Lisa van Dijk
S1985280
Lisa.vdijk@hotmail.com
Supply Chain Management
Supervisor/University:
Prof. dr. J. de Vries
Co-assessor/University:
Dr. N. van Foreest
Supervisor/University Medical Center Groningen:
Drs. J. Scholma
Dr. D. Allersma
How to approach supply chain
problems in a hospital setting?
Table of content
Abstract ... 2
1. Introduction ... 3
2. Theoretical framework ... 4
2.1 What is supply chain management? ... 4
2.2 The supply chain control approach ... 5
2.3 Applying the supply chain control approach ... 7
2.4 Testing the usefulness of the approach ... 8
3. Method ... 9
3.1 Introduction of the focal hospital ... 10
3.2 Data collection and analyses ... 10
3.3 Limitations... 13
4. Results ... 14
4.1 Step 1: Creating overview of the supply chain ... 14
4.2 Step 2: Diagnosing the performance gap ... 16
4.3 Step 3: Diagnosing causes of the performance gap ... 17
4.4 Step 4: Suggesting improvements ... 22
5. Evaluation of the usability of the approach ... 24
6. Conclusions ... 26
References ... 28
Appendix A: Measurement form ... 31
Appendix B: Coding tree ... 32
2
Abstract
3
1. Introduction
In the past few years, the healthcare sector has faced some major changes. This has resulted in an increasing pressure to deliver patient-oriented care in a more effective and efficient way. Therefore, several healthcare organisations have started projects, such as the implementation of electronic patient record systems, care pathways, and patient logistics. In addition, there is an increasing awareness that accurate supply chain management can increase the efficiency and effectiveness of healthcare systems. Several studies have confirmed that creating efficient supply chains can significantly reduce costs and improve the performance of the organisation (Dacosta-Claro, 2002; De Vries & Huijsman, 2011; Drupsteen, Van der Vaart, & Van Donk, 2013).
Supply chain management is comprehensive and consists of various aspects. Supply chain performance is, for example, influenced by material management, but it is also impacted by factors such as managerial practices and the organisational structure (Shortell, Zimmerman, Rousseau, Gillies, Wagner, Draper, Knaus, & Duffy, 1994; Kaplan, Brady, Dritz, Hooper, Linam, Froehle, & Margolis, 2010). Maintaining a multidisciplinary perspective during supply chain analysis seems necessary to create a thorough understanding of how supply chains function and to optimize supply chain systems (Harland, 1996; Croom, Romano, & Giannakis, 2000).
Maintaining a multidisciplinary perspective seems especially important in the healthcare sector. Many authors have described the unique aspects of the healthcare sector that have influenced hospitals’ processes (De Vries & Huijsman, 2011; Boonstra, Versluis, & Vos, 2014; Van Hoeve & De Vries, 2014). For example, there is a delicate power balance between different stakeholder groups within hospitals and there is a variety of stakeholders that try to influence decision-making processes (De Vries, Betrand, & Vissers, 1999; De Vries, 2011; Lega & DePietro, Drupsteen et al., 2013). Furthermore, hospitals are often characterized by their functional department design, which is built around discipline-based specializations (Lega & DePietro, 2005; Vissers, Bertrand, & De Vries, 2001). Communication and integration between departments is rare; each department mostly functions as an independent unit (Drupsteen et al., 2013). These features probably influence how the supply chain works, and they must be taken into account during supply chain improvement projects. However, not many approaches are available to support the systematic analyses of supply chains in healthcare setting by incorporating a multidisciplinary perspective.
4 of the underlying mechanisms of a supply chain system. Additionally, it supports the formulation of improvement suggestions. Although De Vries’s approach has proven to be helpful within the industrial sector, the usefulness of his model in the healthcare sector is still unclear.
The aim of this study is to assess the usefulness of De Vries’s approach in a hospital setting. Outcomes will be discussed, and recommendations will be given to improve the usefulness of this approach in the healthcare sector. This paper will contribute to developing an approach that supports the systematic analysis of supply chain systems in healthcare. Using a systematic approach can increase the effectiveness of supply chain improvement projects. In addition to this theoretical contribution, this paper will increase the understanding of how supply chains within hospitals work and of how different elements of the approach relate to one another within hospitals.
A combination of a case study and action research has been used to carry out this study. The approach of De Vries (2007) was applied to assess and improve the supply chain of a cytostatic drug chain in a Dutch university hospital. Cytostatic drugs are highly customized drugs; different chemicals are mixed to optimize each specific patient’s treatment. Each drug can only be used for the patient for whom it is made; it is a make-to-order system. Multiple departments and stakeholders are involved in this chain. Therefore this case represents the complex features of the healthcare setting and is suitable for this study.
The structure of the paper is as follows: the next section discusses the theoretical background of the approach. Next, the applied methodology and the case study researched will be presented. These will be followed by the results of the study. The last section evaluates the usefulness of the approach and presents a conclusion.
2. Theoretical framework
2.1 What is supply chain management?
5 preferences (Skjott-Larsen et al., 2007). Supply chain management is a multidisciplinary concept. It manages logistics, planning, control, and material issues, and it also contains organisational issues (Croom et al., 2000; De Vries & Huijsman, 2011).
2.2 The supply chain control approach
Because of the multidisciplinary characteristics of supply chain systems, it is important to incorporate different viewpoints in managing the supply chain (De Vries & Huijsman, 2011). De Vries (2007) developed an approach for optimizing supply chain systems by incorporating various aspects related to supply chain management (Figure 2.1). The main aim of the approach is to create an alignment between performance objectives and the arrangement of the supply chain system. Applying the approach will contribute to a more comprehensive understanding of how the supply chain system works and will support efforts to accurately redesign the supply chain (Zomerdijk & De Vries, 2003). The underlying philosophy of this approach is supported by contingency theory. Contingency theory describes that the best way to manage or organize processes will depend on the different antecedents within the environment (Lawrence & Lorsch, 1967; Grotsch, Blome, & Schleper, 2013). These antecedents have to be aligned to achieve optimal performance (Lawrence & Lorsch, 1967; Grotsch et al., 2013).
Figure 2.1 – Supply chain control approach (De Vries, 2007)
6 Johnston (2010) state that performance objectives, both internal and external, are defined in terms of quality, speed, dependability, flexibility, and cost (De Vries, 2007).
7 relates to the formal structure of the hospital. Structure of positions involves, for example, power structures. Power is associated with the capacity to exert will over others in order to realize certain intended benefits (De Vries, 2009). Powerful stakeholders can decide what decisions will be made. However, the structure of positions also relates to responsibilities. It determines which stakeholder is responsible for the dysfunction of a certain process and which stakeholder has to take action to change the situation. Another example is the existence of various informal networks that also influence decisions. Because of the amount of stakeholders involved in healthcare (De Vries & Huijsman, 2011), it is important to thoroughly assess organisational embedding in this case.
The elements and sub-elements involved in De Vries’s approach, as based on the above definitions, are given in Figure 2.2. This list of sub-elements is not exhaustive, but it is helpful in ensuring a complete overview of all the aspects affecting the supply chain system.
Elements and sub-elements
Performance objectives (internal and external)
Quality Cost Speed Flexibility Dependability Physical infrastructure Process steps Throughput times Transportation methods Planning structure Order quantities Order intervals
Planning system (planning rules) Information architecture Information systems Quantity of information Quality of information Timeliness of information Organisational embedding Formal structure Informal structure Responsibilities Political process
Figure 2.2 – Overview elements and sub-elements involved
2.3 Applying the supply chain control approach
8 within the organizations to achieve the company’s objectives. Simultaneously, performance gaps become clear. The next step contains in-depth analyses of the physical infrastructure, planning system, information architecture, and organisational embedding. Because they are closely interrelated, these elements have to be analysed separately as well as together to determine how they relate to each other and how they influence the supply chain system. The performance gaps can be used as input for this in-depth analysis. The aim of this phase is to identify the underlying mechanisms behind the system’s current performance to design an accurate action plan to improve the system. Finally, the fourth phase consists of implementing the suggested improvements. This study will also use this approach.
Figure 2.3 – Applying the supply chain control approach (De Vries, 2007) 2.4 Testing the usefulness of the approach
9
Sub-elements usability
Learnability
The approach should be easy to learn and understand so the user can rapidly start working
Memorability
The approach should be easy to remember so the user can return to it after a period of not having used it and not have to relearn everything.
Efficiency
The approach should be efficient to use so productivity is high during the application of the approach.
Errors
The approach should have a low error rate so users make few errors in using it. Satisfaction
The approach should be pleasant to use so users are subjectively satisfied when using it.
Figure 2.4 –Definitions sub-elements usability (Nielsen, 1993: 26)
Compared to the industrial sector, the hospital sector has some unique features which can influence the usefulness of De Vries’s approach. For example, the physical infrastructure of hospitals is more influenced by complex and variable patient demands than that of the industrial sector (De Vries & Huijsman, 2011). As a result, hospitals have many integrated process structures that can be difficult to understand (Van Hoeve & De Vries, 2014; Boonstra et al., 2014). In addition, hospitals have a unique workforce that includes medical professionals who often possess high levels of expertise, power, and autonomy (Boonstra et al., 2014). Furthermore, supply chains in hospitals are not organized around expenses because quality is a more important performance indicator (Jack & Powers, 2004; De Vries, 2011). Thereby, a distinction can be made between physical material flows and patient flows within the healthcare sector. These flows are highly integrated with each other. Even though this research focuses on a material flow, the cytostatic drug chain, the question arises if it is necessary to add patient flow as additional element to create a more comprehensive understanding of how the supply chain works.
3. Method
10 Therefore, this study is a combination of an explorative case study and action research. Because of time limits and the scope of the research, a single case study was performed. Due to the short time span, a more thorough understanding of a single case was preferred over a multiple-case study. The unit of analysis of this study is the cytostatic drug chain within the focal hospital.
3.1 Introduction of the focal hospital
The case study was performed in a Dutch university medical centre. Within this medical centre, the cytostatic drug flow chain was analysed. This medical centre has the largest oncology centre in the northern part of the Netherlands, and it treats the most complex forms of cancer. In 2014, around 25,000 cytostatic drug infusions were delivered to patients. The cytostatic drug chain involves every process step from the moment the physician prescribes the drugs until the moment the drugs are provided to the patient at the internal day care centre. This chain was chosen because the medical centre indicated they had problems delivering cytostatic drugs to patients on time. Not delivering cytostatic drugs on time can increase patient waiting periods and patient dissatisfaction. The hospital would like to improve this situation to improve patient service. Therefore, the hospital was very willing to cooperate in this study. In addition the chain contains various department and stakeholders. Further, the material flow is highly related to a patient flow. As a result, this case represents the complex features of the healthcare setting and is suitable for this study.
3.2 Data collection and analyses
De Vries’s approach (2007), as explained in the previous section (Figure 2.3), was used to guide this study. The first step in this research was to create an overview of the cytostatic drug chain and to introduce the research to the different stakeholders involved. In the second step, the performance objectives of the chain were identified and current performance based on these objectives was measured. This analysis revealed a gap between achieved performance and desired performance. The third step contained an in-depth analysis of the current arrangement of the supply chain system. During the fourth and final step, suggestions were made on improving the supply chain system. Different information sources were used during this research, including quantitative measurements, interviews, observations, and informal conversations to increase data triangulation (Karsson, 2009). It assured that the cause and effects identified were supportable and were not side effects of other relationships and it improved the internal validity of this research (Karlsson, 2009). In addition, during and after the analyses, formal meetings with various stakeholders were held to review the analyses and to confirm the described relationships. This increased the construct validity of this study (Yin, 1994). The methodology used is summarized in Figure 3.1. These steps will be explained in greater detail in the following sub sessions.
Step 1: Introduce the study and create an overview of the cytostatic drug chain
11 was given to the heads of all the involved departments: outpatient clinic, pharmacy, and the internal day care centre. Next, process descriptions were analysed to get an overview of the different steps of the cytostatic drug chain. Additional information was obtained from informal conversations with both department heads and other staff members. Also, some short observations were done to increase understanding of the cytostatic drug chain. During this first step, the influence of patient flow on the supply chain performance became clear. Information gathered showed the relationship between patient flow and decisions related to the arrangement of the supply chain. Since the aim of the approach was to get a comprehensive view of the supply chain, it was necessary to add patient flow as fifth element that influences the arrangement of the supply chain system.
Step 2: Performance analyses
The second step was compromised of two phases:
1. Determining the performance indicators of the cytostatic drug chain
2. Determining the current performance of the system based on the performance indicators defined in step 1
The performance indicators were derived from short interviews with the department heads of the outpatient clinic, internal day care centre, and pharmacy. Department heads were asked to define the performance objectives of the chain in terms of quality, cost, speed, dependability, and flexibility, as presented in Figure 2.3. As a result, some important milestones became clear.
The next step was to determine performance based on the defined milestones. Quantitative measurements were made with the help of the day care centre. A meeting of fifteen minutes took place in which the researcher explained the measurements to the nurses. The nurses were then able to ask questions about the measurements to ensure they understood what they had to do. Based on this session, a measurement form was developed. It is enclosed in Appendix A. The measurements were taken over the course of eight days (two Mondays, two Wednesdays, two Thursdays, and two Fridays) to increase the reliability of the results. Based on these measurements, a quantitative analysis was done in Excel. The main aim of this quantitative analysis was to determine the performance of the supply chain through the use of Pivot Tables and time series analyses. These analyses indicated some patterns and relationships that could be used as input for further research.
Step 3: In-depth analysis
12 influencing the supply chain. In addition, fourteen semi-structured interviews were held with different stakeholders from different departments within low, middle, and high managerial levels of the organisation. The main aim of selecting interviewees was to enrich the data by interviewing as many stakeholders from different levels and departments as possible. For each semi-interview, an interview protocol was used. The protocols were developed based the elements described in Figure 2.3, and they aimed to ensure the completeness of the interviews. Additionally, these interview protocols enabled the comparability of answers within the case and improved the reliability of the study (Yin, 1994). Further, outcomes of the observations and measurements were mirrored during these semi-structured interviews.The advantage of semi-structured interviews is that they provide the flexibility to ask for more details and explanation if necessary (Forza, 2002). This allowed for a deeper understanding of how the arrangement of the supply chain worked, which fits the exploratory character of this research. Notes from the interviews and observations were directly written down and were elaborated upon as soon as possible after the observations, which increased the reliability of this study (Karlsson, 2009). If additional questions or ambiguities emerged during this step, interviewees were directly asked for clarification or confirmation.
13 Step 4: Suggesting improvements
During the fourth step three other hospitals of different sizes were visited to compare the chain of the focal hospital and to benchmark performance. Observations were done, and some informal questions were asked. Together with the extensive analyses of step 3, improvements were suggested.
3.3 Limitations
14
Figure 3.1 – Overview research method
4. Results
In this chapter, first an overview of the supply chain will be given. Next, the performance objectives of the cytostatic drug chain will be defined and the current performance based on these objectives will be described. Then, an overview of the in-depth analysis of the current arrangement of the supply chain system will be given. In the last section suggestions are made on improving the supply chain system.
4.1 Step 1: Creating overview of the supply chain
In the future, more people will be diagnosed with cancer, but more people will also survive. Medical developments will have a major impact on the facilities, stakeholders, and resources involved. Capacity has to be enhanced by improving the logistic processes and efficiency within the chains. In this study, therefore, the cytostatic drug chain of the University Medical Center Groningen (UMCG) was analysed. The UMCG plays an important role in oncological care in the northern part of The Netherlands. Around 3.4 million inhabitants live in this region, which is 20 percent of the total population of The Netherlands. In addition to regular oncological care, the UMCG functions as the patients’ ‘last resort’ and is a leader in the research of applying new methods to treat cancer.
15 cannot be used for another patient. Currently, around twenty-five infusions are delivered to the internal day care centre each day. The shelf life of cytostatic drugs varies greatly between several hours to several days. Three main departments are involved in the cytostatic drug chain analysed in this case: the outpatient clinic, the internal day care centre, and the pharmaceutical department. These departments are parts of other chains as well. The physician at the outpatient clinic and the day care centre treat various other patient groups in addition to cytostatic drug patients. The pharmacy makes cytostatic drugs for other departments as well (overall about 120 infusions each day). In addition, as result of the academic character of the hospital, a distinction can be made between regular treatments and studies.
Figure 4.1 represents the flow diagram of the cytostatic drug chain. Within the chain, three sub-chains can be identified: flows 1, 2, and 3. In the case of flow 1, all activities take place within one day, the day the treatment is provided. The patient starts with blood sampling at the lab. Then the patient has an appointment with the physician in which the result of the blood tests and current progress are discussed. If the blood test results are sufficient, a cytostatic drug request will be sent to the pharmaceutical department, and the pharmaceutical department will start preparing the drug. When the drug is ready, the treatment will be provided to the patient at the day care centre. In the case of flow 2, blood sampling is scheduled a day before the treatment is scheduled. This way the physician can assess the results and, if they are sufficient, can send the request to the pharmaceutical department the day before the treatment. The patient’s appointment with the physician is still the day that treatment will be provided. The benefit of earlier blood sampling is that it allows the pharmaceutical department more time to prepare the cytostatic drugs. In case of flow 3, no blood sampling or outpatient clinic visit are necessary. Requests are received, registered, and controlled in advance. Patients directly receive their treatment at the day care centre. Cytostatic drugs can be prepared in advance if their shelf life is sufficient. However, the pharmaceutical department never starts producing the cytostatic drug until a pharmacist has verified the request. Patients can follow flows 1, 2, or 3 alternately, depending on various circumstances.
16 The day care centre is responsible for scheduling patients at both outpatient clinic and day care centre. In case of flow 1, blood sampling is scheduled half an hour before the appointment at the outpatient clinic. The appointment at the outpatient clinic is scheduled 1.5 hours prior to the appointment at the day care centre. In case of flow 2, the appointment at the outpatient clinic is scheduled half an hour prior to the appointment at the day care centre. Process times within the day care centre vary per treatment. Fifteen minutes are scheduled per consult at the outpatient clinic.
Various information systems are evident within the chain. Poli+ is a mini, electronic record system. Different types of information are captured within this system, including contact information, lab results, treatment schemes, and additional patient information written by physicians. Cato is the order management system of the pharmaceutical department. Protocols and treatment schemes are recorded in this system. Cairo, the precursor of the Cato system, is still used to enable other departments to check the progress of their orders at the pharmaceutical department. Furthermore, X-care is the planning system used at the day care centre. This list is not exhaustive, but it provides the most important information systems. Requests are handwritten by the physician on standardised formats and send by fax.
Furthermore, the outpatient clinic can be separated into four sub-clinics: IAOP, IHMP, ILOP, and IREP. The physician prescribes drugs through a request form. However, the administration of the outpatient clinic is responsible for sending the requests to the day care centre and pharmacy. Within the pharmaceutical department, there is a distinction between pharmaceutical technicians and pharmacists. The pharmacists are responsible for verifying that each cytostatic drug will be made correctly for the right patient in the right department. Within the described chain, in flows 1 and 2, pharmacists execute their control after the administration of the request in Cato. In addition, the pharmacist is the contact person for pharmaceutical questions. Pharmaceutical technicians carry out all other steps within the pharmaceutical department.
4.2 Step 2: Diagnosing the performance gap
17
Total number infusions delivered
Number of infusions delivered before the time of the
patient’s appointment at the day care centre
Flow 1 126 48
Flow 2 11 6
Flow 3 106 98
Studies 18 3
Total 243 152
Table 4.1 – Performance overview
Additionally, in case of flow 1, the outpatient clinic must send the request to the pharmacy as soon as possible after the appointment. Because the pharmaceutical department cannot start producing the drug until the request is received and controlled, the time available to prepare the cytostatic drug is impacted by when the request is sent. Even though ‘as soon as possible’ is not measurable, the gap between time needed to prepare cytostatic drugs and the time available to prepare cytostatic drugs became clear during this diagnosing phase. On average, the pharmacy had 59 minutes available to prepare the cytostatic drug. However, they have an actual average throughput time of 1:10 hour. There is a gap of eleven minutes between those average times, which means requests should be sent earlier to ensure on-time delivery.
4.3 Step 3: Diagnosing causes of the performance gap
The most important causes and effects are visualized in Figure 4.2. In the following section, these relationships are explained more in detail.
Patient flow
18 are involved, which makes it important to spend more time with them. This increases the variability of process times and makes it difficult to tightly stick to the scheduled agenda without allowing for large buffer times, which lower the utilization rate.
Physical infrastructure
A significant cause of the bad performance of flow 1 is delays in requests being sent to the pharmacy by the physician. One of the most important underlying problems is that the physician often runs out of time during the consultation hour. As explained, patient flow is one of the causes of this problem. It is further influenced by the time required to receive the results of patients’ blood tests. Sometimes the results are not available on time. This is also partly due to patient flow, but it also occurs because determining the results take longer than scheduled. Furthermore, an important determinant that influences the time needed to produce the drugs is the pharmacists’ availability during a day. When the pharmacist is not available, orders grow into a stack because pharmaceutical technicians are not allowed to start filling an order before the pharmacist has authorized it. When this occurs, the workload later on in the process is also influenced. For example, assistants often wait for work first, but then work overload occur when pharmacists authorizes all the requests in a short time period. This results in the late delivery of cytostatic drugs. As a pharmaceutical technician indicated, ‘Sometimes we cannot do anything, because the pharmacist is not there, and then we have not enough capacity to handle everything’. Observations made clear that this is especially the case during mornings. Pharmacists start their working day later than pharmaceutical technician start producing. In addition, reasonable delays arise within the registering step of the pharmaceutical department. Some technicians are quite skilled, while others have to frequently ask colleagues for help and often made mistakes. This influences the time needed to prepare the drugs.
Planning structure
19 appointments at the outpatient clinic are not optimally scheduled for physicians. This is another reason why it was difficult for physicians to stick to their schedules.
Information architecture
The current information architecture also seems to have a large influence on performance. The lack of information sharing, the low level of digitalization, and the low quality of information created several difficulties, as visualized in Figure 4.2. The lack of digitalization increases the amount of paperwork within the chain. When combined with the amount of information transfers, the probability increases that something will get lost or go wrong within the chain. For example, physicians compose prescription requests, which the administration faxes to the pharmaceutical department. The technician registers the request in Cato and Cairo. Lastly, the pharmacist controls if the request in Cato is correct. The more steps, the higher the probability that erroneous transmissions will occur, and the higher the probability that information is not delivered on time to the right stakeholder. In addition, using paperwork increases the probability of an item being incomplete or illegible, which lowers the quality of information. When this occurs, the pharmacy technician must call the physician, who is frequently unavailable, for a new, complete, and legible prescription to be able to correctly enter the request into Cato. These extra steps take a considerable amount of time. This is especially true in the case in studies where the digitalization of processes is the least advanced. Furthermore, the low quality of information about the progress of producing cytostatic drug orders within Cato also negatively affects the performance of the system. Because of the low quality of information, departments are forced to call one another to ask for additional information or to correct errors, which disturbs the process and increases delays. This situation especially occurs when the cytostatic drug is delivered late and patients are already waiting. This is further influenced by the lack of information sharing. Departments do not regularly share information about changes in daily processes. This increases the ambiguity about the progress of processes and results in more frequent phone calls asking for additional information and clarity. These issues disturb processes, and as a result, less time is available for preparing the cytostatic drugs.
Organisational embedding
20 problems. An example of this are flow 2 patients. Flow 2 patients are scheduled before 9:30 am because the request for their medication can be sent a day in advance (Figure 4.1). However, blood test results are not available before 4:00 pm, and the physician has to evaluate those results before he/she can send the requests. As a result, requests are often sent after 4:30 pm the day before the treatment is scheduled. The pharmacy, however, closes at 4:30 pm. Therefore, the cytostatic drugs are prepared the next morning, but the treatments are also planned early in the morning. The cytostatic drugs are therefore not delivered on time. If stakeholders meet more frequently and discusses problems more often, these timing errors might have been avoided. Furthermore, stakeholders indicated that due to a lack of alignment between departments, relationships between departments easily worsen. Because of misunderstanding, questions and phone calls are frequently answered in an unfriendly manner, which reinforces the negativity between departments and further impacted information sharing.
21
Figure 4.2 – Cause and effect diagram
22
4.4 Step 4: Suggesting improvements
During the analyses in section 4.3, some critical issues presented themselves. These issues appear to have a major impact on the performance of the supply chain, and they are highlighted in Figure 4.2.
- The amount of time to prepare cytostatic drugs - The insufficient information architecture
The lack of information sharing between departments The lack of digitalisation
The low quality of information - Difficulties in implementing changes
This has resulted in the inability to deliver cytostatic drugs on time, frustrated patients, and low utilization rates at the day care centre.
The amount of time to prepare cytostatic drugs
Section 4.3 pointed out the relationship between the time available to prepare cytostatic drugs and the time needed to prepare them. The more time available to prepare cytostatic drugs, the more frequently the cytostatic drugs are delivered on time. This finding was reinforced by visits at other hospitals. The amount of time available to prepare drugs consistently influences the system’s performance. This finding is supported by Table 4.1, since flow 3 patients perform much better than flow 1 patients. Doing blood tests earlier, as performed in one of the visited hospitals, proved to be a good option for allowing for more drug preparation time. Most preferably, blood tests should be done two days before the patients get the treatment. To be as patient-friendly as possible, blood sampling should be done at a local blood sampling lab to reduce patients’ travelling time. Even though this seems costly, it will reduce the impact of other underlying problems. When the blood test is done two days before the treatment is scheduled, the physician can evaluate the results and send the drug request that day or the next morning. Then the pharmaceutical department can register the request in Cato a day before the treatment is scheduled, and the pharmacist can authorize the request a day before the treatment is scheduled as well. This means the pharmacist’s unavailability will have a lower impact, and there will be more time to, if necessary, clarify obscurities such as incomplete and illegible information. As a result, all orders can be prepared by the day of the treatment without any barriers. By doing this, the pharmaceutical assistants will have a better idea of how many cytostatic drugs have to be produced and when. They can better divide their work during the day and more efficiently use their capacity. They will be less dependent on the planning of the day care centre and outpatient clinic.
23 times for patients. However, it is important to note that to maximize the advantages of this system, the same strategies must also be implemented in other departments that send requests for cytostatic drugs to the pharmaceutical department. Otherwise peak moments will still occur within the pharmacy during the day, which will negatively influence the performance of the cytostatic drug chain.
The insufficient information architecture
The current information architecture has also caused several of problems. The lack of digitalization, the low quality of problems, and the lack of information sharing enhanced problems. Digitalization would reduce the possibility of incompletely or illegibly filling in forms. Then paperwork could be abolished, and the number of information transfers necessary to send requests would be significantly reduced. As a result, the probability of information getting lost somewhere would decrease as well. However, solely increasing the digitalization of processes would not likely be sufficient. One of the visited hospitals that had digitalized many more processes still had problems in delivering cytostatic drugs on time. Furthermore, Figure 4.2 revealed that the lack of information sharing between departments negatively influenced the ability to improve the supply chain. If department heads would meet more frequently, they could increase their understanding of each other’s processes. During meetings, issues of previous periods could be discussed and new action points can be formulated. In this way minor problems could be more quickly solved. A better understanding of the impact of problems in each department will likely increase workers’ willingness to implement changes as well as reduce friction between departments. Thereby, more information sharing between departments will probably also positively influence the planning system. Departments will be better aware of each other’s preferences. As a result, the planning system can better anticipate on this. Effective use of the available capacity will then increase.
Organisational embedding
24
5. Evaluation of the usability of the approach
Overall, the approach of De Vries (2007) was helpful in diagnosing problems within the cytostatic drug chain. Using the approach allowed for insights into different mechanisms as well as a thorough understanding of various cause-and-effect relationships within the chain. Stakeholders confirmed the correctness of the analyses. In addition, the approach made it easier to recognize the main problems within the supply chain. It also acted as a basis for improvements. A more detailed evaluation of the usefulness of this approach, based on the criteria of Nielsen (1993), will be given below.
The first element of usefulness is the utility of the approach. Using the approach led to a comprehensive overview of the various problems within the supply chain. It supported systematically analysing the chain from different viewpoints, and the approach triggered suggestions for improvement, as it was intended to do. The influence on performance of all four elements of the supply chain system became clear. However, to create a holistic view, it was necessary to include the patient flows as well as expected at the beginning of this research. As explained in section 4.3, patient flow had an influence on the performance of the system that could not be ignored. Even though the patient flows and physical flows are highly integrated, they are not the same concept. Therefore it is necessary to incorporate the patient flow to increase the utility and completeness of the approach. To increase the awareness the patient flow is not part of the physical architecture, the physical architecture is redefined as material flow. The modified approach is visualized in Figure 5.1. Furthermore, during this study, the organisational embeddings impact on supply chain performance became clear. Figure 4.2 shows the impact of the organisational embedding on the performance of the system. It appears that power plays and bureaucracy within the hospital are important causes why many improvements were not fully or successfully implemented in the past. This was also referenced as a barrier by some of the other hospitals visited during this research. Therefore, the question arose as to how effective this approach will be if it does not take into account an accurate and thoughtful implementation strategy.
25 to be defined in terms of physical infrastructure, information architecture, planning structure, and organisational embedding. This study only focused on dependability and, to a lesser extent, on speed, and it was already very extensive. Therefore, paying attention to all these elements within one study would make the study even more time consuming and it would likely come at the expense of the level of detail and accuracy of the research. Therefore, it seems better to choose the main performance elements and to keep others as conditions to increase the usability of this approach. Furthermore, the four-step method of the approach supports researchers in correctly and comprehensively doing analysis. However, within the various steps, the researcher still has high agency to act. Therefore, the extent to which the approach will be applied without errors highly depends on the skills of the researcher. Researchers must develop their own methods and techniques to fulfil the steps. The researchers’ skills will directly influence the quality, scope, and accuracy of the methods and techniques used. Finally, because the approach and the steps within the approach were clear and easy to understand, it was enjoyable to work with the approach. However, as explained in section 4,1, within this chain many departments and stakeholders were involved. In addition, process steps were highly dependent on each other and therefore difficult to understand. This increased the level of difficulty of using this approach within a hospital setting. It also increased the probability of getting lost in too much information and thereby destroying the satisfaction of working with the approach. Therefore, it is preferable to use a cause-and-effect relationship model, like the one used in this research, to represent relationships in a structural and clear way. A cause-and-effect relationship model will increase the understanding of different relationships within the chain, and it will ease the complexity of dealing with the healthcare sector.
26
6. Conclusions
This study confirmed the complexity of healthcare supply chains as described in literature. As visualized in Figure 4.2, many causes related to the patient flow, the physical infrastructure, the planning structure, the information architecture, and the organisational embedding had an important influence on the performance of the supply chain system. This further underpinned the multidisciplinary nature of supply chains and the need to incorporate different viewpoints during the supply chain analysis. In addition, the study showed the usefulness of applying the approach to a hospital setting. The approach was helpful in creating a clear overview of the mechanisms of the supply chain system within a healthcare system. It supported the creation of a thorough understanding of cause-and-effect relationships within the chain, and it gave better insights into which elements should change to optimize the supply chain.
Although the approach is simple and consists of clear steps, some changes have been made to increase the usefulness of the approach. The original approach did not include patient flow. Patient flow did influence the supply chain significantly and could not be separated from the physical supply chain. Therefore the patient flow is added as fifth element of the arrangement of the supply chain system. In addition, physical architecture was modified into material flow to highlight the difference between patient flow and this element. Furthermore, there are also some other remarks that have to be taken into account when applying the approach in a hospital setting in the future. The approach mainly guides the phases of analysis and improvement suggestion. It does not provide much attention the implementation phase. This research showed that the implementation phase is at least equally important to successfully improving the supply chain. In addition, many elements are included within the approach, which makes defining research boundaries important to efficiently executing the research. Finally, the complex nature of the hospital sector increases the need to use cause-and-effects relationship models to support the researcher. That said, keeping in mind these shortages and additions, it can be concluded that the approach is useful in a hospital setting.
27 Furthermore, this study approached the unique power balance within a hospital in terms of bureaucracy and physicians’ positions. These organisational characteristics are an important reason why improvement implementation has been difficult in the past. This is underpinned by the publications of De Vries (2011) and Boonstra et al. (2014). One of the main challenges of top management is balancing the power relationships between stakeholders within hospitals. New problems always arise. Therefore, it is essential to be able to anticipate and quickly react to problems to keep the supply chain efficient and to continuously improve performance. Change management perspectives and accurate project management have an important role in the successful implementation of supply chain improvements in the future. Furthermore, it is noteworthy that even though supply chain management within hospitals became more important in recent years, hospitals still lag behind the industrial sector. For example, most hospitals visited during this study did not pay much attention to optimizing planning systems to produce as efficient as possible. The throughput times of patients and materials are mostly only known within each department; the total throughput times of patients or materials throughout the chain are typically unknown. In addition, this case revealed a lack of integration between departments; the departments within this case mainly worked as separate kingdoms. Managers within hospitals have still much progress to make in the area of supply chain management.
28
References
Bonney, M.C. 1994. Trends in inventory management. International Journal of Production
Economics, 35 (1-3): 107-114.
Boonstra, A., Versluis, A., & Vos, J.F.J. 2014. Implementing electronic health records in
hospitals: a systematic literature review. BMC Health Services Research, 14: 370-394.
Bryson, J.M., Ackermann, T., Eden, C., & Finn, C.B. 2004. Visible thinking: Unlocking
causal mapping for practical business results. Chichester: Wiley & Sons, Inc.
Croom, S., Romana, P., & Giannakis, M. 2000. Supply chain management: an analytical
framework for critical literature review. European Journal of Purchasing & Supply
Management, 6 (1): 67-83.
Dacosta-Claro, I. 2000. The performance of material management in health care
organisations. Internal Journal of Health Planning and Management; 17 (1): 69-85.
De Leeuw, A.C.J. 2000. Bedrijfskundig Management. Assen: Van Gorcum.
De Vries, J. 2007. Diagnosing inventory management systems: An empirical evaluation of a
conceptual model. International Journal of Production Economics, 108 (1-2): 63-73.
De Vries, J. 2009. Assessing inventory projects from a stakeholder perspective: Results of an
empirical study. International Journal of Production Economics, 118 (1): 136-145.
De Vries, J. 2011. The shaping of inventory systems in health services: A stakeholder
analysis. International Journal of Production Economics, 133 (1): 60-69.
De Vries, G., Bertrand, J.W.M., & Vissers, J.M.H. 1999. Design requirements for health care
production control systems. Production Planning & Control, 10 (6): 559-569.
De Vries, J, Commandeur, H.R., & Huijsman, R. 2007. Management en organisatie in de
gezondheidszorg: Overzicht en reflectie. M&O: Tijdschrift voor Management en
Organisatie, 2: 5-21.
De Vries, J., & Huijsman, R. 2011. Supply chain management in health services: an overview.
29
Drupsteen, J., Van der Vaart, T., & Van Donk, D.P. 2013 Integrative practices in hospitals
and their impact on patient flow. International Journal of Operations & Production
Management, 33 (7): 912-933.
Eisenhardt, K.M. 1989. Building theories from case study research. The Academy of
Management Review, 14 (4): 532-550.
Forza, C. 2002. Survey research in operations management: a process based perspective.
International Journal of Operations and Production Management, 22 (2): 152-194.
Grotsch, V.M., Blome, C., & Schleper, M.C. 2013. Antecedents of proactive supply chain risk
management – a contigency theory perspective. International Journal of Production
Research, 51 (10): 2842-2867.
Harland, C.M. 1996. Supply chain management: Relationships, chains, and networks. British
Journal of Management, 7 (1): 63-80.
Jack, E.P. & Powers, T.L. 2004. Volume Flexible Strategies in Health Services: A Research
Framework. Production and Operations Management, 13 (3): 230-244.
Kaplan, H.C., Brady, P.W., Dritz, M.C., Hooper, D.K., Linam, W.M., Froehle, C.M., &
Margolis, P. 2010. The influence of context on quality improvement success in health care: A
systematic review of the literature. The Milkbank Quarterly, 88 (4): 500-559.
Karlsson, C. 2009. Research Operations Management. New York: Routledge.
Lawrence, P.R., & Lorsch, J.W. 1967. Differentiation and integration in complex
organisations. Administrative Science Quarterly, 12 (1): 1-47.
Lega, F., & DePietro, C. 2005. Converging patterns in hospital organisation: beyond the
professional bureaucracy. Health Policy, 74 (3): 261-281.
Miles, M.B., & Huberman, M.A. 1994. Qualitative Analysis. London: Saga.
Nicholson, L., Vakharia, & A.J., Erenguc, S.S. 2004. Outsourcing inventory management
decisions in healthcare: models and applications. European Journal of Operational research,
154 (1): 271-290.
30
Schmenner, R.W., & Swink, M.L. 1998. On theory in operations management. Journal of
Operations Management, 17 (1): 97-113.
Shortell, S.M., Zimmerman, J.E., Rousseau, D.M., Gillies, R.R., Wagner, D.P., Draper, E.A.,
Knaus, W.A., & Duffy, J. 1994. The performance of intensive care units: Does good
management make a difference? Medical Care, 32 (5): 508-525.
Skjott-Larsen, T., Schary, P.B., Mikkola, J.H., & Kotzab, H. 2007. Managing the Global
Supply Chain (3th ed.). Gylling: Copenhagen Business School Press.
Slack, N., Chambers, S., & Johnston, R. 2010. Operations Management (6th ed.). Harlow:
Pearson.
Van Hoeve, J., & De Vries, J. 2014. Ontwikkeling en implementatie van zorgpaden:
Resultaten van drie exploratieve case studies. M&O: Tijdschrift voor Management en
Organisatie, 2: 5-22.
Vissers, J.M.H., Bertrand, J.W.M., & De Vries, G. 2001. A framework for production control
in health care organisation. Production Planning & Control: The Management of
Operations, 12 (6): 591-604.
Voss, C., Tsikriktsis, N., & Frohlich, M. 2002. Case research in operations management.
International Journal of Operations & Production Management, 22 (2): 195-219.
Yin, R.K. 1994. Case Study Research: Design and Methods. Beverly Hills: Sage Publication.
31
Appendix A: Measurement form
Patiëntnummer: ……….
1
ezakje
2
ezakje
Geplande starttijd toediening Cytostatica
*Daadwerkelijke starttijd toediening Cytostatica
**Indien uitloop, reden
(Meerdere mogelijkheden mogen worden aangekruist): Cytostatica te laat geleverd
Uitloop vorige patiënt
Patiënt te laat
Patiënt te laat door uitloop spreekuur
Meer voorbereidingstijd nodig dan gepland
Drukte op te afdeling
Anders, namelijk………
………..
*: Het gaat hier om de geplande starttijd wanneer de cytostatica daadwerkelijk toegediend had moeten worden, nadat alle voorbereidingen afgerond hadden kunnen zijn.
**: Het gaat hier om de starttijd wanneer de cytostatica daadwerkelijk toegediend werd
.
2e32
Appendix B: Coding tree
Category Sub-category Explanation
Physical infrastructure Process steps All steps within the process necessary to produce cytostatic drugs
Throughput times Time related to all process steps necessary to produce cytostatic drugs
Transportation methods Vehicles used transport physical materials to their destination within the chain
Patient flow Process steps All the steps patients take during the process
Service times Time that different steps within the chain take and the waiting time related to those steps Planning structure Order quantities The amount of orders during a
specific time span
Order intervals (Scheduled) time between orders
Planning rules Conditions a schedule must fulfill
Information architecture Information sources The way information becomes available for different
stakeholders Information sharing The extent to which
departments and stakeholders share information with each other
Quality of information The extent to which information is correct, complete, and
readable
Quantity of information The extent to which the needed information is available
33 the right place
Organisational embedding Formal structure The formal distribution of various tasks according to job descriptions
Informal structure Distribution of various activities beyond job descriptions
Responsibilities the formal and informal job duties of employees
Political process Stakeholders’ power, and the way they use that power
34
Appendix C: Detailed performance evaluation
Table C1 shows the performance of the cytostatic drug flow based on the performance indicator: ‘The pharmaceutical department has to deliver the cytostatic drugs before the time of the patient’s appointment at the day care centre’. The performance of the system greatly varies between flows. As seen in Table C1, the performance of flows 1 and 2, as related to these performance indicators, is very poor. However, the performance of flow 3 seems to be quite good. Furthermore, it is remarkable to note that almost no study was delivered on time
.
Total amount infusions delivered
Amount of infusions delivered on time Flow 1 126 48 Flow 2 11 6 Flow 3 106 98 Studies 18 3 Total 243 152
Table C1 – Performance overview
Total infusions delivered Delivery cytostatic drug on time Delivery cytostatic drug on time + Provided on time Delivery cytostatic drugs too late + Provided on time
Total number infusions 243 152 141 43
Flow 1 126 48 44 32
Flow 2 11 6 6 5
Flow 3 106 98 91 6
Studies 18 3 3 3
Table C2 – Infusions delivered on time versus provided on time
However, not every infusion that was provided to the patient on time was delivered on time to the day care centre (Table C2). Therefore, more infusions were provided on time compared to being delivered on time. This variance occurs because some treatments need preparation time, which can involve pre-medication and the insertion of the infusion. If this took longer than the delivery of the cytostatic drug, the drug was still provided on time to the patient. Therefore patients did not always notice when the cytostatic drug was not delivered on time.
35
Day Total number of infusions flow 1 Delivery cytostatic drug on time Delivery cytostatic drug on time + Provided on time Delivery cytostatic drug too late + Provided on time Friday 10 4 4 1 Wednesday 22 9 9 7 Friday 13 3 3 6 Wednesday 16 6 6 3 Thursday 10 6 6 2 Monday 22 8 8 6 Thursday 11 3 2 2 Monday 20 8 5 5 Total 124 47 43 32
Table C3 – Performance overview flow 1
Table C4 – Average throughput time available for preparing cytostatic drugs flow 1
Day Total number of infusions flow 1
Delivery
cytostatic drugs on time
Actual average throughput time preparing cytostatic drugs
(Average time between delivery request at pharmacy and actual delivery of cytostatic drug at day care centre)
Friday 10 40% 1:16 Wednesday 22 41% 1:10 Friday 13 23% 1:00 Wednesday 16 38% 1:12 Thursday 10 60% 1:04 Monday 22 36% 1:11 Thursday 11 27% 1:01 Monday 20 40% 1:17 Total 124 38% 1:10
Table C5 – Actual average throughput time needed for preparing cytostatic drugs flow 1
On average, the pharmacy department has 59 minutes to prepare the cytostatic drug (Table C4). However, Table C5 shows an actual average throughput time of 1:10 hour. There is a gap of 11
Day Total number of infusions flow 1
Delivery cytostatic drug on time (%)
Average throughput time available to prepare cytostatic drug
(Average time between delivery request at pharmacy and scheduled appointment of patient at day care centre)
36 minutes between those average times, which results in a below-standard performance.
Figure C1 and Figure C2 visualize how the available throughput time and the actual throughput time for preparing the cytostatic drug influence the on-time delivery. When the average available throughput time for preparing cytostatic drugs was low, for example 45 to 50 minutes (Table C4), the percentage of infusions delivered on time was relatively lower (Table C4 and Figure C1). However, if the average available throughput time to prepare cytostatic drugs was higher, for example 1:05 hour (Table C4), the percentage of infusions delivered on time was not immediately higher (Table C4 and Figure C1). The reason why the percentage of infusions delivered on time was still quite high is explained by the relatively high actual average throughput time for preparing the cytostatic drugs (Table C5 and Figure C2).
37
Figure C2 – Percentage infusions delivered on time compared to actual average throughput time to prepare cytostatic drugs
The relationship between outpatient clinic and performance are visualized in Table C6.
Outpatient clinic Total number infusions Infusions Flow 1 (%) Infusions Flow 2 (%) Infusions Flow 3 (%) Infusions delivered too late (%) IAOP 122 79% 0% 21% 53% IHMP 95 23% 0% 77% 15% ILOP 19 32% 58% 11% 53% IREP 7 0% 0% 100% 29%
Table C6 – Relationship outpatient clinic and late delivery
Outpatient clinic IAOP has a relative high percentage of infusions flow 1 and a relative high percentage of infusions delivered too late. Nevertheless, the outpatient clinic ILOP has a relative high percentage of infusions delivered too late but a low average percentage of infusions flow 1. However outpatient clinic ILOP has a relative high percentage of infusions flow 2. As presented in Table C1, this flow has very poor performance. IHMP has the lowest percentage of flow 1, no infusions for flow 2, and the lowest percentage of late-delivered infusions.
38
Figure C3 – Relationship between late delivery flow 1 and time span
39
