TRIAGING PATIENT COMPLEXITY

(1)

1

TRIAGING PATIENT COMPLEXITY

A case study at University Medical Center Groningen - Gastroenterology department

MSc. Technology & Operations Management

University of Groningen, Faculty of Economics and Business

June 22, 2015

Author: H. Wijnsma

Student number: S2580713 E-mail: h.wijnsma.1@student.rug.nl

Supervisor: University of Groningen Dr. ir. D.J. van der Zee

Co-assessor: University of Groningen Dr. T. van der Vaart

(2)

2

ABSTRACT

Purpose – Health insurers have a significant influence on how many and which type of

patients should be treated in an Academic Health Center (AHC). Although, the medical policy plan of the UMCG-MDL prescribes that focus is on high complex care, the existing triage system is insufficient in doing so. Therefore, the objective of this study is to design a triage system for AHC-MDLs which minimizes LC patients entries and existing LC patients in the system by diverting these patients to GPs or community hospitals.

Method – In order to address the research objective, a design science approach is used in

which the gastroenterology department at the University Medical Center Groningen serves as a case example. First literature is reviewed in order to identify insights in elements affecting a triage system. Next, the UMCG-MDL triage system is described, analyzed and redesign propositions are formulated with the use of interviews, meetings and documentation analysis.

Findings – The UMCG-MDL triage system practices results in 35,3% low complex patients

of which the majority can appointed to M. Crohn, Idiopatic proctocolitis and irritable bowel syndrome diseases. Root cause analysis identified three main causes which are responsible for the current performance, subsequently, lack of interaction and cooperation between GPs and MDL-physicians, insufficient admission procedures and insufficient discharge procedures. Opportunities for triaging patient complexity are identified based on the defined causes, use of literature and input from domain experts. The opportunities chain management, patient admission and discharge procedures should be used in order to improve the UMCG-MDL triage system performance and to triage patient complexity. The opportunities are evaluated for applicability by defining barriers for implementation. The patient discharge procedure is selected for the UMCG-MDL triage system redesign. The patient discharge procedure contains four elements which are the discharge planning, discharge criteria, discharge process and discharge letter.

Conclusions –A patient discharge procedure is an adequate first step in redesigning

AHC-triage systems which triaging patient complexity. Results of the pilot group shows that 633 patients (27,5%) of the LC patient population can be discharged to their initial referrer.

Keywords – Triage, patient complexity, triage system, gastroenterology, referral process,

(3)

3

PREFACE

This master thesis is the final project for my Master of Science in Technology and Operations Management at the University of Groningen. The goal of this thesis was to define proposals for a triage system which is triaging on patient complexity. Insights on triage systems design were obtained by doing a case study within the UMCG-MDL department.

Performing a research based on practice has been a pleasant experience. I would like to thank Mr. Borgers for his guidance and valuable insights on the complexity of hospital practice. Also, I would like to thank Mr. Kleibeuker, Mr. Dijkstra, Mr. Talsma and Mr. van Waning Bolt for their time, insights and cooperation.

In addition, I would like to thank my supervisor Mr. Van der Zee for his experience and guidance for writing a thesis and setting up a research. I appreciated the two-weekly feedback session with Mr. van der Zee en my fellow students. During these sessions I received many valuable insights and ideas about doing my research. Finally, I would like to thank my co-assessor Mr. van der Vaart for his feedback on the research proposal.

Groningen, June 2015

(6)

6

1 INTRODUCTION

Academic Health Centers (AHCs) have a primary responsibility for maintaining and developing the scientific basis of clinical medicine and are thus widely viewed as setting the standard for high-quality care (Fisher, Wennberg, Stukel, & Gottlieb, 2004). AHCs patient care costs are generally higher than those of nonteaching community hospitals due their subspecialty orientation and highly specialized service capacity (Mechanic, Coleman, & Dobson, 1998). Dutch health insurers adequately exploited this issue for AHCs by forcing them to focus more on treating high-complex (HC) patients by limited funding of low-complex (LC) care in the insurance policy 1 (Baars & Kooreman, 2014). Whereas, the perspective of low and high complexity means the severity of a disease.

The limited funding of LC care implies that AHCs should be able to identify LC patients and be capable of diverting these patients towards alternative caregivers, i.e. General Practitioners (GP) or community hospitals to avoid significant financial losses. This means that AHCs may have to reorganize their current triage system.

Current literature on triage system design and associated procedures is mainly focused in the context of the Emergency Department (ED). Fitzgerald, Jelinek, Scott, & Gerdtz (2010) argue that there are two main purposes of triage: “(1) To ensure that the patient receives the level of quality of care appropriate to clinical need (clinical justice) and (2) that the departmental resources are most usefully applied (efficiency) to this end”. Most commonly used ED triage systems around the world are Canadian Triage Acuity Scale (CTAS), Manchester Triage Scale (MTS) and the Australian Triage Scale (ATS) (Wulp, 2010). These systems focus on an urgency-based classification for patient streaming and therefore focus on the first purpose (clinical justice) of triage.

However, the second purpose of triage (efficiency) is largely overlooked in current triage systems. The Emergency Severity Index (ESI) opts to include the second purpose, although Saghafian et al. (2011) argue that in practice these patients are still sorted and prioritized on the basis of urgency. Road (2011), argues that current literature is insufficient in addressing patient complexity in triage and opts to define principles to guide assessment of patient complexity. These principles should complement to triage process and support decision-making and streaming of patients at triage. Also, Wuerz, Milne, Eitel, Travers, & Gilboy

1

(7)

7 (2000) state that existing ED triage methods are deficient, particularly for identification of non urgent patients who may be referred for care elsewhere.

(8)

8

2 RESEARCH OBJECTIVE AND DESIGN

This chapter clarifies research objective and it’s setup. Section 2.1 describes the problem background, section 2.2 introduces the objective of this study, section 2.3 describes the research design and section 2.4 describes the information resources.

2.1 PROBLEM BACKGROUND

The problem that AHCs faces is that health insurers want to exclude reimbursement for LC patients. Health insurers use triage data when reviewing the “medical necessity” of services (Wuerz et al., 2000). AHCs patient care costs are generally higher than those of nonteaching community hospitals due to their subspecialty orientation and highly specialized service capacity (Fisher et al., 2004). Also, experts agree that going to a GP for minor treatment rather than a hospital can be much cheaper 2. Therefore, health insurers opt that LC patients are treated at GPs or in community hospitals, whereas HC patients should be treated in AHCs. Health insurers fund hospitals by considering the number of patients which can be treated for a set of diseases. This implies that health insurers have a significant influence on how many and which type of patients should be treated in which hospital. Because AHC are more expensive, health insurers only fund a small amount of low complex treatment as this may be affected by less expensive community hospitals or GPs. This has a considerable financial impact on the UMCG-MDL department such that financial rewarding is excluded when the purchased amount of treatment, is reached and surpassed. The medical policy plan of the UMCG prescribes that the UMCG-MDL department should focus on complex care (Jaarplan, 2015). However, current practices on the UMCG-MDL department do not focus on sorting, filtering or excluding referred patients which entails that every referred patient is accepted for care also when the patient limit is reached and surpassed. Therefore, the UMCG-MDL department is representative for other departments within the UMCG and academic hospitals in general.

2.2 RESEARCH OBJECTIVE

To date, there is little known on how to define a triage system which structurally and consistent diverts specific MDL-patients towards alternative caregivers. The definition of a triage system used in this research is depicted in Fig. 1 and described in section 2.3. By using

2

(9)

9 such a triage system, the UMCG-MDL has a certain degree of control over the in- and outflow of MDL-patients such that the needs and interests of the UMCG-MDL and health insurers can be balanced. Therefore, the following objective is defined:

“Design a triage system for AHC-MDLs which minimizes LC patients entries and existing LC patients in the system by diverting these patients to GPs or community hospitals.”

2.3 RESEARCH DESIGN

The structure of elements of triage systems in a Dutch setting will be briefly introduced and depicted in a conceptual model (Fig. 1).

The foundation of a triage system design are three decision phases in which the patient flows. The GP or physician of a community hospital assesses patients complexity in the first decision phase. In principal, HC patients are referred to the UMCG and LC patients are treated by the GP or physician of a community hospital. The physician in the AHC-MDL department again checks the patients complexity in the second decision phase. Also, referrals from other UMCG departments towards UMCG-MDL proceed in this assessment. HC patients are accepted for treatment and LC patients are diverted towards the GP or community hospital. The second decision phase is required due to limited knowledge and/or diagnostic equipment in the Primary Care Provider (PCP) setting. The accepted HC patients undergo treatment and the physician assesses if patients are stabilized till LC, such that they can be discharged to the GP or community hospital. HC patients continue to receive treatment.

Low complexity patient? No Low complexity patient? Low complexity patient? AHC-MDL No No GP or community hospital AHC-department Yes Patient treatment Patient treatment Yes In-process Patient Yes Patient

(10)

10 In order to address the research objective a design-science approach is used in this study. The objective is to propose designs of an artifact: a triage system as described in section 2.2. Table 1 shows the steps of the research design. Chapter 3 consists of a literature review on triage design which implies identifying insights in the triage process and on triage systems. The case-specific chapters 4, 5 and 6 are defined according to the steps taken in the regulative cycle of Van Strien (1997). In these chapters the triage system of the UMCG-MDL is described, analyzes and suggestions for redesign are proposed. Chapter 7 contains the discussion and limitations of the research, whereas the implications of the results for AHC-MDLs are described. Chapter 8 contains conclusion of the research.

Table 1 Research design

Steps Chapter Research question

1. Literature review

3 Literature review on triage system design. Identify insights on elements affecting a triage system.

2. System description

4 What are the main elements of the system under study and how is the triage system set up?

3. Analysis 5 How does the current triage system perform and what causes this performance?

4. Redesign 6  What opportunities for triaging patient complexity can be identified?

 How should the opportunities for triaging patient complexity be addressed?

Discussion 7 Formulate the main findings and limitations of this study.

Conclusion 8 Formulate the summary of main findings and determine directions for future research.

Step 1: Literature review

(11)

11

Step 2: System description

The second step of the research entails defining and describing the triage system of the UMCG-MDL department. The triage system is divided into three decision phases shown in the conceptual model (Fig. 1) and described accordingly. Therefore, the conceptual model functions as an overlaying framework to assure a comprehensive description of the triage system.

Step 3: System analysis

The third step entails analyzing the triage system of the UMCG-MDL department. First, the performance of the UMCG-MDL triage system is identified. Secondly, causes of the identified performance is explored by observations. Finally, in-depth interviews are used in order to determine causes and literature is used to obtain an detailed understanding of the causes.

Step 4: System redesign

The fourth step starts with identifying opportunities for triaging patient complexity based on the identified root causes in step 3. Next, the opportunities are evaluated and the most preferred option(s) is determined. The preferred opportunity is elaborated and a pilot group is set up which functions as a small scale test. The opportunity is designed for the pilot group and performance improvement are determined accordingly. Concluding, barriers for implementation are defined which should be taken into account.

2.4 INFORMATION SOURCES

(12)

12

3 TRIAGE SYSTEM DESIGN – A LITERATURE REVIEW

This chapter consists of a literature review on triage system design. Literature is found making use of the following databases; PubMed, Google scholar and Health Affairs. In the collection of relevant literature, the following question has been taken into account; What do we need to know before we can triage patient complexity? A comprehensive review of literature is obtained by embracing two perspectives of triage system design, a chain and system level perspective. Section 3.1 entails the chain level perspective by describing the triage system (Fig. 1) linking care services activities along the chain. Section 3.2 zooms in on the structure and elements of individual ED triage systems. Subsequent, the implications of both sections on triaging patient complexity is evaluated. This chapter concludes with a summary of main findings in section 3.3.

3.1 TRIAGE SYSTEM

This section evaluates literature concerning triage which consists of the history and definition and the process of triage.

3.1.1 Definition

The term "triage" is derived from the French word trier, which literally means ‘to sort’. The origin of triage and the first triage system dates back to around 1792, by Baron Dominique Jean Larrey, Surgeon in Chief Napoleons Imperial Guard (Robertson-Steel, 2006). Larrey recognized a need to evaluate and categorize wounded soldiers promptly during battle. His system was to treat and evacuate those requiring the most urgent medical attention, rather than waiting hours or days for the battle to end before treating patients, as had been done in previous wars (Iserson & Moskop, 2007). Triage can be defined as “The process used by clinicians to rapidly prioritize a person’s need for treatment based on symptoms and chief complaints” (Fitzgerald 2000; Zimmermann 2001; NVSHV, 2004). The principal purpose of triage is to ensure that the patient receives the level and quality of care appropriate to clinical need which entails clinical justice and that departmental resources are most usefully applied to this end which entails efficiency (Fitzgerald et al., 2010). Clinical justice and as well efficiency ensures that a patient receives care appropriate to need and in a timely fashion.

3.1.2 Decision phases

(13)

13

First decision phase. The primary care provider (PCP) is the stakeholder in the first decision

phase which are GPs and physicians in community hospitals. Coulter, Noone, & Goldacre (1989) identified main reasons of PCPs for making a referral which are (with percentages, where available, from (A. Bowling & Redfern, 2000);

 To establish the diagnosis (52%).  For a specified investigation (48%).  For treatment or an operation (33%).

 For advice on management and referral back (32%).

 For a second opinion to reassure them (the general practitioners) that they had done all that was required (17%).

 For a second opinion to reassure patients or their families that they (the general practitioners) had done all that was required (7%).

 For other reasons (11%).

Piterman & Koritsas (2005) analyzed expectations in the GP-specialist referral process and defined that specialists expect the GP to provide information about the problem to be addressed and adequate patient history. Results from a study of Tejal K. Gandhi, MD, MPH, Nancy L. Keating, MD, MPH, Matthew Ditmore, David Kiernan, Robin Johnson, Elisabeth Burdick, MS, and Claus Hamann, MD (2008) shows that about half the time, the reason for dissatisfaction is a delayed or missing referral letter and reports which complies with the defined expectation of Piterman and Koritsas (2005). Other reasons include missing information in the referral communication, time required to write a referral note, difficulty in finding a specialist (Cummins, Smith & Inui, 1980; Lee, Pappius & Goldman, 1983; Bourguet, Gilchrist & McCord 1998; Gandhi et al., 2008).

(14)

14 practitioners in using minor procedures, direct booking to inpatient or day case care for some common operations (Coulter et al., 1989).

Second decision phase. The physician of an AHC is stakeholder in the second decision phase

which has to evaluate the referral letters and decide to accept or refuse a patient for treatment. Meaningful assessment of referral performance requires a measure of the quality of the clinical referral, including the appropriateness of the referral decision and an assessment of the information contained in the referral letter (Jenkins, 1993).

A study from Jenkins (1993) found that specialists judged that the percentage of referral letters containing errors or omissions ranged from 5.4% to 28.2%, according to the information category. In 22.8% of referrals preliminary investigations that specialists would have expected to have been carried out before referral were not mentioned in the referral letter and this points to either a general under use of the investigations available to general practitioners before referral, or an under-reporting of potentially useful information in the referral letter (Jenkins, 1993). Also, the percentage of information items that were incorrect on letters for referrals that were assessed as possibly or clearly inappropriate was almost three times higher, than on letter for referrals assessed as appropriate. Jenkins (1993) suggested to develop a standard referral letter which includes the reasons and objectives of the referral such to improve the appropriateness of the referral decision, by training general practitioners to think through the exact reasons for referral, the questions they want answered and what they want the patient to gain from the referral.

Third decision phase. The physician of an AHC is also stakeholder in the third decision

phase in which discharging patients for treatment is the main activity. The Active Discharge from Specialists Outpatient Services (ADSOS) Guideline developed by the Health system performance (2012), gives a proper definition of the organization and meaning of the third decision phase. This entails ensuring that timely discharge from specialists outpatient services and return to primary health care providers occurs for patients whose episode of care is complete. Timely discharge of patients back to their PCP is critical to ensuring appropriate use of specialists clinics’ services, streamlining patient flow, and increasing the capacity of specialist clinics to treat new patients (Performance Health System, 2012). The following principles are defined by the ADSOS guideline;

(15)

15  Strengthening linkages with GPs as the primary carer and central to the outpatient

episode and following discharge will be communicated to patients.

 Streamlining the patient journey by introducing effective discharge policies, protocols and practices which are sensitive to the needs of patients.

 Monitoring discharge practices including discharge rates and related aspects of service demand and capacity.

 Using the workforce effectively to support discharge practices.

Burkey et al. (1997) defined patient discharge criteria which are clinical criteria and take the form of the following three questions;

1. Is the exploration and formulation of the patient’s problem(s) complete? [Yes] 2. Is the treatment regime (if any) stable? [Yes]

3. Does the patient’s condition/problem require continuing hospital follow-up? [No]

3.2 ED TRIAGE SYSTEMS

This section identifies and evaluates ED triage systems. Eitel (2003) defined an ED triage system as: “The preliminary screening step that sorts patients at ED presentation into urgency categories to prioritize patients for evaluation and treatment” (Eitel, 2003). Section 3.2.1 identifies most used ED triage systems. Section 3.2.2 evaluates the identified ED triage systems and determines implications for triaging patient complexity.

3.2.1 Systems

This paragraph defines and described ED triage systems.

Manchester Triage System (MTS), Canadian Triage and Acuity Score (CTAS), Australian Triage Scale (ATS). The objective of these triage systems is relative similar and

(16)

16 categories. The systems can all be characterized as urgency-based triage systems which solely focus on clinical justice purpose of triage.

Emergency Severity Index (ESI). The ESI is a triage system which stratifies patients in

categories based on acuity (Wuerz et al., 2000). The ESI is a five-level triage instrument that attempts to predict “not only when should this patient be seen, but also what does this patient need?” (Elshove-Bolk, Mencl, van Rijswijck, Simons, & van Vugt, 2007). Patients are stratified with a flowchart-based algorithm (Fig.3) into five categories, ESI-1 being the most unstable, urgent and resource intensive, and ESI-5 being the least (Table 4). The ESI does not define time intervals for review by a physician but provides a method for categorizing patients by both acuity and resource needs, which is a unique feature of this tool (Wulp, 2010).

3.2.2 Evaluation

This paragraph contains an evaluation of the ED triage systems defined in section 3.2.1. First, evaluation criteria are defined followed by the evaluation of the identified triage systems.

Criteria. Healthcare systems are characterized by interrelated components of input, structure,

processes and outputs (Busse and Wismar, 2002; Belcon & Ahmed, 2009). Although, healthcare systems share certain common features, the design is country-specific. This constitutes the first criteria for evaluation, Applicability in the Netherlands. Christ, Grossmann, Winter, Bingisser, & Platz (2010) defined general triage system evaluation criteria which are Validity and Reliability. Validity entails that the triage system should function to which it is designed for whereas reliability entails how univocal the system is working in various circumstances and how consistent the procedures can be interpreted. The fourth evaluation criteria is applicability of including patient complexity considerations. Systems are evaluated to which extent the severity of the medical condition of a patient can be included in sorting patients. Summarizing the following criteria are used to evaluate the triage systems;

1. Applicability in the Netherlands 2. Validity and reliability

3. Patient complexity considerations

(17)

17 NVSHV, (2004) performed a questionnaire to all the ED in the Netherlands. Results from this questionnaire shows that the triage system; Manchester Triage System (MTS) is been used 55,3% by the respondents, the Emergency Severity Index (ESI) 7,1%, an own triage system 5,9% and 31,8% does not uses a triage system. The MTS, CTAS, ATS are similar systems and therefore grouped and evaluated next, followed by the ESI.

MTS, CTAS and ATS. Validity: A study of Speake et al. (2003) evaluated the systems on

validity. This study aimed at sensitivity and specificity with regards to symptoms of pain on the chest. The systems scored 86.8% on sensitivity and 72.4% on specificity which implies that these are valid systems. Reliability: A study of Storm et al (2007) evaluated the systems on reliability in the Academic Medical Centrum (AMC). This study showed that these systems shows an excellent level of intra and inter rater reliability. Another study by Cooke et al (1999) evaluated the systems and defined that this system shows proper levels of reliability with respect to critically ill patients. Patient complexity considerations: The urgency-based triage systems aim to determine maximum waiting times for patients to be treated. The criteria do not include the severity of the medical condition of the patient. Therefore, the urgency-based triage systems do not show beneficial implications for triaging patient complexity.

ESI. Validity: Studies of (Wuerz et al., 2000; Wuerz et al., 2001; Eitel et al., 2003; Tanabe

(18)

(Elshove-18 Bolk et al., 2007). The ESI system does show elements which might be beneficial in triaging patient complexity.

3.3 SUMMARY OF MAIN FINDINGS

The GP or physician of a community hospital are stakeholder in the first decision phase. The main role of this stakeholder is referring patients to specialists in hospitals from a variety of reason which ranges from unclear diagnosis till known diagnosis and second opinions. Current practices indicate that collaboration between GPs and physicians of a community hospital, and specialists could be improved by developing protocols in which the former takes over more care. AHC physicians are stakeholders is the second decision phase. The main role of the AHC physician is evaluating the referral letters and decide whether to accept or refuse the referred patient. This entails assessing the quality of the clinical referral subsequently the appropriateness of the referral decision and completeness of information contained in the referral letter. The AHC physician is also stakeholder in the third decision phase. The main role is discharging patients for treatment. The ADSOS guideline prescribes a set of principles to assure a timely discharge of patients back to their PCPs to ensure appropriate use of specialists clinics’ services, streamlining patient flow, and increasing the capacity of specialists clinics to treat new patients. Burkey et al. (1997) identified a number of barriers for this discharge as perceived by physicians which range from uncertainties about GP´s ability to provide care to specialists’ perceptions of their responsibilities. Concluding, Burkey et al. (1997) defined patient discharge criteria which consists of three predefined questions.

(19)

19

4 TRIAGE SYSTEM DESCRIPTION

This chapter describes the current triage system of the UMCG-MDL department. In order to define a comprehensive triage system description, the following question has been taken into account;

 What are the main elements of the system under study and how is the triage system set up?

This is achieved by interviewing domain experts of each decision phase of the triage system and informal meetings with the UMCG-MDL manager. Section 4.1 contains an overview of the UMCG-MDL triage system whereas section 4.2, 4,3 and 4,4 describe the system in more detail by the first, second and third decision phase.

4.1 UMCG-MDL TRIAGE SYSTEM

Figure 2 shows the UMCG-MDL triage system which is structured in three decision phases.

Figure 2 UMCG-MDL triage system

(20)

20 accepts the patient for diagnosis. Otherwise, the MDL-physician calls the referrer to obtain additional information. When patient’s received treatment, the MDL-physician assesses if continuing treatment is required and subsequently refers the patient back to the PCP or community hospital.

4.2 FIRST DECISION PHASE

Figure 3 shows the first decision phase in the triage process.

Figure 3 Detailed system description - first decision phase

The first decision phase starts by the patient approaching the PCP. The patient either calls the clinic and makes and appointment by the PCP or enters the clinic directly. The PCP assesses the patient’s complaints and decides if the patient needs hospital referral for advice, diagnosis or treatment. If so, the PCP makes a referral letter which contains the following information;

 Reason of referral.

 Anamnesis, examinations, diagnosis and established treatments.  Past relevant problems of the patient.

 Topical medication.

 Request of referrer which consists of the following categories; - Advice and referral back.

- Takeover of treatment.

- Expected disease, request for diagnosis and referral back. - Request for diagnosis based on additional patient complaints. - Others.

(21)

21 the patient should be referred to. This depends on two criteria which are travel distance to the hospital and admission times. The PCP selects a hospital which has the minimum travel distance and the lowest admission time. Next, the PCP writes a referral letter containing the patient referral information and send the letter to the medical administration department in the UMCG-MDL by using “Zorgdomein”. Zorgdomein is an electronic system which is used to interchange referral letters between care givers. If the patient does not require hospital referral, the PCP either discharges the patient or performs treatment and/or prescribes medication.

4.3 SECOND DECISION PHASE

Figure 4 shows the second decision phase in the triage process.

Figure 4 BPMN Detailed system description - second decision phase

(22)

22 special endoscopies or unknown are also accepted for treatment in the UMCG-MDL. The remaining of patients is discharged to their initial referrer.

4.4 THIRD DECISION PHASE

Figure 5 shows the third decision phase of the triage process.

Figure 5 BPMN Detailed system description - third decision phase

(23)

23

5 TRIAGE SYSTEM ANALYSIS

This chapter contains an analysis of the UMCG-MDL triage system as described in chapter 4. Section 5.1 describes the approach of this chapter. Section 5.2 defines performance of the triage system. Section 5.3 explores causes of the current performance and in section 5.4 detailed causes are defined. This chapter concludes with a summary of main findings shown in section 5.5.

5.1 APPROACH

In order to give a structured analysis of the UMCG-MDL triage system on triaging patient complexity, the following question have been taken into account:

 How does the current triage system perform and what causes this performance?

The first step in analyzing the triage system is obtaining a view of the performance in terms of the number of LC patients in the UMCG-MDL department. The performance is determined based on documentation analysis of patient history information, subsequent the Gupta file (2013). The second step is exploring causes of the identified performance. The stakeholders roles, tasks, decision making on referrals and supportive information systems are considered for their contribution in diverting LC patients. Observations and informal meetings are performed to gather information. The third step consists of an in-depth analysis by detailing the identified causes. A cause and effect diagram is used in order to obtain a proper understanding of the root causes of accepting and performing low complex care. Therefore, in-depth interviews with two GPs and two MDL-physicians, shown in appendix B-D, are performed which assures a comprehensive coverage of the conceptual model (Fig. 1).

5.2 PERFORMANCE

This section defines and describes the performance of the UMCG-MDL triage system. The perspective of performance employed in this study is the number of LC patients which receive treatment in the UMCG-MDL department. First, the share of LC patients in the UMCG-MDL patient population is determined followed by an understanding of which sets of diseases have a large share in the LC patient population.

(24)

24

Figure 6 UMCG-MDL patient complexity. Guptafile (2013)

Figure 7 shows a pareto chart of the LC patient population of the UMCG-MDL department. A pareto chart is used to graphically summarize and display the relative importance of the difference between groups of data 3. This figure show that a small set of diseases have a large share in the LC patient population of the UMCG-MDL department.

Figure 7 Pareto chart LC patient population. Guptafile (2013)

The LC patient population depicted in figure 7 is analyzed in more detail in order to determine the most important diseases, i.e. the diseases which have the largest share in the LC patient population. Figure 8 shows the LC patients per disease. This figure shows that three of 115 diseases in total have a substantial share in the LC patient population. This contains the diseases; M. Crohn, Ulcerative colitis and the Irritable bowel syndrome which represent 45,1% patients in the LC patient population.

3

http://www.isixsigma.com/tools-templates/pareto/pareto-chart-bar-chart-histogram-and-pareto-principle-8020-rule/

2299 (35,3%)

4207 (64,7%)

Patient complexity of UMCG-MDL

Low complex patients

High complex patients

0% 20% 40% 60% 80% 100% 0 100 200 300 400 500 1 7 ₁₃ ₁₉ ₂₅ ₃₁ ₃₇ ₄₃ ₄₉ ₅₅ ₆₁ ₆₇ ₇₃ ₇₉ ₈₅ ₉₁ ₉₇ 103 109 115 F re qu ency Number of diseases

Pareto chart LC patient population

(25)

25

Figure 8 LC patients by disease. Guptafile (2013)

5.3 EXPLORING CAUSES

This section explores causes of the identified performance in section 5.2. Observations are formulated of each stakeholder and its effect on triaging patient complexity is determined.

5.3.1 First decision phase

GPs and physicians of a community hospital are stakeholders in the first decision phase. The stakeholders make the following two decision in the referral process which are discussed next.

1. Does this patient need diagnosis in a hospital? 2. Which hospital to select for referral?

GPs has a stable patient file in which the GP aims at treating these patients as long as possible. The GP underpinned that he has a degree of responsibility of treating the patients himself. However, when a GP’s knowledge, skills and diagnostic equipment is insufficient to assess a patients complaint, the GP decides to refer this patient to a hospital which entail first line referrals. So, we can say that the first line referrals regarding the need for hospital care is appropriate. However, the following limitation has to be mentioned. The GP, physician of community hospital and academic physician have different perceptions of the necessity of referral. This can be explained by the capacity and capability of care givers which increases upwards the chain from GP to physicians of a community hospital and physicians of an AHC. The second question in the referral process is selecting a hospital for patient referral. GPs use two criteria for selecting a hospital which is travel distance to a hospital and admission times. GPs do not include patient complexity criteria in the assessment of hospital selection and therefore do not make a distinction between selecting a community or academic hospitals for patient referral. We can say that hospital selection regarding first line referral is inappropriate.

469 (20,4%)

397 (17,3%)

170 (7,4% 1263 (54,9%

LC patients per disease

(26)

26 Physicians of community hospitals receive first line referred patients. The physician accepts every referred patient and performs diagnosis. Based on the outcome of the diagnose, the physician in a community hospitals performs treatment accordingly. However, if a physician in a community hospitals has insufficient capabilities in terms of knowledge, skills, diagnostic equipment or treatment capabilities, the decision is made to refer patient to a specialist of an AHC which entail second line referrals.

Observation 1: The GP effectively assesses if a patient requires hospital diagnosis. Observation 2: The GP is less effective in selecting an appropriate hospital.

Observation 3: The physician of a community hospital effectively assesses if a patient requires AHC diagnosis and is effective in selecting an appropriate hospital.

5.3.2 Second decision phase

The academic specialist, subsequent the UMCG-MDL physician is stakeholder in the second decision phase. This stakeholder makes the following main decisions in the referral process which is discussed next.

1. Does the referred patient need diagnosis and/or treatment in the UMCG-MDL?

The MDL-physician evaluates referral letters and determines if the patients requires diagnosis and/or treatment in the UMCG-MDL. The MDL-physician checks if a patient can be categorized in IBD or live diseases. If so, the patient is directly accepted for treatment in the UMCG-MDL. The remaining referrals consist of a description of complaints which can be allocated to various diseases. The MDL physician notified that he is unable, based on this description, to determine what the disease of the patient is. The MDL-physician preliminary accepts these patients for diagnosis. If a diagnosed patient can be categorized in oncology, special endoscopies or unknown, the patient is also accepted for treatment. The remaining patients are discharged to their initial referrer.

(27)

27 Observation 4:No detailed admission criteria are used in the UMCG-MDL.

5.3.3 Third decision phase

The UMCG-MDL physician is stakeholder in the third decision phase. This stakeholder makes the following main decision in the referral process which is discussed next.

 Does this patient require continuing care in the UMCG-MDL?

The MDL-physician has regular conversations with patients which receiving treatment in the UMCG-MDL. The MDL-physician evaluates in these conversations if patient’s health is stable and if the patient requires continuing treatment. If so, the MDL-physician discharges the patient to their initial referrer. IBD patients which are chronic, unstable patients, stay under treatment. However, there are no detailed criteria for discharging patients which means that patient discharge mostly depends on the subjectivity of the physician. This entails that the time patient receive treatment varies considerably so far that some patients do not get discharged at all. The consequence of this observation is shown in figure 6-8 in section 5.2 which shows that LC patients have a significant share in the UMCG-MDL patient population. Observation 5: No detailed discharge criteria are used for discharging patients in the UMCG-MDL.

Table 2 shows a summary of the observations structure by decision phases with the effect of triaging patient complexity.

Table 2 Observations UMCG-MDL triage system Decision

phase Observations

Effect on triaging patient complexity

1

The GP effectively assesses if a patient requires

hospital diagnosis. +

The GP is less effective in selecting an appropriate

hospital. -

The physician of a community hospital effectively assesses if a patient requires AHC diagnosis and is effective in selecting an appropriate hospital.

+ 2 No detailed admission criteria are used in the

UMCG-MDL. -

3 No detailed discharge criteria are used in the

(28)

28

5.4 DETAILING CAUSES

This section elaborates on the observations which have a negative effect on triaging patient complexity defined in section 5.3. Causes are identified and structured in the three decision phases. In-depth interviews are performed with GPs and MDL-physicians in order to get a detailed understanding of the causes. Also, literature is used to clarify and obtain an understanding of the causes.

5.4.1 First decision phase

Patients and their wellbeing are the prime concern and responsibility of PCPs. Therefore, GPs organized their decision criteria for hospital selection accordingly. GPs mentioned that patients are not willing to travel long distances to hospitals. Also, the main reason of hospital referral is to establish the diagnosis. This also complies with results from a study of A. Bowling & Redfern (2000) which identified the establishment of the diagnosis the main reason of referral based on eight main reasons (Coulter et al., 1989). The need for diagnosis implies a degree of uncertainty which requires rapid assessment and so low admission times. However, apparently GPs and MDL-physicians roles are not clear such that differences exists in what should be included in the decision criteria for hospital selection. Therefore, the first main cause is:

Main cause 1: Lack of interaction and cooperation between GPs and MDL-physicians.

5.4.2 Second decision phase

The quality of referral letters is of key importance in order to be able to apply detailed admission criteria. MDL-physicians checked referral letters which often appeared to be incomplete. This complies with a study of Jenkins (1993) which found that the percentage of referral letters containing errors or omissions ranged from 5.4% to 28.2%. This complicates the use of detailed admission criteria such that diagnosis is needed in order to guarantee safety of patient’s health. Also, MDL-physicians notified that they have no proper understanding of what constitutes low and high patient complexity. PCPs have to be able to provide treatment for low complex patients which might be difficult for some sets of diseases. So, there is no clear understanding of patient complexity criteria per set of disease such that these criteria is not used either. The second main cause is:

Main cause 2: Insufficient admission procedures, i.e. criteria, roles and responsibilities.

5.4.3 Third decision phase

(29)

29 referrer during the process of patient treatment. Existing health condition and arising issues are not discussed between hospitals and PCPs which would have ensured continuity of care while the patient is in hospital (HSENC Advisory Group, 2012). (2) Lack of Estimated Length Of Stay (ELOS) and Predicted Date of Discharge (PDD) criteria for patients accepted for treatment. The lack of this information affects structured and consistent patient discharge over MDL-physicians. (3) Patients expectations are not managed properly. Patients are not informed if their referral consist of single advice consults, diagnosis or treatment activities which creates a resistance to discharge. (4) The lack of detailed discharge criteria which affects the duration patient receives treatment. This is strongly affected by the subjectivity of the involved physician. A study of Burkey et al. (1997) identified a number of barriers for discharge which includes the physician’s view about the limits of their responsibilities. Physician’s view differ about who should decide when it is appropriate to continue to see patients (Burkey et al., 1997). The third main cause is:

Main cause 3: Insufficient discharge procedures, i.e. criteria, roles and responsibilities.

5.5 SUMMARY OF MAIN FINDINGS

The performance of the UMCG-MDL triage system is defined as the number of LC patients which receive treatment in the UMCG-MDL department. Analysis of the patient’s complexity shows that 35,3% of the patient population consists of LC patients. The diseases M. Crohn, Ulcerative colitis and the Irritable bowel syndrome have a large share in LC patient population, respectively 45,1%.

In order to clarify the identified performance, explorative step is performed by describing observations. The following observations are made; The GP effectively assesses if a patient requires hospital diagnosis but is less effective in selecting an appropriate hospital. The physician of a community hospital effectively assesses if a patient requires AHC diagnosis and is effective in selecting an appropriate hospital. The UMCG-MDL physician does not use formal admission and discharge procedures with detailed admission and discharge criteria. Finally, stakeholders in each decision phase make inadequate use of supportive systems which results in a lack of interaction and cooperation between stakeholders.

(30)

30

6 TRIAGE SYSTEM REDESIGN

This chapter describes proposals for redesigning the UMCG-MDL triage system. Section 6.1 describes the approach of this chapter. Next, section 6.2 identifies and describes opportunities for triaging patient complexity. Section 6.3 describes the proposed redesigned UMCG-MDL triage system. Section 6.4 determines performance improvements. This chapter concludes with the identification of possible barriers for using the redesigned UMCG-MDL triage system described in section 6.5.

6.1 APPROACH

In order to give a structured redesign of the UMCG-MDL triage system the following questions has been taken into account;

 What opportunities for triaging patient complexity can be identified?  How should the opportunities for triaging patient complexity be addressed?

The first step in redesigning the UMCG-MDL triage system is identifying and describing opportunities for triaging patient complexity. The opportunities are identified with the use of the following three different inputs; domain experts, literature and the identified causes in chapter 5. Next, the applicability of the identified opportunities is evaluated by identifying possible barriers. The most preferred opportunity is selected in order to redesign the UMCG-MDL triage system. The selected opportunity for redesign is primarily validated by the domain experts. Next, a pilot group is set up in order to examine the feasibility of an approach that is intended to be used in a larger scale study (Leon, Davis, & Kraemer, 2011). Proposal are defined for the UMCG-MDL triage system redesign for the pilot group by meetings with domain experts and the manager of UMCG-MDL. The effect of the redesign on triage performance is estimated by using the Guptafile (2013). Finally, possible barriers for implementation of the redesigned triage system are identified and validated by domain experts.

6.2 OPPORTUNITIES FOR TRIAGING PATIENT COMPLEXITY

This section identifies and describes opportunities for triage redesign by triaging patient complexity. This consists an identification and evaluation of opportunities.

6.2.1 Opportunities

(31)

31

I. Chain management

The first opportunity is chain management which aims to address the lack of interaction and cooperation between GPs and MDL-physicians in the first decision phase. Chain management should be either initiated or improved in order to cover this cause. This consists of clear division of roles, tasks and responsibilities for GPs and MDL-physicians in the triage system.

 GP. The GP should alter the decision criteria for selecting an hospital for referral. This entails that the GP should be able to assess patient complexity and refer accordingly. If the GP is unable to assess patients complexity, community hospitals should be selected for patient referral. The GP should also be trained in using “Zorgdomein” in a uniform way such that physicians of a community hospital and academic specialists are able to assess the referral’s appropriateness.

 Physician of a community hospital. The physician of a community hospital should also be trained in using “Zorgdomein” in a uniform way for patient referrals to academic hospitals.

 UMCG-MDL physician. The UMCG-MDL physician should know how to handle referral letters and what its role and responsibilities are towards GPs and physicians of community hospitals.

II. Patient admission procedure

The second opportunity is a patient admission procedure which aims to address the current insufficient patient admission procedures in the second decision phase. A structured patient admission procedure enables the MDL-physicians to have a clear view on which patients to accept and refuse for care. The procedure should provide clear criteria, roles and responsibilities within the UMCG-MDL department.

The admission criteria should include detailed patient complexity criteria. These criteria should be used to evaluate referral letters and triage accordingly. By using a patient admission procedure, the UMCG-MDL department has a certain degree of control over the inflow of patients which is desirable concerning the influences of health insurers on provided care. The UMCG-MDL department may decide to refuse any patients surpassing the set patient limit such that financial targets are achieved.

III. Patient discharge procedure

(32)

32 provide clear criteria, roles and responsibilities within the UMCG-MDL department. The use of a patient discharge procedure with a consistent and structured approach enables MDL-physicians to discharge patients from specialist outpatient services towards PCPs. Effective hospital discharge can only be achieved when there is cohesive joint working between hospital and primary care (Johnson & Nile, 2011). The ADSOS guideline (2013) describes how such a procedure should be organized which includes the following four elements; discharge planning, discharge criteria, discharge process and the discharge letter.

6.2.2 Evaluating opportunities

The opportunities described in section 6.2.1 aim to address the causes identified in section 5.4. Ideally, all opportunities should be implemented such that all causes which complicate triaging patient complexity are addressed. However, the following barriers for implementation has to be noticed for chain management and the patient admission procedure.

Initiating and improving chain management requires extensive medical knowledge transfer between the UMCG-MDL physician and GPs. Also, initiating chain management requires involvement of a substantially number of GPs. The patient admission procedure requires admission criteria. These criteria would include a detailed understanding of patients complexity. However, in the stage of the second decision phase, patients are not yet diagnosed and haven’t received treatment. In order to determine patient complexity, the MDL-physician has to assess patient complexity based on complaints and symptoms. Relationships have to be established between complaints and diseases which is unknown in the UMCG-MDL department. Thereby, information containing in referral letters appeared to be incomplete which complicates the situation even more.

(33)

33

6.3 PATIENT DISCHARGE PROCEDURE

This section describes the patient discharge procedure which is selected for redesigning the UMCG-MDL triage system. This consists of a description of a patient discharge procedure, the set up of a pilot group and a proposal for a redesigned triage system of the UMCG-MDL department.

6.3.1 Description

The patient discharge procedure is described next as defined by the ADSOS guideline (2013).

 Discharge planning. The discharge planning is the critical link between the specialist and the referring GP. This process begins at the initial specialist outpatient service appointment and ends with the patient’s return to their referring GP. Discharge planning aims to identify issues relevant to each patient’s discharge back to the referring GP and to initiate action to address these issues so that discharge is not delayed. An important tool to address these issues is managing patient’s expectations. Considering, the patients have “Vrije artsenkeuze” 4

and are allowed to refuse discharge if they do not agree with the discharge. This is supported by results of a study of Burkey et al. (1997) which defined perceptions of patient expectations and feeling of loss as one of the barriers for patient discharge. Also, Alper, O’Malley, & Greenwald (2015) state that discharge planning is the development of an individualized discharge plan for the patient prior to leaving the hospital, to ensure that patients are discharged at an appropriate times and with provision of adequate post-discharge services. The discharge plan is developed with the patient and carer in order to explore options for the patients care post hospitalization, including other healthcare providers. The discharge plan is communicated to primary and community care providers and the discharge plan is documented in the healthcare record (HSENC Advisory Group, 2012).

 Discharge criteria. The discharge criteria assists in identifying the point at which the episode of care is complete, so as to expedite discharge from service. Defining clear discharge criteria promotes consistency of practice and aid decision making across all employees working in the specialty. This reinforces the use appropriate use of specialist outpatient services.

 Discharge process. In the discharge process the physician has to identify patients during consultation who are ready for discharge based on the defined discharge criteria. The

4

(34)

34 determination of a Estimated Length of Stay (ELOS) enables to formulate a Predicted Date of Discharge (PDD). The PDD may function as a trigger for evaluating patients on the discharge criteria (HSENC Advisory Group, 2012). A PDD is assigned to each patient after diagnosis and or treatment. The physician should evaluate the patient on the PDD and decide whether the patient required continued treatment or is ready for discharge.  Discharge letter. The final step is writing a discharge letter which should be provided to

the initial referrer.

6.3.2 Pilot group

A pilot study can be defined as “A small-scale test of the methods and procedures to be used on a larger scale (Porta, 2008). The fundamental purpose of conducting a pilot study is to examine the feasibility of an approach that is intended to be used in a larger scale study (Leon et al., 2011). The pilot study is used by selecting a small set of diseases in order to determine how the patient discharge procedure should be organized.

The sets of diseases which have the largest share in the LC patient population are selected for the pilot group, as defined in figure 8. This concerns M. Crohn, Ulcerative colitis and the Irritable bowel syndrome which represent 45,1% patients in the LC patient population of the UMCG-MDL department. The selected disease are described in more detail next.

 M.Crohn and ulcerative colitis: Crohn’s disease primarily causes ulcerations of the small and large intestines, but can affect the digestive system anywhere from mouth to the anus 5. Ulcerative colitis only involves the colon 6. These diseases have no medical cure and tend to fluctuate between periods of inactivity and activity (Fig. 8).

Figure 8 Characterization of chronic diseases: M. Crohn and Ulcerative colitis (Scheffer, 2013).

5_{http://www.medicinenet.com/crohns_disease/article.htm} 6

(35)

35  Irritable bowel syndrome: This disease is a gastrointestinal disorder characterized by the presence of a cluster of symptoms and signs that include cramping, abdominal pain, increased gas, altered bowel habits, food intolerance and bloating 7.

6.3.3 Proposed redesign

The patient discharge procedure consists of the following four elements and is organized and described next. In-depth interviews and meetings are performed in order to define the content of the patient discharge procedure.

 Discharge planning

The discharge plan is documented in the healthcare record, reviewed and updated after consults, in response to changing needs (Johnson & Nile, 2011). Next, the discharge plan is communicated to the referrers, primary and/or community care providers and contains the following information;

- Estimated date of discharge

- Documentary evidence of communication with the relevant GP or physician of a community hospital.

- Follow-up plan

- Early warning signs of relapse and risks.  Discharge criteria

The following questions should be asked in the discharge process to determine LC patients and discharge these patient accordingly.

M.Crohn:

- Is an abscess, pancolitis or fistula present? [No]

- Does patient require a low amount of prednisone? [Yes] Ulcerative colitis:

- Are primary sclerosing cholangitis present? [No] - Is the disease restricted to the rectum? [Yes]

- Is the disease left sided in which the use of mesalazine is sufficient? [Yes] Irritable bowel syndrome:

- Disease irritable bowel syndrome diagnosed? [Yes]

7

(36)

36  Discharge process

M.Crohn and Ulcerative colitis patients are evaluated during the consultations after diagnosis and/or treatment. LC patients are discharged after the first evaluation meeting after diagnosis and/or treatment towards physicians in community hospitals. The remaining HC patients continue to receive care and further meetings are determined based on the prescribed medication. Figure 9 (see appendix E) shows the types of medication prescribed for M. Crohn and Ulcerative colitis patients with corresponding review periods. These periods function as a trigger to re-evaluate the patient by the discharge criteria.

The irritable bowel syndrome is a complete low complex disease. Patients diagnosed with irritable bowel syndrome should be discharged accordingly.

 Discharge letter

The discharge letter is made by the UMCG-MDL physician which has diagnosed and treated the patient. The discharge letter is send to the primary care provider and uses a pro forma consisting of the following elements;

- Date of first visit

- Reason for referral to specialist outpatient services

- Summary of interventions provided and their outcomes including any diagnosis derived.

- Reason of discharge - Date of discharge - Relevant risks

- Ongoing management plan

6.4 PERFORMANCE IMPROVEMENT

(37)

37 Figure 10 shows the patient population of the UMCG-MDL per diseases of the pilot group. The use of the patient discharge procedure results in a decrease of 171 M. Crohn, 183 ulcerative colitis and 279 irritable bowel syndrome patients.

Figure 9 Patient population of the UMCG-MDL

Figure 11 shows the UMCG-MDL triage system performance in terms of the amount of LC patients in the system. This consists of a decrease of 633 patients.

Figure 10 UMCG-MDL triage system performance

6.5 POSSIBLE BARRIERS

Procedures should be implemented properly in order to assure its effectiveness. This raises the importance to identify and address possible barriers that may arise in implementing the procedure (SURE, 2011: Chap.5). Two perspective are addressed in order to address possible

802 581 279 631 398 0 0 100 200 300 400 500 600 700 800 900 M. Crohn Idiopatic proctocolitis Irritable bowel syndrome Num ber o f pa tient s Type of disease

Patient population of the UMCG-MDL

UMCG-MDL triage system UMCG-MDL triage system with patient discharge procedure

2299 1666 0 500 1000 1500 2000 2500

UMCG-MDL triage system UMCG-MDL triage system with

patient discharge procedure

Num ber o f pa tient s

LC patient population of the UMCG-MDL

(38)

38 barriers for implementation which is the view from AHC and PCPs. Meetings with domain experts are initiated in order to determine and validate barriers for implementation.

6.5.1 Perspective of the UMCG-MDL

Results from a study of Burkey, Black, Reeve, & Roland (1997) identified a number of barriers to discharge as perceived by the physicians of AHCs discussed next.

1) Uncertainties about GP´s ability to provide care. Physicians expressed their concerns about the abilities of their colleagues in general practice and about their own lack of knowledge of the facilities available to patients in the community. Also, a study of Bowling, Stramer, Dickinson, Windsor, & Bond, (1997) shows that a lack of confidence in GPs ability to manage their patients is a reason for the provision of continuing care in hospitals. UMCG-MDL physicians supported this opinion by notifying that GPs have insufficient knowledge of immunosuppressive medication. Also, GPs ability to treat patient outbreaks of M. Crohn and Ulcerative colitis is insufficient because they rarely occur in general practice.

2) Specialists’ perceptions of their responsibilities. Specialists’ views about the limits of their responsibility for the care of patients varied such that some had a clear view of their role whereas others resented the suggestion that it might be for other to decide when it was appropriate for them to continue to see patients.

3) Perceptions of patient expectations and feelings of loss. Some physicians cited patient expectations as a barrier to discharge, because patients who are being discharged often feel a sense of loss or that they are being abandoned because their condition will no longer be reviewed at a hospital. This should be properly management by the MDL-physicians by interacting with patients and providing patients with adequate information.

6.5.2 Perspective of PCPs

Results from a study of Reeve et al. (1997) identified a number of barriers to discharge as perceived by PCPs which are discussed next.

(39)

39 2) Ability to refer back easily for specialist advice. PCPs expressed their concerns about the waiting times for re-referral and access back into the system. This can also have considerable consequences for the UMCG-MDL considering the waiting list for patients is approximately 70 days.

3) The need for guidance on future management. PCPs appeared to be prepared to take over the care of discharged patients in many cases. However, they felt that needed guidance outlining on-going treatment: on-going support was considered very important. A clear protocol of care would facilitate change (Reeve et al., 1997). The patient discharge procedures uses patient specific discharge plans which consists of a follow-up plan such that guidance outline on-going treatment is assured.

(40)

40

7 DISCUSSION

This chapter contains the main findings and discusses the outcomes of the study. Section 7.1 describes the main findings and section 7.2 contains the limitations of this study.

7.1 MAIN FINDINGS

This section describes the main findings of the research. The objective of this study was to design a triage system for AHC-MDLs which minimizes LC patients entries and existing LC patients in the system by diverting these patients to GPs or community hospitals. A design science approach is used in order to achieve the research objective and the following questions are answered.

 What are the main elements of the system under study and how is the triage system set up?

The UMCG-MDL triage system is structured according three decision phases in which patient flows. The first decision phase consists of the first and second line patient referrals from the PCPs towards the UMCG-MDL department. The main activity is evaluating if a patient requires hospital treatment and selecting a hospital based on travel distance and admission times. The second decision phase takes place in the UMCG-MDL department and the main activity is evaluating referral letters and accepting/refusing patient for care. The MDL-physician checks if the referral letter is complete, the patient is known in the UMCG and if the patient is a Live diseases, IBD or oncology patient. The third decision phase also takes place in the UMCG-MDL department and the main activity is discharging patients back to their initial referrer. The MDL-physician writes a specialist letter which contains a description of the activities performed and subsequent steps which need to be taken by the initial referrer.

TRIAGING PATIENT COMPLEXITY