Supporting Strategic Staffing Decisions in Emergency Medical Services

(1)

Supporting Strategic Staffing Decisions in

Emergency Medical Services

University of Groningen

Faculty of Business and Economics

MSc Technology and Operations Management

Thesis

Student: Stef Lensink

Student number: S2517892

E-mail: s.j.w.lensink@student.rug.nl

University supervisor: dr. ir. D.J. van der Zee

(2)

(3)

Abstract:

Purpose: The burden on Emergency Medical Service (EMS) providers is increasing. Demand for nursing staff rises due to changes in the healthcare environment, social changes and innovations in the EMS process. Simultaneously, there is an increasing shortage on the nursing labor market. More pressure is put on analyzing supply as well as demand of nurses. More insight as to where to hire nursing staff and whether hiring numbers are sufficient is needed as well as more insights on what is required given current staff developments. A wide range of methods combined with inaccurate projections of population growth and lack of comprehensive databases have not improved forecasting accuracy yet for strategic staff requirements.

Method: This study developed a high-level architecture for strategic staff planning in the EMS industry, with subsequent implications for use of Decision Support Systems. Design science, a method in which an artifact is created through an iterative process in practical setting was used.

Result: This study resulted in a high-level architecture for strategic staff planning that guides as a migration path for the case companies to improve their strategic staff planning process. Important extension suggestions were use of qualitative demand-, and supply analysis, further extending supply analysis within and outside of organizational borders, the use of scenario analysis and the process’s embedding into the organization. Second contribution of this study is a maturity model which may aid in systematically improving strategic staff planning capabilities.

Conclusion: The Strategic Staff Planning process can be structurally improved by following the high-level ar chitecture and maturity model designed in this study. The architecture was found valid by company stakeholders and can effectively contribute in developing more insight in staff supply and demand on a strategic level. A limitation of this study is not considering all activities of the strategic staff planning process, providing a future research direction improving strategic staff planning in the EMS industry.

(4)

Preface

Before you lies the final project of my Master Technology & Operations Management at the University of Groningen. The process of writing this thesis was long but a fulfilling challenge. Researching the subject of Decision Support for Strategic Staff Planning in the Emergency Medical Service field was a pleasure, as it gave me a feeling of contributing knowledge assimilated during my studies at the University of Groningen and putting it to practice. I recognize the problem of nursing staff shortages in the EMS industry and contributing my academic knowledge to a societal problem gave me a sense of accomplishment, for which I would like to thank several persons that were involved in the process.

First and foremost, I would like to express my sincere gratitude to my University supervisors dr. ir. Durk-Jouke van der Zee and dr. Nick Szirbik for their wise guidance and constructive feedback throughout the process of writing my thesis for the past 5 months.

Furthermore, several company representatives deserve recognition for their valuable insights in the strategic staff planning process regarding Emergency Medical Services. I would like to thank Bram Oosting, Lilian Janssen, Albert Schut and Yvonne Zagers from Ambulancezorg Groningen, Roel Bouwhuis and Carien van Well from Kijlstra Ambulancezorg and Jaap Hatenboer, Eline Kiewiet and Petra Nijmeijer from UMCG Ambulancezorg for providing me with helpful insights. I feel deeply grateful for being welcomed into their organizations and having gained the trust to help contribute to their performance.

Lastly, I would like to thank my family, friends and fellow students for all their warming support and motivation during the process of writing.

S. J.W. Lensink

(5)

List of Abbreviations

EMS Emergency Medical Services SSP Strategic Staff Planning

AA Anglo-American

FG Franco-German

RAV Regional Emergency Medical Service Provider AZN Dutch National Emergency Medical Services DSS Decision Support System

BPMN Business Process Model and Notation

FTE Full-time Equivalent; In this study used as full-time equivalent of employees PESTLE Political, Economic, Social, Technological, Legal, Environmental

(6)

1. Introduction

The health burden on Emergency Medical Service (EMS) is increasing as patient expectations are rising. Additionally, all organized care systems have to cope with the problems of increasing demand, increasing financial pressures, limitations on staff, and an ability to apply ever more complex medical processes to save the lives of patients (Robertson-Steel, 2006). Pressure on Emergency Medical Services (EMS) increase through patient expectations, financial pressure, limited availability of nursing staff, overall increased demand and more complicated medical processes (Robertson-Steel, 2006).

In The Netherlands this is demonstrated, among others, in the number of ambulance dispatches per year, showing a growth of 16% over the period of 2011-2015 (Kommer et al., 2015). Such an increase dispatches results in more demand for EMS resources, especially nursing staff, as they are considered valuable resources (Heidari et al, 2017). While demand for EMS is rising, staff supply is found to be cumbersome in many countries due to labor market shortages that are not easily resolved in the near future (Oulton, 2006; Drennan & Ross, 2019).

This study is motivated by three regional ambulance service providers (RAVs) serving the Northern Netherlands which will be indicated as case company 1 through 3. Combined, the RAVs staff about 370 emergency nurses and serve 1.7 million inhabitants (AZN, 2019). For all RAVs demand is expected to rise in coming years, resulting in an expected shortage of nursing staff. More pressure is put on the RAVs to analyze demand and supply of nurses, like where to hire and whether numbers are sufficient. Currently, such decision support is not available, causing possible mismatches in supply and demand of nursing staff.

Decision support for Strategic Staff Planning (SSP) entails four key activities, i.e. analyzing supply, demand and the gap between supply and demand and ways to resolve the gap (Anderson, 2004; Beitel et al, 2020; Huber, 2012). In the past years, a considerable amount of research has been devoted to decision making in an EMS setting, however, most of the research related to station locations and fleet dimensioning. Little attention was given to SSP in literature, resulting in underdeveloped SSP processes (Bélanger et al., 2018).

(8)

simulate the effects of adding an extra nurse on patients’ time spent in the system. Therefore, Zeltyn et al.’s (2011) paper will be used as an example of how gaps can be assessed in this study. Furthermore, Tayana et al.’s (2007) multi-criteria DSS allocates funds to close gaps between workforce supply and demand through analysis of current workforce supply and forecasted workforce requirements outside the healthcare industry. This study will focus on analyzing supply, future requirements and the gap between them, excluding solution analysis. Relevant efforts have been made towards SSP through use of DSS in and outside of healthcare. In EMS though, this remains an unexplored topic despite the labor market shortage.

The aim of this study is to design a high-level architecture for a Strategic Staff Planning process that provides accurate estimates on staff demand and supply, to enhance Strategic Staff Planning for Emergency Medical Services. Through a Design Science methodology, which defines and evaluates an artifact in its context (Wieinga, 2014), this study focusses on Decision Support Systems to reach this study’s objectives.

Furthermore, through designing the SSP architecture and implications for DSS, knowledge will be obtained which will contribute to EMS literature. Contributions are made about relevant data sources and analysis techniques to use and how these may enhance SSP. This research adds value to literature by examining decision support for SSP in an EMS environment, which has not been studied before. This study provides an architecture of the SSP process, seeking to effectively support SSP in an EMS context.

(9)

2. Theoretical Background

To address the research objective, a literature review is conducted concerning decision support for SSP in EMS. First, a context is provided by discussing EMS nursing staff activities and requirements on their educational backgrounds. Next, the challenges faced in SSP in EMS are considered. Challenges in balancing staff demand and supply are considered as are challenges in improving. Finally, decision support systems are discussed by highlighting its components, use in EMS and related domains for analyzing staff demand and supply.

2.1 Ambulance nurses – activities and education

Two main types of EMS systems exist, namely the Anglo-American system (AA) and the Franco-German system (FG) (Dick, 2003). Both systems, and the Dutch system, which is an adaptation of the AA system, are described in detail in Table 2.1. Information from multiple sources (Roessler et al., 2006; Dick, 2003; Academie voor Ambulancezorg, 2019) is combined and represented in Table 2.1. The main differences are how service is delivered and how staff is addressed. In Anglo-American systems the patient is brought to the doctor, where in Franco-German systems the doctor is brought to the patient (Dick, 2003). This difference in how service is delivered sets different requirements to be performed by EMS nursing staff.

(10)

(11)

2.2 Strategic Staff Planning 2.2.1 Definition and objectives

Strategic planning is the brick and mortar of an organization and addresses structural decision making (Hans et al., 2012). SSP is defined by IPMA-HR (2002, p. 10) as “the strategic alignment of an organization’s human capital with its business direction. It is a methodological process of analyzing the current workforce, identifying future workforce needs, establishing the gap between the present and the future, and implementing solutions so the organization can accomplish its mission, goals and objectives”, and this definition is used in this study. In order to provide high levels of care, decision-making models need to steer human capital management (Kabene et al., 2006). This can translate strategy into designing, dimensioning and the development of healthcare processes.

The objective of strategic planning is often to improve either structure quality, process quality, outcome quality, or a combination of these (Reuter-Opperman, 2017). To assess strategic planning performance of a company, five key indicators for quality in EMS have been identified; availability, reliable access, demand/workload, rate of critical conditions and level of care (Fischer et al., 2011). These indicators are, as the SSP process advances, more integrated with the process as an accurate SSP contributes to improvements in one or more indicators.

SSP consists of four steps, a visual representation of the process is depicted in Figure 2.1. The steps consist of supply, - demand-, and gap analysis, and a solution analysis (Anderson, 2004; GIPD, 2018). Input for supply and demand analyses is internal and external data collected by the analyst or decision maker or is readily available within the organization. Supply analysis generally considers the current workload and workforce and starts with a snapshot of workforce demographics after which employment trends are discovered through analysis (Shukla, 2009). In this study a broader scope is used for supply analysis as nursing staff is considered to be amongst the most valuable assets in healthcare. As supply analysis matures, more differing data sources are used and a longer planning horizon is incorporated (Beitel et al., 2020). Demand analysis uses projection of workforce requirements through simply being aware of shortages or through use of sophisticated models (Choudhury, 2007). After this the gap is determined between supply and demand through analysis of quantitative and qualitative data (Cotton, 2007). The steps, their purpose, activities and data sources are further explained in Table 2.2, which summarizes findings in sections 2.2 and 2.3.

(12)

Table 2.2: SSP Summary

2.3 Improving SSP – Maturity Models

Maturity models are used to describe process improvements, originally proposed for the software industry (Crosby, 1979), but has since extended to many areas of interest (Lahti et al., 2009; Sarshar et al., 2000; Rosemann & de Bruin, 2005). To achieve best practices through improvements, a simple but effective tool for process analysis was introduced. By analyzing processes, maturity models are capable of identifying improvement opportunities and foster achievement of improvement through guided measures (Becker et al., 2009). Maturity models consist of two elements. Multiple maturity levels are established which describe the competence of the organization in a specific domain, and simultaneously provide opportunities for improvement (Rosemann & de Bruin, 2005). The second element concerns the activities and processes that are completed in order to perform that level of the process successfully.

(13)

2.4 Challenges in staff planning 2.4.1 Supply

There are multiple problems in performing supply analysis of workforces in healthcare, for instance, systemic delays and dynamic complexity (Brailsford & Da Silva, 2014; Vanderby, 2014). Systemic delay in an emergency medical service setting represents the time it takes until nursing staff is developed to be part of the working pool, e.g. the length of specific EMS education as well as admission requirements as discussed in Chapter 2.1.

“Dynamic complexity refers to the movement in the workforce, for example attrition rates, training schemes and movement between training paths cause a disbalance between supply and demand” (Willis, 2018; Friis et al., 2008). According to Friis, Ekholm and Hundrup (2008) “nurses are aging and too few graduated nurses are entering the profession to replace those retiring or leaving. p. 241” other reasons are nurses leaving the working environment before retirement age voluntarily, or in forced retirement (Friis et al., 2008). In general, the expulsion from the labor market is because of the demanding physical working environment and changing working hours (Friis et al., 2008). Additionally, individual factors as well as job characteristics, organizational factors play a role (Brennan & Ross, 2019).

The third challenge in analyzing supply, is that little research has been devoted to the subject of workforce forecasting techniques. Organizations feel difficulty selecting workforce forecasting techniques that suit the available data and their strategic direction (Shukla, 2009). One or more combined methods may be best suited, but how to choose the technique or techniques to use remains unanswered (Shukla, 2009).

2.4.2 Demand

Little is known about what influences demand for EMS services (Andrew et al., 2019), troubling precision of findings in analyzing demand. Three main causes have been identified for the growing number of ambulance dispatches (AZN, 2016). The first cause consists of changes in the healthcare environment like closing emergency care centers and hospitals, shifting more specialized care outside of hospitals (AZN, 2019).

(14)

Third main cause is innovation of the EMS process. Demand for care becomes more complex as (chronically ill) patients stay living at home for longer and the role of EMS in the health care supply chain changes. All causes contribute to an increasing uncertain demand, which makes demand for EMS nursing staff harder to predict.

2.4.3 Gaps and Solutions

In a gap analysis the deficit or surplus of staff is identified. To do this, multiple scenarios for staff levels are created, based on changeable parameters. To be able to consider the scenario that is closest to reality, the right parameters and in which direction they change have to be carefully selected (SHRM, 2015). When gap analysis is underdeveloped, little effort is spent on use of analytical tools like extrapolation of trends or scenario analysis (Beitel et al., 2020). The challenge here is to implement use of analysis tools which the end-user is not experienced with (Huber, 2012) as well as overemphasizing a desired future, neglecting alternatives (OPM, 2019). A thorough understanding of what to change and in what direction to change requires time and effort.

Another challenge in gap analysis is the change in competencies asked from a staff group. As the required competencies change due to e.g. emerging technologies a challenge emerges. Assessment of in what way required competencies evolve is necessary to accurately plan for the change.

As to solution analysis, this is mainly done by developing alternatives to perform activities more effectively or with more efficiency. The solution analysis emerges from the gap analysis after which alternatives are developed to close the gap (Cotten, 2009). Solution analysis is done outside of the architecture to be developed in this study. Therefore, solution analysis and its challenges are outside of the scope of this study as is explained in Chapter 3.

2.5 Decision Support Systems 2.5.1 Terminology

(15)

Multiple types of DSSs exist, which can be grouped into five categories; communication driven DSS, data driven DSS, document driven DSS, knowledge driven DSS and model driven DSS (Marin, 2008). This study uses a model driven DSS, mostly used to analyze a situation by using limited data and parameters provided by makers (Power, 2007). Model driven DSSs utilize simple quantitative models to aid in decision-making, for example through what-if scenarios (Power, 2007). These kinds of models are mostly used by managers, staff or people who interact with an organization (Marin, 2008). Model driven DSSs use quantitative models as main components, of which spreadsheets have become increasingly popular in recent years (Power & Sharda, 2007).

2.5.2 Decision support in EMS

Despite the potential of advanced statistical models to offer accurate forecasts, most ambulance service providers still use rudimentary prediction methods. Prediction methods are mainly used for supply analysis and to a lesser extent demand analysis when developing strategic plans (Reuter-Opperman et al., 2017). Within the EMS industry a variety of applications for DSS is used. Clinical DSSs are used for healthcare-related subjects. Vicente et al. (2013) developed a knowledge driven DSS that guides EMS nursing staff in assessing elderly patients’ required level of care through 11 medical conditions. A second use for DSS in EMS is for logistic purposes, for instance through fleet dimensioning, finding optimal station locations and dispatching (Reuter-Opperman et al., 2017). Often logistically oriented DSSs have a mathematical problem-solving focus and use complicated prediction tools. Within EMS, decision support exists on the strategic, tactical and operational level and is used for decision making in logistics, medical decision assessment and policy making. Decision support for staff planning exists on a tactical level, assessing demand based on statistical forecasts and translating it to crew planning (Reuter-Opperman, 2017), but on the strategic level this remains unexplored.

(16)

2.5.3 Efforts towards development of DSS in related work

Thus far no previous studies have been known to support decision making for SSP for EMS. Within related healthcare disciplines though, decision support does exist for SSP. The RAFAELA system (Ruland, 2014) is a nursing staff planning tool developed and implemented in Finland to aid in staff planning within a hospital. Even though RAFAELA is meant for operational and tactical staff planning, lessons can be learned that can be applied in this study. In order to accurately plan human resources with RAFAELA, patient’s care intensity was considered, which can be considered demand for care. Analysis of daily available nursing resources translates to supply of nurses in the current study, after which the optimal level of workload was established. This process was repeated when decision-makers wanted to provide answers in meetings and wanted to back this with data. This study is useful as it provides directions about how governance regarding SSP should be structured as well as what sources of data to use in the process. A Third useful insight is that data is graphically represented at times of decision-making, which is considered in the current study as well.

Another interesting study where a DSS was developed for SSP purposes is Mohany et al.’s (1997) paper. Here three approaches to SSP are discussed; resource-, process-, and intellectual-support approach (Bidgoli, 1987). The resource support approach was chosen, which led to an architecture that included a database management subsystem, a model-base subsystem and a dialog-generation management subsystem, much like the components described in 2.4.1. What is particularly useful for this study is the way the components of the DSS have been further described. The model base, for example, used a forecast module using regression, time-series and Markovian analysis (Mohany et al., 1997). Both the structure of how the architecture and its components is presented is helpful for this study as are the particular tools used to fill in the components.

(17)

2.6 Conceptual model

In this section the conceptual model, found in Figure 2.2, including components generally found in a model driven DSS will be explained. The SSP process consists of steps, guided by governance, have a collective aim and are supported by a DSS.

User: the human component of the decision-making process. This is the person who’s task it is to produce the SSP, and usually is the person that is also responsible for it. His or her goal is to provide descriptive insights to support management decision-making (Keen, 1980). The user is connected with the DSS through the user interface.

Interface: The user interface is where the decision-maker can have dialogues with the system. This interaction consists of what the decision maker can see, can do and must know (Keen, 1980). The interface consists of all the actions a user can perform or witness (Holsapple et al., 1993). Here the decision maker can ask questions where the DSS will provide an answer through performing analysis of the data.

Model Base: The model base for a model driven DSS consists of the set of analysis techniques and assumptions that provide a simplified representation of the real situation (Power & Sharda, 2007). Multiple types of forecasting models exist, varying in techniques used, which influence the level of sophistication of analysis and forecasting accuracy. In SSP, the analysis techniques are the tools used to perform the steps discussed in Section 2.2.1.

(18)

Figure 2.2: Conceptual model

2.7 Summary of findings

(19)

3. Research Methodology

In this section the methodology used for this study will be discussed. The methodology is chosen to fulfill the research objective of designing an architecture for DSS for SSP in an EMS context. To achieve this goal the chosen methodology will be justified, the case companies will be introduced, and the research design will be detailed. Finally, the way data is collected will be discussed.

3.1 Design Science

In order to achieve the research objective a Design Science approach is used. A Design Science study is concerned with developing a solution to a practical, real-life problem, making it a better method for achieving this study’s objectives, in contradiction to a case study or survey (Hevner et al, 2004; Holmström et al., 2009). In this study the practical objective is to enhance SSP by developing an SSP architecture and implications for a DSS. The scientific contribution is the investigation of SSP in an EMS context as well as the generic design of a DSS for SSP. The outcome of the investigation is particularly important for how decision-making on SSP may be supported in a discipline where such support is weakly developed. This will further establish the initial development of SSP in any new area of research as specific insight as to how development of SSP in EMS happens.

This chapter builds on the design cycle by Wieringa (2014) which consists of three steps of the design cycle, namely problem investigation, treatment design and treatment validation. Due to the fact that this study’s goal is to design a generic architecture as well as due to time constraints, the fourth phase of evaluation and implementation of the artifact is not considered.

3.2 Case introduction

This study will be conducted at three RAVs in the Northern Netherlands. All of these RAVs provide EMS in their own region and in all organizations, problems are experienced in estimating staff demand and supply due to uncertainties and little to no available decision support tools. This dilemma results in, decision-makers either under- or overestimating staff requirements with implications for quality of service delivered. Undersupply of staff will result in less availability of ambulances with nursing staff, reliable access and an increased workload. Oversupply of staff will result in a worse financial performance, however it is a virtually non-existent option as supply of new EMS nursing staff is scarce.

3.3 Research Outline 3.3.1 Process Description

(20)

stakeholder goals will be assessed through qualitative semi-structured interviews with the case company experts and other domain experts. If available, planning data input and output as well as relevant literature will be used to describe the system and its components. A systematic description will be proposed, including aims, parties involved and activities, linked to a constructed maturity model. A Business Process Model and Notation (BPMN) will be used to visually support the description.

The constructed maturity model combines efforts of multiple existing maturity models (Beitel et al., 2020; Gov’t of Queensland, 2019; Government of New South Wales, 2019), into a model suited for the EMS industry. The model, consisting of levels of maturity for corresponding capabilities can be found in Appendix A. The capabilities Strategy Development and Implementation Planning & Execution were excluded from this study. They are out of the scope for developing more insight in the shortages of nursing staff and have a more creative, problem solving focus. Reporting is added to the Governance capability and the SSP process is, as it matures, more integrated with other subdomains within the organization to which reporting is done.

3.3.2 Process Analysis

The second step is to analyze and evaluate the way SSP is currently realized in the individual case companies. All capabilities found in the maturity model in Appendix A are analyzed, and the result of the analysis is a level on which the process is currently performed by the case companies. The discrepancy between the current level of capabilities and the desired capability levels acts as requirements for the artifact design. The desired capability levels are logically derived from analysis as well as from stakeholder goals.

3.3.3 Design and Validation

In this study the artifact that is designed is an architecture for the Strategic Staff Planning process, with more specific focus on implications for a model driven DSS. Requirements for the artefact are specified after analysis of the capabilities and based on stakeholder goals. Furthermore, a treatment in the form of an architecture including relevant DSS components is designed. This is called a treatment, because treatment suggests that this artifact interacts with a problem context to solve a real-world problem (Wieringa, 2014). The design of the treatment will be an iterative process, using multiple methods. Semi-structured interviews with decision-makers within the case companies as well as interviews with external domain experts are used to design the process architecture and subsequent DSS components as displayed in Figure 2.2. Company and domain experts act as source for design directions as well as for validation of the designed artifact.

(21)

design prediction and effect valuation (Wieringa, 2010). To determine if the design is externally valid, interviews with company experts have been conducted. Here is considered what the effect of the artifact is when considered in a different EMS context and what assumptions are made about the context in the design phase.

3.4 Data Collection

The three case companies described in chapter 1 were used for data collection. Qualitative data needed for analysis & treatment design was collected through interviews with stakeholders from these companies as well as interviews with external domain experts. To verify the treatment designed, multiple unstructured interviews were conducted with stakeholders. The stakeholders were identified for their stake in the problem and the contribution they could deliver with their expertise.

(22)

4. Description and Analysis of the Strategic Staff Planning Process

4.1 Approach

In this chapter an overview of how the case companies currently perform the SSP process is provided. Firstly, the purpose, activities, parties involved and their use of DSSs is described to gain a full understanding of what the SSP process currently entails for all organizations. As each case company conducts the SSP process in a different manner, individual representations of the process will be provided. Visual representation through the process mapping technique Business Process Model and Notation (BPMN) is included. Secondly, an analysis of the maturity of the current SSP is performed by assessing individual capabilities with the maturity model, found in Appendix A.

4.2 Strategic Planning Process 4.2.1 Case Company 1

Aims: When conducting the SSP process, case company 1 aims to develop more insight in long term staffing situations to support decision-making. This is done to have enough staff capacity in the future.

Parties involved: The planner is mainly responsible for the SSP process, supported by an HR official, making the HR and operational department ultimately responsible for the planning. Management and nursing staff team leaders are involved secondarily by approving changes and checking planning accuracy.

Their activities: Activities undertaken to perform the SSP process are displayed in the BPMN. Within the BPMN, supply related activities are colored blue, demand related activities are colored green and gap determination activities are colored orange. An SSP is composed yearly, and its accuracy is assessed every three months during designated staff planning meetings. Input of supply analysis activities as seen in Figure 3.1, 3.2 & 3.3 consist of data internal to the organization, where demand analysis data consists mostly of data external to the organization. Output is a necessary number of nursing staff inflow, represented in a number of people to hire and educate.

(23)

Use of DSS: Excel is the main decision support system used by case company 1. Internal workforce data is stored in the spreadsheet and the workforce variables are deducted from the total number of FTE available. Limited use of statistical tools is used through functions available in Excel. The total number of FTE available per period are met against the budgeted hours by the insurance companies. This results in a gap to fill with inflow in education or readily available staff. This gap is checked with results of the third-party model to check forecasting accuracy. No advanced analysis tools, scenario analysis or graphical representation methods are

(24)

(25)

4.2.2 Case Company 2

Aims: Company 2 aims to develop more insight in workforce capacity to use as evidence in yearly budget negotiations with insurance companies.

Parties involved: In case company 2 the Head of HR and Development is responsible for SSP. The process is assisted by the planning department, who deliver data to be used in the process. Management have yearly negotiations in which the budget for hours is determined for the next year. Next to this are they responsible for developing solutions when there is expected to be too much or too little capacity.

Their activities:

Their use of decision support systems in executing activities: Case com

pany 2 makes use of Excel as a decision support system. In Excel, formulas are used to subtract variables from the total number of hours available. Data used in the process is internal to the organization and are already stored in a workforce system or derived from analyzing data in the workforce system.

(26)

(27)

4.2.3 Case Company 3

Aims: Case company 3 develops SSP to develop more insight in staffing situation on the longer term. This is done to support management in decision making in order to deal with expected changes in staffing levels and the healthcare environment.

Parties involved: The schedule policy officer develops SSP in collaboration with an HR advisor. Internal information is provided by the Vice-President. Management provides internal data and obtains the result of SSP to develop solutions.

(28)

Their use of decision support systems in executing activities (data sources, model-based support (which, to do what)): Excel is the main decision support source used in the SSP of case company 3. Supply and demand are met in a spreadsheet which results in a shortage or surplus of staff. Some simple statistical methods are used to extrapolate supply to future time periods.

(29)

4.3 Process Analysis 4.3.1 Analysis Approach

The way the case companies currently perform SSP is analyzed through assessing the designed capability maturity with the maturity model in Appendix A. All capabilities will be assessed individually per case company and will be scored on a scale from 0 to 5. Level 0 means no efforts are directed to SSP, whereas level 5 implies that the planning process is incorporated fully with the strategic management of the organization. This implies that the projected time window is longer, the planning uses more varied data sources and is more complete, making it more accurate. Assessment will be done by analyzing whether and to what extent the companies currently perform activities as described in the model. A summary of all scores is depicted in Table 4.4.

4.3.2 Demand Assessment

Demand is assessed in two ways, through the third-party model (Capaciteitsorgaan, 2019) and through budgets allocated by health insurance companies. The third-party model provides an amount of additional staff capacity needed based on expert and internal opinions, and the additional staff capacity can be calculated with this model. The health insurance organizations calculate budgets based on national sociodemographic data collected by the Dutch National Institute for Public Health and the Environment (RIVM). They prescribe a number of ambulances to use simultaneously, budgets for nursing staff are calculated based on this estimate. Both ways of analyzing demand use objective data, but assumptions are made based on The Netherlands as a total, which leads to inaccurate forecasts. As the model used to calculate demand is not fit for the regional situation, demand assessment is scored at level 1.

4.3.3 Supply Assessment

(30)

4.3.4 Gap determination and prediction

All case companies recognize the immediate need of shortages in their staff. Two case companies produce gap determination and prediction on an initiative level, resulting in level 1 for both companies. The third organization only produces shortfalls and surpluses for personal insight. Some linear extrapolation of turnover is considered by case company 3, making this company the only to score their gap determination and prediction capability at level 2.

4.3.5 Governance

Currently the SSP process is not defined as a separate activity/function in case company 2, but it is in company 1 and case company 3. Even though it is defined as a separate process, little attention is given to it. In case company 3, stakeholders stress that meetings with time designated for strategic staff planning quickly change to more ad hoc planning subjects. No champions have been defined for certain business areas who have responsibilities within the SSP process, implying the governance capability is at level 1 for all case companies.

Company 1 Company 2 Company 3

Demand assessment 2 2 2

Supply Assessment 2 2 2

Gap determination 1 1 2

Governance 1 0 1

Metrics & Feedback 0 0 0

(31)

5. Strategic Staff Planning Architecture Design

5.1 Approach

The maturity analysis from section 4.3 guides as a starting point for improvement opportunities. Suggestions to improve the SSP are developed based on outcome of the analysis. A migration path for developing SSP capabilities is established by proposing both short-term and long-term suggestions. Because of the homogeneity in the current levels of staff planning in case companies, a single architecture is created that suffices for all case companies. Individual differences are discussed in text. In this architecture the current level of SSP is addressed firstly, signifying a starting point for extension suggestions from higher levels. Hereafter, directly applicable extension suggestions are proposed per capability followed by long-term extension suggestions. Extension suggestions are summarized in Table 5.1. After designing the architecture, implications for DSSs are provided. A more detailed version of Table 5.1 is provided in Appendix B.

5.2 Design requirements

To cope with the experienced labor market shortages, more insight in what drives the expected shortages is needed. Thus, improvements need to be made to the SSP process, which have to be carefully selected on their expected contribution to the SSP process. To systematically select contributive improvement requirements, desired levels of capabilities are chosen to which improvement efforts will be directed. The requirements found are listed below.

The First requirement is to improve demand analysis, mainly for regionality. By analyzing demand regionally as opposed to nationally, assumptions that are made in determining demand are filtered out. A Good example of an assumption is efficiency increase through economies of scale when more nursing staff capacity is available. This applies in posts where multiple nursing staff members are available, but this might not apply in smaller regions where less economies of scale are achieved from adding the same capacity. By analyzing demand regionally, accuracy of demand analysis could be enhanced.

Second, extending supply analysis, within and across organizational boundaries. As currently supply is solely analyzed quantitatively, and limited extrapolation of variables is applied, forecast accuracy is limited. By selecting the right workforce variables and the most appropriate statistical tools, the most accurate quantitative supply analysis is performed. Incorporation of qualitative analysis of supply could lead to improved insight in unexpected contract terminations and sick leave.

(32)

changes cannot always be predicted beforehand, this uncertainty is an important point for the strategic staff planning process.

The last requirement is to include SSP into business as usual. Analysis showed that the process is currently not formalized, and responsibilities and roles are meagre defined. In one of the case companies SSP is only used for personal forecasting, without any implications for business. When SSP is not recognized by other business areas, little consideration can be given to the outcomes of the process. Even when analysis is accurate, when SSP results are not used, no anticipation occurs within the organization regarding future staff levels. By incorporating the process into business as usual, the process with its inputs and outputs are deeper anchored within the organization. Therefore, its implications are communicated with other business areas, creating more urgency for the implications of staff shortages.

(33)

5.3 Extensions of Strategic Staff Planning 5.3.1 Demand Suggestions

For demand assessment, improving analysis is required, mainly for regionality. As currently only a national forecasting model is used, inaccurate projections are used. To improve accuracy, a near future extension is to improve insight in analysis steps. This can be achieved by using primary data as opposed to secondary data, which improves reliability and validity. Examples of primary data are regional sociodemographic data, for instance demographics sorted per age group. Analysis of the number of dispatches per age group, readily available within the organization, against expected growth or decline per demographic group could indicate the future total number of dispatches. This translated to a number of nurses needed to staff these dispatches, and acts as an objective indicator of demand. Second demand-related extension lies in qualitative data, obtained by analyzing environmental changes or policy changes by interviews with domain experts, surveys or PESTLE-analyses (US Department of Energy, 2005). A more long-term goal is to conduct demand analysis in a consistent manner. By having a model that is continuously used to translate demand requirements into necessary capacity a more robust way of demand analysis is used. This stimulates continuous improvements as the model can be updated when suggestions arise. Consistently using demand analysis models allows for scenario analysis to be incorporated for demand related variables.

5.3.2 Supply Suggestions

(34)

5.3.3 Gap Determination Suggestions

For the Gap Determination capability, a direct extension suggestion to be prepared for unexpected situations is to incorporate scenario analysis. Even though scenario analysis is performed on supply and demand related variables, it is considered a suggestion for gap determination. This is because it directly influences the gap that results from both supply and demand analysis. By incorporating scenario analysis, for instance through Microsoft Excel’s ‘What if?’ function, multiple future realities are created. By leaving room for supply or demand related variables to change, insight is developed into how future conditions could affect the organization (OPM, 2019). A longer-term suggestion is to further utilize scenario analysis. At first, scenario analysis will be performed with limited variables and few scenarios are created. After using scenario analysis for some time, this can be intensified through incorporating variables which are expected to be more likely to change. Observation here is that this depends on earlier performed demand analysis, of which likely environmental changes are used as input for scenario analysis.

5.3.4 Governance & Reporting Suggestions

(35)

5.3.5 Metrics & Feedback Suggestions

For the Metrics & Feedback capability, direct suggestions are to incorporate metrics & feedback mechanisms. Metrics are quantitative assessments standards to be used for systematically assessing performance over time. By selecting and consistently using metrics a more standardized way of monitoring the process is incorporated. As the process is discussed in quarterly meetings, adding metrics creates more structure for direct and indirect stakeholders about the result of the process over a longer period. Metrics should ideally relate to any of the quality indicators, identified in section 2.2.1. Feedback mechanisms are loops embedded within the process to check accuracy of input, output and assumptions made within the process. Generally, the feedback is provided by another stakeholder that did not perform the activity but has understanding of the activity. An additional suggestion is by benchmarking performance to similar RAVs. More insight in whether metrics used in the SSP process match that of related RAVs is a good indicator to assess whether certain values are outliers and need addressing, or if they are common in the industry or environment and thus need a broader approach. Indirect suggestion is to know impact on organization. Feedback loops should be incorporated during the planning process to assure whether input and output of the process is feasible in practice.

(36)

(37)

5.5 Implications for DSS 5.5.1 Model

In this section the implications for a DSS used within the SSP process are structurally detailed. The model’s goal is to support demand analysis, supply analysis and gap determination related activities as accurate as necessary to be able to estimate future staff capacity needed, i.e. to determine how many of readily available EMS nursing staff and inflow in education is required. A spreadsheet-type database, like Microsoft Excel, may be used that incorporates workforce characteristics and supply and demand related variables. This spreadsheet suffices as database as well as model base in which all calculations are made. Within this spreadsheet the planning process should be reflected by statistical tools and optimization tools, which are further described in 5.3.2. The supply and demand related variables are used to determine a number of FTE and are brought together to determine the gap, which is the number of inflows needed for that period. This spreadsheet-type database represents the process by storing values in cells, containing formulas that are locked-in. The formulas automatically generate subsequent variables to ultimately assess the gap between supply and demand for a predetermined period. The model should provide output on a yearly basis, with a window that increases as the planning process advances, ultimately covering the total number of years the longest educational path of becoming an EMS nurse takes.

5.5.2 Tools

A variety of tools can be used to enhance SSP. The most used tool is an advanced spreadsheet-type deterministic model using Microsoft Excel, with built-in formulas that automatically generate subsequent variables. Statistical tools may be used to determine the variables, of which a variety is available within Microsoft Excel. Examples of tools available in Excel and applicable to characteristics and variables is Linear Regression, Moving Average, Exponential Smoothing or Autoregressive Integrated Moving Average. For every supply and demand characteristic and variable, the most appropriate tool should be selected. Another tool to be used for demand analysis is scenario analysis. As demand is subject to unforeseen changes, scenario analysis, readily available through the ‘What if?’ in Excel, can be used to forecast multiple future scenarios with help of managerial input and environmental scanning (Ward, 1996). For demand as well as for supply analysis, a simulation tool may be used. Simulation is an alternative to forecast under uncertainty that provides easy scenario analysis. Disadvantage of simulation is the costs to build a model as well as that results are hard to validate.

5.5.3 Database

(38)

a spreadsheet suffices as database. As the process becomes more advanced and integrated with an organization’s strategy it may be necessary to use a database for the process, but to systematically develop the SSP directly a spreadsheet-type database may suffice.

5.5.4 Data sources

To achieve an accurate SSP, multiple data sources need to be utilized to develop a full understanding of the organization’s future staffing situation. Data to be used can be acquired internally as well as externally at government bodies and third-party organizations. Quantitative data is widely used as a starting point in the SSP process to determine available and required numbers of FTE. Qualitative data on the other hand, is a type of data that is often overlooked in assessing SSP for both supply and demand analysis.

On the supply side, mainly workforce characteristics and subsequent variables are considered. An extension to supply data is the incorporation of future supply. By monitoring educational yields with internal data about numbers of education entries and graduates and external data about number of nursing school graduates more insight is obtained on future supply. Future supply data can thus be acquired externally from Central Bureau of Statistics (CBS) or internally when considering data on the internal education. The qualitative side of supply has not been assessed thus far. By assessing supply qualitatively more insight is provided on workload and potential of workforce. Qualitative data is to be stored within the same spreadsheet but does not require processing for incorporation of calculation in the number of FTE needed. Therefore, qualitative data can be stored in a separate sheet.

Demand data is acquired internally as well as externally. As current demand sources are obtained as secondary data. Demand data can be acquired through publicly available data on socioeconomics and demographics. Population, age/sex distribution, births/deaths and population projection can be used to determine demand for EMS staff. This data is acquired externally by Central Bureau for Statistics and Central Planning Bureau (CPB) or RIVM.

5.5.5 Interface

The DSS’s interface allows the user to perform multiple actions. Data can be entered into the spreadsheet, analytical tools can be used to manipulate data, scenario analysis can be performed through altering parameters and the interface allows for graphical representation of the data. Spreadsheet-type databases often have the reputation of being sensitive to errors, which is experienced as ineffective and frustrating. By locking cells and formulas this problem is countered, as cells that contain formulas cannot be accidentally altered, changing outcomes of other cells. Furthermore, the interface should be easy to understand, which can be achieved by colorizing certain cells to indicate subgroups like supply or demand related variables.

5.5.6 User

(39)

goals, understanding spreadsheet-type databases and how they function is essential, as is knowledge on the different types of analysis to be performed.

Because spreadsheets can only be accessed by one person simultaneously, it is advised that one person within the organization is responsible for gathering and entering data as well as data analysis, which is the process champion.

5.6 Redesign

The designed SSP architecture is presented to case company stakeholders, and after consideration some alterations have been made to improve the suggestions. The incorporation of future supply was a notion which is shortly addressed as a long-term goal, but it appears that this is a suggestion that is able to implement directly and even is already done to some extent. Second redesign consideration is that defining roles and responsibilities particularly within regional teams of nursing staff is a consideration to implement directly. This is a more detailed suggestion as it directly indicates the persons to become owner of a part of the process.

5.7 Validation

The final step in the design cycle (Wieringa, 2014) is validation. Here, the developed artefact and maturity model are internally and externally validated. This is done by assessing design prediction (Wieringa, 2010). By assessing the effect of the artefact within the intended context, its internal validation is assessed. In the design cycle, validation is done before implementing (Wieringa, 2014). In order to validate the architecture, unstructured interviews were held at all case companies. Domain experts and company stakeholders firstly assessed the designed maturity model and later assessed the redesigned model from their organization’s perspective. Internal validity measures whether the architecture meets the requirements for enhancing strategic staff planning at the case companies.

Suggestions regarding demand showed moderate internal validity, as one of the case companies perceived the RIVM calculations to be sufficiently accurate. Suggestions regarding Supply Assessment showed strong internal validity, as did Gap Determination suggestions, mainly for analyzing demand. Governance & Reporting showed strong internal validity, as did Metrics & Feedback. This confirmed that the SSP architecture may enhance strategic staff planning, as they would systematically use more sophisticated data from appropriate sources and more appropriate analytical tools leading to a more rigid and accurate planning. Particularly, incorporation of future supply, qualitative analysis of demand and supply, scenario analysis and governance & reporting were found to benefit the planning process.

(40)

From interviews it appeared that the designed architecture and maturity model are externally valid as their design. Limitation here is that externally validity could only be assessed from the perspective of company stakeholders, making it limitedly valid.

5.5 Summary of Design

In sim, a spreadsheet-type deterministic database like Microsoft Excel can be used in which internal and external data from multiple sources can be stored. Spreadsheet-type databases store data in cells and rows on which alterations can be performed through the mode. MS Excel functions simultaneously as model through using readily available formulas and functions. The formulas and functions can be used to perform statistical analysis to project forecasts, as well as scenario analysis and a basic simulation function. Through a user interface, a user can utilize formulas and functions to perform analysis. By using the interface to graphically visualize data, information about the SSP process is easier adopted by stakeholders. Scenario analysis can as well be performed through the user interface. The user accessing the interface needs to have basic knowledge of functions and formulas present in the spreadsheet, as well as knowledge of tools used to aid the SSP process. A visual representation of the Design is provided in Figure 5.1.

(41)

6. Discussion

This chapter has several uses. Firstly, it discusses research contributions in section 6.1, elaborating on theoretical and practical implications. Secondly, it discusses main research limitations in section 6.2 and lastly it provides opportunities for future research efforts in section 6.3.

6.1 Contributions

In this study a high-level architecture was designed in response to a need felt for more advanced decision support for SSP given labor market shortages. This study builds on a maturity analysis that identifies shortcomings in current SSP processes. This study contributed overview and direction for improvement suggestions for SSP in the EMS industry. Shortcomings were found according to SSP capabilities. Found shortcomings were little use of quantitative and qualitative demand analysis, limited use of quantitative and qualitative supply analysis, basic use of gap determination and little use of governance and feedback. To mitigate shortcomings, extension suggestions were made on the short- and on longer-term, providing a migration path for development of the SSP. Requirements developed were to improve demand analysis, mainly for regionality, to extend supply analysis, within and across organizational boundaries, to be prepared for unexpected situations and finally to include SSP into business as usual. Finally, implications for DSSs were conducted from extension suggestions. This study filled in the gap of improving accuracy of SSP trough use of a DSS in the EMS industry.

This study has several contributions to literature. First and foremost is it a first exploration to the subject of strategic staff planning EMS literature. By proposing a high-level architecture, this study provides overview and direction for research regarding improvement of strategic staff planning in the EMS industry. By analyzing the level of SSP capabilities through using the maturity model, a structured way of determining improvement requirements for SSP in the EMS industry is developed. Hereafter, improvement suggestions are proposed and detailed, which may improve accuracy of the SSP process.

Second contribution is the use of a maturity model for improving strategic staff planning, combined with implications for DSSs. Being an underdeveloped area of research, this study contributed to bridging the gap between maturity models used for strategic staff planning and their implications. Most literature regarding maturity models for SSP is emphasized on a high hierarchical level, i.e. integrating the process with strategy. This study bridges the gap between high hierarchical goals of implementing the strategic staff planning process with low-level DSS implications.

(42)

EMS industry, important elements are added to the maturity model found in Appendix A. Elements from multiple industries were combined, and sector-specific elements were excluded or replaced by elements specific to healthcare or EMS. Empirical validation is needed to further establish the contribution of this model for EMS as well as for other similar industries.

Some practical implications follow from this study as well. Firstly, this study provides an architecture with extension suggestions for the strategic staff planning process of the case companies, found in Table 5.1. Suggestions were made in the short term as well as in the long term, signifying a migration path to take for each capability. A more detailed version of the architecture is found in Appendix B.

A Second practical implication is the development of the maturity model, found in Appendix A, which was specifically constructed for EMS service providers operating in the Anglo-American system. Hence, this model acts as a guideline for analyzing capabilities and subsequently improving capabilities for all EMS service provers operating in Anglo-American systems. Extension of this model’s application to other EMS systems and possibly outside of the Anglo-American system remains unexplored, as characteristics of other systems and industries may have effects on the applicability of the model. This remains a topic for future research.

6.2 Limitations

This study’s goal was to provide an architecture for strategic staff planning, more specifically focusing on implications for decision support needed to do so. This study and this study’s findings were critically reflected on, and following limitations have to be observed with respect to this study.

First, and main limitation is that in this study only a part of the strategic staff planning process was considered. The parts of the strategic staff planning process incorporated contribute to developing more insight and to enhancing forecasting accuracy. As other steps do not contribute to improving forecasting accuracy but have a more creative, problem solving nature, developing solutions and implementation of solutions are not considered. As outcomes of solution development can have implications for earlier activities of the process, this is an important aspect to consider. By incorporating only a part of SSP an important limitation is addressed. Subsequent limitation, following from time constraints, the design cycle is used instead of the engineering cycle (Wieringa, 2014). By not implementing the artefact, additional insights or unexpected redesign opportunities are not incorporated.

(43)

staff would be selected to hire and train, shortages could seem much less urgent, diminishing the need for more decision support. More study effort is required for this study’s architecture to be validated externally.

Third limitation is that it is assumed that the shortages can be solved with means within scope of the strategic planning process. By assuming this, only a part of the total strategic staff planning process were considered. By using the design cycle instead of the engineering cycle (Wieringa, 2014), implementation of the artefact is not considered and insights regarding the practical feasibility of extension suggestions are not considered.

6.3 Recommendations for future research

As this study was an exploration towards shaping the context of SSP within the EMS industry, several recommendations can be made towards future research directions regarding this subject. Limitations of section 6.2 provide suggestions for future research efforts.

A recommendation for future research is to incorporate the total SSP process in the design of the architecture. By doing so, a more complete representation of SSP is considered, and implications for supply-, and demand analysis resulting from solution analysis are considered. Thus, a more comprehensive architecture is created, which may lead to more accurate decision support for the SSP process.

A second recommendation is to further research implications for DSSs for strategic staff planning purposes in the EMS industry. As this study acts as a first effort, more extensive research can be devoted to the use of DSSs and the maturity of the strategic staff planning process within the EMS industry.

(44)

7. Conclusion

(45)

References

Academie voor Ambulancezorg 2019, Initiele opleiding ambulanceverpleegkundige. Available at: https://www.academievoorambulancezorg.nl/initieel/initiele-opleidingen/ambulanceverpleegkundige

[Accessed 23-10-2019]

Ambulancezorg Groningen 2019, Opleiding tot Ambulanceverpleegkundige. Available at: https://www.ambulancezorggroningen.nl/medewerkers/werken-bij/opleiding-tot-ambulanceverpleegkundige [Accessed: 23-10-2019]

Anderson, M. (2004). The metrics of workforce planning, Public Personnel Management, 33(4): 363-378.

Andrew, E., Nehme, Z., Cameron, P., & Smith, K. (2019). Drivers of increasing emergency ambulance demand, Prehospital Emergency Care, 1–11.

AZN, (2016). AZN Factsheet viewed 1-10-2019. Available at https://www.everyoffice.nl/cmspanel /userfiles/235/files/AZN_Factsheet_volumestijging_ambulancezorg_07-10-2016.pdf [Accessed 16-10-2019]

AZN, (2019). Urgentierapport. Available at: https://www.ambulancezorg.nl/nieuws/nivel-onderzoek-naar-urgentie-indeling-inspireert-ambulancezorg [Accessed 06-10-2019].

Bundesministerium der Justiz und für Verbraucherschutz, Notfallsanitätergesetz – NotSanG, Notfallsanitätergesetz §8 Voraussetzungen für den Zugang zur Ausbildung. Available at: http://www.notfallsanitaeter-aktuell.de/gesetz-notsang/zugangsvoraussetzungen-ausbildung.html, [Accessed 10-11-2019]

Beitel, G., Coppus, G., Whitehead, T., (no date). A Capability and Maturity Model for Workforce Planning,

Calgary. Available at:

http://www.dynawise.com/documents/Workforce%20Planning%20Maturity%20and%20Capability%20Model .pdf. [Accessed: 02-01-2020]

Bélanger, V., Ruiz, A., Soriano, P., (2019). Recent optimization models and trends in location, relocation, and dispatching of emergency medical vehicles, European Journal of Operational Research, 272 (1), pp. 1-23

(46)

Booker M.J., Shaw A., Purdy S, (2015). Why do patients with ‘primary care sensitive’ problems access ambulance services, A systematic mapping review of the literature. BMJ Open

Brailsford, S.C., De Silva, M.D., (2015). How many dentists does Sri Lanka need? Modelling to inform policy decisions. Journal of the Operational Research Society. 66: pp. 1566-1577

Capaciteitsorgaan (2019). Capaciteitsplan 2021-2024. Available at:

https://capaciteitsorgaan.nl/app/uploads/2019/12/Capaciteitsplan-2021-2024-Hoofdrapport.pdf. [Accessed: 01-11-2019].

Choudhury, E. H. (2007). Workforce Planning in Small Local Governments. Review of Public Personnel Administration, 27(3), 264–280.

Cotten, A., (2009). ‘The Seven Steps of Effective Workforce Planning’, Office of the Director of National Intelligence, Intelligence Community Equal Employment Opportunity and Diversity, Arlington, 2009.

Drennan, Vari & Ross, Fiona. (2019). Global nurse shortages-the facts, the impact and action for change. British Medical Bulletin.

Friis, K., Ekholm, O., & Hundrup, Y. A. (2008). The relationship between lifestyle, working environment, socio‐demographic factors and expulsion from the labour market due to disability pension among nurses. Scandinavian journal of caring sciences, 22(2), 241-248.

Government of New South Wales, Strategic Workforce Planning Framework, 2019, Public Service Commission. Available at: https://www.psc.nsw.gov.au/ArticleDocuments/3692/PSC_Strategic-Workforce-Planning-Framework.pdf.aspx?Embed=Y. [Accessed: 27-12-2019].

Government of Queensland, Strategic workforce planning framework, Government of Queensland. Available at: https://www.forgov.qld.gov.au/use-strategic-workforce-planning-framework. [Accessed: 01-01-2020].

Hevner, A.R., S.T. March, J. Park and S. Ram, 2004. Design science in information systems research. MIS Quarterly, 28: 75-105.

(47)

Hans, E., & van, Mark & Hulshof, P., (2012). A Framework for Healthcare Planning and Control.

Heidari, M., Seifi, B., & Gharebagh, Z., (2017). Nursing staff retention: Effective factors. Annals of Tropical Medicine and Public Health. 10. 1467.

Holmstrom, J., Ketokivi, M., & Hameri, A.-P. (2009). Bridging Practice and Theory: A Design Science Approach. Decision Science, 40(1), 65–87.

Holsapple, C.W., Jacob, V.S., Zaveri, J.G. (1993). Learning by problem processors: adaptive decision support systems, Decision Support Systems, 10, pp. 85-108

Huber, C., (2012). Strategic workforce planning in Dutch organizations – A contingency perspective, Master’s thesis, Tilburg University, Tilburg.

International Personnel Management Association (IPMA), (2002). Workforce Planning Resource Guide for Public Sector Human Resource Professionals, Alexandria Virginia. Available at: https://www.ipma-

hr.org/docs/default-source/public-docs/importdocuments/pdf/hrcenter/workforceplanning/cpr-wforce-plan.pdf?sfvrsn=4f892a78_0 [Accessed 16-10-2019].

Kabene, S., Orchard, C., Howard, J.M., (2006). The importance of human resources management in health care: a global context. Human Resource Health 4, 20

Keen, P.G.W., and Scott Morton, M.S., Decision Support Systems: An Organizational Perspective, Addison-Wesley, Reading, MA, 1978.

Kommer, G.J., Gijsen, R., Van Gils, P., (2015) Trendanalyse spoedeisende ambulancezorg. Bilthoven: RIVM.

Lowthian J.A., Cameron P.A., Stoelwinder J.U., Curtis A., Currell A, Cooke M.W., McNeil J.J., (2011). Increasing utilisation of emergency ambulances. Australian Health Revisited. 35:63–69

Mann, R.I., and Watson, H.J., "A contingency model for user involvement in DSS development", MIS Quarterly, March 1984, 27-38.

Supporting Strategic Staffing Decisions in Emergency Medical Services