Embedding a health care Case Management System into its surrounding quality framework

(1)

UNIVERSITY OF AMSTERDAM

Faculty of Science

Thesis Master Information Science

Business Information Systems

Embedding a health care Case Management System

into its surrounding quality framework

Elekta grants permission to view this thesis by means of library placement. For publication of this thesis, complete or partially, permission from Elekta is required.

Author: Dorothea Langer _{Student number: 10865446} Supervisor: Prof. Dr. H. A. Reijers Signature:

Second examiner: Dr. Peter Weijland Signature:

(2)

Abstract

The lack of integration of separate quality initiatives is a frequent challenge in health care quality management that puts compliance and user acceptance at risk. Similarly, the iso-lation of Case Management Systems leads to a disregard of information that could prove valuable. In order to remedy this situation, the challenges of a health care quality framework were investigated empirically within the radiotherapy department of the University Medi-cal Center Utrecht using interviews. Subsequently, integration strategies towards a solution of the found challenges were developed and evaluated by a focus group inside the depart-ment. Finally, a prototype of the solution deemed most appropriate by the focus group was implemented in order to prove the technical feasibility of the proposed solution. This proto-type based on the open-source software Camunda integrates the case management system with the internal Clinical Practice Guidelines (CPG). It introduces a workflow element pre-senting the radiotherapist with recommended treatment options based on the current pa-tient data. If confirmed by the radiotherapist the selected option is directly translated into work flows by the Case Management System ensuring that the patient can be treated only in accordance with the selected protocol. This implementation encourages guideline com-pliance through monitoring and through the presentation of a recommended option. The solution is expected to improve user acceptance by removing the need for reading through text-based guidelines. Finally, data on deviations from guidelines and reasons for these are valuable information on the side of the Case Management System. It can uncover parts of the guidelines that require improvement. The thesis at hand thereby presents a solution towards a more holistic view in health care quality.

(3)

List of Figures

1 Line of reasoning of the paper presented . . . 5

2 Quality framework of the radiotherapy department according to interviews. . 8

3 ICT Infrastructure of the radiotherapy department according to interviews . . 17

4 Portfolio of integration strategies . . . 25

5 Software architecture of the proposed solution . . . 26

6 BPMN diagram of Treatment Decision Flow Dosage Hematoma . . . lxxxvi 7 Workflow element Get Treatment Recommendation . . . lxxxvii

List of Tables

1 Examples of hospital indicators (Eggli and Halfon, 2003) . . . 4

2 Interview dates . . . 6

3 Participants of the focus group . . . 7

4 Ranking of strategies according to benefits (by focus group) . . . 24

(6)

1 Introduction

1.1 Health care quality

The significance of quality is widely recognized in many industries. However, it is attributed a special importance in health care because health care quality is concerned not only with financial matters, but mainly with human lives:

Health care quality is optimal care from the appropriate provider in the most appropriate setting in the most appropriate manner for the patient’s unique cir-cumstances. (Nash, 1998 via Ross (2014))

Ideally, a provision of such health care prevents the “Five D’s of Health Care Quality": Death, Disability, Disease, Discomfort, and Dissatisfaction (Ross, 2014). Unfortunately, the quality of health care is not optimal currently. Errors in health care are frequent and have a large impact, both on the patient and from the financial perspective. The often-cited re-port “To err is human" published by the US-American Institute of Medicine in 1999 states that health care errors in the United States are estimated to cause between 44,000 and 98,000 deaths annually. The national costs were estimated to be between 37.6 billion and 50 billion $US when including lost income, lost household production, disability and health care costs themselves (Kohn et al., 1999). 16 years have passed since the publication of this report, yet following reports have found only moderate improvement despite a risen awareness (Wachter, 2004), and health care quality remains an important issue.

1.2 Case Management Systems

According to Ross (2014), one of the main aspects of health care quality is the standardiza-tion of treatment, while at the same time being able to respond variably to unique situastandardiza-tions. One possible solution for this is the use of Case Management:

A dynamic and systematic collaborative approach to providing and coordi-nating health care services to a defined population. It is a participative process to identify and facilitate options and services for meeting individuals’ health needs, while decreasing fragmentation and duplication of care and enhancing quality, cost-effective clinical outcomes. The framework for nursing case management includes five essential functions: assessment, planning, implementation, evalua-tion and interacevalua-tion. (ANCC, 2004, p. 10 via Hoeman (2008))

In addition to its cost saving nature case management is considered to be “a key factor in increasing quality of care, patient and family satisfaction, patient compliance, and increased quality of life" (Harrison and Nolin, J., Suero, E., 2004).

Case Management supports four major processes: Care coordination, quality manage-ment, outcomes management and cost management. (Hoeman, 2008).

Recently, Case Management Systems (software implementing case management pro-cesses), have been widely used especially in the field of cancer treatment (Ozcelik et al., 2014).

1.3 Problem statement

One of the common reasons for the failure of quality improvement initiatives is the lack of “an overall quality strategy or long-term evaluation plan [that leads to] many disconnected, half-evaluated projects that never seem to achieve their objectives." (Strome, 2013)

(7)

A large variety of different approaches to health care quality improvement have promised to solve this lack of a holistic view. These approaches include Continuous Quality Improve-ment (Lee et al., 2006), Total Quality ManageImprove-ment and Business Process Re-engineering (Lee et al., 2006, Millar, 2013), Benchmarking, Supply Chain Management, Six Sigma (Lee et al., 2006) and Lean (Millar, 2013).

However, Lee et al. (2006) state that “although the contemporary performance measure-ment and quality improvemeasure-ment models in health care identify quality issues and suggest im-provements, it lacks a uniform and integrated framework of quality improvement projects." Lee themselves and several other (Sewell, 1997) (Boon et al., 2004) subsequently focused on the production of abstract frameworks. They did not consider actual integration strategies.

Neither the above mentioned approaches nor the abstract frameworks had a significant effect on the implementation of holistic quality improvement programs. This could be due to the limitations of their implementations. An example for this is provided by Mozzacato et al. (2010) who found in their literature research that “while lean theory emphasizes a holistic view [on quality improvement], most cases report narrower technical applications with limited organizational reach".

A concrete method of integration could therefore add value to the health care quality framework by removing the gap between framework and implementation. In this thesis, Case Management Systems are chosen for concrete integration into the quality framework due to their strong isolation. Case Management Systems are built to support single processes without incorporating the surrounding quality framework. This isolation of the system in-troduces a number of issues, namely the disregard of valuable information on the side of the case management system as well as a low compliance and user acceptance on the side of the surrounding quality tools. These three issues are deepened in the following.

1.3.1 Business intelligence

Certain information of quality tools could prove valuable for the quality process in case management, for example:

• Time required for working through quality-related tasks is not known: No information on the efficiency of measures.

• The impact of quality-related tasks is not known: No information on the effectiveness of measures.

1.3.2 Compliance

Widely, compliance is not enforced by quality management tools. For example, the use of checklists is usually not monitored by the implementing organizations. Instead, its users are merely asked to sign that they have followed the given steps after the execution. This can lead to errors and ultimately incidents that the tools were implemented to prevent in the first place. (Degani and Wiener, 1993, Helmreich, 2000)

1.3.3 User acceptance

Research shows that the agents exposed to quality tools such as checklists tend to have neg-ative feelings towards this approach. Because they are guided in their own field of expertise where they often have years of experience, they may see the tool as a sign of doubt in their capabilities. (Gawande, 2011) This problem interacts closely with the second issue, compli-ance.

(8)

1.4 Research questions

This research aims to improve the lack of business intelligence on the Case Management System side and the lack of compliance and user acceptance on the quality framework side. This is to be done by integrating the Case Management System with its surrounding quality framework. Therefore, the following research questions are answered in this thesis:

1. What quality framework is currently used in hospitals?

2. Which aspects of this quality framework are suitable for integration with a Case Man-agement System?

3. Is an integration of the Case Management System and the selected quality framework aspect technically feasible?

2 State of knowledge on hospital quality frameworks

According to Davies et al. (2000), a quality framework is a “detailed set of interlocking strate-gies, tactics, and supporting initiatives" that consists of three phases:

1. Defining appropriate quality standards

2. Deliver health care according to quality standards 3. Monitor the delivery

Davies et al. (2000) also show that the outcome of the monitoring phase will again be used to improve the previous phase of implementation. Since the health care delivery is meant to implement the quality standards, a change of health care delivery must be pre-ceded by an adaptation of the quality standards. It thereby completes the Plan-Do-Check-Act cycle which is the basis for continuous process improvement. (Pavleti´c et al., 2008)

This section discusses the different methods of health care quality management catego-rized into the phases of the quality framework’s development.

2.1 Defining appropriate quality standards

A method for defining quality standards in health care are Clinical Practice Guidelines.

2.1.1 Clinical Practice Guidelines

Clinical Practice Guidelines (CPG) aim at making care more consistent and efficient. Often, there is a gap between what clinicians do and what is supported by scientific evidence. This is potentially damaging to patients. (Woolf et al., 1999) However, the effectiveness and ef-ficiency of care modeled after guideline recommendations strictly depends on the quality of the guideline. “Guidelines that promote interventions of proved benefit and discourage ineffective ones have the potential to reduce morbidity and mortality and improve quality of life, at least for some conditions." (Woolf et al., 1999) On the other hand, if scientific evi-dence is lacking or misinterpreted, if recommendations are influenced by the opinions of the guideline development group or if development priorities are focused on costs or personal interest, the promotion of flawed guidelines may cause harm. (Woolf et al., 1999)

2.2 Deliver health care according to standards

In order to ensure the health care delivered complies with the previously defined standards, organizations can utilize Care Pathways and Continuing Medical Education.

(9)

Table 1: Examples of hospital indicators (Eggli and Halfon, 2003)

Types of Indicators Definition and examples

Relevancy Association of patients’ and activities’ data: • Mean length of stay standardized by case mix • Proportion of Caesarian sections among deliveries • Proportion of cancer screening in eligible patients Productivity Association of activies and resource data:

• Number of procedures per working operation room (adjusted for the severity of the procedure) • Mean bed occupation rate

• Number of full time equivalent nurses per unit of required nursing services • Mean cost per unit of radiological activity

Adequacy Association of resources and patient’s data: • Mean cost per homogeneous patients’ groups • Ratio of full time equivalent physicians per patient Efficacy Association of patients’ and effects data:

• Premature hospital mortality rate adjusted by case mix • Patient satisfaction

• Emergency waiting time, adjusted by risk scores Effectiveness Association of activities and effects data:

• Post-operative nosocomial infections rate

• Proportion of pain complaints among patients under antalgy treatment • Potentially avoidable readmissions rates (discharge process)

Efficiency Association of resources’ and effects data:

• Mean cost of a saved year of life without handicap • Staff satisfaction scores (motivation)

• Mean cost of the reduction by one minute of waiting time

2.2.1 Care Pathways

Care Pathways are personalized care plans that increase the adherence of treatments to CPG recommendations by coordinating the activities of health care professionals involved in the treatment of a patient. (Sánchez-Garzón et al., 2014)

2.2.2 Continuing medical education

Continuing Medical Education is focused on improving patient outcomes. “Models of con-tinuing medical education (CME) seek not only to impart knowledge but to change physi-cians’ behavior and even play a role in facilitating organizational improvement" (Shojania et al., 2012). Typically, conferences for continuing medical education aim to transfer knowl-edge of new treatment methods and advances in diagnostic technologies, but also of practi-cal strategies. (Shojania et al., 2012)

2.3 Monitor the delivery

Finally, the delivery of health care can be monitored using performance measures, clinical peer reviews, hospital accreditation and public performance reports.

2.3.1 Performance measures

Typically performance metrics are used (Eggli and Halfon, 2003) in conjunction with audit tools (Millar, 2013). Examples of such performance metrics are listed in table 1.

2.3.2 Clinical peer review

The clinical peer review is a method of internal control (Eggli and Halfon, 2003). It is “the process whereby doctors evaluate the quality of their colleagues’ work in order to ensure that prevailing standards of care are being met." (Vyas and Hozain, 2014) These professional

(10)

Figure 1: Line of reasoning of the paper presented

standards must therefore be well-established in the institution and known to the reviewer in order for the peer review to be successful. (van Weert, 2000)

2.3.3 Accreditation of hospitals

The accreditation of hospitals is a method of supervision. The external evaluation of health care quality is done by many national accreditation organizations and an international body, the International Society for Quality in Health Care and has been shown to promote change in health organizations. Accredited hospitals have shown significant improvement in sev-eral areas, most notably in nursing organization and safety. (Greenfield and Braithwaite, 2008)

2.3.4 Public performance reports

Publicly released reports are aimed at fostering competition in the health care sector on the basis of quality. Schneider and Epstein (1998) show that even though the behavior of the patients is hardly influenced by such publications, the care providers adapt their quality management to the criteria of the reports.

3 Methodology

This study used a qualitative approach to answer the posed research question. Its subques-tions were answered consecutively with the answer of each question arising from the previ-ous step as shown in figure 1. The current quality framework and challenges were uncov-ered using interviews. From these challenges, possible integration strategies of the quality framework and the Case Management System were deducted. Then, these were evaluated for their benefits and the complexity of the implementation by a focus group. Finally, one of the strategies was chosen and further investigated by development of a prototype.

3.1 Setting

The research project was embedded in a project of Elekta, a company providing radiation therapy, radiosurgery, related equipment and clinical data management for cancer and brain disorder treatment. The project was executed in cooperation with the radiotherapy depart-ment of the University Medical Center Utrecht (UMCU). One of the project’s aims was to develop a case management approach that enables flexible guidance of treatment manage-ment processes.

3.2 Method for research subquestion 1

The first subquestion of this research, “What quality framework is currently used in hospi-tals?", was answered using semi-structured interviews. The main goal of these interviews was to uncover the current quality framework of the UMCU’s radiotherapy department. The research questions aim at integrating this framework with the case management system that was in development at the time. Such an integration is to benefit the implementing

(11)

Table 2: Interview dates

Organization Interviewee Position Date IKNL Xander Verbeek Coordinator Health Care Informatics March 9, 2015 UMCU Jan Stout Application Manager multiaccess April 9, 2015 UMCU Robert Tersteeg Radiotherapist April 20, 2015 UMCU Iris van Dam Radiotherapist April 28, 2015 UMCU Marij Mooring Radiotherapy laboratory assistant April 30, 2015 UMCU Wilfred Tulling Radiotherapy quality staff May 5, 2015 UMCU Judith Roesink Radiotherapist May 21, 2015

hospital. Therefore, a special focus was on pressing quality issues inside of the department that may leave room for improvement through integration.

3.2.1 Participants

The participants included several members of the UMCU’s radiotherapy staff as well as Xan-der Verbeek, an employee of the Integraal Kankercentrum NeXan-derland (IKNL). The IKNL is an organization which is responsible for the production of guidelines concerning the treat-ment of cancer.

3.2.2 Questions

Due to the different perspectives that external influences (IKNL) and the participants of the treatment processes (the radiotherapy staff) have on the quality framework of hospi-tals, different sets of questions have been used. The radiotherapists were expected to have knowledge on the operative aspects of the overall quality framework. On the other hand, the external influences were expected to have knowledge of the guideline development and implementation. The respective questions can be found in Appendix A.

3.2.3 Interview conduction

The interviews have been conducted in the work places of the interviewees in order to cre-ate a setting that resembled their work habitat as closely as possible. This was to support the reflection on their work and to create a known and comfortable environment to the in-terviewees. The list of questions used was varied in its order or was slightly adapted when deemed necessary by the interviewer.

The dates of the interviews are displayed in table 2.

3.2.4 Analysis

The interviews were analyzed according to thematic analysis. The coded transcripts of the interviews can be found in Appendix C.

The current quality framework and the list of quality challenges inside the radiotherapy department were used to answer the second subquestion, namely “which aspects of this quality framework are suitable for integration with a Case Management System?". Possible integration strategies were deduced by investigating how the challenges could be overcome through integration. This focus on the solution of challengeswas to ensure that all proposed strategies provide additional value to the department.

The possible integration strategies were then evaluated by a focus group consisting of three members of the UMCU’s radiotherapy staff on June 2, 2015.

(12)

Table 3: Participants of the focus group

Participant Position Robert Tersteeg Radiotherapist

Wilfred Tulling Radiotherapy quality staff Alexis Kotte Radiotherapy IT staff

3.3.1 Participants

In order to include differing views into this ranking, the participants of the focus group were chosen from different job positions as displayed in table 3.

3.3.2 Questions

The focus group was given one scenario per possible integration strategy. Using this sce-nario, they discussed which benefits they saw for the radiotherapy department and how complex they would estimate the strategy’s implementation. The respective scenarios can be found in Appendix B. The participants were then asked to rank the proposed strategies according to the perceived benefit and the perceived complexity of its respective implemen-tation.

3.3.3 Conduction

The focus group was conducted in an office at the UMCU’s radiotherapy department that was known to all participants. This was to ensure a comfortable surrounding for the partic-ipants.

3.3.4 Analysis

The focus group discussion was analyzed according to thematic analysis. The coded tran-scripts of the focus group can be found in Appendix D.

The third subquestion, “Is an integration of the Case Management System and the selected quality framework aspect technically feasible?", was approached through the development of a prototype. This implements the integration strategy that had been evaluated to be most suitable in subquestion two.

4 Results

The results of the interview analysis are presented in the following. The boxed text displays potential areas of improvement.

The quality framework as depicted in figure 2 was derived from the interviews. In the diagram, the pink layer shows the different views on quality taken by the interviewees. These lead to a focus on different aspects of the quality framework as displayed in the blue layer, which in turn are reflected in an emphasis on different methods and tools as shown in the bottom-most, green layer. The different metrics of quality are appended in green text to their appropriate view, method or aspect of the framework.

The hierarchical structure shows a number of different quality-focused initiatives that are currently in place in the UMCU’s radiotherapy department. This is because the con-cern for quality in the radiotherapy department is “huge" (Robert Tersteeg) and the focus

(13)

Figure 2: Quality framework of the radiotherapy department according to interviews. on quality improvement is “ingrained in the nature" of a radiotherapy department (Marij Mooren).

At the same time the department is developing fast. It is growing, and quality im-provement initiatives are executed “very quickly after each other, or parallelized". (Wilfred Tulling)

There is a large number of different quality processes and initiatives that are not connected enough.

The fast growth of the department led to a large number of isolated initiatives where “people are waiting on each other to change or have difficulties getting the right person to do something about [the quality problem]". (Robert Tersleeg) Marij Mooren believes that the quality initiatives “could be connected way more".

Quality concerns are present at “all kinds of angles [...] in all layers" (Marij Mooren). The main views on quality that are taken by the interviewees are focus on treatment, on customer interaction and on department organization respectively.

4.1 Treatment

The interviewees stated that treatment quality is a major driver in radiotherapy depart-ments. While Marij Mooren says that quality for radiotherapy is about “doing the right thing with treatments", Robert Tersleeg asserts “quality is about outcome, it is about side effects and cure rates. That is one aspect of quality for us, we treat patients and we want them to benefit from the treatment." and Iris van Dam defined quality for herself as patient outcome.

(14)

is a combination of so many things. It is about communication, reproducibility, medical equipment, possibilities for medications,..." Methods focusing on the treatment of patients can be split up into the three aspects of a hospital quality framework as described before: Quality standards, quality delivery and quality monitoring.

4.1.1 Quality standards

Quality standards are set through both work instructions and clinical practice guidelines.

4.1.1.1 Work instructions Some work instructions are department-based and written by either the radiotherapists themselves or by interdisciplinary teams (Robert Tersleeg). Others are hospital-wide in operation. In general, work instructions are not meant for the radiother-apists but rather for the technicians in the department (Judith Roesink). However, there is also a project that is meant to redesign the way work instructions are used and accessed (Marij Mooren).

4.1.1.2 Clinical practice guidelines Clinical practice guidelines can either be maintained by an external organization or adapted internally in the radiotherapy department.

External guidelines External guidelines are “more or less describing the path of the patient and they are interdisciplinary. So they describe for each discipline, for instance the medical oncology and the surgeons and for us what the standard of care in the Netherlands should be." (Robert Tersteeg). Judith Roesink states concerning the importance of external guideline organizations:

We have the NVRO, the Dutch group of radiation oncologists. Within the NVRO there are all the specific tumor-groups and specific tumor-sites. They are leading in writing all the guidelines. We have Oncoline, that is a national website where all these guidelines are public for everyone.

The organization developing Oncoline guidelines is the Integraal Kankercentrum Neder-land (IKNL). Their guidelines are “[mandated] by the representatives from the different dis-ciplines. Once they are finished they get reviewed [and] then the guideline gets a signature [...]."

(Xander Verbeek)

The national centralization of guidelines is hindered due to the slow update process of external guidelines that leads to the use of internal adapted guidelines.

Both the IKNL and the radiotherapists stated that the external guidelines are not updated often enough.

Obviously the dilemma always is that new knowledge that has been gener-ated, new clinical trials have not been incorporated in the guideline of 2012 but you see that the hospitals have changed that care process already. So the older the guideline gets the more deviations from the guidelines you see because the doctors from the hospital consider the guideline outdated. So there is a huge amount of quicker revisions and a change of the process that we are going through right now is that instead of revising the entire guideline we want to go to a process of module revision. (Xander Verbeek)

(15)

Internal guidelines Due to the lack of updates, an adapted version of external guide-lines is used inside the department for the actual treatment:

We use those guidelines from the IKNL to translate the national guidelines to our own treatment guidelines. So we have our own set of protocols and treat-ment guidelines which most often are more strict and very much more in detail than the IKNLâ ˘A ´Zs national guidelines" (Robert Tersteeg).

These are updated regularly:

[Our internal guidelines change] at least every five years. But for some guide-lines five years is stated when there is not much change, but some change twice a year. That happens. For breast we have a very active group developing new guidelines and when there is new literature or new equipment with more pos-sibilities or more insight in what you do, for instance setup verification or new delineation guidelines we just adapt the guideline. And we cannot do this just by ourselves so I make the decisions for hematology with another colleague. The two of us, we know the most about that, and the others trust that. (Robert Ter-steeg)

There is a process in place to ensure quality with the guideline updates.

So when we think this is better for our patients we change the guideline, but then the guideline has to go through a whole process: Protocol committee, protocol meeting with all the other colleagues and everybody can have his or her say about this, such as: When you make this change, have you thought about this or that aspect. So even when the group that is responsible for a guideline decides on change, still there are a lot of more people involved that have the opportunity to say something about this. (Robert Tersteeg)endquotation The update cycle is reviewed by the Quality Manager: “I am with a few other people responsible for making the latest version available. There is sometimes a little bit of a gap but that is work time" (Wilfred Tulling).

The internal guidelines are used extensively.

[Internal guidelines are used] daily. I think most things we do are according to the guidelines. For every patient started on treatment you have to state in the file: Do I treat this patient according to the proto-col, according to the guideline, or not? And if you don’t you have to motivate exactly why you don’t treat according to standard protocol. We strive to do 95% according to guidelines! (Robert Tersteeg)

A different fraction is given by Judith Roesink, who proposes guideline devia-tion in 20% of all cases.

Compliance to guidelines is only partially supported through the current guideline format since radiotherapists often refer to memory instead.

All radiotherapists interviewed state that they have the guidelines internalized partly and do not always refer to the text-based version.

“[I use the guidelines] every once in a while because they are very complicated. The biggest part I have internalized and I don’t have to look it up every time I treat a patient." (Iris van Dam)

(16)

The hospital-wide centralization of guidelines is hindered because of the large number of partially conflicting guidelines.

Currently, the guidelines are rather department-specific in the UMCU. There is, however, an initiative to make the guidelines more central: “Only since about two years, they try to get all the guidelines into one central database. [...] It is a disaster!" (Robert Tersteeg)

4.1.2 Quality delivery

4.1.2.1 Care Pathways Care pathways are the implementation of the guide-lines in the specific patient’s case. They are “related to decision points that doc-tors actually have to make in routine practice". (Xander Verbeek)

4.1.2.2 Discussions The making of these decisions is facilitated through a num-ber of discussion points in the process.

Daily radiotherapists’ meeting The most important discussion point in the UMCU is the daily radiotherapists’ meeting, where every new patient of the department is discussed. Robert Tersteeg asserts that “it is important to have colleagues from other disciplines to think about what you do." The “other dis-ciplines" present in the radiotherapists’ meetings are subspecialties of radiation oncology, such as colleagues focused on the treatment of breast cancer or brain tumors. If the colleagues agree that the proposed treatment of a patient is accord-ing to guideline or well-argued then in case of complaints “as a department, we take responsibility towards the patient." (Robert Tersteeg) If the group had con-cerns about the treatment that the radiotherapist chose to disregard then they are “completely on [their] own" (Robert Tersteeg) in terms of responsibility.

Multidisciplinary meeting In addition to these meetings, also interdisci-plinary meetings are held regularly, at least weekly (Judith Roesink):

We have these interdisciplinary meetings with the medical oncolo-gists, with the radiolooncolo-gists, with the cardiolooncolo-gists, whatever. All spe-cialists where the patients first come in. We always have the second place in the treatment. The indications and the treatment plan is de-signed in the MDO, the multidisciplinary meeting. (Judith Roesink)

Video conferences Another discussion method are video conferences “where the surgeons and the oncologists in the [referring] hospital are sitting in a room with a camera and the radiation oncologist from our department calls in and they have a video meeting, about once or twice every week." (Robert Tersteeg) Additionally, there are ad-hoc discussions whenever they seem necessary: “We discuss a lot with colleagues [...] When we are in doubt, we are not sure, we send an e-mail to everyone [in order to] get to a consensus treatment." (Iris van Dam)

4.1.2.3 Continuous medical education Another method of insuring quality delivery from a treatment perspective is continuous medical education. There are systems of schooling in place in the department. Here, schooling has to be repeated frequently: “[The employees] keep doing [schooling] to keep proficient

(17)

at it, and we record [their participation]." (Marij Mooren) Next to medical educa-tion, these schooling also include technical systems: “Every half a year we give the doctors that need it refreshing in how to use the different systems [...]". (Marij Mooren) Also, the daily radiotherapists’ meetings are valued as an exchange of latest treatment options and new studies. Robert Tersteeg states that “learning from each other is also very important in a quality-minded department group". Other groups within the department have similar points of exchange: “Every day of the week we have a technicians meeting where we get a little bit of ex-tra education and we have time to ask questions or to propose changes." (Marij Mooren)

4.1.2.4 Quality assurance A fourth method of quality delivery is quality as-surance itself. It concerns rather the technical side: “We have quality asas-surance for the physics side, technicians have a lot of quality steps ingrained in the pro-cess." (Marij Mooren)

4.1.3 Quality monitoring

The approach to quality monitoring can be separated into internal control and external control.

Quality improvement initiatives focus too much on quantitative data which can not completely capture all aspects of quality.

For both kinds of control, Judith Roesink has raised concerns over the completeness of used indicators:

I don’t know [whether this is good quality. The used indicators] are just several items that were found to have a relationship with quality. Of course. But if you have all these things done, it doesn’t mean you have good quality. It just gives a hint that you are on a good way." She attributes this development to a certain attitude of the people responsible for deciding on quality monitoring: “I think they want to score everything. They want quantitative data on things that are not quantitative! (Judith Roesink)

4.1.3.1 Internal control

Mistakes committee From the treatment point of view, the number of treat-ment mistakes is a measure for quality inside the departtreat-ment. In case of the ra-diotherapy department these are reported by the employees of the UMCU them-selves: “Everyone has the obligation to fill in a form, an electronic form about mistakes. We get about 300 to 400 a year on average. It can vary monthly." (Wil-fred Tulling)

Wilfred Tulling also asserts that the hiding of mistakes is not a problem inside the radiotherapy department:

I can’t say that there is a culture of not disclosing mistakes for any reason except for sometimes time. On average, I think we have quite a good view on the mistakes that are made in this department.

(18)

These mistakes are then investigated within the mistakes committee, a mul-tidisciplinary monthly meeting where “we look at all the mistakes and some improvements may evolve from that. (Wilfred Tulling)

The outcome of these mistakes varies greatly.

There are sometimes little things like getting people on the right track. Sometimes you see that people don’t know exactly what things involve, for example with position verification. [...] We check whether that is part of the procedure, that it is not right or written down, or is it skills and we need to give a few extra lessons to a few people. [...] Sometimes [our measures involve] cutting back on the amount of work that is done at a certain space. (Wilfred Tulling)

Committees like this exist in every department of the hospital. There is also a central committee.

[The central committee] looks into what we do with [the mistakes]. They have an overview over the mistakes we make. If something really bad happens (it doesn’t happen very often but it could happen any day) we have to report it to them immediately so that they can take the measures that are necessary to the central government. (Wilfred Tulling)

These can then include legal measures:

There is another committee in between the mistakes committee and the calamities committee and they are sort of the link to the outside. If there are legal measures to be taken for punishment, it is a safeguard for the people who have committed the mistake. (Wilfred Tulling)

Quality Manager In addition to the mistakes committee, there is also a Quality Manager exerting internal control. He is “maintaining the whole pro-cess". (Marij Mooren) She also states that the quality manager is responsible for ensuring the formal correctness of patient files:

[The Quality Manager] also checks the systems at the back-end of it when the patients are already through the system. I think every month [the quality manager] has to make a report and for that he also checks if everything is registered right, if the waiting codes are correct [and] if they have a physician assigned in the database.

Tumor group-based monitoring Finally, there is monitoring on a tumor group-basis.

We take some patients as example but in the system we look at accumulated measures for elderly people, or for persons from a certain hospital. You can compare the different referring hospitals and the differences, and then you can try to drill down to the cause. (Robert Tersteeg)

While the overall structure of the groups is the same, the actual standards differ according to each group:

Every tumor-group has its own evaluation of treatment. I am in a quality group and we have evaluations of ourselves about cancer

(19)

treatments according to a European group where all the cervical can-cer patients are put in and evaluated: Survival rate and that kind of measures. There are a lot of groups that have kind of the same: Every-body checks their treatment, but according to different national and international committees. (Judith Roesink)

4.1.3.2 External control In addition to the internal controls mentioned above, there are also external methods of quality monitoring.

Hospital accreditation Hospital accreditation is a quality control process. In the UMCU, the Joint Commission International Accreditation Standards for Hospitals (JCI) are used. They are “a quality control system to measure the qual-ity of the hospital in general" (Judith Roesink).

Hospital accreditation is not seen as a good representation of health care quality.

The radiotherapists stated that the hospital accreditation does not touch their work life directly. While Robert Tersteeg stated that the accreditation process is “not real intrin-sic quality for us, not what treatment is about. There is a gap between the evidence-based medicine that we try to achieve and the quality demands that we have to meet.", Iris van Dam muses “There are probably a lot of things that I don’t get to see or that I don’t notice all the time". Judith Roesink also deplores that there is a large number of items to be monitored by JCI, and that these are frequently changing.

On the other hand, Wilfred Tulling as Quality Manager sees that JCI “does have an impact" which is supported by Judith Roesink:

I think [hospital accreditation influence me] since JCI was intro-duced quite a lot. There are quite a lot of rules and things we have to commit to. There is still some place for our own division’s policy, but these are general rules that we need to follow. About how you deal with high-risk medication, about what you do with your waste..." (Judith Roesink)

Furthermore, Robert Tersteeg mentioned previously that the commission of the international accreditation process caused a project to centralize the hospitals guidelines.

Health inspection The second method of external control is health tion authority. Bad mistakes must be reported by the department to the inspec-tion of health.

The system is on principle blame-free, but only up to a certain level. If you don’t check the identity of patient according to the rules and by accident treat the wrong patient or the wrong part then you are responsible for that and you can be punished for that. If you fill your medicine pump and you didn’t do the medical calculation but you still [give the medicine], you are also responsible for your actions and you could be punished. (Wilfred Tulling)

(20)

4.2 Customer interaction

The second view taken on quality inside the radiotherapy department is cus-tomer interaction. The main measure of the quality of interaction is cuscus-tomer satisfaction as mentioned by the radiotherapists interviewed. Iris van Dam men-tions several aspects that in her opinion play into the satisfaction of patients, namely a limited number of caretakers (“[Quality from the patient side means] not too many faces. There should be one doctor, but also the same people treating the patient[during the all the treatment fractions]."), sufficient patient informa-tion (

Everything should be clear for the patient so that they don’t have to ask too many questions. Quality for the patient is good information from the doctor but also from the people on the unit and from the assistants behind the counter.

and the friendliness of UMCU employees (“I think [friendly] people are very im-portant for the patient").

Patient privacy is endangered by external organizations trying to acquire patient data without central coordination by the hospital.

Judith Roesink mentioned patient privacy as a currently growing concern:

All the external organizations who are asking us to give all different param-eters for registrations. We don’t really know what they are doing with all this information. If the government asks us to register quality and they want to know the registrations time and the number of patients. Do we treat enough patients to have enough expertise? That is OK for me, that is quality. But if there is not a government organization asking the same information, and numbers, and spe-cific questions about that patient and you don’t really know what these external organizations are going to do with this data, that is really difficult. Right now, there is a mess of organizations who are focusing on quality, and all with the best intent, but there is no coordination. And that is difficult. What do we really need? And what do they only need because of the name of their organization? What are they going to do with it? Are they going to discuss this with the gov-ernment, or with insurance people? It is really not clear what is going on in this world. There is no clear leading role on division level that states what needs to be done.

4.3 Department organization

The third and final view on quality mentioned by interviewees is the department organization. It is separated into process flow and information flow.

4.3.1 Process flow

The process flow inside the department is seen to impact overall quality by the interviewees.

(21)

Fragmentation of cases leads to lost time in remembering the case’s background information and increases the risk of errors in treatment.

Iris van Dam mentions that the temporal fragmentation of cases may lead to a de-crease in quality:

If you have less moments when you have to do something with the patient, when it is all combined in one day you don’t get to make the errors because you are focused on the patient. Right now you see the patient, then a few days later you don’t have the focus all the time.

Similarly, interruptions “affect the quality of your work". (Iris van Dam)

4.3.1.1 Waiting times However, the measure monitored by the quality im-provement initiatives is waiting times. Marij Mooring states that for her “[Qual-ity] is also about [...] having everything fit together so that you don’t have to wait any time between steps of the processes when it is not necessary for the medical side of it." This [metric] is looked at by the radiotherapists themselves (Robert Tersteeg) as well as by the Quality Manager who investigates the waiting times of two different groups:

What [the quality manager] mainly does is look at people which are classified as W3 patient. These are the basic patients where nothing is wrong, and they should fit into the normal time frame. He looks at the mean of the waiting times of those. He also looks at W8 patients, those who have different reasons for being delayed. So we know it could be medical reasons for why they could not fit into the time frame that we want, but it could also be something such as the patient wants to go on holiday. (Marij Mooring)

4.3.1.2 Process Improvement Project In order to ensure shorter waiting times a process improvement project has been initiated which consists of three working groups looking at different fragments of the process:

There were three working groups, really. One that looked at the palliation patients, we called them SPUT emergency patients. They have to be done on the same day. So we checked how we could op-timize that process that was one working group. The other working group was from the time they were known here until the first ment with the doctor and the third group was from the first appoint-ment with the doctor until the start of the treatappoint-ment. (Marij Mooring) The group implemented several quick fixes and currently aims at implementing the more fundamental changes. The project has become a constant part of the hospital’s quality framework. (Marij Mooring)

4.3.2 Information flow

Marij Mooring mentions that quality is “also about getting the knowledge to all the right people". This interacts with the process flow in that communication is a large part of resolving imperfect process flow as claimed by Robert Tersteeg: “[In order to handle imperfect process steps] I would look who is responsible for

(22)

Figure 3: ICT Infrastructure of the radiotherapy department according to interviews that part of the process and make sure he or she knows what is wrong with this aspect of his or her part of this process."

4.4 ICT Infrastructure

In addition to the quality framework itself, a number of ICT components were mentioned by the interviewees that enable the quality framework. The hierarchi-cal structure of remarks concerning the ICT infrastructure can be found in figure 3.

(23)

4.4.1 Data Collection

There is a conflict between the viewpoint of the radiotherapists and the process-focused UMCU employees concerning the availability of data. While the latter believe that all data is available, the former are lacking data they deem necessary for analysis.

The understanding of the gap between necessary and available data differed strongly be-tween Marij Mooring and Robert Tersteeg. From the process perspective, Marij Mooring concludes that

“from the first appointment on we actually keep track of everything. We have records of the days of CT scans, MRIs, the first radiation, the whole process, really." On the other hand, Robert Tersteeg feels that there is missing data from the treatment perspective:

There is data I would like to know more about that is very hard to get, for example complication control for tumor groups with a very short survival. You can’t look at it because the patient isn’t there anymore. If you look at intervals from different times there is a registration problem because I don’t have the data from the referring hospitals. Only 20% of our patients are from this hospital, the rest are from referring hospitals that don’t have radiation oncology so they send their patients to us. But the whole process before and after radiation is in another hospital and I am very limited with the data.

4.4.2 Data Integration

Another difficulty stated by Robert Tersteeg is that required data resides on dif-ferent ICT systems that are not integrated with one another:

The lack of data stated above includes both the collection of data and its integration.

I am missing] a lot [of data]. Especially days when patients are being seen, and there is now a project research IT that is now trying to get all the data from all the different systems into one integrated system to have a better look at all the quality aspects. (Robert Tersteeg)

In addition to this there is a temporary challenge with data integration due to a change in the ICT system: “Since one month we have a new IT system. Here we have a breast cancer patient at the end of March. In our old system we have the data from the referring hospital." (Robert Tersteeg)

4.4.3 IT Landscape

The UMCU’s radiotherapy department currently works with a number of differ-ent programs.

(24)

Usage of too many different IT systems that require radiotherapist interaction is causing annoyance with them.

This number of software necessary leads to vexation with the radiotherapists: So we work with eight different programs. I have seven inboxes to check everyday. I have to check my physical inbox, I have to check an inbox in this system, in that system, in that system: Seven inboxes. And if I forget one it has direct impact on the quality because a patient will not get noticed and will not get a letter to come into the department to start the treatment. (Robert Tersteeg) We also have to different agendas. I have, of course, my outlook agenda and I have my clinical agenda. When another colleague looks at my appointment in my clinical agenda, he sees an empty list, and he thinks he has lots of time to talk to me. But this is my other agenda which is very full. I can press a button here to get the appointments from my clinical agenda to the outlook agenda, but this is not public. I cannot put my outlook agenda into the clinical agenda because the secretary of the clinic has to find open spots and if I put my personal one into the clinical one, I could never see a patient again. It is all fully booked. But if you ask me whether I have time for a chat I always have to check two agendas. (Robert Tersteeg)

Most time is spent not thinking about what treatment to give, but for all the administrative stuff, to put things together and to make the EPD clear for someone who doesn’t know that patient, to present what is happening. (Judith Roesink)

So I scanned this mail which is weird, you should be able to send this e-mail to your own department. (Robert Tersteeg)

Iris van Dam also sees a challenge in being guided by ICT:

A resident has to look the solution up by himself and think why we treat the patient in a specific way. If you get it presented by an ICT system it could be that you stop thinking by yourself. If the system [stops working], no one will know what to do.

On the other hand, Marij Mooring believes in a positive effect of new ICT tools: “So today we implemented a new program on the back office. [...] This helps them, I hope, with getting the distribution more even."

4.4.3.1 Mosaiq Mosaiq is a system that “keeps track of all the time steps in the process [to ensure quality]". (Marij Mooring) With Mosaiq it is possible to make series of appointments, check the availability of material, give doses of radiation and send letter to patients. (Jan Stout)

Mosaiq also includes quality control for dose giving: “If you give 200 cGy today and tomorrow, suppose there is a crash during radiation, a power fail-ure. And you only delivered 150 cGy. The system will register that." (Jan Stout) Finally, Mosaiq also includes a simple version of a work flow manager called Quality checklist. They work with so-called task sets. "A task set is a number of tasks. In these task sets I can also say that there is a parameter ‘required date’. What is the due date from this item compared with the start of the treatment?

(25)

The start of the treatment is day 0. And we start counting back. On day -5 we need to complete plan stereo taxing." (Jan Stout)

Mosaiq’s work flow manager is not up to the department’s requirements.

Jan Stout concludes “It is not a bad system for making appointments, it is a good system for radiating. But everything could be working better with the correct work flow manager in the background."

4.4.3.2 Patient Follower System The patient follower system “follows the pa-tient every step of the way and gives the next responsible person in a color scheme, whether it was important to do it today, or whether it could wait until tomorrow." (Wilfred Tulling) The responsible person sees a location-dependent list of patient for which to-dos are open: “If you are at the linear accelerator you see a different list than if you are at the central input check." (Marij Mooring)

5 Discussion

5.1 Possible integration strategies

In the previous section, several areas were exposed which may have improve-ment potential:

• There is a large number of different quality processes and initiatives that are not connected enough.

• Compliance to guidelines is only partially supported through the current guideline format since radiotherapists often refer to memory instead. • The hospital-wide centralization of guidelines is hindered because of the

large number of partially conflicting guidelines.

• The national centralization of guidelines is hindered due to the slow up-date process of external guidelines that leads to the use of internal adapted guidelines.

• Hospital-accreditation is not seen as a good representation of health care quality.

• There is a conflict between the viewpoint of the radiotherapists and the process-focused UMCU employees concerning the availability of data. While the latter believe that all data is available, the former are lacking data they deem necessary for analysis. This includes both the collection of data and its integration.

• Quality improvement initiatives focus too much on quantitative data, which can not completely capture all aspects of quality.

• Patient privacy is endangered by external organizations trying to acquire patient data without central coordination by the hospital.

• Usage of too many different IT systems that require radiotherapist interac-tion is causing annoyance with them.

• Fragmentation of cases leads to lost time in remembering the case’s back-ground information and increases the risk of errors in treatment.

(26)

• Mosaiq’s work flow manager is not up to the department’s requirements. This section consequently describes different solutions that demonstrate improve-ment in at least one, possibly several of the areas of potential improveimprove-ment de-scribed above. They include “Translation of internal guidelines into work flow fragments", “Integration of decision support based on external guidelines", “Pro-cess mining to prove guidelines compliance", “Production of additional data through the Case Management System" and “Integration of additional data sources". For each solution, the relation to areas of improvement potential and remarks on a potential implementation are stated.

5.1.1 Translation of internal guidelines into work flow fragments

One possible integration strategy would be to translate the department’s inter-nal guidelines into computer executable guidelines that will be used to make treatment recommendations to the user of the Case Management System at the appropriate time.

5.1.1.1 Relation to areas of improvement potential By offering the recom-mended treatment option to the radiotherapist the system would ensure that any deviation from the guideline was intentional. It could also provide a field were the treating doctor has to motivate why they deviated from the recommended option.

Such an integration could also be beneficial to the hospital-wide centraliza-tion in case of acceptance there since the resulting work flow fragments would be much easier to compare to one another than text-based guidelines.

Both should raise the quality of care in terms of treatment mistakes.

It can also be expected that there will be a slight increase in efficiency when the radiotherapists do not have to skim through the long text-based guidelines, however, this effect is most likely rather small due to the memory recall that is currently happening often.

5.1.1.2 Implementation The process of translating all relevant guidelines would be tedious and long. The documents in question are often-times hundreds of pages long and must be seen through and translated carefully, not too mention the necessary consolidation with other departments. Lyng (2010) states the fol-lowing:

Computerization of [Clinical Practice Guidelines] for application in clinical practice is a complex job. [...] An important issue [...] is that a wide array of functionality and features are applied in the transforma-tion. Cooperation across a wide range of research areas will therefore be required.

5.1.2 Integration of decision support based on external guidelines

The IKNL is currently working on providing a decision support system that al-lows to get treatment decisions based on the external guidelines translated into flowcharts. In the envisioned system, the hospital sends the patient data re-quired to the IKNL, they then respond with the recommended treatment. The IKNL imagines that at each point, the radiotherapist making use of their DSS system will have to decide at which point in the treatment process they currently

(27)

are. Instead, it would be another framework integration strategy to connect the steps from the Case Management System to certain decision points from the DSS. Should such a system be implemented either by the IKNL or by another provider of external guidelines, the system would exchange its data at that moment with the DSS of the external provider, and the recommended treatment option would be highlighted in the radiotherapy’s CMS. If the doctor does not wish to comply with this recommended treatment option, they would be asked by the system to motivate why. This option is important in order to counter the fear of losing au-tonomy as a professional as well as for the sake of the patients whose situations are not fully covered by the guidelines.

5.1.2.1 Relation to areas of improvement potential This integration strategy entails the possibility to increase the standard of care, nation-wide in the long run. This is because the focus on external guidelines instead of internal guide-lines may lead to radiotherapy departments everywhere sharing there newly dis-covered knowledge on cancer treatment with other hospitals by proposing these changes to the external provider instead of implementing them locally. Also, this solution would produce evidence for the guideline compliance inside of the hospital. However, this strategy could have a negative influence on the privacy challenge. Since the provider of the guidelines would require patient data in or-der to come up with a treatment recommendation, some data would have to be exchanged.

5.1.2.2 Implementation Such a solution would require a high level of coordi-nation with the external provider. In order to be up-to-date, they need to have a quick process of discussing changes proposed by radiotherapy departments. This could prove to be difficult due to the different specializations and equip-ment in hospitals. To reach a standard that is acceptable to all involved hospitals in a moderate time frame seems very complex.

However, should such an acceptable system by the external provider be taken for granted by the department, implementing the strategy would become fairly simply. It would rely on one protocol for message-based exchange between the hospital and the external provider.

5.1.3 Process mining to prove guidelines compliance

Instead of integrating with an external provider directly, guideline compliance could also be encouraged by assessment through process mining methods. Here, integration with both internal and external guidelines would be possible.

5.1.3.1 Relation to areas of improvement potential By establishing a system that can check for guideline compliance the radiotherapists would be given an incentive to stick to the guidelines as closely as possible.

Other than with the previous strategies, this does not yield a treatment rec-ommendation.

5.1.3.2 Implementation The guidelines in question are not fixed business pro-cesses to which compliance can be proven comparatively easily. Rather, they are a large number of recommendations for specific situations. There is a number of possible guideline-compliant patient pathways that is too large to model in a

(28)

static business process. An applied process mining tool would have to be able to handle this kind of flexibility.

5.1.4 Production of additional data through the Case Management System

Additional data could be generated through the Case Management System, for example by the use of detailed forms for the registration of patients or through making the steps in the case management system to be of a very fine granularity in order to get more data on process steps.

5.1.4.1 Relation to areas of improvement potential The currently unspecified need for more data could be resolved.

Should a high granularity of process steps be necessary this would require a tedious use of the system. Secondly, in the long run it may lead to longer registration times. Both could potentially be a risk to user acceptance. These drawbacks must be weighed against the benefits of improving data availability.

5.1.4.2 Implementation Currently, the need for data is not defined precisely. In order to understand what additional data is actually required inside the de-partment, some analysis would be necessary before the production of additional data. Depending on which data needs are actually uncovered, the complexity of the project will change. Should the import be required to be more detailed, this will first require a change to the current patient data system Dataline.

5.1.5 Integration of additional data sources

In order to solve the lack of data integration, the information of several sources could be combined in one of the systems. This could be either different IT sys-tems in the hospital or even the department, or syssys-tems from other hospitals.

5.1.5.1 Relation to areas of improvement potential The currently unspecified need for more data integration could be resolved.

However, this strategy also includes a risk if global data consistency cannot be maintained. Actions that are based on outdated or erroneous data could affect patient safety negatively.

5.1.5.2 Implementation Similar to the production of more data, this would first require some initial research on what data is actually required by the radio-therapists that the current system cannot provide them with. The complexity of the project will vary according to the results of the initial analysis of data needs. The more systems will need to be integrated, the more expensive the implemen-tation project will get. The integration will also increase the software mainte-nance complexity since the integration of the systems will from then on have to be taken into account when updating either of the software products.

5.2 Evaluation of integration strategies

The focus group had been asked to perform a ranking of the previously de-scribed integration strategies both according to benefits and according to com-plexity. The resulting ranking can be found in tables 4 and 5. The first rank in the table concerning benefits implies that the strategy is expected by the focus group

(29)

Table 4: Ranking of strategies according to benefits (by focus group)

Strategy Rank Reasoning Translation of internal guidelines

into work flow fragments

1 • rule out guideline misinterpretation • less time spent looking guidelines up

• risk of decreased thinking for treatment decision Process mining to prove

guide-lines compliance

2 • will require the production of "proper guidelines" (Robert Tersleeg) • The current number of guideline deviations is taken from the treating

doctor’s statement. Integration of additional data

sources

3 • missing data is mainly concerning imaging and follow-up

• "For the data I know I’m missing integration would be very nice." (Robert Tersleeg)

• this data could also be misinterpreted since it is not following the require-ments of a medical study

Production of additional data through the Case Management System

4 • "For study purposes of course there is a gain." (Robert Tersleeg)

• this data could also be misinterpreted since it is not following the require-ments of a medical study

• too many questions may impact user acceptance because of boredom and privacy concerns

Integration of decision support based on external guideline

5 • guidelines are evidence-based and discussed in a group,

leading to compromises that the hospitals still adapt for themselves • reduces opportunities for specialization of academic centers • only useful on a shallow, high level

• "By the time you reach agreement nation-wide the protocol is already too old." (Robert Tersleeg)

Table 5: Ranking of strategies according to complexity (by focus group)

Strategy Rank Reasoning Production of additional data

through the Case Management System

1 • rather simple to implement since the Case Management System is only local.

Integration of decision support based on external guideline

2 • if perfect system is taken for granted, easy to implement

• if building such a system democratically is also considered, discussions would be very complex

• possible if guidelines are imposed by specialized centers instead. Would, however, lead to acceptance problems.

Translation of internal guidelines into work flow fragments

3 • depends on percentage of non-doable cases that is accepted • fairly complex if all exceptions should be built in

• requires high amount of discussion Process mining to prove

guide-lines compliance

4 • currently no experience in data mining: new software development

Integration of additional data sources

5 • requires a lot of communication with other hospitals, etc.

to provide most benefits. Respectively, a first rank in the table concerning com-plexity implies that the strategy is evaluated to be least complex to implement.

These rankings are also displayed in figure 4. In the graphic, two groups of strategies can be clearly discerned: The strategies “Translation of internal guide-lines into work flow fragments", “Process mining to prove guideline compliance" and “Integration of additional data sources" are all higher in benefits but also higher in complexity. On the other hand, “Production of additional data through the Case Management System" and “Integration of decision support based on external guidelines" are lower in benefits and lower in complexity.

While the focus on benefits or on complexity depends strongly on the circum-stances, inside the groups there is a uncontroversial sequence of suitability. For the group of high complexity and high benefits, this suitabality is

1. Translation of internal guidelines into work flow fragments 2. Process mining to prove guideline compliance

3. Integration of additional data sources

Embedding a health care Case Management System into its surrounding quality framework

UNIVERSITY OF AMSTERDAM

Faculty of Science

Thesis Master Information Science

Business Information Systems

Embedding a health care Case Management System

into its surrounding quality framework

Abstract

Contents

List of Figures

List of Tables

1

Introduction

2

State of knowledge on hospital quality frameworks

3

Methodology

4

Results

5

Discussion