What’s Known

Different approaches are used in health workforce planning (HWFP) models, each with the country- and context-specific components. To the best of our knowledge, there is no systematic review study on the components of HWFP models.

What’s New

Approaches and components of HWFP models are investigated. Specific components related to each approach are identified.

Introduction

Workforce planning (WFP) was developed during the 1960s and early 1970s in a period of relative economic stability. ¹ It remained a major performance evaluation in the human resources (HR) departments until the economic recession hit in the 1980s, when the financial crisis undermined the process and economic value of WFP. Since then, WFP has again become the top priority for many organizations. ² The simplest and most commonly used definition of WFP is “employing the right number of people with the right qualifications, at the right job and at the right time”. ^{2 , 3} However, the most comprehensive definition that explicitly describes its procedural perspective is “WFP is a process in which an organization tries to estimate labor demand and assess the size, nature, and sources of supply to meet that demand”. ¹

According to the 2006 World Health Organization report, human health resources (also known as HR for health or health workforce), is defined as “all individuals whose job is to protect and improve the health of their communities”. ⁴ Health professionals play a major role in providing care to a community and are among the least readily accessible workforce. Health managers, both at national and local levels, continuously struggle with the task of managing health professionals in terms of increasing their effectiveness by distributing the workload fairly and increasing their productivity. ⁵

There are well-defined methods and procedures to address WFP across various professions. However, despite the development of several methodologies and approaches for health workforce planning (HWFP), there is as yet no uniformity in processes that accurately address the needs of physicians. ^{6 , 7} However, two main forecasting approaches have been defined, namely conceptual and analytical. The conceptual approaches include various factors influencing HWFP. The main approaches include demand-based, supply-based, target-based, needs-based, and benchmarking. Each approach relies on specific factors to predict the supply and demand of the health workforce. However, these do not take into account the dynamics within the system and between variables. Simulation methods and what-if scenarios can be used when determining the impact of different policy options on forecasts. ⁸ The analytical approach is designed to present analytical data in a suitable format for the health system decision-makers to support their workforce plans and policies. ^{9 , 10} Main approaches include system dynamics, trend modeling, artificial neural networks, Markov models, linear programming, nonlinear programming, simulation, and workload indicators of staffing need (WISN). ⁸

Different HWFP models are used in various countries. Some of these models are solely designed for a given context and many organizations have adapted an existing model to fit their specific context. ^{11 , 12} Forecasting the future health workforce depends on the selected forecasting model. Moreover, its appropriateness depends on factors such as the characteristics of the target population, prediction time period, the way each factor affects changes over time in care, the way of funding of the healthcare financial system, ¹³ and changes in workforce productivity. ¹⁴ Other factors, which are country-specific, include socioeconomic status, ^{15 , 16} level of education, ^{17 , 18} population age structure, ¹⁴ age and sex distribution, and even ethnic distribution. ^{19 , 20}

HWFP is the main responsibility of the Ministry of Health of each country with the important task of making the best use of national resources by training the right number of qualified people to promote health in society. Divergent methods are used in different countries, each with different contexts and specific variables. In addition, methodologies, approaches, and components used in HWFP are very diverse and at times even specific to an organization. In recent years, due to the importance of HWFP, Iran has developed an HR system and established a dedicated research center at the Iranian Ministry of Health and Medical Education. ²¹

Several studies have applied the above-mentioned models and approaches. However, to the best of our knowledge, none specifically investigated the various unique components of HWFP. Hence, the present study aimed to identify the approaches (dimensions) and components of HWFP models using a systematic review. Our findings will help health policymakers and experts devise an effective and evidence-based HWFP.

Materials and Methods

Search Strategy

A systematic review and meta-analysis of studies published between 2004 and 2021 were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist. ²² It is worth mentioning that despite the fact that WFP was initially developed in the 1960s and early 1970s, most of the research studies in this field have been published during the last 16 years. ²³ Therefore, we limited our search to the period 2004-2021.

A comprehensive search was performed on Web of Science, Scopus, Eric, PubMed Central, MEDLINE, and Elmnet. Medical subject heading (MESH) keywords in English and equivalents in Persian were used to search the databases. Using the syntax of various databases, two researchers independently searched the following terms: (“workforce” OR “workforces” OR “human resource” OR “human resources” OR “staffing” OR “staffings” OR “manpower” OR “manpowers” OR “womanpower” OR “womanpowers” OR “labor supply” OR “labor supplies”) AND (planning OR forecasting OR estimating OR estimation OR model) AND (health). For example, the syntax for the Web of Science search strategy was: (TI=[workforce OR “Human Resources” OR “Human Resource” OR Manpower OR Womanpower OR Staffing OR “Labor Supply”] AND TI=[planning OR estimating OR estimation OR forecasting OR model] AND TI=[health)) AND LANGUAGE:(English). Indexes=SCI-EXPANDED, SSCI, A&HCI, ESCI Timespan=2004-2021

The inclusion criteria were all studies published in Persian and English on the assessment of HWFP models, with a specific focus on health services delivery using input-output models, a clear formulation process, and with acceptable quality in accordance with the strengthening the reporting of observational studies in epidemiology (STROBE) checklist. The exclusion criteria were review articles, meta-analysis, and articles of poor quality.

Data Extraction

Duplicate articles were removed using EndNote software, version 20 (Clarivate Analytics, USA). Two researchers (S.B and Sh.Y) independently reviewed the articles according to the inclusion and exclusion criteria. In the case of rejection of an article, the reason was stated by the reviewer. Any disagreements were resolved by the third researcher. The extracted data included the name of the author, year of publication, country, target group (study group), model or formula, approaches (dimensions), and components (table 1).

Quality Assessment

The 22-item STROBE checklist was used to evaluate the quality of the articles. Each item is scored based on its relevance, ranging from 15 to 33. Articles that scored ≥20 points were considered acceptable for inclusion in our review study. ²⁴

Results

A total of 10,646 articles were identified, of which 2,295 were removed for duplication, an additional 8,294 after eligibility assessment, and 37 for failure to meet the inclusion criteria. Eventually, 20 articles were included for qualitative synthesis (figure 1).

Figure 1. The flow diagram depicts the selection process of studies in accordance with the PRISMA guidelines.

As presented in table 1, different HWFP approaches and methodologies have been used in different countries. The most commonly used in these studies were the mixed supply- and demand-based approaches (n=11, 55%), ^{25 - 35} and the least frequently used were the needs-based (n=3, 15%), ^{36 - 38} and target-based (n=1, 5%) approaches. ³⁶ Only one study used the skill mix of HR for assessment. ²⁷ One study used the artificial neural network WFP, ³⁹ and another study explicitly stated the use of a benchmark to estimate demand. ²⁵ In terms of analytical approaches, the most used approach was the system dynamics (n=5, 25%), ^{29 , 31 , 33 , 34 , 39} and the least frequently used the artificial neural network (n=1, 5%). ³⁹ The main target populations in the included studies were nurses (n=9, 45%) and general practitioners (n=5, 25%).

Supply-Based Approach

Of the included studies, supply was assessed in 14 studies, ^{25 - 35 , 38 , 40 , 41} from which 42 components used for modeling, forecasting, and estimation were extracted. The components were current stock, number of students, employment issues, choice of career, training capacity; study duration, student mobility during study years, new graduates and graduation rates, immigration, internal migration (e.g. transfer from one city to another or from one medical center to another); transfer from public to private sector, change of policies and training programs by the government, staff recruitment costs, hours spent per person in the clinic, number of inactive people; number and rate of active people, issues related to license (board), population growth, reduction of working hours, education dropout; job change, admission to higher education, mortality, illness, disability, retirement; emigration, average workforce to patient, workforce per service ratio, maternity leave, military service; attrition rate, full-time equivalent, target number of employees, average working hours per workforce (activity rate), average level of services provided per hour of work (participation rate); salary level, creation of collaborative networks, dialogue between employees and staffs, absence from work, productivity, physical time lag, and workplace (urban/rural).

Employment issues include employment rate, ^{40 , 41} number of new hires, ^{29 , 32} and number of public and/or private employment positions. ³¹ Training capacity includes the announced capacities or quotas, ^{31 , 32} number or rate of admissions or registrants in the field, ^{32 - 34 , 38 , 41} and number of educational programs. ⁴⁰ Internal migration includes people who are native to a country, but study at universities abroad and return to their home country after graduation and foreigners who enter the country and occupy a job position (non-native). Inactive in the specialized field includes unemployed people, ³¹ job change, ^{25 , 34 , 38} and people who have not been employed in their field of study for any reason. ^{27 , 35} Moreover, issues related to licensing that affected the supply include not renewing the license, ⁴⁰ and the acceptance/rejection rate in receiving the license. ³²

In several studies, the age and sex of health workers, ^{34 , 35 , 38} population, ³⁸ and movements among age groups of health workers ^{33 , 35} were considered. Of these, the most frequent components in supply-based modeling included graduates, current stock, emigration, training capacity, mortality, and retirement.

Demand-Based Approach

Of the included studies, demand was assessed in 16 studies, ^{25 - 36 , 39 , 42 - 44} from which 32 components used for modeling, forecasting, and estimation were extracted. The components were population changes, a workforce to population ratio, the number of insured people, the standard number of workforce per service, information time lag; economic issues, growth of the health service industry, change in the size of hospitals and health organizations, decision-making factors, need coefficient; the number of services and workload, activity standard, available working time per year, number of the workforce working hours per day, service coverage; the number of beds, bed occupancy rate, number of workforce per bed, time spent by the workforce per patient or time spent per visit or activity; the average number of visits per year, allowance factor, level of service utilization or demand rate, the prevalence of diseases, change in services provided by health insurance, number of healthy and ill people; treatment duration or sickness absence in days, health literacy, quality of life, service accessibility, per capita procedure rate by sex, labor market characteristics, and race.

The population changes include birth and mortality rates, ³¹ economic issues including economic conditions, ^{26 , 35} people’s income level, ³⁵ service costs, ³¹ and staff salaries. ^{31 , 32 , 35} Workload refers to the number of activities that each workforce can perform per unit of time, and activity standard refers to the time required to perform a standard activity by a fully trained person in the workplace. Moreover, available working time per year is equal to the number of days a person works per week multiplied by the number of weeks in a year, minus the number of days of voluntary and compulsory absences from work.

There are two types of allowance factors, namely the category of allowance factor (CAF) and individual allowance factor (IAF). CAF is a coefficient used to calculate the number of health workers required for health care services and support activities. IAF is the number of personnel required to perform additional activities. ⁴⁵ Other factors used in some other studies to estimate demand were the classification of diseases and treatments, ²⁷ classifications of age, ^{27 , 32 , 34} and sex ^{32 , 34} of health workers, and the age ^{34 , 35} and sex of the population. ^{33 , 34} Among these, the most frequently used components were population changes, level of service utilization or demand rate, and the number of services or workload of activities.

Needs-Based Approach

Three studies examined needs assessment models, ^{36 - 38} from which 11 components used for modeling, forecasting, and estimation were extracted. The components were need coefficient, population, number of services, activity standard, available working time per year, service coverage, prevalence or incidence of disease, cost-effective interventions, income, service delivery level, and health status. Among these, the most common were population, number of services, and prevalence of diseases. In the models developed by Tomblin Murphy and colleagues, an assessment of needs was made based on the age group and sex of the health workers, as well as the patient population. ³⁸

Target-Based Approach

One study used the target-based approach, ³⁶ extracting six components that were used for modeling, forecasting, and estimation. The components were the needs coefficient, size of the target population, number of services to be provided annually to each person, activity standard, available working time per year, and service coverage.

Discussion

In a systematic review, 20 studies published in English and Persian between 2004 and 2021 were reviewed. Components and approaches of WHFP models were identified. The forecasting approaches used in these models were conceptual with four components (supply, needs, demand, and target), and analytical with five components (system dynamics, simulation, regression, WISN, and artificial neural network). In one study, a benchmark was explicitly used to estimate demand. ²⁵

The results showed that HWFP models contain many unique components. The diversity of these components depends on the approach and context modeling used. The variety of components in the models confirms the fact that each country faces its own issues and conditions for HWFP. Each of these models is designed in a specific time and place, and thus may not apply to other communities with different conditions. Predicting the future workforce in the health sector depends on the choice of a forecasting method. The appropriateness of the prediction approach depends on the variables such as the characteristics of the target population, time period, factors affecting change in care over time, funding of the health financial system, and changes in workforce productivity. ¹⁴ Other variables include socioeconomic status, ^{15 , 16} level of education, ^{17 , 18} age and sex distribution, ethnicity distribution, ^{19 , 20} population age structure, and type of health care and treatment. ¹³ The education system, health policies, health literacy, and many other factors that are country- and context-specific must be taken into account. Therefore, the available models are not fully applicable in the context of every country, including Iran.

The diversity in the available models is indicative of the lack of a comprehensive and universally accepted approach. Most models tend to use common approaches to supply and simply replace the factor “people” with the “subject of interest”. Furthermore, it is clear that most forecasting models tend to predict the demand for services based on population growth rather than needs. Our results showed that only three studies used the needs-based approach, and those conducted in Iran did not use this approach at all. Tomblin Murphy and colleagues stated that the word “needs” is often used interchangeably with other words such as “demand” and even “use”, ⁴⁶ and it is, therefore, difficult to compare unique forecasting models developed with different default values and methods. ⁴⁷ The needs-based approach to HWFP, especially if evidence-based, is an area of innovation. ⁴⁸

Most of the forecasting models do not include the dynamic relationship between components, ^{49 , 50} for which the use of system dynamics modeling is most appropriate. ⁵¹ On the other hand, the impact of different decisions and policies on HWFP can be predicted using simulation. ³⁹ Therefore, a combination of system dynamics modeling and simulation systems can provide the best results. Regression, one of the components of the analytical forecasting approach, allows the estimation of the effect of different parameters and their impact on the supply, demand, and needs of health care professionals.6 However, regression methods do not manage sensitivity analysis. ⁵²

The main limitation of this study was the exclusion of articles in other languages than Persian and English. Therefore, their data and findings are not included in our systematic review.

Conclusion

Different approaches are used in health workforce planning models, each with specific strengths and weaknesses. Depending on the context, the use of the most appropriate model or a combination of models is recommended. Given that each model contains context-specific components, the inclusion or exclusion of a component has a significant impact on the forecasting accuracy and power of the HWFP model used. On the other hand, HWFP should be based on context and evidence. It is strongly recommended to include the specific conditions and context of a country when using a HWFP model. In the case of Iran, inclusion of socioeconomic status, epidemiological situation, goals of the health system, and the actual needs of health services is recommended.

Acknowledgment

The present manuscript is part of a PhD thesis in Medical Education at Shahid Beheshti University of Medical Sciences (Tehran, Iran). The authors would like to thank all researchers for their collaboration.

Authors’ Contribution

All authors were fully or partially involved in the study concept, design, data acquisition/interpretation/analysis, and draft version/critical review of the manuscript. The authors have read and approved the final manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Conflict of Interest:

None declared.

References

1. Reilly P. Human Resource Planning: An Introduction. ERIC: Report 312; 1996.
2. Sinclair A. Institute for Employment Studies: Workforce planning: A literature review; 2004.
3. Adano U. Collection and analysis of human resources for health (HRH) strategic plans. Chapel Hill: Capacity Project. 2006.
4. Organization WH. The world health report 2006: working together for health. World Health Organization: Geneva; 2006.
5. Doosty F, Maleki MR, Yarmohammadian MH. An investigation on workload indicator of staffing need: A scoping review. J Educ Health Promot. 2019; 8:22. Publisher Full Text | DOI | PubMed
6. Roberfroid D, Leonard C, Stordeur S. Physician supply forecast: better than peering in a crystal ball?. Hum Resour Health. 2009; 7:10. Publisher Full Text | DOI | PubMed
7. Ono T, Lafortune G, Schoenstein M. Health workforce planning in OECD countries: a review of 26 projection models from 18 countries. OECD iLibrary. 2013.
8. Rafiei S, Abdollahzade S, Hashemi F. Health manpower forecasting: A systematic review for models and approaches. Bangladesh Journal of Medical Science. 2019; 18:458-72. DOI
9. Klarman HE. Economic Aspects of Projecting Requirements for Health Manpower: A Review Article. The Journal of Human Resources. 1969; 4:360-76. DOI
10. Markham B, Birch S. Back to the future: a framework for estimating health-care human resource requirements. Can J Nurs Adm. 1997; 10:7-23. PubMed
11. Committee AHWA. Australian Health Workforce Advisory Committee: Nursing Workforce Planning in Australia: A Guide to the Process and Methods Used by the the Australian Health Workforce Advisory Committee; 2004.
12. Keet MP, Henley LD, Power HM, Heese HD. Medical manpower--South African situation models for planning and recommendations. S Afr Med J. 1990; 78:591-7. PubMed
13. Wendt C. Mapping European healthcare systems: a comparative analysis of financing, service provision and access to healthcare. Journal of European Social Policy. 2009; 19:432-45. DOI
14. Evashwick C, Rowe G, Diehr P, Branch L. Factors explaining the use of health care services by the elderly. Health Serv Res. 1984; 19:357-82. Publisher Full Text | PubMed
15. Agerholm J, Bruce D, Ponce de Leon A, Burstrom B. Socioeconomic differences in healthcare utilization, with and without adjustment for need: an example from Stockholm, Sweden. Scand J Public Health. 2013; 41:318-25. DOI | PubMed
16. Van der Heyden JH, Demarest S, Tafforeau J, Van Oyen H. Socio-economic differences in the utilisation of health services in Belgium. Health Policy. 2003; 65:153-65. DOI | PubMed
17. Hulka BS, Wheat JR. Patterns of utilization. The patient perspective. Med Care. 1985; 23:438-60. DOI | PubMed
18. Machry RV, Tuchtenhagen S, Agostini BA, da Silva Teixeira CR, Piovesan C, Mendes FM, et al. Socioeconomic and psychosocial predictors of dental healthcare use among Brazilian preschool children. BMC Oral Health. 2013; 13:60. Publisher Full Text | DOI | PubMed
19. Park E, Kim J. Gender- and age-specific prevalence of metabolic syndrome among Korean adults: analysis of the fifth Korean National Health and Nutrition Examination Survey. J Cardiovasc Nurs. 2015; 30:256-66. DOI | PubMed
20. Loh PT, Toh MP, Molina JA, Vathsala A. Ethnic disparity in prevalence of diabetic kidney disease in an Asian primary healthcare cluster. Nephrology (Carlton).. 2015; 20:216-23. DOI | PubMed
21. Asghari Sarem A, Markazi Moghaddam N. Analysis of success crucial factors for human resource development at health sector. Quarterly Journal of Public Organizations Management. 2015; 3:27-50.
22. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009; 339:b2700. Publisher Full Text | DOI | PubMed
23. Ahern S, Woods N, Kalmus O, Birch S, Listl S. Needs-based planning for the oral health workforce - development and application of a simulation model. Hum Resour Health. 2019; 17:55. Publisher Full Text | DOI | PubMed
24. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014; 12:1495-9. DOI | PubMed
25. Bazyar M, Noori Hekmat S, Rafiei S, Mirzaei A, Otaghi M, Khorshidi A, et al. Supply-and-demand projections for the health workforce at a provincial level from 2015 to 2025 in Ilam, Iran. Proceedings of Singapore Healthcare. 2021; 30:18-27. DOI
26. Chung SH, Jung DC, Yoon SN, Lee D. A dynamic forecasting model for nursing manpower requirements in the medical service industry. Service Business. 2010; 4:225-36. DOI
27. Harper P, Kleinman E, Gallagher J, Knight V. Cost-effective workforce planning: optimising the dental team skill-mix for England. Journal of Enterprise Information Management. 2013; 26:91-108. DOI
28. Juraschek SP, Zhang X, Ranganathan V, Lin VW. United States registered nurse workforce report card and shortage forecast. Am J Med Qual. 2012; 27:241-9. DOI | PubMed
29. Koichubekov B, Kharin A, Sorokina M, Korshukov I, Omarkulov B. System dynamics modeling for general practitioner workforce forecasting in Kazakhstan. Ann Ig. 2021; 33:242-53. DOI | PubMed
30. Landry MD, Hack LM, Coulson E, Freburger J, Johnson MP, Katz R, et al. Workforce Projections 2010-2020: Annual Supply and Demand Forecasting Models for Physical Therapists Across the United States. Phys Ther. 2016; 96:71-80. DOI | PubMed
31. Leerapan B, Teekasap P, Urwannachotima N, Jaichuen W, Chiangchaisakulthai K, Udomaksorn K, et al. Correction to: System dynamics modelling of health workforce planning to address future challenges of Thailand’s Universal Health Coverage. Hum Resour Health. 2021; 19:44. Publisher Full Text | DOI | PubMed
32. Oh Y. The future prospects of supply and demand for urologists in Korea. Investig Clin Urol. 2017; 58:400-8. Publisher Full Text | DOI | PubMed
33. Vanderby SA, Carter MW, Latham T, Feindel C. Modelling the future of the Canadian cardiac surgery workforce using system dynamics. Journal of the Operational Research Society. 2014; 65:1325-35. DOI
34. Lodi A, Tubertini P, Grilli R, Mazzocchetti A, Ruozi C, Senese F. Needs forecast and fund allocation of medical specialty positions in Emilia-Romagna (Italy) by system dynamics and integer programming. Health Systems. 2016; 5:213-36. DOI
35. Nooney JG, Lacey LM. Validating HRSA’s nurse supply and demand models: a state-level perspective. Nurs Econ. 2007; 25:270-8. PubMed
36. Fakhri A, Seyedin H, Daviaud E. A Combined Approach for Estimating Health Staff Requirements. Iran J Public Health. 2014; 43:107-15. Publisher Full Text | PubMed
37. Gailey S, Bruckner TA, Lin TK, Liu JX, Alluhidan M, Alghaith T, et al. A needs-based methodology to project physicians and nurses to 2030: the case of the Kingdom of Saudi Arabia. Hum Resour Health. 2021; 19:55. Publisher Full Text | DOI | PubMed
38. Tomblin Murphy G, Birch S, MacKenzie A, Alder R, Lethbridge L, Little L. Eliminating the shortage of registered nurses in Canada: an exercise in applied needs-based planning. Health Policy. 2012; 105:192-202. DOI | PubMed
39. Ghosh B, George C. Computer-assisted physician requirement planning. Journal of Health Management. 2006; 8:157-66. DOI
40. Hooker RS, Muchow AN. Supply of physician assistants: 2013-2026. JAAPA. 2014; 27:39-45. DOI | PubMed
41. Tjoa A, Kapihya M, Libetwa M, Schroder K, Scott C, Lee J, et al. Meeting human resources for health staffing goals by 2018: a quantitative analysis of policy options in Zambia. Hum Resour Health. 2010; 8:15. Publisher Full Text | DOI | PubMed
42. Seo S, Spetz J. Demand for temporary agency nurses and nursing shortages. Inquiry. 2013; 50:216-28. DOI | PubMed
43. Azimi Naibi B, Mohebbifar R, Rafiei S. Estimating the number of required nurses in an emergency department of a hospital in Qazvin: Application of WISN method. Journal of Inflammatory Diseases. 2018; 22:28-37. DOI
44. Sanaeifar E, Mohaghegh B, Moghri J, Tabatabaee SS. Human Resources Estimation of the Linear Accelerator Sector in Reza Radiotherapy-Oncology Center Based on Workload Indicator of Staffing Needs (WISN). Journal of Health Administration. 2021; 24:23-33. DOI
45. Organization WH. Workload indicators of staffing need. WHO: Geneva; 2010.
46. Tomblin Murphy G, Birch S, MacKenzie A, Bradish S, Elliott Rose A. A synthesis of recent analyses of human resources for health requirements and labour market dynamics in high-income OECD countries. Hum Resour Health. 2016; 14:59. Publisher Full Text | DOI | PubMed
47. Tomblin Murphy G, MacKenzie A, Guy-Walker J, Walker C. Needs-based human resources for health planning in Jamaica: using simulation modelling to inform policy options for pharmacists in the public sector. Hum Resour Health. 2014; 12:67. Publisher Full Text | DOI | PubMed
48. Tomblin Murphy G, Gilbert JH, Rigby J. Integrating interprofessional education with needs-based health workforce planning to strengthen health systems. J Interprof Care. 2019; 33:343-6. DOI | PubMed
49. Dreesch N, Dolea C, Dal Poz MR, Goubarev A, Adams O, Aregawi M, et al. An approach to estimating human resource requirements to achieve the Millennium Development Goals. Health Policy Plan. 2005; 20:267-76. DOI | PubMed
50. Goodman DC, Committee on Pediatric W. The pediatrician workforce: current status and future prospects. Pediatrics. 2005; 116:e156-73. DOI | PubMed
51. Saraji MK, Sharifabadi AM. Application of system dynamics in forecasting: a systematic review. International Journal of Management, Accounting and Economics. 2017; 4:1192-205.
52. Lopes DF, Ramos AL, Castro EA. The health workforce demand: a systematic literature review. Cien Saude Colet. 2021; 26:2431-48. DOI | PubMed