Develop a scheduling protocol for the residents’ activities

(1)

Develop a scheduling protocol for the residents’

activities

Master thesis

Supply Chain Management & Technology and Operations

Management

By Erik van Beek, s3033597

University of Groningen

Faculty of Economics and Business

Supervisor University:

Prof. dr. J. Riezebos

Co-assessor University:

Dr. ir. D.J. van der Zee

University Medical Center Groningen Supervisor:

(2)

ii

Abstract

Background: Medical education is being transformed by a novel method using Entrustable Professional

Activities, EPAs, introduced by Ten Cate (2005). EPAs give an overview of the competencies including knowledge, attitude, and skill in a clinical environment. Due to EPAs a holistic view of the residents' clinical competencies can be operationalized and assessed. The EPA scale consists of five entrustment levels for the residents, where each increase in EPA level gives an increase in the autonomy of handling of the residents. The research of EPAs currently focuses on the implementation of EPAs in a medical environment, while the effects of EPAs on operations in the healthcare scene are still unknown.

Objective: Developing a scheduling protocol which is able to incorporate the EPAs, resulting in a more

effective way of educating and providing care in a healthcare environment.

Method: A design science research is conducted at the Intensive Care Unit (ICU) of the University

Medical Centre Groningen (UMCG) to test the usability of EPAs in an operational way. First, the literature on EPAs and scheduling in a healthcare scene is reviewed to attain insight into the current status of EPAs is and how this could contribute to a better and more efficient way of educating. Following, the scheduling at the ICU is described and analysed. At last frameworks incorporating EPAs are designed and tested with the use of interviews with relevant stakeholders.

Results: A framework is designed on how to implement EPAs in the scheduling process.

Conclusion: Suggestions are done on how to incorporate EPAs in the schedule on an operational level

and a tactical level.

Keywords: Resident’s scheduling, Entrustable Professional Activities (EPA), workforce scheduling

(3)

iii

Preface

This thesis is the final assignment of completing my masters Technology and Operations Management and Supply Chain Management. Writing this thesis in a healthcare organization really helped me understand the complexity of healthcare systems regarding on one side the healthcare providing aspect and on the other side the strive for efficiency and improvements. The process of writing my thesis in a healthcare organization provided me with valuable new insight and interesting challenges, possibly for future career opportunities.

(4)

iv

List of abbreviations

AIOS: Residents in a residency programme to become a specialist in a health discipline, in Dutch

called AIOS (Artsen In Opleiding to Specialist).

ANIOS: Graduated physicians providing only the regular care task, they are not participating in the

residency programme (yet). In Dutch called ANIOS (Artsen Niet In Opleiding tot Specialist).

Entrustable Professional Activities (EPAs): A novel assessment tool to assess the resident's

competency level based on skill, knowledge and attitude in a clinical environment.

FTE: the fulltime-equivalent, ratio of 1.0 FTE means that the ratio number of worked hours divided by

the hours considered as a fulltime job is one.

ICV: Intensive Care for adults, department in the UMCG where they provide intensive care for adult

patients. In Dutch called Intensive Care Volwassenen, ICV).

P&C: Planning and Control

(5)

v

1. Introduction

"A workweek of fewer than 48 hours is uncommon". A headline a newspaper about the work circumstances of a graduate physician in training to become a Gynaecologist. To become eligible for a specialist function, postgraduate physicians are required to participate in a postgraduate residency-training programme. For three to seven years, residents gain experience to become a specialist. During this training, many residents are coping with too much work pressure resulting in burnout complaints. Different research stated that 18 to 82 percent of all the residents is coping with burnout symptoms (Prins et al., 2007). Cause of this high work pressure is that residents have to provide patient care, thereby attend educational activities and lastly manage their family life. Those three factors together result in a complex list of restrictions. This stresses the importance for departments to design feasible schedules that satisfy both the demand of providing patient care and educational activities and takes the preferences of the residents into account. This raises the question of how to integrate their educational activities within their provision of patient care tasks. Leading to a more simple and efficient way of scheduling for the resident’s (educational) activities.

The training of residents is done by a new education method, introduced by Cate and Scheele (2007) called Entrustable Professional Activities (EPAs). This method is able to operationalize the residents’ competencies and milestones in the context of actual clinical work (Hauer et al., 2013). The EPA method tests the resident on; skills, knowledge, attitudes and judges the professional task with a measurable or recognizable output which can be observed and judged by an expert. EPA consists of five competency levels, where each level gives an increase in autonomy for the resident. Resulting in a holistic overview of the performance of the resident. That eventually will lead to a well-trained specialist for the specific field. Although the literature on the assessment of EPAs is growing, the operational aspect lags behind.

There are no frameworks available in the healthcare supporting the scheduling process of the residents considering their double role and EPA level. If the resident has accomplished a higher EPA level, the resident is able to accomplish more tasks independently. However, this increase in flexibility and capacity are difficult to integrate into the schedule, due to the aggregation level of these schedules. Alternatively, if there is misalignment between the actual EPA level and the expected EPA level (by the scheduler), it could influence the provision of care for the patient. Since residents could make potentially (small) preventable medical errors, which would not happen under (more) supervision (Holmboe et al., 2016). Incorporating EPAs in the schedule of resident’s results in more transparency for the scheduler. So, the scheduler could better align the learning objectives of the residents with the demand for patient care.

(9)

2 The structure of this paper is as follows: chapter 2 provides a theoretical background for this

(10)

3

2. Theoretical framework

The aim of this research is to establish a scheduling protocol for the residents considering their preferences, double role (providers of care and getting educated) and Entrustable Professional Activities (EPAs). This chapter establishes a robust theoretical base related to those factors, by reviewing previous literature related to EPAs and scheduling methods for staff and resources in healthcare.

2.1 Entrustable Professional Activity, EPA

This section addresses the function of EPAs in the education of medical personnel. First, the EPAs will be explained. Second, the relevance for the EPAs will be examined. Third, the characteristics of an EPA will be defined. Fourth, the advantages and disadvantages of EPAs will be discussed. At last, the logistical challenges of EPAs in healthcare are elaborated. This will lead to an understanding of what factors to consider implementing EPAs in the schedule of the residents.

2.1.1 What is EPA?

EPA is a novel medical education programme introduced by Ten Cate (2005) which reflects a holistic overview of the trainee competencies including knowledge, attitude and skill in a clinical environment. Due to EPAs a holistic view of the residents' clinical competencies can be operationalized and assessed. The EPA scale consists of five entrustment levels for the residents; (1) has knowledge, but is not allowed to do anything, (2) acts under full supervision, (3) acts under moderate supervision, (4) may act independently and (5) may act as a supervisor (Ten Cate and Scheele, 2007). The competencies are logged in the personal portfolio of the residents. Each resident is responsible for his own portfolio. If the resident has reached a certain competency, supervisors need to assess the competency. They decide if the resident has reached a certain EPA level. If the resident has proven to be competent for a certain EPA level, the resident will proceed to another EPA level. The competency is rewarded with a Statements of Awarded Responsibilities (STAR). This will give the resident a gradual increase of autonomy and responsibilities in a safe and justifiable way (Cate, Snell and Carraccio, 2010a). Overall, the goal of EPAs is to educate and train residents to a well performing professional, which is able to carry out its critical EPAs in a clinical environment. Eventually, it should lead to better care for individuals and society (Cate, Snell and Carraccio, 2010a).

2.1.2 Relevance for EPAs

(11)

4 2.1.3 Characteristics EPAs

Due to the fact that the EPAs are new in the medical education, research mostly focuses on how to determine and assess EPAs (Ten Cate, 2005; Dijksterhuis et al., 2009; Cate, Snell and Carraccio, 2010a; Hauer et al., 2013; Van Loon et al., 2014; Touchie and Ten Cate, 2016), but not on the characteristics of the EPAs. However, there is a change in the literature on EPAs. Since EPAs are getting applied in the health profession education, such as Gynaecology and Obstetrics (Scheele et al. 2013), Family Medicine (Goodell et al., 2013), Psychiatry (Boyce et al., 2011), Haematology and Oncology (Karnad et al., 2014) and Pulmonary and Critical Care (Pastores et al., 2014), the characteristics of a well-described operational EPA becomes more clear. Most of the literature uses the framework provided by Ten Cate (2005) to specify the attributes of EPAs, which consist of the eight attributes represented below.

1. Are work activities that are essential in a given (clinical) context.

2. Requires specific skills, knowledge and attitudes acquired through training. 3. Leads to a recognised output of professional labour.

4. Usually restricted to qualified personnel. 5. Can be executed independently.

6. Executable within a timeframe.

7. The output should be measurable and observable during the process and lead to a conclusion. 8. At last, it must reflect one or more competencies that need to be acquired to be entrusted

with the professional activity.

In addition to the eight attributes described above, Van Loon et al. (2014) stated that EPAs should not be too detailed resulting in an administrative burden. Where assessing EPAs becomes assessing a checklist, instead of assessing competencies of the resident. Thereby the workplace must provide some flexibility for individual learning trajectories.

2.1.4 Pros and cons of EPAs

Now the EPAs being implemented in the healthcare, the (dis)advantages become notable compared to the traditional master-apprentice teaching method. A huge benefit of EPAs is that they are flexible. EPAs take previous training and experience of the trainee into account. EPAs also facilitate time flexibility. Instead of fixed rotations where trainees get a restricted amount of information in a certain time, EPAs are more flexible. Learners with a steep learning curve will still be rewarded using EPAs, and learners with a less steep learning curve can take more time to accomplish a certain EPA level. Thereby it distinguishes the competencies needed for a specific clinical environment (Cate, Snell and Carraccio, 2010a). However, this also results in location-specific EPAs, trainees learning for the same specialism at different locations could have a different set of EPAs (Mulder et al., 2010). Universal EPAs will be needed to be established universal education and assessment for each medical health professional education programme.

(12)

5 trainees need to be reminded to complete their expected activities, the faculty needs to be informed about the progress of the trainee and the administration needs to be checked to ensure the trainee has achieved a certain level of entrustment (Hauer et al., 2013).

Besides working with EPAs results in more transparency of the trainees’ competencies instead of working with single diplomas and uniform registrations (Cate, Snell and Carraccio, 2010a; Ten Cate et

al., 2015). An EPA can be "nested" which means that it consists of smaller EPAs. Resulting in a better

understanding of the educational goals and better tracking of the progress of the trainee. Some EPAs overlap each other, this counter redoing of already accomplished learning goals. Figure 1 below is a visual representation of a nested EPA.

Figure 1 Example of nested EPAs (Ten Cate et al., 2015)

Additionally, EPAs assess the trainees on observable competence. While the majority of internships are mostly organised in a fixed rotation system, based on time, not on learning outcomes (Ten Cate et

al., 2015). Although the flexibility in the duration of education is resulting in more complexity from an

operations perspective. Less standardized throughput times (of the educational programmes), more flexibility (in choice for education department) resulting in a more complex schedule.

Nevertheless, EPAs also have some drawbacks, which make them hard to implement. Due to the clinical work environment, is it hard to implement EPAs assessment tools. The service expectations are high, inpatient rotations, different learning potentials for the trainees, the capacity of supervisors and nursing staff and at last the culture (Cate, Snell and Carraccio, 2010a). Immediate work duties have more priority than education (Hauer et al., 2013), since the trainees are assessed by the supervisor they are sometimes obligated to prioritize work duties over educational goals. As a consequence, residents struggle to balance work and learning (Deketelaere et al., 2006).

(13)

6 2.1.5 Logistical challenges of EPAs

Since EPAs are new, the effects of EPAs on operations are unknown (Ten Cate et al., 2015). The literature with some operational focus is on the digitalization of the portfolio of the trainees (Mulder et al., 2010; ten Cate et al., 2017). Although logistical problems would occur for implementing EPAs in the workplace. Since residents having a double role, acting as employees providing patient care and acting as trainees who must accomplish its learning goals, their scheduling becomes complex. Depending on the EPA level the resident is in, it needs a supervisor to assess its work. The resident probably needs to accomplish its EPAs within a certain period, at a certain ward. Causing all sorts of restrictions resulting in a complex schedule. All those factors stressing the importance of a scheduling framework, which is able to cope with the flexibility of the EPAs.

2.2 Scheduling Residents

This section provides a Planning and Control framework (P&C) for the residents’ activities. Thereby it discusses different factors affecting the framework.

2.2.1 Planning and control residents

P&C can be described as the "Integrated coordination of resources (staff, equipment and materials)

and product flows, in such a way that an organization’s objectives are realized." (Anthony, 1965). The

urgency for providing more cost-efficient healthcare is rising. Due two reasons: decreasing healthcare budgets and ageing population resulting in an increase in demand for health care services (Poksinska, 2013). However, the P&C of the healthcare lags behind compared to the manufacturing P&C because of the unique character of the healthcare scene (Hans et al., 2011). The reasons listed below are explaining why the P&C in healthcare is lacking behind compared to the manufacturing P&C. Specified on the residents' situation. However, these problems are general for the healthcare scene.

▪ Lack of cooperation between involved parties, to illustrate, residents have different objectives (i.e. education) compared to the scheduler (satisfy the demand for care). Each department has its own objectives, which could be conflicting (Hans et al., 2011).

▪ The information systems are not supporting the P&C function. The information systems are purely focussed on the financial and clinical information (Khoumbati, Themistocleous and Irani, 2006). The schedule of the resident is not able to log the residents' competencies, misalignment between the entrusted activities and performed activities by residents could occur. Which could affect the provision of patient care (Holmboe et al., 2016).

▪ P&C is fragmented among autonomously managed departments. Each department has no knowledge beyond the P&C of its own department (Porter and Teisberg, 2007). One department could have too many residents during a day, while another department is coping with a capacity shortage. Exchanging the capacity is not possible due to autonomously managed departments.

▪ Most of the staff is educated in medicine, thus especially focussed in providing the best care, however, they lack training related to operational knowledge to schedule the resources in the most effective and effective way (Carter, 2002). Many residents themselves are responsible for providing the operational schedule. Besides the lack of training, they have to do it in addition to their already busy tasks. Which affects the quality of the schedule.

(14)

7 Resources, like staff, equipment are shared among different departments all affecting the P&C. Since residents must do internships at different departments, they are shared resources. The residents' programme is dependent on the patient population, they sometimes can only perform activities on certain patients to train their EPAs. Thereby shifting capacity occurs due to the fact of emergency activities, affecting the educational activities.

In short, there are still sufficient improvements to make regarding the overall P&C for residents and the healthcare P&C in general.

2.2.2 Planning and control framework healthcare

The literature of Hans et al. (2011) provides a P&C framework adapted to healthcare. This framework is based on the classical decomposition of hierarchical levels provided by Anthony (1965). He decomposed the P&C into three levels; strategic (structural/ strategic decisions), tactical ("translating" the strategic decisions into operational objectives, such as the number of staff needed yearly to fulfil demand) and operational level (short-term, decisions related to the execution of daily operations). Hans et al. (2011) extended this by the online and offline operational level. Where online concerns reacting to unforeseen or unanticipated events, and offline concerns the decisions related to the planning of operations in advance.

To make a distinction in the P&C of different resources in the healthcare system, Hans et al. (2011) divided the healthcare into four managerial areas. The four areas concerning medical planning (decisions related to i.e. treatments, medical protocols, mostly done by clinicians), resource capacity (P&C of renewable resources, such as equipment, staff, rooms), materials planning (consumable resources/ materials) and financial planning (financial decisions).

The decisions per managerial area and hierarchical level are affected by internal and external environments characteristics as can be seen in Figure 2. Internal characteristics are the boundaries of the organization, while external characteristics are outside the organization. External characteristics could be expressed according to the STEEPLED analysis an abbreviation for Social, Technology, Economic, Environmental, Political, Legislation, Ethical factors and Demographics (Johnson, Scholes and Whittington, 2008). The relevant external factors for the residents P&C are elaborated below.

Social

Social factors are related to the lifestyle and social characteristics of the system. Work-life balance, adherence to schedule preferences, illness of residents and hierarchy in the healthcare system are social factors influencing the P&C of the residents.

Technical

The P&C of the residents is influenced by technical factors, such as the scheduling programme, new medical treatments developed affecting the education of the residents, use of IT systems for delivering care, the software used to supervise the residents.

Economic

(15)

8

Political

Political change is influencing the P&C of the residents. Since they decide to decrease the funding for the residency programme and shorten the length of the education with six months (Bremers, 2011; Bremers, 2019).

Legal

The legislation is one major influencing factor for the P&C for the residents. The residents are only allowed to work a limited number of hours each month, during their work they have to adhere to the safety protocols.

In the table below are the relevant external factors for the P&C for the residents for this research represented.

Table 1 Relevant external STEEPLED factors applied to the residents’ situation based on Johnson, Scholes and Whittington (2008)

Factor Examples

Social • Work-life balance

• Adherence to schedule preferences • Illness

• Hierarchy in the healthcare system

Technical • Scheduling programme

• New medical treatments developed • Use of IT for providing care

• Software for supervising

Economic • Funding education residents

• Operational budgets hospitals

Political factors • Decreased funding for residents

(Bremers, 2011)

• Shorter educational programme (Bremers, 2019)

Legal • Working hours regulations

• Safety Protocol adherence

(16)

9

Figure 2 Framework adapted from Hans et al. (2011)

2.2.P&C framework residents

For developing a scheduling protocol for the residents' activities, the scope of this research is on the managerial area, resource capacity planning on a tactical and operational level (indicated by the red block in Figure 2 above). The framework provided by Hans et al. (2011) needs to be adapted for suiting the unique situation for the residents.

Shared resources

One of the reasons why scheduling residents is difficult is that they are a shared resource in the healthcare system. Residents provide patient care and attend educational activities. Hoekstra and Romme (1992) defined a shared resource as a common-capacity source that is used in two or more ways. Examples of shared resources in healthcare are beds, equipment, staff, since they are used for multiple objectives or patient groups. The main reason to share a resource is scarcity, due to cost (i.e. expensive equipment) or knowledge (i.e. specialised staff). The P&C of these shared resources is already complex, but becomes even harder if it requires high utilization (Drupsteen, Vaart and Donk, 2015). However, shared resources also have advantages. As stated by Hoekstra and Romme (1992), (1) it increases resource utilization, (2) flexible capacity to overcome temporary overcapacity/ under capacity, (3) it enhances specialisation (i.e. more specialised surgeries performed due to sharing resources among different departments) (4) and it enhances synergy in the system.

Current literature on shared resources are focussed on the allocation of hospital beds (Ridge et al., 1998), X-ray (Drupsteen, Vaart and Donk, 2015) and nurses (Maenhout and Vanhoucke, 2013). While the allocation of residents as shared resource considering their EPA level is still missing in the literature.

Competency level

(17)

cross-10 trained, resulting in full flexibility of resources. Therefore, resources can be widely applicable in the system, they “float” between peaks in demand, resulting in less capacity needed (Bam et al., 2017). In the Operations Research literature there has been a lot of mathematical models incorporating the skills of the employees in the healthcare (Maenhout and Vanhoucke, 2013; Paul and MacDonald, 2014; Bam et al., 2017), service operations (Pinker, Shumsky and Simon, 2000; Bard, 2004; Campbell, 2008) and manufacturing (Bokhorst and Gaalman, 2009; Othman, Bhuiyan and Gouw, 2012). Nonetheless, these models are lacking to incorporate information about the actual situation on the floor regarding employees’ condition and production. As a consequence, there is a gap between the theory and the practice of production planning models (Othman, Bhuiyan and Gouw, 2012; Bokhorst and Gaalman, 2009).

A structured way to represent workers skills can be done by a bipartite graph (Jordan and Graves, 1995; Bokhorst and Gaalman, 2009; Easton, 2011). This graph links employees’ abilities to a certain task. A task can be identified by three constituent elements: labour, entities and resources (Hopp and Van Oyen, 2004). Labour is related to who is accomplishing the task. Entity refers to the output of the system and resources refers to the associated equipment required to perform work in the system. For this research labour implies the residents and supervisors, entities are the patients and resources are equipment used during the care, such as hospital bed, X-ray. Those three elements are all incorporated in the EPA level. In the figure below is a bipartite graph represented specified on the resident’s EPA level. As can be seen from Figure 3 below, resident 1 is capable of doing task up to EPA level 4, while resident 2 is only capable of doing task up to EPA level 2. Resulting in that resident 1 is more flexible to schedule. EPA level 5 EPA level 4 EPA level 3 EPA level 2 EPA level 1 Resident 2 Resident 1

Figure 3 Bipartite graph specified on the residents

(18)

11

Figure 4 Skills level (Lx) for Maintenance Operators (MOx) on Specialities (Sx) (Hadjaissa et al., 2016)

2.2.4 Performance resident's schedule

To measure the performance of the P&C protocol, performance indicators are needed. Maenhout and Vanhoucke (2013) measured the performance of their nursing schedule based on the performance indicators: efficiency (such as cost), job satisfaction (schedule desirability) and effectiveness (providing quality care). These performance measurements are also suited for this research. The three performance indicators are decomposed in operational outcomes in the section below. Those operational outcomes are specified on the EPAs.

Efficiency

Efficient use of the resources and especially the Bottleneck capacity. An indicator of efficiency in the P&C of the residents is (Maenhout and Vanhoucke, 2013):

• Supervisor capacity (ability to achieve educational activities is dependent on the supervisors, while supervisors have a maximum capacity of supervising residents).

Job Satisfaction

For this research, job satisfaction considers:

• Working conditions (Aziri, 2003), such as the fair distribution of shifts among residents (Topaloglu, 2006; Rose et al., 2015; Bam et al., 2017), even distribution of day, night and weekend shifts. The supervision and workers capacity differ over the shifts, during the night and weekend shifts there are fewer supervisors available for providing feedback. Which could make it harder to accomplish EPAs during these shifts.

• Social relationships (Aziri, 2003), with patients and colleagues.

• Autonomy (Mueller and Mccloskey, 1990), the degree of feeling capable of performing tasks independently.

Effectiveness

(19)

12 residency programme, the objective is to expose the residents to learning goals during the shift. Thus, the following components will be considered:

• Exposure to learning activities for the Residents: since the education occurs especially on the job.

2.3 Summary theoretical background

The theoretical has shown that the current literature on EPAs is mostly focussed on establish assessment criteria and implementing EPAs in the residency educational programme. The use of EPAs results in more flexibility for the resident, better transparency in assessing the resident's skills and provides more structure in getting entrusted with the professional task. Logistical challenges of implementing EPAs in the residents' schedule, however, is missing in the current literature. Operational problems such as scheduling regular care task and educational activities considering their EPA level, scheduling EPAs in a flexible time frame are not discussed in the current literature. This stresses the importance of developing a scheduling framework for the residents considering their double role and EPA level.

(20)

13

3. Research framework

This chapter is structured as follows: section 3.1 problem description and stakeholders, 3.2 the research objective, 3.3 conceptual model, and lastly, 3.4 the research design and different phases of the study.

3.1 Problem description

This section will give the problem description well as the relevance of the problem. This section also addresses the stakeholders involved including their requirements.

3.1.1 Problem statement

UMCG has three main tasks. They deliver patient care, conduct research, and lastly, they provide education. This research specifies on the educational task of the UMCG. The UMCG is providing medicine residency programs, which allow graduated physicians in medicine to become a medical specialist in a certain health discipline. These residents are called AIOS in Dutch (Dutch abbreviation: "Artsen In Opleiding tot Specialist"). The medicine residency programme consists of workshops, multidisciplinary meetings, training and internships. Resident’s competencies are assessed by a new method, called Entrustable Professional Activities, EPAs. Which allows residents to train competencies (under supervision) in a specific clinical environment. While the residents are trained to become a specialist, they also have to provide patient care task. If the resident has accomplished the training of three to seven years (dependent on the specialism), he or she becomes a specialist in a health discipline.

However, during the residency programme, residents have a double role. They act as both a learner and a provider of care services. During the training, residents have to be supervised by a supervisor, while during regular shifts they act independently (they can ask for support from a supervisor when needed). This results in a complicated schedule for the residents. Major hospitals are facing this problem of designing feasible schedules that cope with the double role of the residents without enormous over hours (Abrass et al., 2001; Day, Napoli and Kuo, 2006; Topaloglu, 2006). Resulting in exhausted residents with an unbalanced work-life balance (Prins et al., 2007).

In addition, there is a decrease in funding for the residency programmes (Abrass et al., 2001), as a result, there are fewer residents available which increase, the already high workload of the current residents (Day, Napoli and Kuo, 2006; Bremers, 2019). Often the providing of patient care is prioritized, which affects the resident's time spent at educational goals. Hence, longer training of residents is required. While since 2012, the Dutch government is trying to reduce the training of residents with 6 months. This stresses, even more, the importance of an effective and efficient schedule for the residents (Bremers, 2019).

Furthermore, EPAs results in a more personalized schedule for the residents. They can choose their own pace and clinical environment to learn the competencies. This is conflicting with the traditional rotation system for residents with set due dates. Besides, there are also the preferences of the AIOS themselves, regulations like the amount of consecutive day and night shifts, staffing requirements of the ward and this all must result in a fair distribution of responsibilities among the residents (Topaloglu, 2006).

(21)

14 multiple adoptions are performed as gaps rise in the schedule (Day, Napoli and Kuo, 2006). This process is (time) costly and is still not meeting the requirements of various stakeholders (Topaloglu, 2006). This complex scheduling problem is present in a lot of hospitals who provide residency programmes (Day, Napoli and Kuo, 2006; Topaloglu, 2006; Anderson and Gamarnik, 2013; Guo et al., 2014).

3.1.2 Case study

For this research, a single case study will be conducted. This case study will take place at the UMCG. The case should have residents and is currently developing or already working with the EPAs. Additionally, the department is looking for ways to cope with the decreasing capacity of residents in terms of numbers and duration of the internship. This should result in more willingness to corporate in this case study and thus better research outcomes.

3.1.3 Stakeholders

Views and requirements of different stakeholders shape the research objective of this project. These different views are simplified into two main views in this research. The patient care view and educational view. The first one is based on providing the best patient care for the patient and the second is based on providing the best education for the residents. However, the residents have a special view, since they act both as learners and as providers in the system. This section provides an explanation of the stakeholder’s role(s) and requirements.

Residents (AIOS)

The main subjects of this research, since the goal is to develop a feasible scheduling protocol for the residents considering the EPAs. Residents have the following requirements:

▪ Have the freedom and flexibility to choose their own EPAs in their own timeframe. ▪ Well supervised during the internship, to accomplish EPAs.

▪ Require a sufficient amount of time for their family life and educational goals. ▪ A fair balance between day and night shifts among the residents.

▪ Fewer changes in the schedule at the last moment, since those, are causing stress. ▪ Will have some freedom to select their own shifts.

Supervisors

The supervisors are responsible for the residents during their educational activities (especially internships). Supervisors have the following requirements:

▪ Want to provide the best education for the resident.

▪ Would like to spend enough time per resident, which requires a maximum threshold on the number of residents.

▪ Are safeguarding the quality of the residents' competencies, since they assess the EPAs of the residents.

▪ Want Fewer changes in the schedule at the last moment, causing extra stress.

Residency programme coordinator

Responsible for the residents during the residency programme and working closely together with the supervisors. The residency programme coordinator has the following requirements:

▪ Responsible for the P&C of the residents.

(22)

15 ▪ Wants to provide the best educational programme for the residents.

Other physicians at the department (ANIOS)

Are physicians providing the same regular care task as the residents, but they are not participating in the medicine residency programme, called ANIOS (Dutch abbreviation: "Artsen Niet In Opleiding tot Specialist"). Often choose this option to gain experience at the department, before they are applying for a residency programme. The physicians have the following requirements:

▪ Fair distribution between day and night shifts among AIOS and ANIOS. ▪ Want fewer changes in the schedule at the last moment, causing extra stress. ▪ Will have some freedom to select their own shifts.

Scheduler

Is responsible for the schedule of the employees at the department. The scheduler has the following requirements:

▪ Wants to have a more robust planning system, which should decrease the changes at the last moment, causing stress.

▪ Wants to have an overview on how to schedule AIOS with respect to their EPAs. ▪ Wants to fulfil the schedule preferences of the employees, if possible.

▪ Wants to have a sufficient amount of residents and other employees to satisfy capacity demand at the ward.

Other employees

Other (medical) employees, such as nurses. They have the following requirements: ▪ Want to have a feasible schedule.

▪ A fair distribution of day and night shifts. ▪ Want adherence to preferences if possible.

▪ Want a schedule, which gives less last-minute mutations.

3.2 Research objective

Considering the problem introduced in the previous paragraph, many hospitals are struggling to establish a P&C system for residents that could cope with EPAs, preferences and restrictions. This stresses the importance to develop a scheduling protocol for the residents. Therefore, the following research objective can be obtained:

"Develop a scheduling protocol for the planning and control of the residents', considering the EPA level of the residents, which is striving to improve the effectiveness, efficiency and job satisfaction at the department, residency program and the residents."

(23)

16 The scope of this research is limited to the residents' perspective. The main objective is to optimize the schedule for the residents at the case study,considering the educational objectives according to the EPAs and the patient care tasks. Other employees scheduling and patient scheduling will be left out the scope of this research due to complexity and time constraints. This research has a technical and social scope. The technical scope is focussed on how to implement the EPA's in the schedule to establish an efficient and effective schedule related to the education of the residents and the provision of patient care task. The social scope is focussed on adhering the preferences of the various stakeholders. Such as providing a P&C system that is more efficient combining educational and patient care tasks, resulting in a better family-work life balance and job satisfaction for the residents.

This research is conducting a case study at the UMCG. Obviously, this results in specific deliverables for the case study. The main deliverables for the case of the UMCG will be:

▪ Provide an overview of the current P&C methodology related to restrictions and limitations and stakeholder requirements.

▪ Provide a scheduling protocol which is able to standardize the use of EPAs in the P&C of the case.

▪ Asses the scheduling protocol incorporating EPAs based on the performance indicators and identify the effects of the proposed P&C protocol.

▪ Asses the usability of the P&C protocol incorporating EPAs for the departments.

3.3 Conceptual model of this research

(24)

17

Figure 5 Conceptual model

As can be seen in Figure 5 the input of this research is the residents’ tasks, since they are both providers of patient care and attending educational activities. Those two perspectives need to be addressed in the P&C protocol for the residents. This research is mostly focussed on the operational and tactical level indicated by the red block in the figure above. After taking those factors into account a schedule can be produced. The scheduling framework is tested on effectivity, efficiency and job satisfaction (Maenhout and Vanhoucke, 2013). Which will eventually lead to trained residents and provided patient care task by the residents.

3.4 Research design

For this research, a design science is used. Design science is applicable for solving problems in practice and is able to generalize the findings (Holmström, Ketokivi and Hameri, 2009). Since the research is coping with a practical problem, and the outcome of the research is possibly a useable planning protocol for other departments or hospitals, makes this research suited for the design science method. In contrast, a case study and survey are not able to create such artefact, and for this reason, not applicable for this research.

(25)

18 After the literature review to form a well-founded theoretical background, the research follows the regulative cycle of practical problem solving provided by van Strien (1997). This approach consists of 6 phases: (1.) problem definition, (2.) system description, (3.) system analysis, (4.) design, (5.) evaluation and (6.) implementation. Due to the restricted amount of time available for conducting this research, phase 6, implementation, is left out in this research.

3.4.1. System description

Provides an overview of the case at the UMCG related to the scheduling protocol of the residents. The system will be described according to the literature provided by Hans et al. (2011). Information is gathered through interviews with the stakeholders, especially the scheduler(s).

3.4.2 System analysis

In extension of the system description, the system will be analysed. The aim of this chapter is, how the current P&C protocol is scoring based on the EPA related factors. Data that is needed to asses these criteria are provided by interviews with the stakeholders and historical scheduling data of the case. Afterwards, directions for improvements will be stated.

3.4.3 Design

This phase of the research will elaborate on the causes found in the previous chapter. Considered causes affecting the P&C of the residents will be taken into account. This, in addition to the literature findings, will be used for designing a protocol incorporating EPAs. Which eventually, will lead to an improved P&C protocol for the residents according to the three performance indicators, efficiency, effectiveness and job satisfaction. During this phase, it is essential to take the generalisation into account. Since the framework should contribute to standardization of scheduling of the resident’s activities at different departments (or hospitals).

3.4.4 Evaluation and implementation

Evaluation and implementation, this is the last phases of the regulative cycle of van Strien (1997). This phase is used to evaluate and implement the suggested improved P&C of the residents. The system will be tested on the same performance indicators, used in the system analysis chapter. However, due to the restricted time available for this research, this chapter will mostly provide recommendations for further research and improvements.

3.4.5 Data

(26)

19

4. System description

This chapter describes the P&C of the residents at the Intensive Care Unit (ICU). First, the patients, employees and the capacity of the departments will be considered. Thereafter the educational programme will be discussed. At last, the P&C of the residents is elaborated based on strategic, tactical, and operational level introduced by (Hans et al., (2011).

4.1 Intensive Care Unit (ICU)

This research takes place at the UMCG. Different departments in the UMCG were contacted and investigated for their usability of this research. As a result, the department Intensive Care (IC) for adults (in Dutch IC voor Volwassenen, abbreviation ICV) suited the requirements.

4.1.1 Patients

The ICU delivers intensive care to patients who need (temporary) support or guarding of their vital organ functions. Different patients are treated at the ICU. There are patients who had surgery, this inflow is estimated around 50% of the total inflow of the ICU. Those patients arrive most of the time in office hours or shortly after. For these patients there can be made some assumptions beforehand of which care they would need, however, complications could always occur changing the need for care. The remaining 50% of inflow of the ICU are patients from other departments/ hospitals who need intensive (specialised) care, those patients arrive through the day. These patients are unknown (short) in advance, resulting in that the required care is assessed at the department. The capacity of the ICU divided over three departments (ICV 2, ICV 3 & ICV 4) in the UMCG, this is due limited amount of space. All the IC departments are capable of treating all kinds of patients, however, every department is specialised in a health discipline. The speciality and capacity per ICU are depicted in Table 2 below. Due to a shortage of nursing workforce, the workable capacity is lower at the departments.

Table 2 capacity of the ICU departments in number of hospital beds

Department Capacity (number

of hospital beds)

Workable capacity in hospital beds (nursing workforce available)

Speciality

ICV 2 13 9 Cardiac surgery

ICV 3 14 12 Complex neurology

ICV 4 18 12 Liver transplants

4.1.2 Employees

(27)

20 Graduate

physician (ANIOS) Resident (AIOS) Fellow

Intensivist (supervisor, staff

members)

Figure 6 Hierarchical structure ICV

4.2 Intensive care medicine residency programmes

There are several education tracks within the ICU. There are Fellows, those are trained to become an Intensivist. The educational background of those Fellows is mostly specialist Internal Medicine or Anaesthesiology, however, some Fellows are specialized in Cardiology, Pulmonary Medicine, Internal Medicine, Surgery or Neurology. Residents or AIOS (interchangeably used throughout this paper), post-graduate physicians who want to specialize in a certain health discipline. The following residents are doing an (short) internship at the ICV: Internal Medicine, Anaesthesiology, Cardiology, Pulmonary Medicine, Surgery, Internal Acute Medicine or Neurosurgery. At last, ANIOS, post-graduate physicians who are not specializing in a health discipline (yet). In Table 3 and Table 4 below are the functions represented including their background, specialism and duration of their internship (depending on their educational background and specialism) at the ICU. Since ANIOS do not follow a Residency or Fellow programme they are not represented in the table below. ANIOS work a maximum of two years (as ANIOS) at the ICU, due to contractual obligations after two years.

During their training programme AIOS and Fellows have the following responsibilities: • Provide patient care during the day, night and weekend shifts.

• Attend multidisciplinary meetings (weekdays), with different health discipline specialist. Discussing the medical status of the patients at the ICU.

• Attend not patient-related educational activities, such as courses, simulation-based training, exams and attend congresses.

Table 3 Fellow programme at the ICU

Educational background (completed residency programme)

Duration fellow programme (based on 1.0 FTE (46 hours pw)

Time spend at department

Anaesthesiology 1 year ICV 2 & ICV 3: 3 months

ICV 4: 6 months

Internal Medicine, Cardiology, Pulmonary Medicine, Surgery

2,25 years (since they need 3 months extra at the Operation Room, OR)

(28)

21

Table 4 Residency programmes and duration of internship performed at the ICU

Residency programme Duration at the ICU based on 1.0 FTE (46 hours pw)

At department

Anaesthesiology 12 months ICV 2: 3 months

ICV 3: 3 months ICV 4: 6 months

Internal Medicine (INT)

3 months ICV 4

Surgery (CHIR) 3 months ICV 4

Cardio-Thoracic Surgery (CTC)

3 months ICV 2

Acute Care (SEH) 3 months ICV 4

Neurosurgery (NCH) 3 months ICV 3

Cardiology (CARD) 0,5 months ICV 2

Internal Acute medicine (INT AG)

2 months ICV 3

For this research, the choice is made to focus only on the schedule of the Fellows. Since Fellows are mostly trained at the ICU department, by EPAs developed by the department (except for the internship at the Operating Room for the Fellows with a specialization in Internal Medicine, Cardiology, Pulmonary Medicine see Table 3), an increase in EPA level on a specific task/ patient group is increasing directly the level of entrustment and autonomy of a Fellow at the ICU department. Another reason is that those Fellows are for a long time at the department, therefore a higher level of EPA can be accomplished. This higher level represents a higher level of autonomy of the Fellow, which can be seen as a more cross-trained employee (more flexible capacity at the department). Which makes it easier to pool capacity (see chapter 2), to cope with under/ or overcapacity at different IC departments at the UMCG. Besides the Fellows are able to train and guide A(N)IOS, thus playing a key role at the ICU. At last, the Fellow training is currently the most advanced using EPAs. However, the designed protocol should be easy to adapt to be useful for the AIOS in the (near) future.

4.2.1 Training of Fellows

EPAsICM

(29)

22 Advocate and Communicator (Mulder et al., 2010). The CoBaTrICE (Competency-Based Training in Intensive Care MedicinE) is the international standard of ICM training in Europe, which determines the level of knowledge, skills and attitude the Fellows should have to conduct medical tasks. Those two are combined in the EPAsICM. The ICU assess the Fellows on 15 different EPAsICM, which are represented in

Table 5.

Table 5 The 15 different EPAsICM and the description of the clinical problem

EPA ICM Clinical problem

1 Postoperative management care of surgical patient 2 Consultation and triage of potential ICU patients 3 Management of patient with sepsis

4 Management of patient with acute abdominal condition

5 Management of patient with cardiogenic shock and/or cardiovascular disorders 6 Management of patient with massive bleeding

7 Management of patient with complex ventilation and oxygenation problems 8 Management of patient with altered consciousness

9 Management of patient with acute (on chronic) liver failure 10 Management of trauma patient

11 Management of patient with acute brain condition (ischemia and intracranial bleeding) 12 Management of patient with renal failure

13 Management of immunocompromised patient 14 Management of patient with metabolic disturbance 15 Management of long stay ICU patient

To illustrate the use of EPAsICM at the ICU, EPA ICM 3, Management of patient with sepsis, will be further explained. In Appendix B can the assessment form of EPA ICM 3, Management of patient with sepsis, be found. The structure of the assessment form is as follows, the first line is to indicate the level of supervision for the Fellow, each increase in level gives the Fellow more entrustment for executing clinical task more autonomously. The degree of entrustment and the corresponding level of educational supervision is depicted in appendix B, Table 13. Each EPA ICM has a main focus, this describes globally the clinical problem and the needed Skills, Knowledge and Attitude. Each assessment form includes a checklist, based on the knowledge, skills and attitudes needed (based on CoBaTrICE programme) and the associated CanMEDS competency role(s). The supervisor can give feedback on the different aspects, which gives him a transparent and holistic overview of the Fellow’s competencies on the EPA ICM 3. Thereafter the supervisor signs the form, after that the Fellow can put this form into his own portfolio. If the Fellow wants to reflect on certain aspects, he or she can make use of the CoBaTrICE programme where all the knowledge, skills and attitudes are explained extensively. Each EPA ICM has a reference list to the corresponding literature in CoBaTrICE programme (appendix B, Table 14).

KPBs

(30)

23 the progress meetings with the supervisor(s) are logged in the online EPASS programme. This online programme is able to log the competencies of the health professional using the CanMEDS method.

Supervision

Supervisors are responsible for the quality of care at the ICU unit and thus the actions of the Fellows. In Figure 7 is the degree of autonomy and the degree of supervision the Fellow needs over a timespan represented. As can be seen from the figure the quality does not change, only the degree of supervision changes if the Fellow accomplishes another EPA level. In the beginning, the supervision rate is high and the degree of autonomy of the Fellow is low. After time exceeds and the Fellow is completing higher EPA levels, less supervision is needed, and the Fellow becomes more independent.

Figure 7 Quality at the ICU and the degree of supervision of the Fellow

(31)

24

Figure 8 Increase in EPA level versus the needed capacity (Fellow + degree of supervision supervisor)

4.3 P&C of ICU

4.3.1 Strategic P&C ICU

The total capacity of the Fellows, A(N)IOS at the ICU department at the UMCG is given in Table 6. The capacity is represented as the average per year since 2015, multiplied by the average

fulltime-equivalent, FTE (1.0 FTE = 46 hours a week), per function. Only 2019 is divided into periods of six months because the capacity schedule of period July-December is still in progress during the period this research is conducted, therefore not all capacity is known. As can be seen from the table, the ICU unit is coping with a decrease of inflow AIOS. The total capacity at the department is set by the hospital based on the available resources at the department and the estimated number of patients. The inflow of AIOS and Fellows at the ICU is based on recommendations of the Capaciteitsorgaan, this organisation is responsible for balancing the need for medical specialist with the number of specialists in training in the Netherlands.

Table 6 Average capacity in fulltime-equivalent (46 hours a week) per function per year since 2015

Function Average FTE Average

(32)

25 AIOS 0,9 17,08 15,5 17,42 11,42 9,5 6,17 Total Fellow ANIOS AIOS 35,23 37,48 39,24 33,39 29,3 23,72

4.3.2 Structural P&C ICU

The number of supervisors, A(N)IOS and Fellows is set by the hospital. The structural planning, incorporating when the Fellows and A(N)IOS are starting at the ICU, is scheduled by the department. This could have a timespan of 5 years in advance. There are two stakeholders responsible for the occupation overview schedule including A(N)IOS and Fellows. An Intensivist and the Secretary’s Office. The Intensivist is responsible for composing the occupation overview schedule for Fellows and A(N)IOS based on experience in a trial-and-error way. The schedule is produced in Microsoft Office Excel. The Secretary’s Office oversees specifying when which A(N)IOS and Fellows are available and the degree of FTE. If there are mutations in this schedule, the Secretary’s Office is in control of interchanging the capacity among the three ICU departments. Ensuring that the capacity at each department is sufficient. An overview of the occupation schedule is represented in the appendix C.1 Occupation overview schedule.

4.3.3 Operational P&C ICU

Monthly schedule

The occupation overview schedule provides the input for the monthly schedule with Fellows and A(N)IOS at the ICU departments. The schedule is made 1,5 months in advance. The schedule is also made by the Intensivist supported by the Secretary’s Office. The Secretary’s Office is responsible for filling the schedule with predefined (educational) activities, such as training, holiday request and other wishes of the Fellows/ residents. The Secretary’s office is supporting the scheduling process to relieve the (scheduler) Intensivist. After the Fellows/A(N)IOS, FTE rate, special periods, holiday request are filled by the Secretary’s office, the file is ready for the scheduler Intensivist.

(33)

26

Restrictions

• Capacity. Number of Fellows, A(N)IOS attend to provide patient care, depending on the department and shift. Influenced by the government decisions, the budget of the hospital and set protocols (Political, Economic and Legislation).

• Sufficient amount of supervisors (Intensivists) due to the variability in patients’ illness/ disease at the ICU during the day there need to be two supervisors, during the evening and night one for department 4 and one supervisor shared among ICV 2 & ICV 3(Legislation).

• After a weekend night shift, the Mondays and Tuesdays are off (Legislation). • Only one type of shift a day, to adhere to the working time law (Legislation). • At least 2 nightshifts are subsequent (Legislation).

• After the nightshifts 2 days off (Legislation).

Wishes and request

• The Fellows and A(N)IOS are able to do three “Hard” requests, those requests are (mostly) honoured. Besides they are able to do three “soft” requests, the scheduler is trying to obey to these wishes, however, this is not guaranteed (Social).

• The resign term for the Fellows/ A(N)IOS is 1 month, however, the schedule is made 1,5 months beforehand. Thus, the request of the planner is to indicate this 1,5 before, resulting in fewer mutations (Social and Technical).

• Schedule the A(N)IOS/ Fellow on the department where they could learn the most, resulting in a more flexible workforce (Social)

• No more than 4 nightshifts are subsequent. • No more than 7 shifts subsequent scheduled. • A Fellow during a weekend day (long) shift.

(34)

27

Table 7 Shifts at the ICU department

Shift Time Minimum

A(N)IOS or Fellow per department

Min Fellows ICV 2,3,4 aggregated or AIOS Anaesthesiology >= 6 month working experience at ICU Min amount of supervisors per department Total capacity Supervision rate = (A(N)IOS+Fellows) / Supervisors Weekdays 8.00-17.30 ICV2: 3 ICV3: 3 ICV4: 3 - ICV2: 1 ICV3: 1 ICV4: 2 14 9/4=2.5 Evening (weekdays) 15.00-0.00 ICV2: 1 ICV3: 1 ICV4: 1 1 ICV2 + ICV3: 1 ICV4: 1 7 4/2=2 Night (weekdays) 23.00-9.00 ICV2: 1 ICV3: 1 ICV4: 1 1 ICV2 + ICV3: 1 ICV4: 1 7 4/2=2 Weekend day short + long 8.30-15.00 (short) 8.30-21.00 (long) ICV2: 2 ICV3: 2 ICV4: 3 - ICV2 + ICV3: 1 ICV4: 1 7 7/2=3.5 Weekend night 20.30-9.00 ICV2: 1 ICV3: 1 ICV4: 1 1 ICV2 + ICV3: 1 ICV4: 1 6 4/2=2.0 Operational online

(35)

28

5. System analysis

This chapter describes the process flow at the ICU, next the system is analysed based on

performance indicators related to EPAs. This is based on schedules from January 2019 up to April 2019, these schedules are validated and verified and by the scheduler and representing the actual situation at the ICU. First the process flow at the ICU is explained, next, the distribution in shifts among A(N)IOS and Fellows are depicted. Thereafter the experience of the team during a shift will be analysed. At last, the dispatching of the workforce among the patients will be analysed.

5.1 Process flow ICU units

The simplified process flow of the ICU unit is depicted in Figure 9.

Dispatch workforce among patients Provide patient care Transfer of patients to next shift

Figure 9 Process flow ICU units

Dispatching workforce

At the beginning of the shift, there is a transfer of patient information from the previous shift to the upcoming shift which takes around 30 minutes. During this transfer, they walk by every patient at the ward and discuss patient illness/ diseases and the treatment plan for the patients. Every A(N)IOS and Fellow has a document where the patients are listed including their relevant medical information. The workforce takes notes for relevant information about the patient. Throughout the patient transfer patients getting dispatched among the workforce. Afterwards, the workforce takes place at a control unit. Here are computers and monitors placed with patient health status such as heart rate, blood pressure, oxygen levels.

Providing care

During the providing of care, the A(N)IOS and Fellows discuss treatment plans with the nursing workforce or supervisor and provide care to the patients. Besides patient information is registered in an electronic patient system (called Elektronisch Patiënten Dossier, EPD, in Dutch). During a shift Fellows and A(N)IOS can check for relevant (background) information, such as information about the pathology of the patient and care protocols, to provide optimal care for the patients. Every weekday from 12.30 until 13.30 there is a multidisciplinary meeting with specialisms of relevant departments. Here are all doctors of the ICU (supervisors, A(N)IOS and Fellows) present. During this meeting (prospective) patients are discussed from different medical specialism to get a holistic view of the patient’s illness/ disease. Leading to a customized care plan for the patient. During a shift, new patients could arrive, these are distributed among the A(N)IOS or Fellow who has the capacity (in terms of beds available and time). If a patient is stabilized and needs less care, he is transferred to a normal department. This needs to be arranged (by mail or phone) with the other department and is dependent on their capacity and beds available. Occasionally, it occurs that the patient cannot directly be transferred due to that the receiving department is full.

Subsequent shift

(36)

29

5.2 Distribution shifts among A(N)IOS and Fellows

The distribution of shifts among A(N)IOS and Fellows should be equally divided to create fairness among the employees (Topaloglu, 2006). Besides the distribution of shift is affecting the learning curve of AIOS and fellows, since differing supervising capacity during the. To test the schedule on the distribution of shifts among the A(N)IOS and Fellows a distinction is made between the different functions of ANIOS, AIOS and Fellows. Thereafter, the percentage of a type of shift (i.e. weekday night shift in a month is calculated by the total hours weekday night shift divided by the total hours worked in the month). At last, the average is calculated over all the A(N)IOS and Fellows. Using this method gives a transparent way of the distribution of shifts among the A(N)IOS and Fellows and is not affected by the FTE rate the A(N)IOS and Fellows are working. Employees who work less than 24 hours (due to leave of absence/ moving to another internship) in the month are left out for analysis.

Table 8 depicts the average shift distribution of A(N)IOS and Fellows of the period of January 2019 up to April 2019. The shifts are mostly equally divided among the A(N)IOS and Fellows, however, the Fellows have relatively more day shifts compared to the A(N)IOS. The ANIOS have relatively the most night shifts, while the Fellows have the least night shifts. Appendix D: Analysis shifts distribution Fellows and A(N)IOS Jan – April 2019 is the average represented per month over the period from January 2019 up to April 2019.

Table 8 Average (AVG) distribution of shifts A(N)IOS and Fellows period January up to April 2019

Jan 19 – Apr 19 AVG % holiday request AVG %Education al activities AVG %Weekend night AVG %Weekend day short AVG %Weekend day AVG %Day shifts AVG %Night shifts AVG % Evening shifts ANIOS 7% 8% 7% 6% 7% 34% 18% 14% AIOS 11% 8% 7% 5% 4% 36% 14% 15% Fellow 9% 4% 8% 4% 7% 43% 11% 15%

During interviews with Fellows, AIOS and supervisors the most difficult shifts for learning new EPAs were named the night shifts. During these shifts, the supervisors are standby, but mostly not physically present at the department. Thus, it becomes hard to assess the Fellow or AIOS on a certain EPA during this shift. However, if the Fellow or AIOS is already at a higher EPA level (more autonomy, i.e. for deciding on the patient care policy during a shift), he or she could sign off the EPA during the transfer if the made decisions are well grounded and the supervisor is entrusting this activity to the Fellow or AIOS autonomously. During those evening and night shifts, the Fellows and AIOS (who are capable of doing consult shift) are assisting the supervisors. They can discuss the decisions made by other AIOS, Fellows, but they are not able to sign off EPAs. This is restricted to the supervisors.

Concluding, the Fellows or AIOS capable of doing consult shifts are better able to sign off EPAs during the night shift. While the AIOS or Fellows with lower EPAs could struggle by accomplishing new EPAs during the night shifts.

Develop a scheduling protocol for the residents’ activities