Data Scan and Assessment: Population
Many comprehensive, high-quality data on population health characteristics are available for use in population needs-based, health workforce planning, though most (if not all) are not designed explicitly for these purposes. The results of the scan for data related to population demography, cultural and socioeconomic diversity, population health status, and health service utilization are shown in Tables 1, 2, and 3.
Multiple sources of data are available to the TC LHIN to support planning. The information is available at various levels of geography and for a variety of time periods.
We identified several populations for which data are not available, or the available data are not of sufficient quality, to allow for quantitative modeling. These populations include Indigenous patients, homeless patients, and non-insured patients. For these populations, we suggest that other sources of data – such as local surveys – be considered and that decision-makers advocate for better data to support planning for equitable provision of healthcare for these groups of patients.
Population Grouping Methodology
We assessed the CIHI Population Grouping Methodology2 as the most promising method to project needs-based primary care service requirements within the TC LHIN. This innovative, made-in-Canada grouping methodology uses individual-level clinical and demographic data to estimate population health needs, including the predicted number of visits to a family physician (Table 4). Application of this methodology to a region, such as the TC LHIN, with plentiful neighbourhood-level data on population health characteristics, including the social determinants of health, would allow decision-makers to develop a thorough understanding of the influence of various characteristics of the local population on the need for physician services, and to plan accordingly.
Data Scan and Assessment: Workforce
Some high-quality data on the health workforce are available for use in health workforce planning. The results of the scan for data related to physicians, nurses, nurse practitioners, and other regulated health professionals, including those in two inter-professional primary care delivery models operating in Ontario are shown in Table 5.
Data related to the health workforce are comprehensive, generally of high quality and may be accessed by regional planners through an application and approval process. Capturing practice activities of different health worker cadres is more challenging, and quantitative data on service provision, and on where practice activities overlap between professions, are less robust.
Building a Health Workforce Planning Model
Using data elements related to population health characteristics and health workforce profiles, as well as the CIHI Population Grouping Methodology, the model examines the alignment of the service requirements of the population with the service capacity of the workforce. An additional module allows for allocation of services across providers before the final examination of alignment. This exercise is an important but often overlooked part of planning. By including this step, our model embeds a systematic approach to minimizing the gap between workforce requirements and supply and optimizing the alignment of the health workforce with the needs of the population.
Step 1: Estimation of Population Service Requirements
The movement of patients across boundaries – between the urban TC LHIN and the four adjacent suburban LHINs, between sub-regions within the LHIN, and between neighbourhoods within the LHIN or beyond – was identified by the TC LHIN as a unique planning challenge. For the purposes of planning, the global population of patients who receive primary care within the TC LHIN is necessarily approached as two distinct populations: (1) residents of the TC LHIN, and (2) patients who do not reside in the geographic area of the TC LHIN but who seek and receive their care within the LHIN’s borders. This approach recognizes that the TC LHIN must plan to provide services to the patients who reside within its borders, while also acknowledging the reality that patients travel from outside the LHIN to access care.
To estimate service requirements, it is necessary to define a geographic planning region. This may be a single neighbourhood, a group of neighbourhoods, one or more LHIN sub-regions, the TC LHIN, or the entire City of Toronto. Service requirements are defined as the number of primary care visits needed for patients receiving care within the planning geography. These patients may be residents or non-residents and may receive a variable proportion of their care in the region in question. Service requirements are derived using the CIHI Population Grouping Methodology, which produces individual-level outputs – predicted number of visits to a family physician in the next year – that are then aggregated to the population level.
Thus, in a given year (y) in a defined geography (g), Service Requirements can be expressed by the following equation:
VyTOT= VRES (PRESg) + VNRES (PNRESg) + Us
VTOT is Total visits required
VRES is Visits required by the resident population
PRESg is % care residents receive within the defined geography
VNRES is Visits required by the non-resident population receiving care within the defined geography
PNRESg is % care non-residents receive within the defined geography
Us is a Subjective measure of unmet need
Unmet health care needs are an important contributor to service requirements, and estimates of service requirements are strengthened by explicit consideration of these needs. As such, a subjective measure of unmet health care needs – informed by data from the Canadian Community Health Survey and through local consultations – can be added to the calculations.
Step 2: Estimation of Workforce Service Capacity
Estimation of the service capacity of the workforce is conducted on a uni-professional basis and begins with identification of the stock of providers available to provide service. Adjustments are then applied to account for each factor that influences the service capacity of the workforce. These factors include inflows (immigration), outflows (emigration), and attrition due to death and retirement. Additional adjustments may be applied to account for activity rates, participation in the provision of care within the TC LHIN, and scope of practice that varies between providers. More detailed adjustments, such as an adjustment to account for changing practice patterns, may be applied, subject to available data. Finally, a subjective productivity variable allows for scenarios related to productivity to be incorporated into the model, adjusting the workforce service capacity according to productivity assumptions.
The end result of these calculations is an estimate of the total service capacity of the workforce. Due to the variability of the data elements in the available datasets, the unit of analysis for physicians is visits per year, while for other regulated primary healthcare professionals, the unit of analysis is hours per year. The unit of analysis of Activity for NPs may be either visits or hours per year, depending on the dataset employed. These estimates would have to be integrated prior to proceeding further in the model.
Step 3: Assessment of Alignment Between Service Requirements and Service Capacity
The next step in the model is an assessment of the alignment between service requirements and service capacity. In the case of perfect alignment, no further action is necessary. It is more likely, however, that a gap exists between the requirements of the population for healthcare services and the capacity of the workforce to deliver these services. In this case, an allocation process is suggested to explore ways of minimizing this gap.
Step 4: Allocation of Services Across Providers
The goal of this step in the planning process is to optimize the distribution of services, such that service capacity approximates service requirements as closely as possible.
Allocation may be conducted quantitatively, qualitatively (descriptively), or both. The leading practice quantitative allocation methodology uses a ‘plasticity matrix’[i]. This methodology compares actual workforce activities with a standard or benchmark activity distribution to examine the sufficiency of the existing workforce, accounting for flexible and overlapping scopes of practice within and between professions and specialties. The methodology enables the shifting of services within a given specialty or between specialties or professions to achieve an optimal distribution of services.
In the context of primary care services in the TC LHIN, shifting visits between family physicians and nurse practitioners presents an opportunity to optimize allocation of services (and the scope of practice of each). In order to accomplish this allocation using a plasticity matrix, detailed quantitative data regarding the scopes of practice of both family physicians and nurse practitioners and an understanding of where these scopes overlap and where they differ is necessary. Unfortunately, quantitative data of this sort is not of high enough quality and sufficient detail at this time. As such, we suggest that a descriptive allocation process be undertaken locally, with the goal of defining unique and overlapping scopes of practice and shifting services to minimize the gap between workforce requirements and workforce supply, to optimize skill mix, and to facilitate all providers practicing to the optimal extent of their scopes of practice. This can be integrated into a scenario analysis at the allocation phase.
The descriptive allocation process in the planning toolkit is inspired by adjusted service target-based planning.[ii],[iii],[iv],[v] The process estimates the requirements of a specific population for a defined package of primary care services. Services are then iteratively allocated to providers with relevant scopes of practice until optimal alignment between service requirements and service capacity is achieved. This approach accounts for factors that are often at play in health systems, including skill mix within the workforce, current or projected workforce availability, service costs, and emerging models of care.
Step 5: Final Assessment of Alignment Between Workforce Requirements and Workforce Supply
The gap between workforce requirements and workforce supply offers important insights into the state of local health systems and can be helpful in informing policy development. If requirements exceed supply in a defined geographic planning region, then plans can be made to supplement existing resources to better meet the needs of the population. Conversely, if workforce supply exceeds workforce requirements, then resources can be diverted to other areas experiencing greater need. In a system with limited resources, such analyses are foundational to effective and equitable distribution of healthcare resources. Local health leaders, primary care providers, and patients themselves can validate the gap analysis, indicate whether the results resonate with their experiences, and provide local intelligence to guide the planning process and offer potential solutions to local health service issues.
As described in Chamberland-Rowe, Simkin & Bourgeault PAPER 1, scenario analyses are designed to simulate the potential implications of changes that could occur within the system. Policy-makers can test the impact of different scenarios, at different points in the model, such as changing population characteristics, changing workforce profiles, or alternate models of care. Altering the value of various parameters within the model and examining the results facilitates decision-making that accounts for a range of possible futures. Scenario analyses may be supported by quantitative or descriptive data and provide an opportunity to engage stakeholders and incorporate local intelligence into decision-making. This promotes greater understanding and acceptance of planning and resource allocation decisions.