Introduction: Long term care (LTC) residents require complete or extensive support, including 24-hour nursing and personal care. LTC residents contribute a greater number of emergency department (ED) visits when compared to community-dwelling older adults. Little is known about which resident-level characteristics at admission are predictive of LTC resident transfer to the ED. The objective of this thesis was to identify which admission characteristics are associated with ED transfers in Ontario, Canada. Methodology: I conducted a population-level retrospective cohort study using the Resident Assessment Instrument Minimum Data Set Version 2.0 (RAI-MDS). The cohort included 56,433 LTC resident admission assessments from January 1, 2017, to December 31, 2018. Logistic regression and 10-fold cross-validation were used to identify adjusted associations between characteristics routinely collected during LTC admission assessment and ED transfers. Model performance was assessed using the area under the receiver operating characteristics curve (AUC). Outcomes of interest included any ED use, potentially preventable, and low acuity ED transfers. Results: A recent change in medical orders, previous ED visitation, female sex, the presence of an indwelling catheter, and the need for oxygen therapy were informative predictors for any, potentially preventable, and low acuity ED transfers. Deterioration in cognitive status and change in behavior was influential to any ED transfers only. Urinary tract infections, pneumonia, indictors of delirium, and change in mood are unique to potentially preventable ED transfers, and antibiotic resistance is unique to low acuity ED transfers. Similar discrimination was reached for any ED use (AUC = 0.630), potentially preventable transfers (AUC = 0.659), and low acuity transfers (AUC = 0.645). Conclusion: The factors associated with ED transfers may be modifiable, and closer attention to these factors may help reduce ED transfers. Although the discriminability of the models was poor, advanced knowledge of informative characteristics can support upstream decision-making for clinicians. Future studies are required to validate these findings, derive risk scales, and demonstrate the utility of this model in health service planning. / Thesis / Master of Science (MSc) / Long term care (LTC) provides residents with 24-hour nursing and personal care. When the care or clinical needs of the resident cannot be met in the LTC facility, they may be transferred to the Emergency Department (ED). However, the ED’s are poorly situated to manage the distinct needs of older adults, given the sole focus on medical acuity rather than geriatric complexity. Unwarranted ED transfers are burdensome for LTC residents and increase their risk for adverse health events, such as nosocomial infections, delirium, and injuries. Understanding characteristics associated with ED transfers can help identify which residents may be at a risk of an ED transfer. The objective of this thesis was to identify which LTC resident characteristics at admission are associated with ED transfers in Ontario, Canada. A recent change in medical orders, previous ED visitation, female sex, the presence of an indwelling catheter, and the need for oxygen therapy were informative predictors for ED transfers.

Little is known about the receipt of cancer screening for new primary cancers among Canadian cancer survivors. The objectives of this thesis are to i) synthesize evidence comparing receipt cancer screening between cancer survivors and non-cancer controls; and ii) analyze breast and cervical cancer screening receipt among Nova Scotian colorectal cancer (CRC) survivors. This thesis consists of a systematic review and meta-analysis, and a population-based cohort study of Nova Scotian CRC survivors. We found that while cancer survivors were more likely to receive cancer screening than the general population, a significant proportion of cancer survivors were not screened. We observed significant heterogeneity between studies, most of which remained unexplained after subgroup and sensitivity analyses. 30.1% and 47.9% of Nova Scotian CRC survivors never received a breast and cervical cancer screen after their CRC diagnosis. Receipt of pre-CRC diagnosis screening was strongly predictive of receiving screening post-diagnosis.

京都大学 / 新制・課程博士 / 博士(医学) / 甲第23067号 / 医博第4694号 / 新制||医||1049(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 今中 雄一, 教授 武藤 学, 教授 妹尾 浩 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・論文博士 / 博士(医学) / 乙第13220号 / 論医博第2167号 / 新制||医||1033(附属図書館) / (主査)教授 高橋 淳, 教授 宮本 享, 教授 伊佐 正 / 学位規則第4条第2項該当 / Doctor of Medical Science / Kyoto University / DFAM

Purpose: To determine the timing of first ambulation during hospitalization among patients admitted for acute myocardial infarction (AMI) and to identify the predictors of first ambulation. Methods: This retrospective cohort study included 500 AMI patients admitted during 2004 to one of three hospitals that form the Hamilton Health Sciences Corporation in Hamilton, Ontario, Canada. The patients were randomly selected from a total of 1,014 charts from the Hamilton Health Sciences Computerized Health Records (SOVERA). Using a chart abstraction tool, the following data were collected from each patient's chart: demographic information, past medical history, treatment, complications, and patterns of ambulation while in hospital. The primary outcome was first ambulation, defined as the first time patients walked during their hospital stay. Secondary outcomes included heart rate at discharge and mortality during hospitalization. The relationship between patient and care-related factors and the time of first ambulation after AMI was explored through a time to event analysis using Cox regression; the associations were expressed as hazard ratios. The fit for the proportional hazard model was assessed and a stratified proportional hazard model was performed for age. Results: Of the 500 charts, 60 were excluded. Of the 440 patients who were included in the final analysis, 340 (77.3%) walked during hospitalization. One hundred fifteen (26.1 %) walked during the first 48 hours (early walking), 98 (22.3%) walked between 49-96 hours (intermediate walking), and 127 (28.9%) walked after 96 hours (late walking). A total of 100 patients (22.7%) were categorized as non-walkers. Factors that emerged in the survival analysis that were positively associated with early ambulation after AMI and that proved the proportionality on the assessment of the fit of the model were: having a family history of cardiovascular disease (HR 1.33; 95% Cl 1.00, 1.44; p=0.05), receiving thrombolysis (HR 1.47; 95% Cl 1.11, 1.49; p=0.007), receiving nitroglycerin (HR 1.51; 95% Cl 1.19, 1.93; p<0.001 ), and taking calcium channel blockers (HR 1.58; 95% Cl 1.22, 2.05; p<0.001 ). Factors that were negatively associated with early ambulation after AMI were age >59 years (HR 0.98; 95% Cl 0.97, 0.99; p<0.001 ), having an arrhythmia in-hospital (HR 0.48; 95% Cl 0.22, 0.94; p=0.04), taking inotropic drugs (HR 0.72; 95% Cl 0.53, 0.98; p<0.001 ), and undergoing coronary artery bypass surgery (HR 0.51; 95% Cl 0.33, 0.78; p=0.002). Conclusion: There is variability in the timing of first ambulation among patients hospitalized with an AMI. Furthermore, those who walked early were more likely to have a family history of cardiovascular disease, have received thrombolysis, and be taking nitroglycerin or calcium channel blockers. Those least likely to walk early were older (>59 years), were more likely to have had an arrhythmia inhospital, to be taking inotropic drugs, and to have undergone coronary artery bypass surgery. / Thesis / Doctor of Philosophy (PhD)