A New Integrated Approach
by Desmond QUEK & John Pastor ANSAH
Fail to plan, plan to fail - the silver tsunami and the need for workforce projection
Health administrators and their human resource teams are often plagued with the difficult task of a timely assessment of future needs for the healthcare workforce. Healthcare forms a large proportion of the national budget in many countries,1 with healthcare manpower costs amounting to 60-70 percent of the healthcare budget. The size of the health workforce directly influences population health, healthcare costs, operation of the healthcare system and access to healthcare.2
The populations of many developed and developing countries are undergoing significant demographic changes, which have significantly increased the demand for healthcare resources and services. This demand is predicted to rise substantially with an aging population, as the prevalence of many common chronic ailments increase with age.3,4 In Singapore, for example, the population aged 65 and above is projected to rise by 207% from 2010 to 2050.5 This aging demographic shift, coupled with population growth and increasing life expectancy, is certain to lead to a substantial increase in eye disease and demand for eye care.
Technological innovations (pharmaceuticals, diagnostic and therapeutic equipment, and techniques) and organisational innovations targeted at improving the performance of healthcare systems, too, affect the demand for services. Additionally, a more highly educated, sophisticated elderly population having greater access to, awareness and expectations of services, will contribute to an increase in the eye care service demand. As the population ages, there will need to be a shift towards services to manage chronic eye conditions such as glaucoma and age related macula degeneration. This would require a careful tailoring of recruitment policies that determines not just the size of the eye care workforce but its composition and required skill sets as well.6
Demand and workforce projections are crucial first steps to design policies to address the “unmet need” for eye care.7 The decision is made more challenging by the training delay for ophthalmologists (who, in Singapore, after 5 years in medical school and 1 year of housemanship, spend 5 years in postgraduate residency training to become a specialist, and an additional 1-2 years in subspecialty fellowship training before being accredited to perform subspecialty work independently). This protracted training delay accentuates the need for proper manpower planning so that appropriate healthcare policies and training requirements can be put in place for the efﬁcient and timely delivery of health services.
Health workforce forecasting is a challenging task. The different approaches to health workforce forecasting are ill-deﬁned and confusing - the Organisation for Economic Co-operation and Development (OECD) broadly outlines ﬁve main approaches;8 the World Health Organisation (WHO) four;9 among many others.10,11,12 A combination of approaches, each with unique underlying assumptions, are customarily used and advocated. Of these, the workforce-to-population ratio, needs based, and utilisation based approaches are the most prominent. The assumptions that exist to simplify the complex health workforce planning process have signiﬁcant impact on forecasted results. Thus, it is crucial to have a thorough understanding of the various assumptions before adopting the use of any forecasting approach.
Existing workforce planning models
The workforce-to-population ratio, needs based approach and utilization based approach are three planning strategies commonly used.13,14,15,16,17,18,19 Each approach makes unique assumptions which have implications for the reliability of the forecast within the context of its application.
The workforce-to-population ratio is a simplistic approach for determining the number of healthcare personnel required to serve a given population, which can be cross referenced with benchmarks or expert opinions. It considers demographic data such as population growth and information on the workforce and adjustments can be made to account for factors such as differing utilization rates by age or gender and attrition rates of the health workforce.13, 20 In this approach, a reference country or region with a slightly more developed healthcare sector than that to be investigated is selected and its ratio assumed to be the bench mark.21 Despite its speed and ease of application, this approach often neglects factors such as productivity, utilization, and distribution ofhealthcare personnel, making interpretation of the results difﬁcult, and resulting in persistent unequal distribution of healthcare workforce with the projected estimates.21,22
As its name suggests the needs based approach determines health workforce requirements by considering the current estimated healthcare needs of a population11 – the number of healthcare professionals or quantity of services required to provide optimal healthcare services to maintain a healthy population. Fundamental to this approach are demographic characteristics such as the disease prevalence, age, gender, and education level of a population.11 This approach assumes that all healthcare needs will be met; economical methods to address the needs can be established; and healthcare resources are consumed according to relative levels of needs. The needs based approach addresses the healthcare needs of the population and is unaffected by current health service utilization. It is logical, comprehensible, consistent with professional ethics and can be employed as an advocacy tool. However, it demands extensive epidemiological data, which is often unavailable. It also overlooks the efﬁciency of the allocation of resources and requires regular updating of certain variables such as the level of technology. As such, projected staff and service targets may be unattainable.21
Current levels of service utilization by the population can be taken as a proxy for satisﬁed demand, and used to estimate the future healthcare workforce requirements in the utilization based approach. Satisfied demand refers to levels of healthcare services a population will seek and have the ability to acquire at the current pricing within a certain timeframe. Like the needs based approach, the utilization based approach relies on demographic information such as disease prevalence, age, gender, and education levels. Additionally, utilization patterns of healthcare services and the market factors that inﬂuence these patterns are also considered.11 This approach assumes that the current level, combination, and distribution of health services adequately meet the current demand for healthcare. Furthermore, age- and gender- speciﬁc requirements are assumed to be held constant into the future; and that demographic changes over time can be predicted based on prevailing trends.23 The assumption that there is little or no change in the population-speciﬁc utilization patterns makes the utilization based approach useful in predicting economically feasible targets,21 and effective in studies of geographical variations, where utilization patterns are stratiﬁed. However, skewed projections can result from this method due to failure to account for changes in future utilization patterns. The unavailability of information on the utilization and demand for healthcare services, especially in the private sector, limits the adoption of this approach, which also overlooks the disparity between demand, utilization, and needs for services.22
An integrated, melting-pot approach
Researchers at the DUKE-NUS Medical School, Singapore National Eye Centre (SNEC) and Singapore Eye Research Institute have collaborated to build a model that predicts the requirements for and supply of ophthalmologists for Singapore under plausible future scenarios. The new model incorporates credible estimates of the prevalence of eye diseases and the demand of eye care service, and how the demand for these services translates into workforce requirements and supply.24 The model adopts an integrated approach that explicitly considers factors such as changes in demographic and healthcare characteristics rather than replacing them with simplifying assumptions, making it more feasible and robust.
The new integrated model estimates healthcare workforce requirements by projecting the healthcare needs or quantity of services required to provide optimal healthcare services to maintain a healthy population (needs based approach). It forecasts the likely current and future demand (utilization plus waiting list) of healthcare services among the population with healthcare needs, using demand data and expert opinions, taking into consideration expected demographic, health policy and technological changes. Demand herein refers to healthcare utilization combined with the time lapse between appointment booking and the actual patient visit (wait list).
Similar to its constituent approaches, the integrated approach combines information on demographic characteristics such as disease prevalence, age, gender, education level, utilization of healthcare services, wait list of patients seeking healthcare and the market factors that inﬂuence usage. Simultaneously, other features such as ﬁnancing, expansion of services and changing expectations of healthcare are also incorporated in the healthcare utilization estimates. While the assumptions of this approach are analogous to those of its constituent approaches, some differences exist due to the complementary nature of the two approaches. By considering those who require healthcare services and their care seeking behavior within the context of the health system and the likely expected future changes, this new model aims to mitigate the limitations of the individual approaches and better project future health workforce demands.
The new integrated model consists of four related modules – population structure, disease prevalence, demand for eye care services and supply of ophthalmologists.
Change in population size and structure. Using publicly available national population statistic data,25,26,27,28 considering births, death, immigration and emigration, we developed a population model that illustrates the aging process of Singapore’s resident population and age distribution of the population disaggregated by 1-year age cohorts by gender, ethnicity and educational attainment. The model predicts, that by 2040, the total population of Singapore is expected to increase by 23% from 5.6 to 6.9 million, with the number of residents greater than 60 year-old doubling to 1.6 million. Residents between 55-70 years of age will form the largest group, and proportion of residents having secondary and tertiary education will increase from 53% to 82%.
Prevalence of disease. The prevalence of eye diseases from the Singapore Epidemiology of Eye Diseases (SEED) study 29,30,31 for resident Singaporeans 40 years and older, disaggregated by age, ethnicity and educational attainment, was applied to the population model of resident Singaporeans.
The eye conditions included cataract, diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and myopia. Based on known prevalence and predicted change in population size and structure, it can be calculated that by 2040, the number of patients with specific eye conditions will increase to 182% (of 2015 numbers) for cataract, 213% for DR, 202% for glaucoma and 155% for AMD.
Care seeking behaviour and demand for eye care services. Information on demand for eye care services in the public sector was obtained by considering the number of visits and patients on the waiting list, as well as the case-mix of patients seen. Assuming that the change in number of all new patients seeking care will be proportional to the change in the prevalence of conditions among residents, we projected the number of new patients entering the system over the time frame of the simulation. Based on current utilization and care-seeking patterns, we applied a rate of 2.4 visits per year per patient and a service uptake factor 9.7%. Scenarios of increased care seeking behaviour and alternate models of care (which may reduce demand) can then be conveniently simulated using this module.
Supply of ophthalmologists. The ophthalmologist supply module is a continuous time compartment model that tracks the changing number of ophthalmologists employed over time in the public sector, as well as the training pipeline of ophthalmologists. The change in the number of ophthalmologists is a result of new hires and attrition, which is a blended value of retirements, deaths, and resignations. The model takes into consideration the workload per ophthalmologist (patient visits per year per ophthalmologist), proportion of time spent on clinical work, as well as the average attrition rate from public service.
By arriving at the demand for eye services based on population statistics, disease prevalence and care seeking behaviour, and matching this with the supply of ophthalmologists, we can calculate the number of ophthalmologists required and hence the number of ophthalmology residents that need to be trained each year.
Application and limitations of the model
This new model enables health administrators to predict, based on current productivity rates and “business-as-usual” operational procedures, the number of ophthalmologists and ophthalmology trainees a year required to maintain the same standard of care and accessibility to care (wait-times, visits per year). By altering the variables in the model such as moderation of workload (number of visits per ophthalmologist per year or percentage of time dedicated to clinical work), or higher uptake of services due to a more educated population leading to increased care-seeking behaviour, similar estimates of the manpower required to support the changes in demand can be predicted.
It is important to bear in mind that the infrastructure (such as clinic space and operating theatres) needed to train and accommodate the projected numbers of ophthalmologists must be simultaneously planned and provided for. If, however, infrastructure or policy limits the number of ophthalmologists to be trained, then other measures need to be instituted to compensate for the shortfall. These may include new models of care that decrease demand (for example, decanting and right siting of patients with stable chronic eye conditions to primary eye care physicians who are non-ophthalmologists), or increase productivity (virtual clinics, or team-based care comprising of an ophthalmologist leading a team of non-ophthalmologists in patient care).
The proposed integrated model offers an additional holistic approach to health care manpower projection. When the differences and appropriateness of each approach, characteristics of the healthcare system and population over short, medium and long terms are carefully considered, the new integrated approach was found to be most versatile and suitable for health workforce planning compared to the first three distinct conventional forecasting approaches.32
About the Authors
Dr Desmond Quek
Singapore National Eye Centre
MMed(Ophth), FRCS(Ed), FAMS, MBA
Dr Desmond Quek is a Consultant at the Singapore National Eye Centre (SNEC). He completed his medical training with the Faculty of Medicine, National University of Singapore (NUS) in 2001, where he was awarded the NUS undergraduate scholarship, and Dean’s list award for academic excellence. He was also awarded the Ministry of Health Post-graduate Training Scholarship in 2007 for specialist training. Dr Quek has held various administrative appointments, including Chief Resident of SNEC in 2008 and Chief Registrar of the Division of Ophthalmology of Changi General Hospital in 2009. He became a Member of the Royal College of Surgeons of Edinburgh in 2007, and was awarded the Masters of Medicine (Ophthalmology) by NUS in 2008. He was subsequently appointed as a Fellow of the Royal College of Surgeons of Edinburgh, as well as a Fellow of the Academy of Medicine, Singapore (FAMS), upon successful completion of his ophthalmology residency in 2011. His practice interests include general ophthalmology, adult cataract surgery and glaucoma.
After successfully completing Glaucoma sub-specialty fellowship training with at SNEC, Dr Quek went on to pursue an executive Masters of Business Administration at INSEAD in 2013, and was appointed Director of Medical Informatics at SNEC from 2014-2017 when he spearheaded the development of electronic medical records. He is currently Chair of the IT committee at SNEC.
Dr Quek has published scientific papers in respected peer-reviewed medical journals, and presented at numerous international and local conferences. His research interests are in glaucoma and cataract surgery. He is active in teaching medical students and ophthalmology residents, and is a clinical senior lecturer at the Yong Loo Lin School of Medicine, NUS, and an adjunct assistant professor at the DUKE-NUS Medical School, Singapore.
Dr John P. Ansah, is a Research Assistant Professor in the Signature Research Program in Health Services and Systems Research at Duke-NUS Medical School and a Faculty Fellow at Residential College 4 in the National University of Singapore. He holds a Ph.D. in the System Science methodology of System Dynamics from University of Bergen. Dr. Ansah is well versed in the application of System Dynamics to national policy making and implementation. He played an instrumental role in developing the “Threshold 21” dynamic simulation model used by UNEP in the Green Economy Report to analyze the impacts of sustainable green investment in supporting the transition to a global green economy. His research interests lie broadly in developing healthcare strategy and planning simulation models that are rigorous, evidence-based and customized to the optimal usefulness of stakeholders to inform policy.
- Bloor K, Maynard A, Hall J. Planning human resources in health care: towards an economic approach: an international comparative review. 2003: Canadian Health Services Research Foundation (Fondation canadienne de la recherche sur les Services de santé).
- Scott A. Alternative Approaches to Health Workforce Planning School of Population Health, University of Queensland. 2011.
- Denton FT, Spencer BG. Chronic Health Conditions: Changing Prevalence in an Aging Population and Some Implications for the Delivery of Health Care Services. Canadian Journal on Aging (Revue canadienne du vieillissement). 2010; 29(1):11-21.
- Wong TY, Loon S-C, Saw S-M. The epidemiology of age related eye disease in Asia. Br J Ophthalmol. 2006;90:506–11.
- Asher MG, Nandy A. Singapore’s policy responses to ageing, inequality and poverty: an assessment. Int Soc Sec Rev. 2008;61(1):41–60.
- OECD. Health at a Glance 2013: OECD Indicators. 2013: OECD Publishing.
- Zheng Y, Cheng CY, Lamoureux EL, Chiang PP, Rahman Anuar A, Wang JJ, et al. How much eye care services do Asian populations need? Projection from the Singapore Epidemiology of Eye Disease (SEED) study. Invest Ophthalmol Vis Sci. 2013;54(3):2171–7.
- Ono T, Lafortune G, Schoenstein M. Health workforce planning in OECD countries. 2013.
- World Health Organization. Models and tools for health workforce planning and projections. Geneva. 2010.
- O’Brien-Pallas L, et al. Forecasting models for human resource in health care. Journal of Advanced Nursing. 2001; 33(1):120 129.
- Roberfroid D, Leonard C, Stordeur S. Physician supply forecast better than peering in a crystal ball? Human Resources for Health 2009; 7(1):10-10.
- Lopes MA, Almeida ÁS, Almada-Lobo B. Handling healthcare workforce planning with care: where do we stand? Human Resources for Health. 2015; 13(1):38.
- Cromwell J, et al. CRNA Manpower Forecasts: 1990-2010. Medical Care. 1991; 29(7): 628-644.
- Grumbach K, et al. Physician supply and medical education in California. A comparison with national trends. Western Journal of Medicine. 1998; 168(5): 412.
- Healy E, Kiely PM, Arunachalam D. Optometric supply and demand in Australia: 2011–2036. Clinical and Experimental Optometry. 2015; 98(3): 273-282.
- Murphy GT, et al. An applied simulation model for estimating the supply of and requirements for registered nurses based on population health needs. Policy, Politics, & Nursing Practice. 2009; 10(4): 240-251.
- Ishikawa T, et al. Forecasting the absolute and relative shortage of physicians in Japan using a system dynamics model approach. Human Resources for Health. 2013; 11(1): 41-41.
- Burke BT, et al. A needs-based method for estimating the behavioral health staff needs of community health centers. BMC Health Services Research. 2013; 13(1): 245-245.
- Zimbelman JL, et al. Physical therapy workforce in the United States: forecasting nationwide shortages. PM&R. 2010; 2(11): 1021- 1029.
- Weiner JP. Forecasting the Effects of Health Reform on US Physician Workforce Requirement: Evidence From HMO Stafﬁng Pat- terns. JAMA. 1994; 272(3): 222-230.
- Dreesch N, et al. An approach to estimating human resource requirements to achieve the Millennium Development Goals. Health Policy Plan. 2005; 20(5): 267-76.
- Dussault G, et al. Assessing future health workforce needs. 2010: World Health Organization Copenhagen.
- O’Brien-Pallas L, et al. Integrating workforce planning, human resources, and service planning. Human Resources for Health Development Journal. 2001; 5(1-3): 2-16.
- Ansah JP, De Korne D, Bayer S, Pan C, Jayabaskar T, Matchar DB, Lew N, Phua A, Koh V, Lamoureux E, Quek D. Future requirements for and supply of ophthalmologists for an aging population in Singapore. Human Resources for Health. 2015 Nov 17;13:86.
- Singapore Department of Statistics. Key indicators of resident households. Singapore: Singapore Department of Statistics; 2010.
- Ansah JP, Matchar DB, Love SR, Malhotra R, Do YK, Chan A, et al. Simulating the impact of long-term care policy on family eldercare hours. Health Serv Res. 2013;48(2 PART2):773–91.
- Eberlein RL, Thompson JP, Matchar DB. Chronological ageing in continuous time. 30th International Conference of the System Dynamics Society; 2012. Albany: System Dynamics Society; 2012.
- Thompson JP, Riley CM, Eberlein RL, Matchar DB. Future living arrangements of Singaporeans with age-related dementia. Int Psychogeriatr. 2012;24(10):1592–9.
- Pan C-W, Wong TY, Lavanya R, Wu R-Y, Zheng Y-F, Lin X-Y, et al. Prevalence and risk factors for refractive errors in Indians: the Singapore Indian Eye Study (SINDI). Invest Ophthalmol Vis Sci. 2011;52(6):3166–73.
- Lavanya R, Jeganathan SE, Zheng Y, Raju P, Cheung N, Tai ES, et al. Methodology of the Singapore Indian Chinese Cohort (SICC) Eye Study: quantifying ethnic variations in the epidemiology of eye diseases in Asians. Ophthalmic Epidemiol. 2009;16(6):325–36.
- Rosman M, Zheng Y, Lamoureux E, Saw S, Aung T, Tay W, et al. Review of key findings from the Singapore Malay Eye Study (SiMES-1). Singapore Med J. 2012;53(2):82–7.
- Ansah JP, Koh V, De Korne D, Bayer S, Pan C, Jayabaskar T, Matchar DB, Quek D. Comparing health workforce forecasting approaches for healthcare planning: The case for ophthalmologists. International Journal of Heathcare. 2017;3(1):84-96.