Metformin has been associated with a reduction in breast cancer incidence, however its effect on mortality following cancer has not been adequately examined. The purpose of this study was to evaluate the impact of metformin therapy on mortality in women with breast cancer. Using Ontario health databases, this retrospective cohort examined the impact of metformin on mortality among women aged 66 years or older with diabetes and breast cancer. After a mean follow-up of 4.5 years, there was no association between cumulative metformin use and either all-cause or breast cancer-specific mortality (HR 0.97, 95% CI 0.92-1.07; HR 0.91, 95% CI 0.81-1.03 respectively per additional year of cumulative metformin use). Though metformin was not associated with a reduction in mortality in our study of older women with breast cancer, there is still a need to examine whether metformin has an effect on mortality in other breast cancer populations.

Metformin has been associated with a reduction in breast cancer incidence, however its effect on mortality following cancer has not been adequately examined. The purpose of this study was to evaluate the impact of metformin therapy on mortality in women with breast cancer. Using Ontario health databases, this retrospective cohort examined the impact of metformin on mortality among women aged 66 years or older with diabetes and breast cancer. After a mean follow-up of 4.5 years, there was no association between cumulative metformin use and either all-cause or breast cancer-specific mortality (HR 0.97, 95% CI 0.92-1.07; HR 0.91, 95% CI 0.81-1.03 respectively per additional year of cumulative metformin use). Though metformin was not associated with a reduction in mortality in our study of older women with breast cancer, there is still a need to examine whether metformin has an effect on mortality in other breast cancer populations.

Background: The efficacy of interventions to prevent in-hospital transmission of extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-E) is poorly defined, particularly for admission screening. Methods: Variability in ESBL-E infection control practices was evaluated with a survey of 15 hospitals. All ESBL-E positive clinical and screening specimens at 12 hospitals (6 screening and 6 non-screening) from 2005-2009 were included and defined as hospital-onset or community-onset using standard definitions. ESBL-E incidence and susceptibility were studied. Screening efficacy was evaluated with a negative binomial model, adjusting for study year and incidence of community-onset cases. Results: Diverse practices in infection control for ESBL-E were found with 53.3% of hospitals utilizing admission screening. Overall incidence and hospital-onset cases increased 4-fold and 2-fold, respectively. Fluoroquinolone susceptibility for E. coli (12.8%) and K. pneumoniae (9.0%) was low. Hospital-onset cases were 49.1% lower in screening compared to non-screening hospitals (p<0.001). Conclusion: Admission screening can reduce the incidence of hospital-onset ESBL-E cases.

Background: The efficacy of interventions to prevent in-hospital transmission of extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-E) is poorly defined, particularly for admission screening. Methods: Variability in ESBL-E infection control practices was evaluated with a survey of 15 hospitals. All ESBL-E positive clinical and screening specimens at 12 hospitals (6 screening and 6 non-screening) from 2005-2009 were included and defined as hospital-onset or community-onset using standard definitions. ESBL-E incidence and susceptibility were studied. Screening efficacy was evaluated with a negative binomial model, adjusting for study year and incidence of community-onset cases. Results: Diverse practices in infection control for ESBL-E were found with 53.3% of hospitals utilizing admission screening. Overall incidence and hospital-onset cases increased 4-fold and 2-fold, respectively. Fluoroquinolone susceptibility for E. coli (12.8%) and K. pneumoniae (9.0%) was low. Hospital-onset cases were 49.1% lower in screening compared to non-screening hospitals (p<0.001). Conclusion: Admission screening can reduce the incidence of hospital-onset ESBL-E cases.

Alcohol is a recognized cause of over 60 injuries and diseases and is consistently in the top 5 most important risk factors for global burden of disease. It is important to be able to measure how drinking alcohol affects our health, and how our risk of getting injured or acquiring diseases caused by alcohol is dependent on how much alcohol we drink. This type of information allows us to make personal choices about our health and is an integral piece of public health evidence to inform how alcohol policy is informed, implemented, or monitored. This analysis will, for the first time, model the effects of alcohol for injury outcomes over the entire drinking lifetime for men and women separately using a new method that aims to improve upon existing calculations by accounting for different patterns of drinking – both acute consumption and average daily drinking. In both cases, both the amount consumed and the number of times it is consumed is taken into account. Within acute consumption, the number of occasions and the amount ii consumed at each occasion was counted. What’s more, for the first time in this field, a lifetime approach was adopted – risks will no longer be seen as discrete, individual events that occur independently of each other. In this study, risks are combined much like other exposures to environmental substances or contaminants – in a cumulative manner over a lifetime of drinking. The method combines data sources from experimental data, from meta-analyses, Canadian mortality and hospital data, and survey data, making this a rich, yet complicated analysis. Its products were dose-response risk curves for each injury outcome, by sex, and age group, and alcohol-attributable fractions and their variance estimation for mortality and morbidity for injury. This study has important implications for forming and planning health policy, represents advancements in absolute risk calculation, and will result in important consumer-level information that will enable development of limits around healthy drinking.

Background: Survivors of severe acute kidney injury remain at high risk of death well-after apparent recovery from the initial event. Methods: We conducted a cohort study of hospitalized adults in Ontario from 1996 to 2008 with acute kidney injury who received temporary dialysis and survived for 90 days following discharge independent from dialysis. The exposure was nephrology follow-up. We used propensity scores to match individuals with early nephrology follow-up to those without. The primary outcome was time to mortality. Results : We identified 3877 patients with acute kidney injury who met the eligibility criteria. A total of 1583 patients had nephrology follow. The incidence of all-cause mortality was lower in those with early nephrology follow-up as compared to those without early follow-up (8.4 vs. 10.6 per 100 person-years, HR 0.76 (95% CI 0.62-0.93)). Conclusions: Nephrology follow-up after hospitalization with acute kidney injury and temporary dialysis was associated with improved survival.

Background: Survivors of severe acute kidney injury remain at high risk of death well-after apparent recovery from the initial event. Methods: We conducted a cohort study of hospitalized adults in Ontario from 1996 to 2008 with acute kidney injury who received temporary dialysis and survived for 90 days following discharge independent from dialysis. The exposure was nephrology follow-up. We used propensity scores to match individuals with early nephrology follow-up to those without. The primary outcome was time to mortality. Results : We identified 3877 patients with acute kidney injury who met the eligibility criteria. A total of 1583 patients had nephrology follow. The incidence of all-cause mortality was lower in those with early nephrology follow-up as compared to those without early follow-up (8.4 vs. 10.6 per 100 person-years, HR 0.76 (95% CI 0.62-0.93)). Conclusions: Nephrology follow-up after hospitalization with acute kidney injury and temporary dialysis was associated with improved survival.

Antihypertensive drugs can cause hypotension in the elderly and such an effect may lead to fall injuries. This thesis examined the association between antihypertensive drugs and hip fracture risk among elderly patients during the initiation of monotherapy. This population-based self-controlled case series study used healthcare administrative databases to identify Ontario residents aged ≥ 66 years with a first prescription for a thiazide diuretic, angiotension II converting-enzyme inhibitor, angiotensin II receptor antagonist, calcium channel blocker or beta-adrenergic blocker. A cohort of newly treated hypertensive elderly was then linked to the occurrence of hip fractures from April 1, 2000 to March 31, 2009. We found that hypertensive elderly initiated on an antihypertensive drug had a 43% increased risk of having a hip fracture during the first 45 days of treatment, IRR 1.43 (95% CI 1.19-1.72). Initiating antihypertensive drugs in community-dwelling elderly should be approached with caution due to increased fracture risk.

Inflammation is believed to play a pivotal role in the development of cancer and in particular lung cancer through both intrinsic and extrinsic pathways. This thesis aimed through 4 manuscripts to determine the role of lung-specific inflammation-related exposures as well as genetic variants (Single Nucleotide Polymorphism (SNPs)) in genes related to inflammation pathways in lung cancer risk. Risk factors of particular interest were lung diseases including emphysema, chronic bronchitis, pneumonia and tuberculosis. Data were collected from a Toronto-based case-control study and combined with data from the International Lung Cancer Consortium (17 studies for the non-genetic analyses and 6 studies for the genetic analyses). For non-genetic data, methods included, but were not limited to, unconditional logistic regression within single studies, Cox proportional hazards regression, meta-analytic techniques and pooled analyses techniques. For genetic data, additive models of association across 7,650 SNPs (selected based on role in inflammation) were pooled across studies and hierarchical modeling iv (HM) was used to incorporate prior functional information from various sources concerning the SNPs of interest. Within the Toronto study (manuscript 1) associations with increased lung cancer risk were observed for occupational exposures, previous lung diseases and a family history of cancer. The meta-analysis (manuscript 2) and pooled analysis (manuscript 3) showed relatively consistent associations between previous lung diseases and lung cancer risk across histology, gender and smoking categories. Results persisted when examining lung disease diagnoses made >10 or >20 years before cancer diagnosis. Our pathway-based analysis of inflammation-related variants and lung cancer risk using HM (manuscript 4) showed that HM is applicable to SNP analysis using pooled designs where heterogeneity can be incorporated into SNP priors. After HM a previously observed variant (rs2736100) and novel variant (rs2741354) were observed to be associated with lung cancer risk at corrected levels and replicated in an independent population. Taken together these results provide further evidence of both intrinsic and extrinsic factors affecting risk of lung cancer through inflammation.

Antihypertensive drugs can cause hypotension in the elderly and such an effect may lead to fall injuries. This thesis examined the association between antihypertensive drugs and hip fracture risk among elderly patients during the initiation of monotherapy. This population-based self-controlled case series study used healthcare administrative databases to identify Ontario residents aged ≥ 66 years with a first prescription for a thiazide diuretic, angiotension II converting-enzyme inhibitor, angiotensin II receptor antagonist, calcium channel blocker or beta-adrenergic blocker. A cohort of newly treated hypertensive elderly was then linked to the occurrence of hip fractures from April 1, 2000 to March 31, 2009. We found that hypertensive elderly initiated on an antihypertensive drug had a 43% increased risk of having a hip fracture during the first 45 days of treatment, IRR 1.43 (95% CI 1.19-1.72). Initiating antihypertensive drugs in community-dwelling elderly should be approached with caution due to increased fracture risk.