Danielle M. Nash

Key Points: Multidisciplinary kidney care eligibility changed from eGFR ≤33.4 ml/min per 1.73 m ² to 2-year kidney failure risk ≥10% (includes eGFR and proteinuria). After the change, we found no difference in outcomes such as dialysis initiation, preparation for dialysis, and hospitalizations. The results of this study help guide decision making with respect to the care of patients with advanced CKD. Background: We examined the effect of a change in the eligibility criteria for multidisciplinary kidney care implemented across Ontario, Canada, on April 1, 2016 (criteria moved from an eGFR ≤33.4 ml/min per 1.73 m ² to 2-year kidney failure risk ≥10%, which incorporates assessment of proteinuria, or an eGFR 2 followed by a nephrologist ( N =97,299). We examined multidisciplinary clinic and nephrologist visits, dialysis-related outcomes, hospital encounters, and mortality by monthly interval between October 1, 2013, and February 1, 2020. Autoregressive integrated moving average models were tested for level (immediate) and slope (over time) changes in outcomes after the intervention and were fit on preintervention change data for projected trends. Results: After intervention, there was a significant monthly decline in the multidisciplinary clinic visit rate of 1.49 visits per 100 person-years (95% confidence interval [CI], −2.83 to −0.14) and a significant monthly increase in the proportion of patients with a referral to multidisciplinary care within 1 year before starting dialysis ( i.e ., late referral; positive slope change, 0.67%; 95%, CI, 0.06 to 1.28). The intervention was not associated with significant changes in the proportion of patients initiating dialysis, initiating dialysis with a central venous catheter or during a hospitalization, using home dialysis, or in the rate of nephrologist visits. Immediately after intervention, there were small, significant increases in mortality and hospital encounters, but importantly, there was no monthly (slope) change in mortality and a significant negative monthly change in hospital encounters (−0.24 encounters per 100 person-years; 95% CI, −0.38 to −0.09). Conclusions: Changing the eligibility criteria for multidisciplinary kidney care from an eGFR ≤33.4 ml/min per 1.73 m ² to a 2-year kidney failure risk ≥10% or an eGFR 2 resulted in significantly fewer patients receiving such care without negatively affecting important clinical outcomes.

Team-based care is a primary care model that involves collaborative care between primary care providers and other professional providers. In Ontario Canada

Key Points: Among 4173 children with AKI, 18% had major adverse kidney events (death, kidney failure, or CKD) during a median 10-year follow-up. AKI survivors were at 2–4 times higher risk of major adverse kidney events, hypertension, and subsequent AKI versus matched hospitalized comparators. This justifies improved surveillance after pediatric AKI to detect CKD and hypertension early and improve long-term kidney health. Background: AKI is common in hospitalized children. Pediatric AKI receiving acute KRT is associated with long-term CKD, hypertension, and death. We aim to determine the outcomes after AKI in children who did not receive acute KRT because these remain uncertain. Methods: Retrospective cohort study of all hospitalized children (0–18 years) surviving AKI without acute KRT between 1996 and 2020 in Ontario, Canada, identified by validated diagnostic codes in provincial administrative health databases. Children with prior KRT, CKD, or AKI were excluded. Cases were matched with up to four hospitalized comparators without AKI by age, neonatal status, sex, intensive care unit admission, cardiac surgery, malignancy, hypertension, hospitalization era, and a propensity score for AKI. Patients were followed until death, provincial emigration, or censoring in March 2021. The primary outcome was long-term major adverse kidney events (a composite of all-cause mortality, long-term KRT, or incident CKD). Results: We matched 4173 pediatric AKI survivors with 16,337 hospitalized comparators. Baseline covariates were well-balanced following propensity score matching. During a median 9.7-year follow-up, 18% of AKI survivors developed long-term major adverse kidney event versus 5% of hospitalized comparators (hazard ratio [HR], 4.0; 95% confidence interval [CI], 3.6 to 4.4). AKI survivors had higher rates of long-term KRT (2% versus <1%; HR, 11.7; 95% CI, 7.5 to 18.4), incident CKD (16% versus 2%; HR, 7.9; 95% CI, 6.9 to 9.1), incident hypertension (17% versus 8%; HR, 2.3; 95% CI, 2.1 to 2.6), and AKI during subsequent hospitalization (6% versus 2%; HR, 3.7; 95% CI, 3.1 to 4.5), but no difference in all-cause mortality (3% versus 3%; HR, 0.9; 95% CI, 0.7 to 1.1). Conclusions: Children surviving AKI without acute KRT were at higher long-term risk of CKD, long-term KRT, hypertension, and subsequent AKI versus hospitalized comparators.

Purpose: To understand staff and health care providers’ views on potential use of artificial intelligence (AI)-driven tools to help care for patients within a primary care setting. Methods: We conducted a qualitative descriptive study using individual semistructured interviews. As part of province-wide Learning Health Organization, Community Health Centres (CHCs) are a community-governed, team-based delivery model providing primary care for people who experience marginalization in Ontario, Canada. CHC health care providers and staff were invited to participate. Interviews were audio-recorded and transcribed verbatim. We performed a thematic analysis using a team approach. Results: We interviewed 27 participants across 6 CHCs. Participants lacked in-depth knowledge about AI. Trust was essential to acceptance of AI; people need to be receptive to using AI and feel confident that the information is accurate. We identified internal influences of AI acceptance, including ease of use and complementing clinical judgment rather than replacing it. External influences included privacy, liability, and financial considerations. Participants felt AI could improve patient care and help prevent burnout for providers; however, there were concerns about the impact on the patient-provider relationship. Conclusions: The information gained in this study can be used for future research, development, and integration of AI technology.

Introduction: The Alliance for Healthier Communities represents community‐governed healthcare organizations in Ontario, Canada including Community Health Centres, which provide primary care to more disadvantaged populations.Methods: In this experience report, we describe the Alliance's journey towards becoming a learning health system using examples for organizational culture, data and analytics, people and partnerships, client engagement, ethics and oversight, evaluation and dissemination, resources, identification and prioritization, and deliverables and impact.Results: Many of the foundational elements for a learning health system were already in place at the Alliance including an integrated and accessible data platform. Leadership championed and embraced the movement towards a learning health system, which led to restructuring of the organization. This included role changes for data support personnel, better communication, and dissemination plans, strategies to engage clinicians and other front‐line staff, restructuring of committees for more collaborative planning and prioritization of quality improvement and research initiatives, and the development of a new Practice‐Based Learning Network for more opportunities to use the data for research and evaluation.Conclusions: Next steps will focus on continued clinical engagement and partnerships as well as ongoing reflection on the transition and success of the learning health system work.

Rationale, Aims and Objectives: A learning health system model can be used to efficiently evaluate and incorporate evidence‐based care into practice. However, there is a paucity of evidence describing key organizational attributes needed to ensure a successful learning health system within primary care. We interviewed stakeholders for a primary care learning health system in Ontario, Canada (the Alliance for Healthier Communities) to identify strengths and areas for improvement.Method: We conducted a qualitative descriptive study using individual semistructured interviews with Alliance stakeholders between December 2019 and March 2020. The Alliance delivers community‐governed primary healthcare through 109 organizations including Community Health Centres (CHCs). All CHC staff within the Alliance were invited to participate. Interviews were audio‐recorded and transcribed verbatim. We performed a thematic analysis using a team approach.Results: We interviewed 29 participants across six CHCs, including Executive Directors, managers, healthcare providers and data support staff. We observed three foundational elements necessary for a successful learning health system within primary care: shared organizational goals and culture, data quality and resources. Building on this foundation, people are needed to drive the learning health system, and this is conditional on their level of engagement. The main factors motivating staff member's engagement with the learning health system included their drive to help improve patient care, focusing on initiatives of personal interest and understanding the purpose of different initiatives. Areas for improvement were identified such as the ability to extract and use data to inform changes in real‐time, better engagement and protected time for providers to do improvement work, and more staff dedicated to data extraction and analysis.Conclusions: We identified key components needed to establish a learning health system in primary care. Similar primary care organizations in Canada and elsewhere can use these insights to guide their development as learning health systems.

Background: Printed educational materials (PEMs) have long been used to inform clinicians on evidence-based practices. However, the evidence for their effects on patient care and outcomes is unclear. In Ontario, despite widely available clinical practice guidelines recommending antihypertensives and cholesterol-lowering agents for patients with diabetes, prescriptions remain low. We aimed to determine whether PEMs can influence physicians to intensify prescribing of these medications. Methods: A pragmatic, 2 × 2 factorial, cluster randomized controlled trial was designed to ascertain the effect of two PEM formats on physician prescribing: a postcard-sized message (“outsert”) or a longer narrative article (“insert”). Ontario family physician practices (clusters) were randomly allocated to receive the insert, outsert, both or neither. Physicians were eligible if they were in active practice and their patients were included if they were over 65 years with a diabetes diagnosis; both were unaware of the trial. Administrative databases at ICES (formerly the Institute for Clinical Evaluative Sciences) were used to link patients to their physician and to analyse prescribing patterns at baseline and 1 year following PEM mailout. The primary outcome was intensification defined as the addition of a new antihypertensive or cholesterol-lowering agent, or dose increase of a current drug, measured at the patient level. Analyses were by intention-to-treat and accounted for the clustering of patients to physicians. Results: We randomly assigned 4231 practices (39% of Ontario family physicians) with a total population of 185,526 patients (20% of patients with diabetes in Ontario primary care) to receive the insert, outsert, both, and neither; among these, 4118 practices were analysed (n = 1025, n = 1037, n = 1031, n = 1025, respectively). No significant treatment effect was found for the outsert (odds ratio (OR) 1.01, 95% confidence interval (CI) 0.98 to 1.04) or the insert (OR 0.99, 95% CI 0.96 to 1.02). Percent of intensification in the four arms was similar (approximately 46%). Adjustment for physician characteristics (e.g., age, sex, practice location) had no impact on these findings. Conclusions: PEMs have no effect on physician’s adherence to recommendations for the management of diabetes-related complications in Ontario. Further research should investigate the effect of other strategies to narrow this evidence-to-practice gap. Trial registration: ISRCTN72772651. Retrospectively registered 21 July 2005.

Background: Learning health systems have been gaining traction over the past decade. The purpose of this study was to understand the spread of learning health systems in primary care, including where they have been implemented, how they are operating, and potential challenges and solutions. Methods: We completed a scoping review by systematically searching OVID Medline®, Embase®, IEEE Xplore®, and reviewing specific journals from 2007 to 2020. We also completed a Google search to identify gray literature. Results: We reviewed 1924 articles through our database search and 51 articles from other sources, from which we identified 21 unique learning health systems based on 62 data sources. Only one of these learning health systems was implemented exclusively in a primary care setting, where all others were integrated health systems or networks that also included other care settings. Eighteen of the 21 were in the United States. Examples of how these learning health systems were being used included real-time clinical surveillance, quality improvement initiatives, pragmatic trials at the point of care, and decision support. Many challenges and potential solutions were identified regarding data, sustainability, promoting a learning culture, prioritization processes, involvement of community, and balancing quality improvement versus research. Conclusions: We identified 21 learning health systems, which all appear at an early stage of development, and only one was primary care only. We summarized and provided examples of integrated health systems and data networks that can be considered early models in the growing global movement to advance learning health systems in primary care.

Background: Clinical practice guidelines recommend routine kidney function and serum potassium testing within 30 days of initiating ACE (angiotensin-converting enzyme) inhibitor or angiotensin II receptor blocker therapy. However, evidence is lacking about whether follow-up testing reduces therapy-related adverse outcomes. Methods and Results: We conducted 2 population-based retrospective cohort studies in Kaiser Permanente Northern California and Ontario, Canada. Patients with outpatient serum creatinine and potassium tests in the 30 days after starting ACE inhibitor or angiotensin II receptor blocker therapy were matched 1:1 to patients without follow-up tests. We evaluated the association of follow-up testing with 30-day all-cause mortality and hospitalization with acute kidney injury or hyperkalemia using Cox regression. We also developed and externally validated a risk score to identify patients at risk of having abnormally high serum creatinine and potassium values in follow-up. We identified 75 251 matched pairs initiating ACE inhibitor or angiotensin II receptor blocker therapy between January 1, 2007, and December 31, 2017, in Kaiser Permanente Northern California. Follow-up testing was not significantly associated with 30-day all-cause mortality in Kaiser Permanente Northern California (hazard ratio, 0.75 [95% CI, 0.54–1.06]) and was associated with higher mortality in 84 905 matched pairs in Ontario (hazard ratio, 1.32 [95% CI, 1.07–1.62]). In Kaiser Permanente Northern California, follow-up testing was significantly associated with higher rates of hospitalization with acute kidney injury (hazard ratio, 1.66 [95% CI, 1.10–2.22]) and hyperkalemia (hazard ratio, 3.36 [95% CI, 1.08–10.41]), as was observed in Ontario. The risk score for abnormal potassium provided good discrimination (area under the curve [AUC], 0.75) and excellent calibration of predicted risks, while the risk score for abnormal serum creatinine provided moderate discrimination (AUC, 0.62) but excellent calibration. Conclusions: Routine laboratory monitoring after ACE inhibitor or angiotensin II receptor blocker initiation was not associated with a lower risk of 30-day mortality. We identified patient subgroups in which targeted testing may be effective in identifying therapy-related changes in serum potassium or kidney function.

End-stage kidney disease is a serious complication of diabetes. We describe the prevalence of chronic kidney disease, prevalence and incidence of end-stage kidney disease and quality of care of early-stage chronic kidney disease for First Nations people with diabetes compared to other Ontarians with diabetes. We conducted a retrospective cohort study in Ontario using linked administrative data at ICES. We included adults with incident diabetes between 1994 and 2014, and used laboratory values to identify kidney disease and quality indicators for care for early-stage disease. We compared measures in First Nations people to those in other people in Ontario, and used direct age and sex standardization. We used Cox proportional hazards regression to compare the incidence of end-stage kidney disease between groups. Our study included 21 968 First Nations people with diabetes. The age- and sex-standardized prevalence of chronic kidney disease was higher for First Nations people than for other Ontarians (20.7% v. 18.4%), as was the prevalence of end-stage kidney disease (2.9% v. 1.0%). The incidence of end-stage kidney disease was higher among First Nations people than among other people in Ontario (9.3 v. 4.7 events per 10 000 person-years; age- and sex-adjusted hazard ratio 2.23, 95% confidence interval 1.72-2.89). The 2 groups were similarly likely to receive recommended medications, but First Nations people were less likely to receive laboratory tests for their kidney disease. Despite receiving similar quality of care for early-stage kidney disease, First Nations people with diabetes had higher rates of end-stage kidney disease than other Ontarians. Further research is needed to better understand contributing factors to help inform future interventions.