary spanning functions of managers. A descriptive, correlational design was used to collect survey and administrative data from employees, managers, and organizations. Managerial time allocation data were collected through self-logging and validated through observation. Acute care hospitals were selected through purposive sampling. For supervision satisfaction, the final sample size was 31 first-line managers and 558 nurses. For teamwork, the final sample size was 30 first-line managers and 754 staff. The Leadership Practices Inventory, the Satisfaction with my Supervisor Scale, and the Relational Coordination Scale were used. Hierarchical linear modeling was the main type of analysis conducted. Raw span interacted with leadership and hours of operation to explain supervision satisfaction. Teamwork was explained by leadership, clinical support roles, hours of operation, total areas, and non-direct reports, but not by raw span or time in staff contact. Large acute care hospitals can improve satisfaction with supervision and teamwork by modifying first-line management positions.

Patient safety experts identify changes in culture as critical to creating safer care (Flin, 2007; Leape, 1994; Reason, 1997; Vincent, Taylor-Adams & Stanhope, 1998). Yet there is limited understanding of how to best study, evaluate and make changes to patient safety culture. The literature on organizational culture, safety sciences and health services research suggests varying perspectives on studying culture and an evolving approach to creating tools to measure culture change. This thesis reports two projects. The first project used the Manchester Patient Safety Culture Assessment Tool, the Modified Stanford Instrument, and qualitative interviews to examine whether safety culture profiles varied by research method and instrument used to assess culture. Comparative assessment of the results suggests that while the quantitative measurement tools provide a high level organizational summary of safety issues, the qualitative interviews provide a more fine-grained understanding of the contextual and local features of the culture. The second research project used a multiple case study design to understand what hospitals have learned from trying to improve patient safety culture. Interviews in three organizations were used to determine how these organizations shifted their cultures. Although each organization had different experiences and used varying methods, they all created culture change through the simultaneous implementation of practice, policies and strategic framing of patient safety culture concepts in their everyday work. The third research paper examined how leaders measured changes in patient safety culture. Both leaders and front line workers look to both process measures (e.g., talking about safety and encouraging patient safety activities) together with outcome measures (e.g., adverse events, infection rates, and culture survey results) to evaluate their success in culture change. Overall this dissertation deepens our knowledge of how methods influence our assessment of patient safety culture and how leaders influence culture change. Future research needs to assess in more detail the roles of leaders and middle managers to understand how these individuals are able to reconcile the practice environment challenges while continuing to create a culture of patient safety.

In emergency medical transport, “time to definite care” is very important. Emergency medical services and transport medicine agencies have several possible vehicle options for interfacility transfers. Use of a land vehicle, helicopter, or fixed wing aircraft will be dependent on patient condition, distance between sending and receiving hospitals, crew configuration and capabilities, and other factors such as weather and road conditions. This thesis lays out the complex process of patient transfers and highlights the challenges in decision making under time pressure; it then describes the behaviour of human operators in estimating time to definite care. To support the operators in choosing a transportation mode, a decision support tool was built, which provides relevant time estimates for interfacility transfers based on historical dispatch and call data. The goal is to enable operators to make evidence-based decisions on vehicle allocation. A prototype interface was generated and was evaluated through a usability study.

In Canada and internationally, increases in Home Care (HC) services for the elderly have been a policy priority in recent decades. HC services include Home Health Care (HHC) and Homemaking/Personal Support (HM). The primary objectives of this study were to explore the interrelationship among publicly funded, privately funded, and informal HC services in terms of potential for substitution, and between publicly funded HHC and HM services; and the determinants of the receipt of each type of HC services. Stabile, Laporte, and Coyte’s family home care decision model (2006) was extended, to develop an understanding of the demand for HHC and HM services separately and to include different household arrangements. The consequential hypotheses were tested in two empirical studies. Individual panel data for those aged 65 and over were derived from 8 biannual waves of the Canadian National Population Health Survey (1994-95 to 2008-09). A Panel Two-Stage Residual Inclusion method was used to estimate the likelihood of the receipt of HC services, adjusting for socio-demographic, health status, disability, dependence on help with Activities of Daily Living (ADLs), and regional characteristics. The results showed that receipt of publicly funded HM is complementary with receipt of publicly funded HHC services after adjusting for functional and health status. Receipt of publicly funded and privately funded HM services did not show an effect on each other. Receipt of publicly funded HM did not affect the receipt of informal HM services. The availability of informal care from a partner or other adult sharing the household reduced the likelihood of publicly funded HM receipt. Age, dependence on help with ADLs, health status and income are determinants of the propensity to receive publicly funded HHC and HM services as well as privately funded and informal HM. Findings in this study suggest that changes in the availability of publicly funded HC services may not greatly affect the provision of informal care in Canada. The complementary effect between publicly funded HHC and HM services and the income effect in the receipt of publicly and privately funded HC services may raise concerns about equitable access to HC services in Canadian jurisdictions.