Advancing the population needs-based health workforce planning methodology: A simulation tool for country application

Although the conceptual underpinnings of needs-based health workforce planning have developed over the last two decades, lingering gaps in empirical models and lack of open access tools have partly constrained its uptake in health workforce planning processes in countries. This paper presents an advanced empirical framework for the need-based approach to health workforce planning with an open-access simulation tool in Microsoft® Excel to facilitate real-life health workforce planning in countries. Two fundamental mathematical models are used to quantify the supply of, and need for, health professionals, respectively. The supply-side model is based on a stock-and-flow process, and the need-side model extents a previously published analytical frameworks using the population health needs-based approach. We integrate the supply and need analyses by comparing them to establish the gaps in both absolute and relative terms, and then explore their cost implications for health workforce policy and strategy. To illustrate its use, the model was used to simulate a real-life example using midwives and obstetricians/gynaecologists in the context of maternal and new-born care in Ghana. Sensitivity analysis showed that if a constant level of health was assumed (as in previous works), the need for health professionals could have been underestimated in the long-term. Towards universal health coverage, the findings reveal a need to adopt the need-based approach for HWF planning and to adjust HWF supply in line with population health needs. In operationalising the conceptual framework, we present an analytical framework in which we identify four interrelated estimations of the supply of health workers; the need for health workers; gap analysis; and cost implications. Two distinct, fundamental mathematical models have been defined to quantify the supply of, and need for, health professionals, respectively. These are then compared to establish the gaps in both absolute and relative terms as well as the cost to be compared with available (and anticipated) budget or fiscal space—the main driver of how needs are translated into actual demand for health workers. The supply of health professionals refers to the pool of qualified health professionals who are willing to find appropriate jobs to offer health services. Thus, the supply (S) depends on the ‘stock’ (T) and ‘flows’ of the health workforce. The stock refers to the current number of the active health workforce, while the flows have two components: inflow and outflow. The inflow represents new entrants to the labour market from domestic training pathways and through immigration, whereas the outflow represents both voluntary exits (loss to other sectors, emigration) and involuntary exits such as retirement, ill health and death [14]. As defined in the National Health Workforce Account (NHWA) [38] adopted by the World Health Assembly in 2017, the stock of any group of health professionals can be categorised into three groups: registered/overall stock; professionally active stock; and the practising stock. The registered/overall stock comprises all those who have registered with the relevant professional regulatory body or authority within a jurisdiction to practice a health profession, irrespective of whether they are within the jurisdiction and practising or not. Within the overall stock (registered), those that maintain good standing with the professional regulatory body by renewing their licenses, are considered the professionally active stock—some of whom may be engaged in other career interests (such as teaching, research and policy) rather than direct health service delivery. Out of the professionally active stock, those that are engaged in (or are willing to find jobs to be engaged in) direct health service delivery are considered the practising stock of health workers. The interest in health service planning is to estimate as close as possible, the current and future stock of the practising health workforce. We modelled the supply using a stock and flow process [29,39] as represented in Figure 1 and illustrated in the following equation: Formulae for attrition rate and inflows as defined in equation one is contained in supplementary material 2. Before the model development, we undertook a scoping review of analytical applications of the needs-based approach of health workforce planning [32] in which we synthesised the main considerations towards methodological harmonisation and increased transparency in needs-based health workforce modelling which mostly informed this empirical framework (see Box 1 for a summary of the critical considerations for needs-based health workforce planning synthesised form the scoping review). The need for health services is a function of three broad parameters, namely, demographic characteristics of the population (size, gender, age distribution and geographical location); their level of health or health status (disease prevalence and risk factors); and the services planned to address the health deficits or otherwise necessary to maintain optimal health. Building on the works of Birch and colleagues [35] which was further developed by Mackenzie and colleagues [29], we introduce an explicit adjustment for the instantaneous rate of change of the health status or level of health of the population. The relationship between parameters determining the need for health services could then be mathematically expressed as: where: Equation (2) represents the evidence-based need for services that address the fundamental equation of how many of the population by age cohort, gender, in a particular location will need a specific type of services? This is vital for any aspect of health service planning—health workforce, essential medicines, infrastructure, or equipment, among others. Translating the need for health services or evidence-based service requirements generated from Equation (2) into health workforce requirements is only feasible by making explicit assumptions about some measure(s) of the productivity of health professionals for the specific services that are planned or otherwise required by the population. We define the measure of the productivity of health professionals borrowing the concept of Standard Workload (SW), which underpins the widely used and well-documented Workload Indicators of Staffing Need (WISN) tool developed by the World Health Organisation [37,40,41]. The standard workload is defined as the amount of work of a particular service delivery task that one health professional who is well trained could perform in a year if the health professional dedicated all his/her working to delivering that service [37]. It is a function of two components: (a) The service standard (SS) for the activity to be performed—the average time that a well-trained and motivated health professional will spend to perform the service delivery activity to acceptable professional standards in the context of the jurisdiction; and (b) the available working time (AWT)—the time a health worker was available in one year to do his/her work, taking into account all absences. Equation (3) illustrates the concept of standard workload and supplementary material 2 contains a formula for calculating the available working time. where: When more than one type of health professional category perform a service delivery task, MacKenzie and colleagues defined a variable for work division [29]. To account for this variable, the estimated need for health services derived in Equation (2) is adjusted for the proportion of work division (which can be represented by W) to get the number of service activity, y to be performed by a health professional of category n for individuals of health status h, age group i, gender j at location g over time t. The workload division adjusted need for health services can then be divided by the standard workload (defined in Equation (3)), as illustrated in Equation (4). It is worth noting that the standard workload is a measure of productivity relating only to direct patient/client service delivery activities that can be counted per person (or patient). Therefore, the needs-based health workforce requirements estimated using Equation (4) relates mainly to direct person services and excludes indirect patient care or catalytic activities of the health professional. As demonstrated by Birch et al. [27] and MacKenzie et al. [29], needs-based models recognise that health workers spend some time performing the activities that are essential for supporting the delivery of direct patient/person services, but such activities are not linked to individual patients/clients. While the previous needs-based models accounted for the phenomenon from a supply perspective by defining and adjusting health workforce supply by activity rates (the proportion of health worker’s time spent on direct care) [29], we take the view that for planning, it is appropriate to estimate the number of health professionals needed to cover such support activities that are catalytic for the direct patient care. For example, the process of handing over from one group of nurses in a shift to another group is a crucial component for continuity of care, but because the activity may not necessarily be counted per patient, it could be ignored in needs-based models. Leveraging on the WISN methodology [37], we define a support allowance standard (SAS) as the proportion of a health worker’s time that is spent on the support (or indirect patient care) activities. When the total SAS (in a proportion) is subtracted from the whole, the difference represents a proportion of the health worker’s AWT that is devoted to direct patient/person services [37]. To incorporate the SAS into Equation (4), an adjustment factor known as the support activities adjustment factor (SAAF) is defined—mathematically expressed as the inverse of the proportion of a health professional time left for direct per person care activities [37]. Thus, the overall need-based requirement with both direct and indirect services can be expressed as: where: The two main quantities estimated in Equations (1) and (5) above (supply and needs-based requirements) can be analytically integrated to compare the current and anticipated gaps in the health workforce in absolute and relative terms. The difference between the projected supply levels and the projected need for a particular health worker category is considered the absolute gap whereby a negative gap is indicative of a supply shortfall, which will be deemed as the number ‘needed to be trained’ by the health professions education institutions [42]. In contrast, a positive gap is indicative of an over-supply from the health professions education institutions in comparison with the need. where: This Is the Ratio of the Needs-Based Health Workforce Requirement That Will Be Met by the Anticipated Supply The SAR shows the anticipated amount of needs-based workload pressure that will be on the current and future health workforce or the proportion of professional standards that can be maintained if interventions are not put in place to influence the supply (and employment) of health professionals. For interpretation, SAR of 1 indicates that the anticipated supply will optimally meet the needs-based requirements, whilst a SAR of less than one shows that the anticipated supply is failing to meet the needs-based requirements. On the other hand, when the SAR is greater than 1, it is indicative of supply outstripping the projected need for health professionals. To understand the cost implications and investments requirements for employing health workers, the following is conservatively specified: where: Similarly, assuming there is sufficient supply, the cost implications for filling the needs-based health workforce requirements is conservatively specified as: