Reducing complexity in a complex situation: How to work with shared resources?

(1)

Reducing complexity in a complex situation:

How to work with shared resources?

An operations management approach

focused on the outpatient clinic dermatology

M. Boltjes B.

(S1937987)

Drachten, October 2011

University of Groningen

Faculty of Economics and Business

Business Administration –

(2)

By M. Boltjes B.

University of Groningen

Supervisor:

dr. J.T. van der Vaart

Co-Assessor:

prof. dr. J. Wijngaard

The hospital

(3)

Foreword

When I finished my HBO in 2009 I thought I was ready for the job market. I was wrong. A Master programme adds so much knowledge. I think that I would not have been a skilled worker with just the HBO degree. I am very happy that I made the decision to start a Masters. In the beginning I had an HBO attitude, as I like to call it. It turned out that for a Masters this was not going to work. Old habits had to be removed from daily routines, and from the moment I made that decision everything went smooth. With this Master Thesis my journey through the academic world at the University of Groningen comes to an end. Two years I have enjoyed being part of the Master programme Business Administration. Especially the second year in which the focus on Operations & Supply Chains got clearer. During these two years I have met some inspiring and smart people. With these people I have conducted several interesting and challenging assignments.

This thesis was conducted in a very interesting setting: a hospital. Until the moment I started working on the business administrative side of the hospital I had a complete different view on hospitals. During this project I got into contact with interesting theories. When I got introduced to the problem at hand I was overwhelmed by the complexity of it. The first few weeks I felt quite insecure about it, because I did not have a feeling with the project. Thanks to my supervisors, both from the university and the hospital, things got clearer and clearer. I want to thank dr. J.T. van der Vaart for his encouraging comments and feedback. They were really helpful, especially in the first few weeks. I also want to thank Pieter Buwalda for his critical looks at and feedback on my work, the helpful consultations and discussions we had, and his overall guidance throughout the project. Also Irene Zonderland, thanks for your help during this complex project. Without the information you have provided me with it would have been impossible to conduct this research project. From the outpatient clinic dermatology I wish to thank Martje and Lous for letting me have a close look at their planning activities. Further I wish to thank Minze de Boer for assisting me with unrevealing data. Also Konstantin Ignatov; thanks a whole lot for (y)our supportive conversations, discussions, and consultations. It was really fun to work alongside you.

Outside the hospital I could always count on one of my best friends Mark Lageman. Thanks for our chats, your helpful comments on my work, and for just being there. Also I want to thank my parents in law, Bé and Elsje, for their care and support in the period I broke my ankle. And, above all, I want to thank my girlfriend Melanie who has supported me while I did not see myself succeeding anymore. Thanks for being there.

(4)

By M. Boltjes B.

Abstract

As hospitals increasingly need to produce more with less, planning becomes an increasingly important business administrative task within hospitals. Processes become ever more complex to plan, many elements affect planning performance. Think of employee skills, length of planning period, use and utilization of resources, patient groups, and planning detail. Reducing planning complexity is a very complex matter. Complexity has two main pillars in a healthcare setting: the healthcare process and the planning process. This research has focused on the planning process of an outpatient clinic dermatology and has unravelled the main causes of planning process complexity in this setting. It has also created a specific method for finding causes of planning performance problems. This method is applicable for other service settings as well, be it with slight changes. The recommendations for the hospital were mainly in terms of restructuring the planning process and training their workforce. In the end benefits of a solid planning process and an optimal work schedule can be significant: a higher planning process performance, a more motivated workforce, a better use of capacity (cost efficient). All covered in this research.

(5)

By M. Boltjes B.

7 1 Introduction

1.1 The hospital and its planning problem

the hospital is located in Drachten and employs approximately 1500 employees. This study focuses on the outpatient clinic dermatology. Approximately 321 of these 1500 employees work at the outpatient clinic dermatology; four dermatologists, 212 medical assistants, and seven administrative workers. Typical treatments that are conducted within this outpatient clinic are light (UV) therapy, removal of varicose veins, skin cancer treatment, allergy tests, and psoriasis treatment.

The reason to conduct a research within the outpatient clinic is that they experience problems in the planning of their resources. These problems mainly find their cause in planning complexity. Planning has become more and more complex. The need to do more with less and greater diversity in provided treatment types are examples of developments that have created this complexity. The complexity makes that it is difficult for the planners of the outpatient clinic to create a schedule in which every treatment is covered sufficiently, contracts of nurses are met (in terms of contracted hours and set days off), skills and resources are optimally utilized, patients receive high quality healthcare, and all required specialists and assistants are available when needed. These complexity causes combined make it even more complex, is also a problem that Ernst, Jiang, Krishnamoothy & Sier (2004) found; they found that it is complex to simultaneously satisfy all these kinds of requirements.

The planners of the outpatient clinic have to put a lot of effort in scheduling, and the amount of required effort is growing. The scheduling problems they experience can easily result in inefficient usage of critical resources and waiting times for patients. This research should provide detailed insight in the scheduling problem and in the consequences for the operational performance of the outpatient clinic.

In general it can be stated that a complex planning process has two main aspects that contribute to this complexity. On the one hand there is the (healthcare) process to be planned (i.e. the puzzle to be made), and on the other hand there are the resources involved in this process (i.e. the different pieces of the puzzle). The way in which these two processes (i.e. the healthcare planning process and the resource planning process) are taken care of influences the level of complexity.

In terms of planning, the way in which the planning process is structured (how the puzzle in made), i.e. what resources are scheduled (what pieces are to put together) first will increase (or decrease) the level of planning complexity (Wijngaard, 2003). In this specific case, the process to be planned concerns the resources involved in this healthcare providing process. These are the physicians, medical assistants, treatment rooms, medical equipment, etc. The treatment of patients, the way in which these patients are scheduled, the variety of routings the different patients have, the explicit way of treatment (such as number of follow-up appointments, aftercare, etc.) form the healthcare planning process. Figure 1 shows the described relations. It also depicts the performance indicators that are affected by the two different processes. The healthcare process affects patient lead- and access times, and the quality of care provided. The complexity of the planning process affects the utilization of resources and the planning effort needed to „solve the puzzle‟.

1

The exact number cannot be given with certainty, since the number of employees fluctuated throughout the period of the project. 2

The exact number cannot be given with certainty, since the number of employees fluctuated throughout the period of the project.

Af fe cts Involves Involves Complexity Resource Pplanning process Healthcare process

 Variety of routings of patients;

 Care path / routing of patients;

 Way of treating patients.

 Planning constraints;

 Way of planning (i.e. priorities,

time distribution);

 Level of detail.

 Patient lead time;

 Patient access time;

 Quality of care provided.

 Planning effort;

 Resource utilization.

Af

fe

cts

(8)

8

At the hospital the management was interested in the complexity of the planning process itself. It wanted to get more insight in the aspects that contributed to the growing complexity of the planning process. More concretely, the hospital management wanted to know what factors contribute to the complexity of the planning problem and what they could do to overcome it. From OM literature we know that planning complexity is the result of several factors: the level of detail of the created schedules (e.g. Burke, Causemaecker, Vanden Berghe & Van Landeghem, 2004) , use of (shared) capacity (skills and resources) (e.g. Hoekstra & Romme, 1992, Vissers, 1994), and composition of workforce (part-timers with specializations) (e.g. Siferd & Benton, 1992). This study has examined whether these aspects were indeed contributing to the planning complexity at the outpatient clinic dermatology of the hospital. The planning process has been observed and analyzed. Detailed examination of the process, the skills, and the use of resources has taken place, and recommendations to overcome the causes of complexity have been formulated.

What is important to mention is that the planning process in this study concerns the planning of resources, treatments, and employees and their skills (the skills form the basis of the schedule and the coverage of it (in relation with routing variety (Burke et al., 2004))). For this reason the availability of treatment rooms, medical equipment, and physicians, the availability of skills in time, patient group composition, and treatment coverage throughout a weekly routine have been elaborately examined and analyzed during this research. In addition these different types of resources have been classified (as leading, following, shared, dedicated) to be able to analyze the influence of each resource and resource type on planning complexity. The differences in these classifications are set out in paragraph 1.2 and the classifications and the analysis can be found in chapter 3 and 4. During the planning process at hand no patients are scheduled. This means that the healthcare process (see Figure 1) is not elaborately discussed and covered in this research. As can be distilled from this, the planning process at hand involves mainly the supply side of the outpatient clinic. During the planning process at hand only little attention is paid to the demand side (what treatment is needed due to higher demand in period x?). This indicates that the planners do not schedule patients (healthcare process), but create the possibility for others to schedule the patients by making the resources available in time (resource planning process).

Often planning problems have an effect on the waiting times and lead times within a process (i.e. Schönsleben, 2007). the hospital has set waiting time standards and the waiting times at the outpatient clinic were in accordance with these standards, these were the actual and up to date waiting times. the hospital‟s outpatient clinic dermatology monitors its waiting times closely. Also, after examining the data on waiting times it got clear that the waiting times are not being (negatively) affected by the planning process at hand (since they are affected by the healthcare process (see Figure 1)). The planning process at hand deals with resources and capacity; the planning process at hand does not schedule patients and the patients are therefore not directly influenced by the planning process at hand. These two facts (waiting times not being negatively affected and planning process not focused on patients) are the reasons for not covering waiting times in this research. Since the planning process at hand focuses on resources, an interesting aspect to research is the utilization of these resources. This research focuses on the causes of planning complexity, the consequences this complexity has on planning performance (mainly in terms of planning effort and resource utilization, planning performance is dealt with in more detail in section 1.2.4), and on possible interventions that lower planning complexity. The research question that has been used to conduct this research:

“What causes planning complexity and what interventions should the hospital take to reduce the effect of these causes on planning complexity?”

Before starting a project it is important to determine the relevance of solving the problems. The main problem is that the outpatient clinic dermatology is having a hard time creating an optimized schedule; it requires more and more effort to schedule. Increased planning complexity and rising planning efforts lead to an increase in costs (time and money). Finding what causes the complexity will give insight in the problem and will lead to recommendable interventions to reduce planning complexity

The research objective has been formulated as follows:

(9)

9 1.1.1 Sub questions

The main research question needs guidance, i.e. some precise research directions. These will be given by the sub questions that will be presented in this section. The first step towards recommendations for reducing the complexity will be to clarify the current way of working:

1. “How does the planning (process) of the outpatient clinic dermatology currently operate?”

This sub question should give insight in the current way of working. What is scheduled during planning (resources, treatments, etc) the amount of effort put in the planning process per planning tasks (creating a schedule, dealing with requests from employees, etc). Also an important thing to research in this phase of the research are the priorities the planners work with; what do they schedule first and what follows? The last thing that is covered by this sub question is the coverage of the schedule: what resources are required for what treatments?

Analysing the resources and the way in which they are utilized requires researching the number of resources, the degree to which these resources are shared and dedicated (differences are explained in section 1.2.3), and which of these resources need to be present at the same time for a specific treatment.

To be able to find an efficient way of using the resources, insight in the current utilization of these resources is required. For this reason, the following sub question with additional questions:

2. “How is capacity at the outpatient clinic dermatology currently used and utilized?”

Important aspects in the usage and utilization of capacity and resources are covered by the following questions: What are the skills of the medical assistants and when are these available? What patient (groups) needs what resources? Are these resources available throughout the week? Or are there specific timeframes in which these resources are available and are they time-shared (Vissers, Bertrand & De Vries, 2001)? Underlying theory about these questions can be found in paragraph 1.2.

Since the complexity makes it difficult to create a schedule in which every treatment is covered, employee contracts are met, skills and resources are optimally utilized, and all required personnel is available when needed it is important to distinguish the aspects mentioned in the questions above . It is important to get insight in the different skills of the medical assistants; what treatments are they allowed to perform and, in relation with their contract, when are these skills available? This is an important step in finding the causes of planning complexity, since these skills are needed for certain treatments. It might be the case that some treatments cannot be scheduled on certain days due to a lack of skills, or it might be that the workforce is not used to its full potential. Sub question two covers this.

Determining the main causes of planning complexity will be an important step towards (possible) interventions Which dermatologic resources are shared between patient groups and does this contribute to planning complexity? :

3. “What are the main causes of planning complexity?

Does the availability of skills contribute to planning complexity? When the main causes are determined and analysed, interventions to overcome these causes need to be given. The main causes of planning complexity will be determined by observing the planning process, examining the resources that are scheduled (mainly desk-research), and short interviews (meetings) with the people involved (mainly the outpatient clinic manager. After determining the main causes of the complex planning process, It should be researched whether the current way of working can be made more efficient in terms of planning effort.

4. “What interventions should be taken to improve the performance of the planning process?”

Wijngaard (2003) states that schedulingstructure is an important factor in reducing planning process complexity. Subquestions one to three should make clear whether there is any planning structure at the outpatient clinic and if this structure can be improved. Beforehand, restructuring the planning structure seems to be the main intervention that should be taken to improve the performance of the planning process at hand.

(10)

10

planning performance? Are specialists a different kind of capacity if compared to medical assistants? And if so, how to deal with this difference? Paragraph 1.2 discusses these kinds of questions. At the end of the paragraph a conceptual model will depict the important relations in this research.

1.2 Underlying theory

Planning complexity, it sounds interesting but what does it mean? Complexity here simply means that something is hard to deal with or understand. What is planning? What steps are involved in planning? What levels of planning can be distinguished? To get a better understanding of this, planning and scheduling will be set out. When scheduling (human) resources on the operational level, planning is the process in which schedules are created. Planning (human) resources is the backbone of this theoretical section, most of the used literature focuses on planning nurses or medical resources.

So, in this paragraph, the underlying theory of this research will be set out. This paragraph will be focused on literature dealing with planning complexity and its causes. The used literature deals with nurses in ward settings, which is not the exact situation at the outpatient clinic dermatology. Nevertheless, there are linkages with this specific case at hand.

Since planning becomes increasingly more complex at the hospital, it is interesting to highlight possible causes of planning complexity found in the past. Highlighting these aspects gives an overview of aspects that have been taken into account in this research.

Planning at the hospital‟s outpatient clinic dermatology deals with the „supply side‟ of the clinic. It schedules resources and capacity. For this reason capacity is shortly introduced: what is capacity and what effect does planning have on capacity? Hospitals work with specialists, dermatologists in this case. These specialists are involved in the vast majority of treatments, meaning that these resources are a special kind of capacity. For this reason specialist-time as a shared resource (resources that are needed and used in two or more networks (Van der Vaart & Van Donk, 2004)) is set out in a bit more detail. Ways to make effective use of this scarce capacity source are set out

Since planning is an important factor in overall performance, this paragraph also discusses performance, a discussion of what performance is, the aspects that form performance in this research are also set out: planning effort and the utilization of resources (skills (employees), machinery, and rooms).

The paragraph finishes with the conceptual model of this research, in which the connections and relations between causes, capacity, complexity, and performance will be depicted.

1.2.1 The planning process

For a hospital to be able to provide healthcare of a high standard and meet rigorous quality standards, it needs to have the right number and type of resources available when needed (Siferd & Benton, 1992). As Vanhoucke & Maenhout (2009, p.457) state “the scheduling of nursing personnel is (…) essential for the delivery of care to patients”. The planning of resources becomes increasingly important and at the same time it becomes increasingly more complex. But what is planning?

What planning is, how it is usually done for (human) resources, and what can make it so complex will be set out after the discussion of scheduling.

Scheduling – Ernst et al. (2004) see scheduling as the process of constructing timetables. According to

them optimal scheduling is scheduling in such a way that an organisation can satisfy its demand for its product or service. They also state that optimized schedules can provide enormous benefits, but that it is complex to simultaneously satisfy requirements such as flexible workplace arrangements, shift equity, staff preferences, and part-time work; exactly the problems the hospital‟s outpatient clinic dermatology has to deal with.

(11)

11

(re)allocation of nurses. This decision is made on the spot, each shift. It is the fine-tuning of the scheduling decision.

This indicates that scheduling is heavily influenced by the way in which the hospital staffs and recruits. Hospital staffing involves determining the number of personnel of the required skills in order to meet predicted requirements (Burke et al., 2004). The reason why this is an important aspect to mention is because the outpatient clinic dermatology works with a set amount of FTE with which it has to meet predicted requirements or meet the production target

In this research we follow Ernst et al. (1992) in their view on scheduling (the process of constructing work timetables for staff and resources in such a way that organisations can satisfy the demand for their product or service), since this fully covers the outpatient clinic‟s planning process.

Planning – As stated, scheduling is part of planning, which indicates that planning is a broader aspect.

Planning is done on different levels and for different time horizons. Hopp & Spearman (2008) distinguish three general time horizons: long range (strategic level), intermediate (tactical level), and short term (operational level). Shorter time horizons are characterized by more detailed planning. Long range planning is concerned with aspects such as determining whether to buy a new machine or not, and it also deals with determining the levels of production, staffing, inventory. Intermediate planning is more detailed, with a shorter planning horizon. It typically deals with rough-cut planning, which needs fine-tuning in the end. This fine-tuning comes with the short-term planning. Short-term planning is concerned with generating schedules.

Manually creating schedules is generally adapted in hospital wards and that this is a labour intensive procedure (Burke et al., 2004) indicating that planning effort is an interesting topic.

Burke et al. (2004, p.422) state that “the importance of a systematic approach to create good timetables is very high, especially in healthcare…”. Planning influences the quality of care (e.g. Burke et al., 2004). Siferd & Benton (1992) add to this that patients are in need of high skilled nursing care, which can only be attained when the right number and type of hospital staff is available when needed. Nij Smellinge‟s outpatient clinic is not a ward setting, nevertheless having the right number and type of hospital staff and medical resources available when needed is also key in this specific setting. In fact, this coverage constraint (Burke et al., 2004) might be an important cause of the planning complexity at the hospital. The problems of the hospital lie in the fact that it is hard to satisfy more than one planning requirement simultaneously. Therefore it is difficult, if not impossible, to create an optimal schedule (Revere, 2004). Tien and Kamiyama (1982, cited in Burke et al. 2004) state that the nurse scheduling problem is more complex than the travelling salesperson problem. At the same time, it has been argued (e.g. Agho, Mueller & Price, 1993) that a quality roster increases employee satisfaction. A benefit that follows from satisfied employees is for example higher productivity.

What can be concluded from the disquisition above is that scheduling is part of planning. Creating a decent schedule is important, if not essential, for a hospital to be able to deliver quality care. At the same time it turns out that it is a very complex process, since there are often several requirements to be satisfied (as described in the introduction). In addition, these schedules are not only influenced by short-term factors such as the direct availability of a resource, but planning complexity may also be caused by longer-term planning decisions concerning the amount of FTE available for an outpatient clinic.

1.2.2 Possible causes of complexity during planning

Planning complexity is not something that just popped up in the last two years. Due to diminishing governmental subsidies and healthcare budgets, hospitals have been forced to reduce costs and increase efficiency; forced to do more with less. This often led to less employees (nurses) being hired, which leads to fewer nurses available for scheduling (less capacity). There is an outcry to cut costs and increase efficiency of healthcare (e.g. Siferd & Benton, 1992; Vissers, 1994; Vissers et al., 2001; Revere, 2004; Green, 2004). The reduction and lowering of budgets and subsidies makes it necessary for hospitals to make decisions about resources that are scarcely available (Vissers, 1994). According to Green (2004) the choices that come along with this need are generally made without an OR (operations research) model-based analysis, while other service organizations do use these kinds of analysis.

(12)

12

diverse the pieces), the more difficult the planning process (puzzle). Also, the way in which the resources are scheduled may contribute to planning complexity. As with an ordinary puzzle, structure in making the puzzle makes it easier to finish the puzzle in the end.

In the past various aspects influencing planning complexity have been examined. Table 1 gives an overview of the relevant causes. The main causes represent four main classes that cause planning complexity followed by sub causes. The last column mentions the authors.

Main cause Sub causes Stated by

Size of problem

Number of nurses to schedule

Vanhoucke & Maenhout (2009)

Number of shifts/tasks to schedule

Planning horizon Burke et al. (2004), Vanhoucke & Maenhout (2009)

Level of detail

Flexibility Burke et al. (2004)

Employee preferences Ernst et al. (2004), Vanhoucke & Maenhout (2009)

Developments in healthcare3

Increasing specialization Plogh, Klazina & Casparie (1999), Broekhuis & Van der Vaart (2002)

Increasing complexity of demand

Broekhuis & Van der Vaart (2002)

Multidisciplinary healthcare

Mollema (2001)

Coverage requirements of schedule

What resource needed when and where

Skill (level) of nurses Siferd & Benton (1992), Gray, McIntire & Doller (1993), Vanhoucke & Maenhout (2009)

Table 1 Causes of planning complexity found in literature

Two causes of planning complexity mentioned in Table 1 deserve extra attention in the given context: the skills of nurses in combination with the number of shifts/tasks to schedule. At the outpatient clinic dermatology of the hospital these two aspects combined are very important. The medical assistants have certain skills and that means that they can or cannot participate in certain treatments because they lack the required skill to perform a task within that treatment. As will be discussed in more detail in chapter 3 and 4, the outpatient clinic schedules on tasks, i.e. relatively detailed.

There is one last aspect that has not been mentioned yet; planning manually. Planning manually costs a lot of effort and is usually very specific in terms of empiricism (Burke et al., 2004). This in contrast to the algorithms that have been developed in the past decades (from Albernathy, Baloff, Hershey & Wandel (1973) up to Vanhoucke & Maenhout (2009) and more recent as well). These algorithms are focused on single problems only, meaning that the algorithms are only applicable in limited settings. As Burke et al. (2004) state, hospitals around the world have very different administrative procedures and planning problems, which leads to a situation in which it is impossible to create an algorithm that is applicable for all (or at least the vast majority) of the hospitals. Moreover, mathematical approaches to the planning problem usually are „optimizing approaches‟ that often optimize one goal or criterion (Burke et al., 2004). This makes automated scheduling not ideal for the hospital‟s outpatient clinic dermatology, since there is more than one goal to be optimized (number of treatments, skills of employees met, rooms utilization, etc).

Some of the above discussed aspects (Table 1) contribute more to planning complexity than others do, but they can all be regarded as constraints. There are two types of constraints that can generally be distinguished: hard and soft constraints. Burke et al. (2004) characterize hard constraints as constraints that should be met no

3

(13)

13

matter what. Soft constraints can be defined as desirables but may need to be violated in order to generate a workable solution for the planning problem (Burke et al. 2004). Planning needs to take both hard (e.g. regulations) and soft (e.g. requests for days off) constraints into account. At the hospital‟s outpatient clinic dermatology this is no different. How these constraints contribute to planning complexity will be discussed in chapter 3 and 4.

Paragraph 1.2.3 will discuss resources (capacity) and will discuss this in four main areas: capacity as a whole, different capacity sources of interest, shared/dedicated resources, and leading/following resources. The differences between these kinds of resources will be set out, but paragraph 1.2.3 will start with capacity as a whole.

1.2.3 Resources

The planning process at hand deals with scheduling resources (capacity). This means that the utilization of these resources is controlled by this planning process as planning is closely related to capacity; planning typically determines when to use which capacity source. Schönsleben (2007) distinguishes several types of capacity, such as theoretical capacity: “Theoretical capacity is the maximum output capacity, determined by the number of shifts, the number of workers or machines, and the theoretically available capacity per shift” (Schönsleben, 2007, p.28). However, this capacity never gets fully utilized for a number of reasons. One of the important causes for this is the availability of resources (Schönsleben, 2007). Schönsleben mentions „maximum demonstrated capacity‟, which is the highest capacity ever shown when all efforts have been made to optimize resources; the schedules of these resources was optimized. This underlines the importance of a solid planning and optimizing the utilization of a resource once more. Nevertheless, managing resources in an optimal way is a complex and difficult task for healthcare providers (Revere, 2004).

So resources have a specific amount of capacity. The better the resource is scheduled and utilized the higher this capacity will be. But scheduling resources is not the same for every resource. There are different types of resources that require different approaches when scheduling them, different pieces of the puzzle need different handling. Shared and dedicated resources are good examples of resources that need different ways of handling (during planning).

Shared resources - Hoekstra & Romme (1992) in their book discuss basic structures (or base forms) of

which the most interesting one in the given context is shared resource (see

Figure 2). A shared resource is a common-capacity source that occurs in two or more different product-market combinations (Hoekstra & Romme, 1992; Van der Vaart & Van Donk, 2004). A shared resource can also be a resource that is shared between patient groups, or between disciplines, for multidisciplinary purposes for example. Shared resources especially contribute to planning complexity if the capacity of these resources is scarce (Huckman, 2009), as is specialist-time of a „shared‟ specialist (Vissers et al. (2001).

Figure 2 Schematic depiction of a shared resource

There are several kinds of shared resources. Vissers et al. (2001) discuss time-shared resources, and the before mentioned specialist-time as a resource. Within the outpatient clinic dermatology these resources are all represented. Paragraph 3.2 deals with this in detail for the hospital.

Dedicated resources - At the other end of the dimension of shared resources, are dedicated resources. A

dedicated resource is a resource that has features that make this resource applicable for one purpose only.

Now that it is clear what planning is, what aspects can possibly contribute to planning complexity, and what types of capacity are important to take special care of, it is interesting to know what the results in the end will be. Planning properly can have a significant impact on performance (e.g. Schönsleben, 2007), since a proper schedule covers all requirements (e.g. Ernst et al., 2004). Performance can be interpreted in many ways. To get a clear direction of what performance in the given context means, section 1.2.4 discusses performance.

Shared resource Group A

Group B

(14)

14 1.2.4 Performance

What adds to the complexity of the problem is the contradiction between the need to reduce cost (e.g. Siferd & Benton, 1992; Vissers, 1994; Revere, 2004; Green, 2004; Vissers et al., 2007), the wish (and demand) for dependability, high quality healthcare (e.g. Siferd & Benton, 1992; Vanhoucke & Maenhout, 2009), and the quest for increased efficiency by optimal utilization of resources (Broekhuis & Van der Vaart, 2002). Vissers et al. (2001) advocate high utilization rates of specialist-time in order to strive for „optimal‟ use of this resource. As Schönsleben (2007) states, high utilization of resources will lead to lower costs and this will also lead to bottlenecks and queues more often. On the other hand, lower utilization of resources will lead to less efficiency and higher costs. For a hospital to achieve lower costs and high(er) dependability is a very complex (planning) problem (Revere, 2004).

The evaluation of performance is a vital managerial function (Mentzer & Konrad, 1991). However, there is no consensus between academics about what performance actually is. Some say that productivity and performance are interchangeable (Mentzer & Konrad, 1991), but this is not fully correct. Performance is a term that should be used at a higher aggregation level than productivity. It is better to see productivity as one of the underlying measures of performance, as are quality, dependability, and speed (Slack & Lewis, 2008) for example. Others argue that performance can be measured by analysing both effectiveness (the degree to which a goal is achieved) and efficiency (the ratio of resources utilized against the results derived) (see for example Mentzer & Konrad, 1991). Otley (1999) states that an organization that is successful in attaining its objectives and has good performance standards is performing well. This means that he links performance to goals and objectives as well. In his work he relates performance to stakeholders and their opinion about organizations. Third parties may experience different performance levels than the organizations themselves do. This research on planning performance will use performance as a construct that represents efficiency and effectiveness by making use of the terms planning effort and resource utilization.

The measurement of performance is important for organizations to monitor and optimize their processes. Martinez, Pavlov & Bourne (2009, p.71) state that “Performance measurement has two broad roles in managing performance. The first is to clarify the objectives of the organisation and communicate them in a way that makes the strategy explicit. The second role is to measure performance against these objectives to provide feedback as to whether or not the goals are being achieved. The first role is fulfilled during the process of designing the measurement system, the second role when performance is reviewed”.

This research has split up performance into measurable indicators; planning effort (the amount of time and effort required creating a schedule) and the utilization of resources (the degree to which this is efficient and effective). The relation between the complexity of the scheduling process and the performance in terms of effort and utilization will now be discussed.

When more constraints (e.g. shared resources) come into play, the planning process becomes more complex. Highly complex processes lead to more planning effort required. Also, when more of these constraints are of influence, it is likely that the optimal way of scheduling and utilizing resources becomes more difficult.

In his study, Revere (2004) investigated the effect of a more efficient use of resources. Accompanied by three other changes in the processes of the hospital on which his research focused, both the efficiency and the effectiveness increased and the hospital increased its performance. This was partly achieved by using the current resources more efficient. The study actually increased this efficiency by creating better schedules for the workers, which indicates the importance of a good schedule once again.

1.2.4.1 Planning effort

Effort is the physical or mental activity needed to achieve something (dictionary.cambridge.org), an exertion of physical or mental power (www.dictionary.com). The more effort required to achieve a goal the harder it is to achieve that goal. For this research, which is about planning complexity, the effort of concern is planning effort. In fact, planning effort both requires physical as well as mental activity for the goal to be achieved. The physical effort can be measured by means of time, the longer a planning session takes, the more effort is required to finish that planning session. The mental effort deals with the complexity of the planning problem, the pressure to create a schedule that fulfils (all) hard and (as much as possible) soft constraints.

1.2.4.2 Utilization of resources

(15)

15

these aspects. As Schönsleben (2007, p.29) puts it “(Capacity) utilization is a measure of how intensively a resource is being used to produce a good or service. Traditionally, it is the ratio of its actual load to its theoretical capacity”. The utilization of a resource relates to the capacity of this resource. The utilization is a percentage of its theoretical capacity. Utilization rates have different effects on performance measurements, and should be set according to the goals and objectives determined by the organization. When the goal of the organization is to reduce costs the utilization rate of resources should be as high as possible (Schönsleben, 2007). This will lead to lower costs through better use of the investment costs of the resource. When dependability is a goal of the organization it should utilize its resources in a different manner. High utilization rates result in less idle capacity. This means that the resource cannot cope with significant fluctuation in demand. When dependability is the objective, the resources should be utilized with this in mind.

What also should be kept in mind according to Schönsleben (2007), is that well-utilized capacity is not only cost-efficient. It can also lead to bottlenecks. Whenever capacity is not available to work it immediately influences the flow of the organization (its performance) in a negative way. It will lead to queues/waiting lines. Since Vissers et al. (2001) stated that specialist capacity has a high probability to be a bottleneck capacity, it is important to monitor the utilization of specialists closely. Specialist capacity is not the only capacity source that is to be monitored closely to achieve an optimal schedule. All (shared resources) have significant influence on the flow of the process (Hoekstra & Romme, 1992); all (shared) resources should be monitored.

1.2.5 Conceptual model

The constructs in Figure 3 represent the body of the research as set out in this chapter. The constructs are shortly introduced and discussed below:

Figure 3 Conceptual model

 Resource limitations – these are resources located and used at the outpatient clinic dermatology and form the basis of what is scheduled. The tangible resources are divided into groups, such as leading, following, shared or dedicated. The intangible resources (skills of the medical assistants) are also to be considered an important factor that contributes to planning complexity. Both the tangible as well as the intangible resources are used for different treatment methods. Every treatment needs a set skill level and resources with specific features. Getting these to be available in the right quantity, at the right time is difficult. The more resources needed or the less resources available, the more complex the planning process becomes. As with an ordinary puzzle, the more pieces the more difficult the puzzle. Also, when some pieces of the puzzle are missing or not available, the puzzle becomes harder to solve or even impossible to complete. Examples of the resources are treatment rooms, medical equipment, and medical assistants, the skill level of employees, specialists, and more.

For this specific case it is important to know the following: if all medical assistants are able to perform everything, the planning process will not be complex (puzzle pieces fit in more than one place). If all medical assistants are able to perform only one treatment, then the planning process will not be complex either. The complexity comes into play when the resources are able to perform a combination of different treatments, when resources are shared between treatments (and patients groups), and when the resource capacity is scarce;

 Planning requirements – this construct represents the constraints planning has to deal with. When to schedule what treatments in such a way that production is achieved and all treatments are covered sufficiently with the right personnel. Also the contractual constraints are represented by this construct. As it turned out, workforce composition has significant influence on planning complexity. The more planning requirements added, the harder and more complex the planning process becomes. The flexibility and level of detail of the planning process are also represented by this construct. One can think of an employee requesting a day off in a particular week – as soon as this request is handed in it becomes a planning requirement, and a (soft) constraint.

(16)

16

 Degree of complexity of the planning process - this construct represents the degree of complexity of the planning process within the outpatient clinic dermatology. As this chapter has pointed out (section 1.2.2) there are many things that can lead to increased complexity, hard and soft constraints (see Table 1 in particular). The higher the planning complexity, so the harder it is to create a solid schedule, the more effort required to create the schedule and the more chance there is for inefficient usage of critical resources. In other words, the more complex the planning process the lower the operational planning performance;

(17)

17 2 Methodology

For guidance of the research, a set of four sub questions was formulated. These sub questions encompass all elements of the given conceptual model (Figure Figure 3, section 1.2.5). Researching and answering these sub questions eventually lead to an answer for the main research question and lead to recommendations by means of a redesign. For convenience, the main research question is repeated in this section:

“What causes planning complexity and what interventions should the hospital take to reduce the effect of these causes on planning complexity?”

To be able to answer the research question it was also important to find out what influenced the degree of complexity within the scheduling process at the hospital. As one can see from the conceptual model and read from literature, there are a number of influencing factors. It was important to get a clear view of all resources, the degree to which these were shared, dedicated, leading, and following and the contributions they had on planning complexity.

The overall method that has been used to find an answer to the research question and come up with sound recommendations (redesign) has been to first find causes of planning complexity. This was done by close observations of the current planning process. After these observations, semi-structured interviews with the planners have taken place, as well as many consultations with the manager of the outpatient clinic dermatology. To further investigate the causes, research after capacity usage and availability has taken place. It has been researched which patients could be grouped together to form patient groups, after which it was researched which patient groups made use of the same key resources (see chapter 3.2, Table 7). After analysing the current planning process and the use of capacity, recommendations have been done.

Per sub question there were specific research methods used to find answers to the questions.

First sub question - The starting point for this research was the functional problem, which in this case is

the complex planning process leading to possible inefficient use of critical resources and high planning efforts. The first step in unravelling the causes of complexity was creating an overview of the current way of working.

1. “How does the planning (process) of the outpatient clinic dermatology currently operate?”

To get a clear picture of how the current planning process flows, the research methodology for this sub question consisted out of semi-structured interviews with the people involved in the planning process. Also, the manager of the outpatient clinic dermatology was consulted several times by means of semi-structured interviews. The most valuable and important information was gathered during observations of the planning process. Observing the process gave insight in how the planners divide their time, what goals (and objectives) they strive for, which processes within the planning process take up most of their time, how requests are dealt with, and how resources are allocated. And lastly, desk-research took place to get an overview of the protocols present, the number of treatments and the resources each treatment requires, and the way in which this has been scheduled in the past.

How much time do planners spent on their core business? From observation this data was retrieved. Analysis on whether this is the correct way to plan, also gave interesting outcomes.

Second sub question - Since planning is about resources (supply side), it is important to know how these

resources are used. As has been set out in paragraph 1.2.3, shared resources contribute to planning complexity. What is important, is to find out how resources are used (utilized) and shared within the outpatient clinic dermatology. The following sub question gave insight into this:

2. “How is capacity at the outpatient clinic dermatology currently used and utilized?”

This question has been dealt with in two chapters; data has been gathered during the observation of the process (chapter 3) and during analysis (chapter 4).

(18)

18

Desk-research has taken place, quite some time was spent on figuring out what influenced what and how. In the end it turned out sufficient to find out which treatments require which resources. Observations and measurements gave insight in the current utilization rate of the resources.

Analysis of the data gathered mainly consisted out of desk research. What are the patient groups, what are the chances for a new patient to end up in either of the nine patient groups? What resources are combined with what treatments? Analysis took place solely by working with the data by hand. Excel helped a lot to create overview.

Third sub question – By answering the first two sub questions, it should become clear what the causes of

the planning process complexity are. This third sub question sums up the causes and determines the main causes.

3. “What are the main causes of planning complexity?

Research method for this sub question was desk-research. Data gathered with the first two sub questions gave insight in the utilization of resources. Analysis of these data was the starting point for this sub question. After analyzing the data, the outcomes gave incentives to change the current process. These where the input for sub question four, the recommended interventions (to be found in chapter 5).

Fourth sub question - When the main causes are determined and analysed, interventions to overcome

these causes need to be given. These will be presented as recommendations at the end of the paper. What should be kept in mind is the fact that creating or developing a new/better way of planning is the last step in a sequence of three. The first step is to simplify current processes by integrating steps, or eliminating obsolete ones. Second a reduction in complexity can be realized by reducing interdependencies, separate process flows where possible. And the third and last step will be an adequate planning system. These steps return in sub question four

4. “What interventions should be taken to improve the performance of the planning process?”

. Sub question three covers the analysis of the current way of working – finding out what are the main contributing factors in terms of planning complexity. A large part of the time spent on this step was used to analyse the medical assistants and their availabilities and skills.

Research strategy - It was important to have a clear „starting point‟ of the research. Since the goal of this

research was to try and solve a problem, it can be considered a problem solving research. In this light, the starting point of the research was a functional problem. In this case, the functional problem was represented by the increasing planning effort (and thus suffering operational planning performance/utilization of resources). How come planners increasingly need more time to create a schedule and roster the resources?

To find an answer to that question, it was necessary to research a total of seven constructs during this project, the vast majority of them relating back to the underlying theory discussed in chapter 1: patient groups, shared resources, skill level of the workforce, utilization of resources, availability of resources (in time), level of detail of planning process, and the planning process itself.

Planning complexity Shared resources Skills of workforce Utilization of resources Planning process Availability of capacity Patient groups Level of detail

(19)

19 As

Figure 4 shows planning complexity is/can be affected by seven aspects. These seven aspects were all researched during this project. A description on how these aspects were researched will be given now.

Shared resources – Section 1.2.3 has described shared resources as common-capacity sources that occur in

two or more different product-market combinations (Hoekstra & Romme, 1992). To find the shared resources within the outpatient clinic dermatology it was key to find out what patients were receiving what type of treatment(s) or operations. Operations differ in resources they need and patients (or patient groups) differ in operations they need. The care profile or routing of patients has been researched, the belonging operations per routing have been set out, and the required resources per operation were linked to the routings of patients. With this information it was possible to find leading and following resources (Vissers, 1994). The leading resources were set out against usage per patient group and this lead to finding the shared and dedicated resources.

Skills of workforce – As Ernst et al. (1992) stated, nurses often times are not allowed to perform all

treatment types the outpatient clinic or ward offers. This is often limited by the diplomas of the nurses/medical assistants. Planning is influenced by this because planners have to take into account that nurses are not freely applicable resources (Newman et al., 1991). To find the skills of the workforce in this particular research was not that difficult, since the planners themselves had researched this just before the start of this project. For researchers that do not have this piece of information available it is important to distinguish high aggregated skills. If a medical assistant is allowed to inject patients with some medicine and is also authorized to see what effects the injection has had, try to find a „skill‟ that covers both operations. When the skills are then available, find out what employees are allowed to perform what kind of treatment. Appendix I shows the diagram that was used for this research.

Utilization of resources – When resources such as treatment rooms and medical equipment are influencing

the planning complexity due to (perceived) under capacity it is useful to research the utilization of these resources. For this to be successful it is important that the hospital or organization monitors this utilization closely. Find the theoretical capacity (Schönsleben, 2007) of resources. This may be difficult in a hospital setting due to human beings participating in the production process (e.g. Fitzsimmons & Fitzsimmons, 2011). One patient may take ten minutes for an ultrasound while another takes five or fifteen. To overcome this situation, this research made use of managerial and specialist input. Specialists know exactly how long a resource is utilized during an operation.

Finding the utilization of resources is more difficult. For this research it was not possible to get a rock solid utilization because the utilization was simply not monitored at the outpatient clinic. What has been used was the amount of treatments that require a specific (leading) resource per period. These treatments are scheduled according to the time of the consultation. This does not necessarily mean that the resource is used during the total time of the consultation. Monitoring and observing the resources would have given a better result. What should be observed or monitored is the point in time a resource is „started‟ and „ended‟.

Planning process – Planning complexity may also be caused by the planning process itself. Planners can

spend time on aspects of planning that are not that relevant when creating a schedule or the planning process may be interrupted a lot what leads to loss of concentration. Observations are the best way to research the planning process and that is also the method that has been used for this research. During observations it is important to focus on the process of planning: where do the planners struggle with, what part of the process takes most of their planning time, and can the planners work in a quiet surrounding without interruptions.

Availability of capacity – The particular outpatient clinic dermatology this research focused on employs

merely part-time medical assistants. This factor has significant influence on the availability of capacity, especially in combination with the skills of the workforce. Finding out when capacity is available is important because it gives insight in when what types of treatment can be scheduled. During this project it got clear that long term strategic planning (production) was the leading factor while planning. When production is leading it is required to schedule a set amount of treatments per period. If the capacity required for these particular treatments is not available is gets complex to schedule them.

(20)

20

Patient groups – It was important to distinguish the above mentioned patient groups and at the same time

it was difficult. It turned out that the outpatient clinic dermatology works with 29 possible diagnoses and 149 routings. Finding a higher aggregation level was key to work with these data. This level were the patient groups. In consultation with specialists and managers patient groups were created. This aggregation was based on the routings. Diagnoses with comparable routings were grouped. This led to nine patient groups. Creating these groups was not considered easy. Given the 149 routings it took quite some time to find a way to work with them.

Level of detail – The last of the seven variables that can have influence on planning complexity can be the

level of detail of planning. Researching this was done by observations of the planning process itself and researching the schedules and rosters of the past. This showed that planning was very detailed, especially for the administrative tasks within the outpatient clinic.

(21)

21 3 Current situation

This chapter sets out the current planning situation of the outpatient clinic dermatology, it describes how the outpatient clinic‟s planners currently „solve the puzzle‟. Next to this, this chapter also describes the resources of the outpatient clinic (specialists, assistants, treatment rooms), it describes the different pieces of the puzzle. Lastly it shortly discusses the patients of the outpatient clinic. The patients and their routings belong to the healthcare process (see Figure 1, page 7). As stated, this side of the complexity causing aspects is not elaborately discussed in this research.

Data for this disquisition has been gathered by doing observations and desk-research. During these observations the current planning process has been observed; where do the planners struggle with? What planning steps are taken and which of these take up most of their time? What priorities do the planners have to take into account? How do they deal with the different resource types? Do the seven aspects mentioned in chapter 2 indeed contribute to planning complexity?

To get some additional data semi-structured interviews have taken place. During these interviews, which can be found in Appendix VIII, the opinion of the planners has been slightly highlighted. As stated, these interviews were conducted for additional data.

This chapter covers the first sub question:

“How does the planning (process) of the outpatient clinic dermatology currently operate?” And partly the second sub question:

“How is capacity at the outpatient clinic dermatology currently used and utilized?

Due to the before mentioned intervention of the manager (see chapter 2), data gathered in later observations has been interpreted slightly different. Changes in planning that were pointed out by the manager have been „eliminated‟ from observation, meaning that the minor improvements due to the intervention are not taken into account.

The set up of the chapter is as follows. First the planning process itself will be discussed; the way planning is done (times per week, number of hours per session, the priorities the planners have, the basic setup of the planning). A schematic overview of the planning priorities will be presented in Figure 5, after which the time distribution will be discussed in a more detailed disquisition.

After this disquisition of the planning priorities and time distribution between planning-tasks, the use of resources will be set out. There are three main groups of resources determined: (1) specialists, (2) medical assistants, and (3) treatment rooms and their inventory. Patient groups are also discussed in this section of the chapter. These four groups form the basis of the second part of this chapter. How are the resources currently scheduled? What does this mean for treatment coverage throughout a weekly routine ? What pieces of medical equipment are present at the outpatient clinic and how are these distributed between the different types of rooms? Are these resources leading or following? The latter part of this section will then set out the utilization of the most important resources (leading resources). And last but not least the dermatologic patient are grouped into nine different patient groups.

3.1 The planning process

At the outpatient clinic dermatology of the hospital planning is done manually. It is done twice a week, on Tuesdays (four hour session) and Fridays (three hour session). This means that the outpatient clinic spends approximately fourteen hours a week on creating schedules. Financially this does not seem like a problem (costs: approximately €35.000,- on a yearly basis). However, the increasing amount of effort the outpatient clinic puts into planning and the increasing complexity of the planning process make this research worthwhile.

(22)

22

schedule based on the available resources and skills. This is most often done on Tuesdays. This treatment schedule is uploaded to an intranet, and patient are scheduled into this schedule (at the outpatient clinic desk).

After each Friday session, the schedule for the week to come is printed out on paper. This printed version is then translated onto a whiteboard each day. This whiteboard represents the planning for the day to come. In the past it has been the case that schedules and rosters were not appreciated and/or accepted by the employees of the outpatient clinic dermatology. This has led to employees shifting their name from working on a certain job to another job (on the whiteboard). Management has taken action to prevent this from happening. However, it is still the case that employees try to adjust the roster to their wishes. This frustrates the planners, because they have the feeling that their effort is not appreciated by their colleagues.

But what priorities do the planners have? Do they schedule with a certain structure? Do they have other priorities during planning, such as planning treatment x more often? Or do they schedule some resources with priorities? The next section will deal with these questions.

3.1.1 Priorities and time distribution

As set out in section 1.2.1 there are different views on where planning takes place, on the strategic level (Warner, 1976), on the tactical level, or at the operational level (Ernst et al., 2004). At the hospital‟s outpatient clinic planning takes place on the operational level. Nevertheless the strategic level is highly influential. On strategic level the management determines the production for each discipline. Hopp & Spearman (2008) call this a long term decision, Warner (1976) also refers to this. He states that on the long term, decisions are made for the amount of FTE per discipline, for example. Both these decisions, the production to produce and the amount of FTE available at a specific discipline, influence the planning process at the outpatient clinic dermatology. It sets the amount of treatments and patients to be scheduled and it sets the available capacity.

For the outpatient clinic dermatology the set production target is the leading factor for planning. Top management knows what the outpatient clinic dermatology is capable of producing and what can be expected from it. Based on this, top management sets the production for dermatology; process a certain amount of patients of each patient group each period.

However, in practice it turns out that the most important factor is the availability of resources and their capacity. When the required assistants cannot be scheduled for a certain consultation hour or when specialists are not available, the planners simply have to decide not to schedule a consultation hour (and thus treatments) for that particular day. Even though theoretically the set production is the leading factor when it comes to planning, in practise availability overrules production. This is where the amount of available FTE per discipline has its influence. Keep in mind that meeting production is what the outpatient clinic is judged on in the end, meaning that reaching the target production is important.

Figure 5 depicts the above described relations, it shows the way in which the planners „solve the puzzle‟ in such a way that it shows the current priorities (what to schedule first). The dotted lines represent the indirect influence, though leading, of the production as it is depicted above the other three boxes. This represents production being the leading factor in the planning process. The bold letters of „available capacity‟ represent the inevitable impact of the availability of capacity.

Reducing complexity in a complex situation: How to work with shared resources?