Development of population-level colon cancer pathway concordance measures and association with survival.

Clinical cancer pathways help standardize healthcare delivery to optimize patient outcomes and health system costs. However, population-level measurement of concordance between standardized pathways and actual care received is lacking. Two measures of pathway concordance were developed for a simplified colon cancer pathway map for Stage II-III colon cancer patients in Ontario, Canada: a cumulative count of concordant events (CCCE) and the Levenshtein algorithm. Associations of concordance with patient survival were estimated using Cox proportional hazards models adjusted for patient characteristics and time-dependent cancer-related activities. Models were compared and the impact of including concordance scores was quantified using the likelihood ratio chi-squared test. The ability of the measures to discriminate between survivors and decedents was compared using the C-index. Normalized concordance scores were significantly associated with patient survival in models for cancer stage-a 10% increase in concordance for Stage II patients resulted in a CCCE score adjusted hazard ratio (aHR) of death of 0.93, 95% CI 0.88-0.98 and a Levenshtein score aHR of 0.64, 95% CI 0.60-0.67. A similar relationship was found for Stage III patients-a 10% increase in concordance resulted in a CCCE aHR of 0.85, 95% CI 0.81-0.88 and a Levenshtein aHR of 0.78, 95% CI, 0.74-0.81. Pathway concordance can be used as a tool for health systems to monitor deviations from established clinical pathways. The Levenshtein score better characterized differences between actual care and clinical pathways in a population, was more strongly associated with survival and demonstrated better patient discrimination.

Trends of Dementia among Community-Dwelling Adults in Ontario, Canada, 2010-2015.

BACKGROUND: There are increasing numbers of people living with dementia (PLWD) and most reside in community settings. Characterizing the number of individuals affected with dementia and their transitions are important to understand in order to plan for their healthcare needs. Using administrative health data in Ontario, Canada, we examined recent trends in the prevalence and incidence of dementia among the community-dwelling population, described their characteristics, and investigated admissions to long-term care (LTC) and overall survival. METHODS: Using a validated case ascertainment algorithm, we performed a population-based retrospective cohort study of community-dwelling PLWD aged 40-105 years old between 2010 and 2015. We assessed crude and age- and sex-adjusted prevalence and incidence, cohort characteristics, and time to LTC admission and survival. RESULTS: Between 2010 and 2015, the adjusted community prevalence increased by 9.5% (p < 0.001), while the incidence decreased by 15.8% (p < 0.001). Demographic and socioeconomic characteristics remained similar over time, while the prevalence of comorbidities increased significantly from 2010 to 2015. There was no difference in the time to LTC admission for individuals diagnosed in 2014 when compared to 2010 (p = 0.06). A lower risk of 2-year mortality was observed for individuals diagnosed in 2015 compared to 2010 (HR 0.93, 95% CI 0.90-0.97, p < 0.001). CONCLUSION: There was an increase in the prevalence of dementia despite decreasing incidence among community-dwelling PLWD. Lower rates of mortality indicate that PLWD are surviving longer following diagnosis. Adequate resources and planning are required to support this growing population, considering the changing population size and characteristics.

Strategic analytics: towards fully embedding evidence in healthcare decision-making.

Cancer Care Ontario (CCO) has implemented multiple information technology solutions and collected health-system data to support its programs. There is now an opportunity to leverage these data and perform advanced end-to-end analytics that inform decisions around improving health-system performance. In 2014, CCO engaged in an extensive assessment of its current data capacity and capability, with the intent to drive increased use of data for evidence-based decision-making. The breadth and volume of data at CCO uniquely places the organization to contribute to not only system-wide operational reporting, but more advanced modelling of current and future state system management and planning. In 2012, CCO established a strategic analytics practice to assist the agency's programs contextualize and inform key business decisions and to provide support through innovative predictive analytics solutions. This paper describes the organizational structure, services and supporting operations that have enabled progress to date, and discusses the next steps towards the vision of embedding evidence fully into healthcare decision-making.

Identifying Breast Cancer Recurrence in Administrative Data: Algorithm Development and Validation.

Breast cancer recurrence is an important outcome for patients and healthcare systems, but it is not routinely reported in cancer registries. We developed an algorithm to identify patients who experienced recurrence or a second case of primary breast cancer (combined as a "second breast cancer event") using administrative data from the population of Ontario, Canada. A retrospective cohort study design was used including patients diagnosed with stage 0-III breast cancer in the Ontario Cancer Registry between 1 January 2009 and 31 December 2012 and alive six months post-diagnosis. We applied the algorithm to healthcare utilization data from six months post-diagnosis until death or 31 December 2013, whichever came first. We validated the algorithm's diagnostic accuracy against a manual patient record review (n = 2245 patients). The algorithm had a sensitivity of 85%, a specificity of 94%, a positive predictive value of 67%, a negative predictive value of 98%, an accuracy of 93%, a kappa value of 71%, and a prevalence-adjusted bias-adjusted kappa value of 85%. The second breast cancer event rate was 16.5% according to the algorithm and 13.0% according to manual review. Our algorithm's performance was comparable to previously published algorithms and is sufficient for healthcare system monitoring. Administrative data from a population can, therefore, be interpreted using new methods to identify new outcome measures.

Forecasting Ontario Oncology Drug Expenditures: A Hybrid Approach to Improving Accuracy.

BACKGROUND: The Provincial Drug Reimbursement Program (PDRP) at Cancer Care Ontario (CCO) is responsible for monitoring actual and projected outpatient intravenous cancer drug spending in the province. We developed a hybrid forecasting approach combining automated time-series forecasting with expert-customizable input. OBJECTIVE: Our objectives were to provide a flexible tool in which to incorporate multiple forecasts and to improve the accuracy of the resulting forecast. METHODS: The approach employed linear and non-linear time-series techniques and a combined hybrid model incorporating both approaches. We developed an interactive tool that incorporated the statistical models and identified the best performing forecast according to standard goodness-of-fit measures. Model selection procedures considered both the amount of historical expenditure data available per drug policy and the individual policy contributions to the overall budget. The user was allowed to customize forecasts based on knowledge of external factors related to policy or price changes and new drugs that come to market RESULTS: A comparison of 2016/17 fiscal year expenditures showed that all policies with a significant contribution to the overall budget were forecast with < 4% error. Forecasting error was reduced by at least $Can5 million for the nine most expensive policies compared with expert opinion. This approach to drug budget forecasting was implemented in Ontario for the first time in the 2017/18 fiscal year, where 1% error was observed for the overall budget, corresponding to an overestimate of expenditures by $Can3.0 million. CONCLUSION: We introduced a pragmatic approach for regular forecasting by budget holders in Ontario. Our approach to isolating 'big budget' from 'small budget' drugs using an 80-20 rule and providing multiple forecasts depending on the length of the drug expenditure histories is transferable to other jurisdictions.

Regional Variation in Palliative Care Receipt in Ontario, Canada.

Background: Delivery of health services in the province of Ontario is organized into 14 Local Health Integration Networks (LHINs), and further into 76 LHIN subregions, making these a natural unit of comparing the regional differences in palliative care receipt among decedents who were identified as having palliative care needs. Objective: To assess the presence and magnitude of the remaining regional variation in palliative care receipt in Ontario after accounting for demographic and socioeconomic differences between the LHIN subregions, and therefore to assess whether the standardized proportion of palliative care receipt as a performance indicator can capture potential performance-related issues. Design: A retrospective cohort study based on Ontario administrative data sources. Setting/Subjects: Ontario residents who died between April 1, 2015 and March 31, 2016 and were identified as having palliative care needs. Measurements: Date of death, diagnostic codes used for determining palliative care needs, and services receipt in last year of life were identified from multiple administrative databases. Demographic and socioeconomic information were derived from linking decedents' postal codes to Statistics Canada Census data and Ontario Marginalization Index. Results: Statistically significant variation ranging from 63% to 75% in palliative care receipt exists between Ontario subregions even after accounting for demographic and socioeconomic differences, including age, sex, rurality, income quintile, and the four dimensions of the Ontario Marginalization Index. Conclusions: Annual directly standardized proportion of palliative care receipt can be used as a performance indicator to detect regional differences in service receipt while adjusting for regional differences in the characteristics of the decedent populations. The factors to be adjusted for can be chosen based on the comparison of interest.