Canadian Institute for Health Information

Distributed analytics enables analyses to be conducted when data is partitioned across multiple sites and cannot be pooled in one location. The primary motivation for using these methods is to protect confidentiality of line-level data. Distributed analytics methods rely on exchanging intermediate numerical outputs generated locally at each participating site. However, the absence of line-level data exchange should not be mistaken for privacy protection. To address this, the GRIIS developed a framework for evaluating privacy, helping users distinguish different levels of privacy preservation when applying distributed methods. This work was especially important, as earlier research demonstrated that an existing distributed method could be reverse-engineered to recover line-level data, which underscores that the distributed nature of an approach does not guarantee data privacy. In the framework, we define data privacy as guaranteed if line-level data cannot be uniquely recovered from the numerical outputs exchanged or disclosed during the execution of the approach. By eliciting line-level data as a set of unknown variables subject to constraints derived from the procedure, we can, in some settings, mathematically prove that privacy is protected under this definition. Achieving this depends on the nature of the variables included (binary, continuous), the results shared at the end of the procedure and the availability of external information. Ultimately, this work plays a critical role in supporting the responsible use of distributed analytics. By establishing governance mechanisms to evaluate and mitigate privacy risks, it helps ensure that these methods can be adopted appropriately and at scale across the country.

Background: Solid organ transplant (SOT) recipients in Canada are particularly vulnerable to adverse hospital outcomes, especially during admissions involving a COVID-19 diagnosis. Limited evidence exists regarding how risks vary across different organ types and the extent to which a COVID-19 diagnosis influences hospital outcomes. This study aims to examine the association of organ subtypes on hospital morbidity and mortality, both in the presence and absence of a COVID-19 diagnosis in a large, nationally representative Canadian cohort. Methods: We used data from the Canadian Organ Replacement Register and the Discharge Abstract Database to examine hospitalization rates and in-hospital outcomes among all available adult SOT recipients with functioning grafts in Canada (excluding Quebec and Manitoba) from January 2021 to December 2022. In-hospital outcomes included transfer to a special care unit (SCU) and hospital mortality. Comparisons between organ subtypes (kidney, liver, heart, lung, and other/multi-organ) were conducted separately for admissions with and without a diagnosis of COVID-19, using kidney transplant (KT) recipients as the reference group. We included all admissions with a COVID-19 diagnosis irrespective of whether it was the primary reason for admission or not. Rates of hospitalization, SCU transfer, and mortality were analyzed using negative binomial or Poisson regression models (adjusted for age and sex) and reported using incidence rate ratios (IRRs) with 95% confidence intervals (CIs). Results: Among 23 497 SOT recipients, the majority (14 628, 62%) were KT recipients. Within this cohort, 2428 individuals (10.3%) experienced a total of 2925 hospitalizations with a COVID-19 diagnosis. In comparison, 7808 (33.2%) individuals experienced 17 656 hospitalizations without a COVID-19 diagnosis. Lung transplant recipients were more likely to be hospitalized (IRR = 1.65, 95% confidence interval CI: 1.52-1.80) and die in hospital (IRR = 1.2, 95% CI: 1.05-1.34) than KT recipients during admissions involving a COVID-19 diagnosis. In contrast, heart and liver transplant recipients were less likely to be hospitalized or experience a poor outcome. For hospitalizations without a COVID-19 diagnosis, lung and other/multi-organ transplant recipients were more likely than KT recipients to be hospitalized (IRR = 1.94, 95% CI: 1.76-2.15; IRR = 1.81, 95% CI: 1.45-2.26, respectively), transferred to an SCU (IRR = 1.89, 95% CI: 1.58-2.27; IRR = 1.81, 95% CI: 1.45-2.26, respectively), and die in hospital (IRR = 2.04, 95% CI: 1.84-2.27; IRR = 1.57, 95% CI: 1.33-1.85; respectively). Conclusion: SOT recipients in Canada, especially lung transplant recipients, experience high rates of hospitalization, SCU admission, and in-hospital mortality. Notable differences observed between organ subtypes for admissions with and without a COVID-19 diagnosis may reflect differences in immunosuppressive medication regimens, informing areas for future research.

As Canada's population ages, understanding older adults' experiences with the healthcare system is essential to ensuring high-quality, equitable care. The 2024 Commonwealth Fund International Health Policy Survey of Older Adults (age 65+) offers a unique opportunity to learn from other countries. Among 10 high-income countries surveyed, older adults in Canada were the least likely to feel satisfied with the quality of care they had received over the past year. This may reflect other challenges they reported, like limited access to timely primary care, long wait times, declining care coordination and out-of-pocket costs.

Opioid agonist treatment (OAT) is the recommended first-line therapy for individuals with opioid use disorder. The availability of different products and prescribing guidance for OAT has rapidly evolved in recent years. Therefore, we conducted a repeated cross-sectional study using pharmacy dispensing data from six Canadian provinces to examine trends in methadone and buprenorphine use between January 2018 and December 2022. Monthly population-adjusted rates of OAT recipients were calculated by province, and annual cohorts were stratified by demographic characteristics and OAT type. Prevalence of OAT use varied across provinces, ranging from 1.03 per 1000 in Quebec to 3.59 per 1000 in British Columbia. Increases in OAT use between 2018 and 2022 were observed in Alberta (3.00-3.75 per 1000) and Manitoba (1.68-2.39 per 1000), while rates remained relatively stable elsewhere. OAT use was highest among adults aged 25-44 years, males, and residents of lower-income neighbourhoods. A notable shift toward buprenorphine prescribing was observed across all provinces, with 45%-74% of OAT recipients dispensed a buprenorphine product in 2022. These cross-provincial differences in OAT rates likely reflect variations in opioid use disorder prevalence, prescribing practices, and access to care.

Reinforcement Learning (RL) offers a promising pathway for aligning Multimodal Large Language Models (MLLMs) with the strict precision requirements of chest X-ray (CXR) analysis. However, even with supervised initialization, the complexity of CXR interpretation creates a sparse reward manifold, often leading to exploration stagnation and training instability. To address these challenges, we propose ChestR1, a sample-efficient RL framework grounded in Group Relative Policy Optimization (GRPO). We introduce a novel Ground-Truth Augmented sampling strategy to guide exploration, coupled with an Entropy-Aware Policy Modulation mechanism to prevent distribution collapse. Furthermore, a Dynamic Failure-Aware SFT mechanism tightly couples supervision with RL, specifically injecting guidance for hard instances where exploration fails. Validated on our curated multi-task dataset, ChestR1 achieves competitive performance with lower training consumption and a smaller model size compared to state-of-the-art baselines.