The purpose of this study was to examine factors identified by patients as relevant to health human resources (HHR) planning for primary healthcare (PHC). Eleven focus groups were conducted in British Columbia and a thematic analysis was undertaken, informed by a needs-based HHR planning framework. Three themes emerged: (a) the importance of geographic context, (b) change management at the practice level and (c) the need for choices and changes in delivery of PHC. Findings suggest that more attention could be focused on overcoming geographic barriers to providing services, change management within office-based practices, and providing support structures that allow primary care providers to work closer to their full scope of practice. That these factors align with many strategic directions set out by government and planners signals the readiness for change in how PHC is delivered and HHR planned.
Administrative databases can be used to measure healthcare performance. This can lead to identification of high-performing practice characteristics and inform innovations. However, a key challenge is that administrative data cannot be easily combined across provinces. Comparable measures must be defined across provinces but operationalized within each province. The purpose of this work is to provide an example of defining a population health concept, osteoporosis screening, and creating measures to examine it across British Columbia, Ontario, and Nova Scotia, Canada.
TRANSFORMATION is a study that seeks to improve the science and reporting of Primary Health Care performance. We used administrative data from the above three provinces to examine osteoporosis screening in those aged 65 years and older. Challenges of databases with different data elements and levels of methods development (e.g. macros) can be overcome for purposes of cross-provincial comparisons. Flexibility of analytic methods and frequent communication is needed.
Background: In 2003, the government of British Columbia, Canada introduced a universal drug benefit plan to cover drug costs that are high relative to household income. Residents were required to register in order to be eligible for the income-based benefits. Given past research suggesting that registration processes may pose an access barrier to certain subpopulations, we aimed to determine whether registration rates varied across small geographic areas that differed in ethnic composition. Methods: Using linked population-based administrative databases and census data, we conducted multivariate logistic regression analyses to determine whether the probability of registration for the public drug plan varied across areas of differing ethnic composition, controlling for household-level predisposing, enabling and needs factors. Results: The adjusted odds of registration did not differ across regions characterized by high concentrations (greater than 30%) of residents identifying as North American, British, French or other European. Households located in areas with concentrations of residents identifying as an Asian ethnicity had the highest odds of program registration: Chinese (OR = 1.21, CI: 1.19-1.23) and South Asian (OR = 1.19, CI: 1.16-1.22). Despite this positive finding, households residing in areas with relatively high concentrations of recent immigrants had slightly lower adjusted odds of registering for the program (OR = 0.97, CI: 0.95-0.98). Conclusions: This study identified ethnic variation in registration for a new public drug benefit program in British Columbia. However, unlike previous studies, the variation observed did not indicate that areas with high concentrations of certain ethnicities experienced disadvantages. Potential explanations are discussed. ; Nursing, School of ; Medicine, Faculty of ; Population and Public Health (SPPH), School of ; Applied Science, Faculty of ; Reviewed ; Faculty
Abstract Background In 2003, the government of British Columbia, Canada introduced a universal drug benefit plan to cover drug costs that are high relative to household income. Residents were required to register in order to be eligible for the income-based benefits. Given past research suggesting that registration processes may pose an access barrier to certain subpopulations, we aimed to determine whether registration rates varied across small geographic areas that differed in ethnic composition. Methods Using linked population-based administrative databases and census data, we conducted multivariate logistic regression analyses to determine whether the probability of registration for the public drug plan varied across areas of differing ethnic composition, controlling for household-level predisposing, enabling and needs factors. Results The adjusted odds of registration did not differ across regions characterized by high concentrations (greater than 30%) of residents identifying as North American, British, French or other European. Households located in areas with concentrations of residents identifying as an Asian ethnicity had the highest odds of program registration: Chinese (OR = 1.21, CI: 1.19-1.23) and South Asian (OR = 1.19, CI: 1.16-1.22). Despite this positive finding, households residing in areas with relatively high concentrations of recent immigrants had slightly lower adjusted odds of registering for the program (OR = 0.97, CI: 0.95-0.98). Conclusions This study identified ethnic variation in registration for a new public drug benefit program in British Columbia. However, unlike previous studies, the variation observed did not indicate that areas with high concentrations of certain ethnicities experienced disadvantages. Potential explanations are discussed.
"The report describes work underway at CHSPR to develop and validate a sector-specific survey regarding adults' experiences with primary health care (PHC) in the province. The resulting survey measures dimensions (or aspects) of PHC identified as priorities through focus groups held with patients across the province, a review of relevant scientific literature, and input from government stakeholders. Understanding British Columbians' Experiences with PHC lays out broad findings based on survey development and validation—a more detailed analysis of the survey items, index and scales is available upon request." -CHSPR website ; Applied Science, Faculty of ; Nursing, School of ; Medicine, Faculty of ; Population and Public Health (SPPH), School of ; Unreviewed ; Faculty ; Researcher ; Graduate
Background: Since the 1960s, the federal government has been providing or funding a selection of community-based primary healthcare (PHC) programs on First Nations reserves. A key question is whether local access to PHC can help address health inequities in First Nations on-reserve communities in British Columbia (BC). Objectives: This paper examines whether hospitalization for Ambulatory Care Sensitive Conditions (1) can be used as a proxy measure for the organization of PHC in First Nations reserve areas; and (2) is associated with premature mortality rates. Methods: In this descriptive correlational study, we used administrative data available through Population Data BC, including demographic and ecological information (i.e. geo-codes indicating location of residence). We used two different measures of hospitalization: rates of episodic hospital care and rates of length of stay. We correlated hospitalization rates with premature mortality rates and the level of care available in First Nations communities, which depends on a federal funding formula based upon community size and, more specifically, the level of isolation from a provincial point of care. Results: First Nations communities in BC that have local 24/7 access to PHC services have similar rates of hospitalization for ACSC to those living in urban centres. This is demonstrated by the similarities in the strengths of the correlation between premature mortality rates and rates of avoidable hospitalization for conditions treatable in a PHC setting. This is not the case for communities served by a Health Centre (weaker correlation) and for communities serviced by a Health Station or with no on-reserve point of care (no correlation). Conclusions: Improving access to PHC services in First Nations communities can be associated with a significant reduction in avoidable hospitalization and premature mortality rates. The method we tested is an important tool that could serve health care planning decisions in small communities. ; Applied Science, Faculty of ; Non UBC ; Nursing, School of ; Reviewed ; Faculty
Abstract Background Since the 1960s, the federal government has been providing or funding a selection of community-based primary healthcare (PHC) programs on First Nations reserves. A key question is whether local access to PHC can help address health inequities in First Nations on-reserve communities in British Columbia (BC). Objectives This paper examines whether hospitalization for Ambulatory Care Sensitive Conditions (1) can be used as a proxy measure for the organization of PHC in First Nations reserve areas; and (2) is associated with premature mortality rates. Methods In this descriptive correlational study, we used administrative data available through Population Data BC, including demographic and ecological information (i.e. geo-codes indicating location of residence). We used two different measures of hospitalization: rates of episodic hospital care and rates of length of stay. We correlated hospitalization rates with premature mortality rates and the level of care available in First Nations communities, which depends on a federal funding formula based upon community size and, more specifically, the level of isolation from a provincial point of care. Results First Nations communities in BC that have local 24/7 access to PHC services have similar rates of hospitalization for ACSC to those living in urban centres. This is demonstrated by the similarities in the strengths of the correlation between premature mortality rates and rates of avoidable hospitalization for conditions treatable in a PHC setting. This is not the case for communities served by a Health Centre (weaker correlation) and for communities serviced by a Health Station or with no on-reserve point of care (no correlation). Conclusions Improving access to PHC services in First Nations communities can be associated with a significant reduction in avoidable hospitalization and premature mortality rates. The method we tested is an important tool that could serve health care planning decisions in small communities.
Background: Public health and primary care are distinct sectors within western health care systems. Within each sector, work is carried out in the context of organizations, for example, public health units and primary care clinics. Building on a scoping literature review, our study aimed to identify the influencing factors within these organizations that affect the ability of these health care sectors to collaborate with one another in the Canadian context. Relationships between these factors were also explored. Methods: We conducted an interpretive descriptive qualitative study involving in-depth interviews with 74 key informants from three provinces, one each in western, central and eastern Canada, and others representing national organizations, government, or associations. The sample included policy makers, managers, and direct service providers in public health and primary care. Results: Seven major organizational influencing factors on collaboration were identified: 1) Clear Mandates, Vision, and Goals; 2) Strategic Coordination and Communication Mechanisms between Partners; 3) Formal Organizational Leaders as Collaborative Champions; 4) Collaborative Organizational Culture; 5) Optimal Use of Resources; 6) Optimal Use of Human Resources; and 7) Collaborative Approaches to Programs and Services Delivery. Conclusion: While each influencing factor was distinct, the many interactions among these influences are indicative of the complex nature of public health and primary care collaboration. These results can be useful for those working to set up new or maintain existing collaborations with public health and primary care which may or may not include other organizations. ; Applied Science, Faculty of ; Other UBC ; Non UBC ; Nursing, School of ; Reviewed ; Faculty
Background: Purposefully building stronger collaborations between primary care (PC) and public health (PH) is one approach to strengthening primary health care. The purpose of this paper is to report: 1) what systemic factors influence collaborations between PC and PH; and 2) how systemic factors interact and could influence collaboration. Methods: This interpretive descriptive study used purposive and snowball sampling to recruit and conduct interviews with PC and PH key informants in British Columbia (n = 20), Ontario (n = 19), and Nova Scotia (n = 21), Canada. Other participants (n = 14) were knowledgeable about collaborations and were located in various Canadian provinces or working at a national level. Data were organized into codes and thematic analysis was completed using NVivo. The frequency of "sources" (individual transcripts), "references" (quotes), and matrix queries were used to identify potential relationships between factors. Results: We conducted a total of 70 in-depth interviews with 74 participants working in either PC (n = 33) or PH (n = 32), both PC and PH (n = 7), or neither sector (n = 2). Participant roles included direct service providers (n = 17), senior program managers (n = 14), executive officers (n = 11), and middle managers (n = 10). Seven systemic factors for collaboration were identified: 1) health service structures that promote collaboration; 2) funding models and financial incentives supporting collaboration; 3) governmental and regulatory policies and mandates for collaboration; 4) power relations; 5) harmonized information and communication infrastructure; 6) targeted professional education; and 7) formal systems leaders as collaborative champions. Conclusions: Most themes were discussed with equal frequency between PC and PH. An assessment of the system level context (i.e., provincial and regional organization and funding of PC and PH, history of government in successful implementation of health care reform, etc) along with these seven system level factors could assist other jurisdictions in moving towards increased PC and PH collaboration. There was some variation in the importance of the themes across provinces. British Columbia participants more frequently discussed system structures that could promote collaboration, power relations, harmonized information and communication structures, formal systems leaders as collaboration champions and targeted professional education. Ontario participants most frequently discussed governmental and regulatory policies and mandates for collaboration. ; Applied Science, Faculty of ; Other UBC ; Non UBC ; Nursing, School of ; Reviewed ; Faculty
Over the last 30 years, public investments in Canada and many other countries have created clinical and administrative health data repositories to support research on health and social services, population health and health policy. However, there is limited capacity to share and use data across jurisdictional boundaries, in part because of inefficient and cumbersome procedures to access these data and gain approval for their use in research. A lack of harmonization among variables and indicators makes it difficult to compare research among jurisdictions. These challenges affect the quality, scope, and impact of work that could be done. The purpose of this paper is to compare and contrast the data access procedures in three Canadian jurisdictions (Manitoba, Alberta and British Columbia), and to describe how we addressed the challenges presented by differences in data governance and architecture in a Canadian cross-jurisdictional research study. We characterize common stages in gaining access to administrative data among jurisdictions, including obtaining ethics approval, applying for data access from data custodians, and ensuring the extracted data is released to accredited individuals in secure data environments. We identify advantages of Manitoba's flexible 'stewardship' model over the more restrictive 'custodianship' model in British Columbia, and highlight the importance of communication between analysts in each jurisdiction to compensate for differences in coding variables and poor quality data. Researchers and system planners must have access to and be able to make effective use of administrative health data to ensure that Canadians continue to have access to high-quality health care and benefit from effective health policies. The considerable benefits of collaborative population-based research that spans jurisdictional borders have been recognized by the Canadian Institutes for Health Research in their recent call for the creation of a National Data Platform to resolve many of the issues in harmonization and validation of administrative data elements.
IntroductionFrailty is a medical syndrome, commonly affecting people aged 65 years and over and is characterized by a greater risk of adverse outcomes following illness or injury. Electronic medical records contain a large amount of longitudinal data that can be used for primary care research. Machine learning can fully utilize this wide breadth of data for the detection of diseases and syndromes. The creation of a frailty case definition using machine learning may facilitate early intervention, inform advanced screening tests, and allow for surveillance. ObjectivesThe objective of this study was to develop a validated case definition of frailty for the primary care context, using machine learning. MethodsPhysicians participating in the Canadian Primary Care Sentinel Surveillance Network across Canada were asked to retrospectively identify the level of frailty present in a sample of their own patients (total n = 5,466), collected from 2015-2019. Frailty levels were dichotomized using a cut-off of 5. Extracted features included previously prescribed medications, billing codes, and other routinely collected primary care data. We used eight supervised machine learning algorithms, with performance assessed using a hold-out test set. A balanced training dataset was also created by oversampling. Sensitivity analyses considered two alternative dichotomization cut-offs. Model performance was evaluated using area under the receiver-operating characteristic curve, F1, accuracy, sensitivity, specificity, negative predictive value and positive predictive value. ResultsThe prevalence of frailty within our sample was 18.4%. Of the eight models developed to identify frail patients, an XGBoost model achieved the highest sensitivity (78.14%) and specificity (74.41%). The balanced training dataset did not improve classification performance. Sensitivity analyses did not show improved performance for cut-offs other than 5. ConclusionSupervised machine learning was able to create well performing classification models for frailty. Future research is needed to assess frailty inter-rater reliability, and link multiple data sources for frailty identification.
Plain English summary Increasingly, health researchers are conducting their research in partnership with non-researchers such as patients and caregivers, advocacy groups, clinicians, and policymakers. The idea behind this partnership is to make research more relevant and appropriate. However, so far there is not much evidence about how this partnership or engagement actually affects research. We conducted an online survey of 12 teams in Canada that have engaged patients and other stakeholders in community based health research, partly as a requirement to obtain funding. We found that in many cases, the teams have engaged a wide variety and large number of stakeholders, and have involved them in many different stages of their research. Teams reported that their overall experience of this approach to research has been positive, but some challenges have been encountered along the way. Some teams found that it was difficult to communicate appropriately with all the stakeholders, and to keep them informed when research was going slowly. Other teams had trouble finding government representatives to work with. Several teams noted that engagement is time-consuming, and requires a lot of effort. Nevertheless, all teams reported that they had learned from the experience, and found it valuable. As a result, Canadian health care researchers are better positioned to engage with patients and other stakeholders in the future. Abstract Background Patient and other stakeholder engagement in research is increasingly important, but there is limited evidence of its impact. In 2013, the Canadian Institutes of Health Research launched a five-year Community Based Primary Health Care (CBPHC) initiative that funded 12 teams for innovative approaches to primary health care involving engagement with patients, communities, decision-makers, and clinicians across jurisdictions in Canada. The present study examines the extent of engagement by these teams, and the factors that affected it, either as challenges or opportunities. Methods We conducted a cross-sectional web-based survey across the 12 CBPHC Innovation Teams, in which we were also participants. We used a data collection tool developed by the Patient Centered Outcomes Research Institute that included both closed and open-ended questions. Results The quantitative data showed that the CBPHC Innovation teams have engaged with diverse stakeholders at different levels and in different stages of research. Almost all teams surveyed engaged with policymakers, most with clinicians and health system representatives, and more than half with patients, mostly at the level of consultation or collaboration. There were very few instances of stakeholder-led research reported. There was a near universal recognition of the importance of communications processes/tools in facilitating engagement, whereas time was the most commonly identified challenge. In almost all cases, challenges encountered were partially if not fully resolved. The qualitative findings showed that each team's engagement was contextualized by factors such as the jurisdictions and geographic scope of the project, the number and type of stakeholders engaged and their level of involvement. These intersected with the researchers' motivations for engagement, to give rise to diverse experiences, but ones that the CBPHC teams assessed positively as an approach to research. Conclusions Over the past five years, primary health care researchers in Canada have been actively engaging with patients and other stakeholders. The wide range, extent and nature of that engagement shows that these researchers have anticipated developments in this approach to research and are thus in a position to support and strengthen future efforts to understand the impact of this engagement on health care outcomes.
Plain English summary: Increasingly, health researchers are conducting their research in partnership with non-researchers such as patients and caregivers, advocacy groups, clinicians, and policymakers. The idea behind this partnership is to make research more relevant and appropriate. However, so far there is not much evidence about how this partnership or engagement actually affects research. We conducted an online survey of 12 teams in Canada that have engaged patients and other stakeholders in community based health research, partly as a requirement to obtain funding. We found that in many cases, the teams have engaged a wide variety and large number of stakeholders, and have involved them in many different stages of their research. Teams reported that their overall experience of this approach to research has been positive, but some challenges have been encountered along the way. Some teams found that it was difficult to communicate appropriately with all the stakeholders, and to keep them informed when research was going slowly. Other teams had trouble finding government representatives to work with. Several teams noted that engagement is time-consuming, and requires a lot of effort. Nevertheless, all teams reported that they had learned from the experience, and found it valuable. As a result, Canadian health care researchers are better positioned to engage with patients and other stakeholders in the future. Background: Patient and other stakeholder engagement in research is increasingly important, but there is limited evidence of its impact. In 2013, the Canadian Institutes of Health Research launched a five-year Community Based Primary Health Care (CBPHC) initiative that funded 12 teams for innovative approaches to primary health care involving engagement with patients, communities, decision-makers, and clinicians across jurisdictions in Canada. The present study examines the extent of engagement by these teams, and the factors that affected it, either as challenges or opportunities. Methods: We conducted a cross-sectional web-based survey across the 12 CBPHC Innovation Teams, in which we were also participants. We used a data collection tool developed by the Patient Centered Outcomes Research Institute that included both closed and open-ended questions. Results: The quantitative data showed that the CBPHC Innovation teams have engaged with diverse stakeholders at different levels and in different stages of research. Almost all teams surveyed engaged with policymakers, most with clinicians and health system representatives, and more than half with patients, mostly at the level of consultation or collaboration. There were very few instances of stakeholder-led research reported. There was a near universal recognition of the importance of communications processes/tools in facilitating engagement, whereas time was the most commonly identified challenge. In almost all cases, challenges encountered were partially if not fully resolved. The qualitative findings showed that each team's engagement was contextualized by factors such as the jurisdictions and geographic scope of the project, the number and type of stakeholders engaged and their level of involvement. These intersected with the researchers' motivations for engagement, to give rise to diverse experiences, but ones that the CBPHC teams assessed positively as an approach to research. Conclusions: Over the past five years, primary health care researchers in Canada have been actively engaging with patients and other stakeholders. The wide range, extent and nature of that engagement shows that these researchers have anticipated developments in this approach to research and are thus in a position to support and strengthen future efforts to understand the impact of this engagement on health care outcomes. ; Applied Science, Faculty of ; Non UBC ; Nursing, School of ; Reviewed ; Faculty
Background. In order to address several health challenges, the Chinese government issued the National Essential Public Health Services Package (NEPHSP) in 2009. In China's large cities, the lack of funding for community health centers and consequent lack of comprehensive services and high quality care has become a major challenge. However, no study has been carried out to estimate the cost of delivering the services in the package. This project was to develop a cost estimation approach appropriate to the context and use it to calculate the cost of the NEPHSP in Beijing in 2011. Methods By adjusting models of cost analysis of primary health care and workload indicators of staffing need developed by the World Health Organization, a model was developed to estimate the cost of the services in the package through an intensive interactive process. A total of 17 community health centers from eight administrative districts in Beijing were selected. Their service volume and expenditure data in 2010 were used to evaluate the costs of providing the NEPHSP in Beijing based on the applied model. Results The total workload of all types of primary health care in 17 sample centers was equivalent to the workload requirement for 14,056,402 standard clinic visits. The total expenditure of the 17 sample centers was 26,329,357.62 USD in 2010. The cost of the workload requirement of one standard clinic visit was 1.87 USD. The workload of the NEPHSP was equivalent to 5,514,777 standard clinic visits (39.23 % of the total workload). The model suggests that the cost of the package in Beijing was 7.95 USD per capita in 2010. The cost of the NEPHSP in urban areas was lower than suburban areas: 7.31 and 8.65 USD respectively. Conclusions The average investment of 3.97 USD per capita in NEPHSP was lower than the amount needed to meet its running costs. NEPHSP in Beijing is therefore underfunded. Additional investment is needed, and a dynamic cost estimate mechanism should be introduced to ensure services remain adequately funded. ; Applied Science, Faculty of ; Nursing, School of ; Non UBC ; Reviewed ; Faculty
