INTEGRATION AND INFORMATION SHARING WITHIN THE CHAIN OF ACUTE CARE: A DIAGNOSTIC CASE STUDY

SHARING WITHIN THE CHAIN OF ACUTE

CARE: A DIAGNOSTIC CASE STUDY

By

Jasper de Jonge

University of Groningen

Supervisor: prof. dr. J. (Jan) de Vries

Faculty of Economics and Business

Content Abstract I List of Figures II List of Tables II Preface III 1 INTRODUCTION 1 2 THEORETICAL BACKGROUND 4 3 METHODOLOGY 12 5 RESULTS 23 6 ANALYSIS 30 7 DISCUSSION 47 8 CONCLUSION 49 REFERENCES 51

APPENDIX A – Interview guide 56

APPENDIX B – Codebook for analysis 60

Abstract

The need for integrated care has grown in the past decade and there is an increasing interest in how healthcare workers, managers and policy makers could implement effective integrated care services. Streamlining patient flows, establishing trustful relationships with partners in the chain and linking planning and information systems are examples of activities within the complexity of integrated care. Many authors stress the importance for more integration and information sharing research within in a healthcare setting. However, the body of knowledge about information sharing between healthcare organisations seems to be rather limited. Understanding the mechanisms that influences information sharing could help care providers to improve the collaboration within the care chain and improve effectiveness of the whole chain. The relationship between the organisational, the policy & legislation and technical perspective was explored by using a single in-depth case study amongst healthcare providers within the chain of acute care. Eleven semi-structured interviews, observations and document analysis were used for data gathering. This research found that the variables within the organisational perspective are significant enablers towards information sharing, that policy is bounded by current legislation and that compatibility and data sharing standards are mediating each other towards information sharing. The three perspectives are interrelated and influence each other in either a positive or negative way. In this study, the gap in literature is filled by exploring and understanding the complex nature and dynamics of the relationships between these enablers and barriers.

List of Figures

Figure 2.1 - Conceptual framework of the research ... 9

Figure 3.1 - Network of stakeholders in acute care ... 13

Figure 3.2 - Simplified overview acute care chain ... 15

Figure 3.3 - Example of theory-driven coding framework ... 21

Figure 5.1 - Path A of the acute care chain ... 24

Figure 5.2 - Path B of the acute care chain ... 25

Figure 5.3 - Patient routing at the ED with triage level 1 & 2 ... 25

Figure 5.4 - Patient routing at the ED with triage level 3, 4 & 5 ... 26

Figure 6.1 - Relational framework organisational perspective ... 43

Figure 6.2 - Relational framework policy & legislation perspective ... 44

Figure 6.3 - Relational framework technical perspective ... 45

Figure 6.4 – Revised empirical framework with relationships ... 46

List of Tables Table 2.1 - Operationalization of the constructs and variables ... 11

Table 3.1 - Inflow facts of the hospital's ED (2017) ... 15

Table 3.2 - Discharge destination after visiting ED (2017) ... 16

Table 3.3 - Details of the organisations ... 17

Table 3.4 - Characteristics of the interviewees ... 18

Table 3.5 - Data collection table... 20

Table 3.6 - List of data-driven codes ... 22

Preface

1 INTRODUCTION

Over the last decades, the healthcare sector has changed rapidly. Increased competition, a growing influence of patients and the necessity of delivering healthcare services in a more efficient and effective manner has enabled the search for integrated services (Aptel & Pourjalali, 2001). Integrated care has become an important development to serve patients’ needs and reduce the fragmentised character of healthcare systems (Minkman, 2011). The redesign of healthcare services and the implementation of integrated care programs are frequently addressed as being critical strategies to decrease resource utilization and improve healthcare quality (de Vries & Huijsman, 2011).

The current stream of patients, and in particular the more chronically ill and elderly, demand specific care needs which require efforts of several healthcare professionals and multiple healthcare organisations. The concept of ‘integrated care’ focuses on the total needs of clients, not only on the services provided by one single professional or healthcare organisation (Minkman, 2011). Integrated care is required when the services of separate and individual healthcare professionals do not cover all the demands of the client (Ouwens, Wollersheim, Hermens, Hulscher & Grol, 2005). The need for integrated care has grown in the past decade and there is an increasing interest in how healthcare workers, managers and policy makers could implement effective integrated care services. However, developing integrated care services remains a complex phenomenon (Minkman, Vermeulen, Ahaus & Huijsman, 2011).

become a necessity for improving effectiveness. The uncertainty surrounding many patient groups requires flawless information integration to enable efficient and timely decision-making (Shih, Rivers & Hsu, 2009). Cooperative attitudes of chain partners provide rapid access to the required information and enables faster response times (Sezen, 2008). Furthermore, in integrated chain care, patients are sent from one care provider to another, each with their own info regarding the patient. Close cooperation and information sharing will thus enhance the healthcare delivery process and align chain members to provide the best quality care and improve operational performance.

However, it is worthwhile to note that information sharing in public organisations is heavily subject to factors that enable or inhibit the information sharing practices between chain partners (Yang & Maxwell, 2011). From government studies (e.g. Yang & Maxwell, 2011; Landsbergen & Wolken, 2001; Zhang & Dawes, 2006) it is well known that factors such as organisational barriers, technical barriers and legislative barriers can impede inter-organisational information sharing and thus inhibit the integrative practices in the chain or network. Current literature only provides research on the enablers and barriers of information sharing within manufacturing supply chains and between governments, but not in a specific healthcare context. Furthermore, most studies about integrated care solely focus on specific diseases and only address internal care chains within hospitals. The main view of this paper is therefore that information sharing should also be evaluated from the entire external chain of acute care.

Despite the positive effects of information sharing and integration in supply chains, it remains unclear if findings from operations research can be translated to a healthcare setting (Drupsteen et al. 2013). Thrasher, Craighead & Byrd (2010) denote the importance of further examining integration and information sharing in service-based settings such as in the chain of acute care. From literature it is unknown how information sharing within a chain of acute care providers is organised and which factors enable or inhibit the integration of several parties within the chain. Drupsteen et al. (2013) acknowledge that contextual factors significantly influence the use of Supply Chain Management (SCM) practices, therefore a healthcare context might pose different and currently underexposed enablers and barriers towards information sharing. This study aims to be one of the first to research the inter-organisational information sharing between chain partners in an acute care setting.

understand how enablers and barriers influences information sharing and affects the healthcare delivery processes within the chain of acute cure. In this study, the gap in literature is filled by exploring and understanding the complex nature and dynamics of the relationships between these enablers and barriers. An in-depth explorative single-case study was conducted within the chain of acute care. This chain consists of the Emergency department (ED) of one hospital, one general practitioner (GP) organisation, the Emergency Medical Services (EMS), one ambulance service organisation, a pharmacist and a district care organisation in the Netherlands.

2 THEORETICAL BACKGROUND

The theoretical background explains the different concepts used within this study. The topics integrated care and acute care are explained and the literature about integration and information sharing is outlined. This section will end by combining the different views of both literature reviews which eventually is visualised into a conceptual model. The constructs and variables of the model are operationalized and presented in a table.

2.1 Features and characteristics of integrated care and acute care

The characteristics of traditional healthcare organisations are changing. Over the past decade, the integration of care in many countries has gained increasing attention from managers, healthcare workers, policymakers and researchers as a strategy to improve healthcare delivery (Minkman et al. 2011). Integrated care appears in a variety of forms in literature and there is no uniform accepted definition. Significant diversity of terminologies has been used to describe integrated care such as ‘disease management’, ‘shared care’, ‘transmural care’, ‘coordinated care’ or just ‘integrated care’ (Ouwens et al. 2005). In this research, integrated care can be referred to as a “coherent and co-coordinated set of services, which are managed

and delivered to individual services users across a range of organisations and by a range of co-operating professionals and informal carers” (Raak, Mur-Vreeman, Hardy, Steenbergen

& Paulus, 2003: 26). It is an organisational process of coordination that seeks to achieve seamless and continuous care, tailored to a patient needs and based on a holistic view.

The focus on integrated care originates from the growing fragmentation and supply-oriented approach in healthcare. There is a widespread belief that integration of care is mandatory to respond to deficiencies in current healthcare settings and that integration will enhance client satisfaction, quality of life, process outcomes and eventually will reduce costs (Ouwens et al. 2005; Minkman et al. 2011). Additionally, an integrated focus on the patient increases the necessity for interaction between several stakeholders within the chain of acute care. Acute care differs in characteristics compared to elective care and is prone to demand uncertainty, fluctuations and the need for care of multiple healthcare providers.

for more integrated care to reduce the existing discontinuity, duplications or absence of responsibilities for the whole continuum of care (Minkman et al. 2011).

2.2 Concept of integration and information sharing

Several studies (e.g. Sezen, 2008; Pagell, 2004; Flynn, Huo & Zhao. 2013) have found empirical evidence that integration leads to higher performance in a SCM context. SCM emphasizes integration as the overall and long-term benefit of all parties on the chain through co-operation and information sharing (Gunasekeran & Ngai, 2004.). From a SCM perspective however, the body of knowledge regarding the healthcare sector still seems to be rather fragmented (de Vries & Huijsman, 2011). Integration is an important phenomenon with respect to aligning several stakeholders in the care chain. Integration entails organisational entities within a chain not acting as functional entities, but as a unified whole (Barki & Pinsonneault, 2005). Increasing the level of integration and information sharing among members of the chain has become essential for improving the effectiveness of supply chains (Sezen, 2008). Many healthcare organisations recognized the importance of adopting supply chain practices and the application of techniques and methods (de Vries & Huijsman, 2011). However, the complexity of technologies, the interests of multiple stakeholders, a dynamic internal and external environment and distinctive characteristics of healthcare operations, often hinder a straight forward application of industrial oriented SCM practices.

2.3 Benefits of information sharing

Information sharing amongst chain partners can be denoted as inter-organisational information sharing. Inter-organisational information sharing refers to the extent to which critical and proprietary information is communicated to one’s chain partner (Li & Lin, 2006). Well- designed and executed information sharing has the potential to streamline data management, improve information infrastructure, facilitate the delivery of integrated services and enhance relationships among the chain partners (Landsbergen & wolken, 2001; Zhang & Dawes, 2006). Information sharing can reduce duplicate data gathering and handling, with more consistent and comprehensive information about the patient, the chain partners can better define and solve joint problems and better coordinate programs, policies and services. Through the process of medical information sharing, generating trust and commitment, creating shared expectations and establishing norms of interaction amongst the chain partners is paramount for the success of the care chain (Minkman et al. 2011).

2.4 Barriers of information sharing

of patient information sharing amongst partners in the chain of acute care. For inter-organisational information sharing to be successful, it depends heavily on understanding the factors that influence cross-boundary information sharing (Yang & Maxwell, 2011).

Organisational perspective: Researchers indicate that sharing of information and knowledge can involve complex interactions between participating organisations because of their different origins, values, and cultures (Yang & Maxwell, 2006). Inter-organisational information sharing relationships rely heavily on trust building between the chain partners involved and the development of commitment over time (Pardo et al., 2006). Next to that, creating shared vision is acknowledged as a critical factor in aligning chain partners and enable information sharing amongst them. Different expectations may reflect each party’s individual and organisational history or simply the variations in the characteristics of the individuals or organisations (Pardo, Cresswell, Dawes & Burke, 2004). However, it is worthwhile to note that the performance of the chain depends on what information is shared, how it is shared and with whom. Such characteristics of the flow of information are largely determined in the design process of the chain. Therefore, the information sharing within the chain of acute care should also be evaluated with the design considerations (Sezen, 2008; Li & Lin, 2006). The design consideration is a critical factor to understand how the acute care chain is organised and for determining the effectiveness of a chain. The process design involves decisions about the proximity to chain partners and contractual terms between chain members (Chopra & Meindl, 2004). According to Minkman et al. (2009) the process design for integrated care (i.e. delivery system) consists of several elements ranging from agreements on procedures for information exchange, using shared client treatment and care plans, developing criteria for assessing clients’ urgency, collaboratively offering client information to the care partners but also aim towards developing a front-office as a single-entry point for generating patient information.

Policy & Legislation perspective: Legislation and policy have a strong influence on the sharing of information and knowledge across organisations, especially for organisations in the public sector (Dawes, 1996; Landbergen & Wolken, 2001; Zhang & Dawes, 2006). The existence of legal and policy regulations can facilitate relationship building, trust development and inter-organisational cooperation. On the opposite, legal factors and a lack of legislative support to assure privacy and confidentiality of shared information can impede inter-organisational information sharing within the public sector (Yang & Maxwell, 2011).

Dawes, 2006). To enable information sharing and provide more accurate exchanged information, the total chain need compatible infrastructures as well as consistent data sharing standards (Dawes, 1996). Different organisations have often various types of hardware and software within their information systems, and in general it is a challenge to integrate heterogeneous information systems of different platforms and data sharing standards. In addition, because of security and confidentiality, it is critical to design a system that can handle access authorization and authentication for shared information (Yang & Maxwell, 2011). Moreover, the challenge of continuously development of technology further complicates the maintenance of adequate levels of knowledge and expertise (Dawes & Pardo, 2003)

2.4 Development of the conceptual model

FIGURE 2.1

Conceptual framework of the research

It can be stated that information sharing among different stakeholders can be influenced by many factors. Achieving an understanding about these factors influencing information sharing is difficult, because information sharing in practice is intimately linked to its context (Pardo et al., 2006). From literature (Sezen, 2008; Li & Lin, 2006) it is suggested that a lack of trust in supply chain partners is considered as a common obstacle towards information sharing. Trust reduces the fear of information disclosure and loss of power in the relation to information sharing. Trust between partners stimulates favourable attitudes and behaviours to ensure the quality of the information shared (Li & Lin, 2006). Furthermore, commitment and shared vision can involve trusting the partners with proprietary information and share a common view of the whole chain. The above arguments lead to the following question: How is the acute care chain organised and which influence have the variables within the organisational perspective on information sharing?

(Pardo et al. 2004). Because of the potential influence of these factors on information sharing the following question arise: In what way do policy, privacy and legislation regulations influence the information sharing amongst the chain partners?

Researchers (e.g. Li & Lin 2006; Sezen, 2008) consider IT as a great enabler for information sharing within networks and supply chains. The changing and expanding use of data in public organisations demand increasing attention to all the components of IT and data quality such as accuracy, timeliness, consistency and completeness (Pardo et al. 2004; Li & Lin, 2006). The effectiveness of public organisations often depends on the data exchanges with partners. IT enables the coordination across organisational boundaries to achieve high levels of efficiency and productivity and opens up new possibilities for increasing value through better communication (i.e. data sharing standards/protocols) and information sharing. Therefore, it is important to explore: How does compatibility and data sharing standards enable or inhibit inter-organisational information sharing within the chain of acute care?

TABLE 2.1

Operationalization of the constructs and variables

Independent variables Dependent variable: Information sharing Organisational perspective Trust Commitment Shared vision Process Design

• Trust is a critical factor in the development and maintenance of the inter-organisational relationships on which information sharing depends;

• Trust reduces the fear of information disclosure and loss of power in the relation to information sharing;

• Allowing a chain partner to view transaction-level data improves trust amongst the chain of acute care;

• Commitment is the variable that discriminates between relationships that continue and that break down;

• Commitment can involve trusting partners with proprietary and sensitive information;

• The lack of shared vision between partners in the chain of acute care will lead to less information sharing;

• Collaboration within the chain of acute care can be achieved only if the chain partners share a common view and goal for the chain of acute care;

• Integrated services with work rules and procedures

combined with shared decision making within the chain of acute care.

Policy & Legislation perspective

Policy &

Legislation • Policy & legislation can facilitate relationship building, risk reduction and trust development in inter-organisational information sharing when specific guidance such as how to utilize medical information is proposed;

• Lack of legislative support to assure the privacy and confidentiality of shared medical information can impede inter-organisational information sharing;

• Inter-organisational information sharing can be hindered because of policy and legislation that prohibit healthcare institutions from sharing sensitive and regulated patient information; Technical perspective Compatibility Data sharing standards

• Compatibility of IT-infrastructure enables coordination across organisational boundaries to achieve new levels of efficiency and productivity and opens up possibilities for increasing value through better communication and information sharing;

3 METHODOLOGY

The goal of this research is to understand how the relationships between enablers and barriers affects the information sharing processes within the chain of acute cure. Given that little knowledge exists on this topic, an exploratory case-study is conducted as suggested by Eisenhardt (1989). Deriving from the previous goal, the initial ‘why’? And the consecutive ‘how? questions are applicable. For this type of questions, a case study suits well. According to Yin (2014) and Eisenhardt (1989) a case study provides better insights in the field of the subject and can help understanding complex relationships. Moreover, this method is being used less frequent in the field of operations management but is widely recognized as being able to contribute to enriching the field (Voss, Tsikriktsis & Frolich, 2002). This section describes how the case study was designed and how the research has been conducted.

3.1 Research setting

This research has been performed at a care organisation in the northern part of the Netherlands. This organisation offers care services to patients both in the cure side and the care side of healthcare. The company owns three hospitals (cure) and has seventeen care institutions which offers services to elderly, disabled people or patients that need after-hospital services. The company has approximately 6300 employees divided over several business units. The research that has been conducted consist of an in-depth single case study within the chain of acute care. According to Saunders, Lewis & Thornhill (2009) a single case can be selected because it is typical or because it provides the researcher with an opportunity to observe and analyse a phenomenon that only a few have considered before.

FIGURE 3.1

Network of stakeholders in acute care

3.2 Case selection and Unit-of-Analysis

There are three reasons why the choice has been made to conduct a single-case study in this specific region. First, the number of organisations involved in acute care chain are numerous. At least 5 different organisations are involved which makes the magnitude of the research wider and also more complex. Secondly, this specific region is an area that has been used for testing and implementing new ideas and pilot projects around integrated acute care since March 2017. Stakeholders in this region are willing to cooperate in such research and were familiar with certain concepts. Thirdly, due to the limited amount of time available and the complex nature of the research, there was simply just not enough time to conduct this study at multiple regions or hospitals.

The reasoning behind this research are the changing trends in demand and also the need for less fragmentised but more integrated healthcare. The acute chain care is a process with several stakeholders and stand-alone professionals and it is of great importance to collaborate to provide the best possible care to the patient. This single-case study is aimed to understand i) how are healthcare delivery processes between care providers organised in terms of patient information sharing and ii) how do enablers and barriers influence the patient information sharing practices within the healthcare delivery processes between the different care providers.

patient files. Figure 3.2 illustrates a simplified overview of how the acute care chain is organised.

FIGURE 3.2

Simplified overview acute care chain

The research has been conducted from the perspective of the ED of the hospital. In table 3.1, inflow facts of the ED of the hospital are presented. The ED were this research has been conducted had almost 14.000 patients in the year 2017, from whom most of the patients (49%) arrived through the GP and the GP Out-of-Hours. These numbers denote the importance of effective and efficient patient information sharing amongst chain partners in acute care.

TABLE 3.1

Inflow facts of the hospital's ED (2017)

Inflow type # of Patients % of Total

112/EMS 2.754 20 Self-referral GP Out-of-Hours 698 3.459 5 25 GP 3.408 24 X-rays 1.197 9 Specialist/Outpatient clinic 2.230 16

Nursing Home/District care 190 1

After visiting the ED, the patients can get discharged to different destinations. Table 3.2 presents an overview of the patients’ destinations that visited the ED in the year 2017. The largest group of patients (66%) were send home after visiting the ED. The second largest (39.63%) were admitted to the hospital in terms of nursing wards. The remaining 3.65% is discharged to a nursing home or another hospital.

TABLE 3.2

Discharge destinations after visiting ED (2017)

Inflow type # of Patients % of Total

Intensive care 232 1.7 Nursing Home Discharge by GP 87 100 0.6 0.7 Discharge ED 7904 57 To Hospital 5523 40 Other Hospital 90 0.6 Total 13.936 100% 3.3 Data collection

In order to establish meaningful conclusions about the results of this study, the method of data triangulation has been used. Triangulation is “the use and combination of different methods to study the same phenomenon” (Voss et al. 2002: 206). Multiple data sources and methods have been applied: interviews, informal conversations, direct observations, content analysis of documents and archival research. According to Blumberg, Cooper & Schindler (2011) it is likely to conduct semi-structured interviews when an exploratory case study is performed. This allows the researcher to explore new emerging themes or issues when they occur within the case study.

researcher to understand the process design and allowed the researcher to conduct observations of the patient information sharing activities of the different stakeholders. Table 3.3 provides an overview of the organisations involved for acquiring empirical data.

TABLE 3.3

Details of the organisations

Organisation Label Type of organisation # of Interviewees

A Hospital 3

B C

GP Out-of-Hours Emergency Medical Services

2 2

D Regional Ambulance Services 2

E District care provider 2

The questions asked during the interviews were set up according to a literature derived interview protocol which is presented in appendix A. Semi-structured interviews were chosen because this can offer flexibility to approach different interviewees in a different method while still covering the same area of data collection (Karlsson, 2016). The used interview protocol was divided and structured into five different parts. Each part had his own function and goal. During the first part of the semi-structured interviews, every participant was asked to describe how the acute care chain was organised from the stakeholder’s perspective in terms of protocols, guidelines and roles in general. Next to that, the stakeholders were asked to identify their most important chain partners and asked them why these stakeholders were the most important to them. In this way, it was possible to see from which perspective stakeholders were reasoning and how certain relationships between chain partners were more developed than others. Once the stakeholders described the acute care chain in their own words and described their relationships with the most important chain partners, the researcher was able to compare outcomes per interviewee and could draw a valid conclusion about the design, definition and goal of an acute care chain.

in turn provided comments and feedback of the interviewees, so the researcher could improve his view and understanding on the design of the acute care chain.

The third part of the interview concerned the policy & legislation perspective, open questions were asked to gain an understanding how these aspects influence the information sharing practices. The answers of these questions were compared with policy documents for verification. The fourth part of the interview consisted of questions about the technical perspective. Because the interviews were held at the location of the organisations, the interviewees could explain the researcher how current IT was used and to which extent software was used to share patient information between chain partners. These observations were used as a check on the different answers the interviewees provided. The answers could be compared to the real-life situation which were useful to validate these answers.

The last part of the interview protocol were open questions related to information sharing. In this way, the interviewee could provide the researcher with more comprehensive data about the current situation and provide a better understanding of the influence that enablers and barriers have on inter-organisational information sharing.

TABLE 3.4

Characteristics of the interviewees

Intervie wee

Organisation Label

Role in acute care chain

Age Working experience in years 1 A Manager 45 25 2 A ED Physician 35 5 3 A Pharmacist 28 4 4 B Manager 54 16 5 B General Practitioner 57 21 6 C Quality Advisor 42 2 7 C EMS Operator 63 30 8 D Manager 57 25 9 D Ambulance Nurse 53 25 10 E Manager 63 12 11 E District Nurse 48 7

Every participant was asked for consent and a signature prior to the start of the interview to guarantee anonymity and confidentiality. The researcher consciously decided to not send the interview protocol to the participants in advance in order to reduce participant bias and politically desirable answers. All the participants in this research were Dutch native speakers, so to prevent miscommunication, incomplete answers and to generate in-depth data as much as possible, the interviews were all held in Dutch. To ensure construct validity, all the interviewees were offered the opportunity to comment on the transcripts of their interviews and check for factual inaccuracies and confirm that the interview was transcribed correctly.

quick scan analysis documents. The documents were provided by the several stakeholders in the care chain and next to that, the database of company was accessible for the researcher.

The observations in the work field were conducted at almost all of the stakeholders in the care chain and focused on how the processes were executed in real-life situations. Work-visits and participating in the process of the acute chain care provided helpful insights and comparison material to evaluate findings and results. Table 3.5 on the next page illustrates an overview of which type of data collection is applied at every organisation involved.

TABLE 3.5 Data collection table

Organisa tion

Stakeholders Interviews

Internal Documents External Documents Observations/Field notes

A • Manager; • ED Physician; • Pharmacist; • ED workflows and procedures; • Hospital protocols; • Quality documents acute care (ROAZ); • SBAR-protocol; • Taskforce group documents;

• Way of working at the ED, handover moments, patient treatment, administrative tasks; Use of IT;

B • Manager;

• General Practitioner • Partners covenants; • Workflows and

procedures;

• ROAZ Reports; • NTS-Protocol; • NHG-Standard;

• Way of working at the GP Out-of-Hours, Registration at front-office, triage of a patient, communicating with ED; C • Quality Advisor;

• EMS Operator; • Workflows/SOP’s; • Partners

covenants;

• Quality

documents acute care (ROAZ);

• Way of working, taking phone calls, triage of patient, registration in system, planning of ambulance; D • Manager;

• Ambulance nurse; • Handover forms; • Ambulance Protocol (LAP) • Quality documents acute care (ROAZ); • Way of working in ambulance, show how consultation forms is generated and being send digitally to chain partners;

E • Manager;

• District Nurse. • Handover forms. • Project Plans with chain partner.

3.5 Data analysis

The data collected was analysed by the three steps that are suggested by Miles & Huberman (1994). These three steps consist of data reduction, data display and the conclusion-drawing/verifications. Data reduction can be achieved by selecting the relevant quotes and codes of the transcribed interviews. For coding, both theory and data driven codes were used. Theory-driven codes, which can evolve from a project’s research goals and questions (DeCuir-Gunby, Marshall & McCulloch, 2010), were derived from the operationalization of the constructs in section 2. As stated by DeCuir-Bunby et al. (2010) coding is the essential step of labelling pieces of data that are connected to a specific subject. “Codebooks are essential to analyse qualitative research because they provide a formalized operationalization of the codes” (DeCuir-Bunby et al., 2010: 138). By comparing each quote with previous quotes in the same category, theoretical properties of categories and the subcategories of these properties were developed (Voss, 2009). Additionally, looking at the collected data in several different ways prevents the danger of reaching incomplete and false conclusions (Eisenhardt, 1989). Figure 3.3 illustrates an example of how the codebook has been established.

FIGURE 3.3

Example of theory-driven coding framework

analysis. These codes were utilized to find new information and insights that was not considered before.

TABLE 3.6 List of data-driven codes

For the analysis of the results, Excel has been used to capture the quotes of the different interviewees. This enabled the researcher to compare quotes with previous quotes in the same categories and subcategories. For example, relationship is a subcategory of the category commitment and eventually contributes to the code family of the organisational perspective. For every quote within the codebook it could be derived to which interviewee the quote was related. In this way, dominant relationships between the different variables were emerging for each stakeholder and this enabled the researcher to make causal maps. The different relationships were either positive or negative towards inter-organisational information sharing and some variables also moderates and mediates the situation. Within the analysis in section 6, the relationships between the variables are supported with “illustrate of events” to make the relationships evidence-based and strengthen the outcome. This way of analysis provides an integral overview of the interdependencies between the variables that were tested, which eventually evolved into a revised empirical framework.

Code Family Code Example

Organisational

Perspective Stakeholders Interest

Quality of Care

“We’ve integrated the x-rays at the GP services but now the ED has less patients to see, which will cost the hospital a Diagnosis Treatment Combination, in other words, money”

“The barriers in information sharing strongly influences the quality of care. A patient has to tell his story like three to four times at every care provider, this certainly influences the experience of the patient in terms of care quality”

Technical

Perspective Information quality “I don’t always get my information on time or in the standard way. If you have GP’s which ignore the protocols, it is hard to treat the patient in the same way when you have been provided with full information”

Information sharing technology “Our systems are not connected, most of

5 RESULTS

This chapter of the report contains the results of this study. The first aim of this study was i) how are healthcare delivery processes between care providers organised in terms of patient information sharing. The results will therefore start with an overview of how the acute care chain is organised and how the processes are executed at the Emergency Department of the hospital. The second aim was ii) how do enablers and barriers influence the information sharing practices within the healthcare delivery processes between the different care providers. The results of the second objective are outlined according to the conceptual model presented in chapter 2.

5.1 Describing the acute care chain

The acute care chain can be described in many ways and is also experienced differently by different stakeholders in the chain of acute care. During the interviews, all the stakeholders were individually asked to describe the acute care chain and the goal of the chain in their own words. The outcome of this question resulted in different explanations, but the essence of the care chain remains the same for all stakeholders which resulted in the following definition: “The acute care chain is an organisation with individual entities whom collectively work together and becomes active when someone has an acute care need which need to be covered by multiple stakeholders”. To provide the reader with a general view of the acute care chain, the healthcare delivery process of the chain will be explained briefly.

means of own transportation or ambulance transportation. Also, when the need for care is of high urgency (e.g. triage level 1), an ambulance is called straight away.

When the patient or bystanders call 112, the EMS is answering the phone and performs triage through the phone with a certain protocol. When the transportation is done by the ambulance, another triage moment and exchange of patient information takes place. The ambulance transports the patient to the ED of the hospital and at that moment another handover process is conducted. If the patient is discharged from the ED there are several possibilities: the patient is either send home, to a nursing home, to a pharmacist to get medicines (also after treatment of the GP) or the patient is in such a condition that it needs to be admitted to the hospital. The process of acute care is visualized in Figure 5.1 and illustrates the different paths a patient can follow within the chain of acute care.

FIGURE 5.1

Path A of the acute care chain

FIGURE 5.2

Path B of the acute care chain

At the specific region were this research is conducted, the GP out-of-Hours is integrated at the ED of the hospital since last year. Both parties made the agreement to use the same triage system. ED processes are triggered by the arrival of a patient and patients have a defined, but variable set of health problems. They are categorised by their level of acuity, this triage system has five levels of urgency ranging from 1 to 5, where triage level 1 represents very high urgency and 5 stands for low urgency. Whether a patient is being transported to the ED by ambulance, came by self-referral, is diagnosed by the GP or GP Out-of-Hours, whenever the acute care need is assessed with triage level 1 or level 2, then the following process will be activated at the ED of the hospital:

FIGURE 5.3

Over the year 2017 at least 1231 patients were diagnosed with triage level 1 and 2. The remaining patients were diagnosed with triage level 3, 4 or 5. Additional diagnostics consists of medical imaging or lab tests. If patients were diagnosed in the higher triage levels, then a slightly different path will be followed which is illustrated in figure 5.4. With the integrated GP services at the ED, one front-office is established in which a patient can register either for the ED or for the GP out-of-hours and where triage is executed only once. This resulted in less rework and positive results towards collaboration and efficiency.

FIGURE 5.4

Patient routing at the ED with triage level 3, 4 & 5

To summarize the whole process, a patient can get involved in the acute care chain in different ways and can follow different paths. This means that the complexity of the acute care chain is high in general and can be subject to inefficiencies continuously. Therefore, it is of great importance that information about the patient and information sharing during handover process is executed in the right way, in the right amount and by the right care provider. The next section will discuss which information is shared and how it is shared when a patient goes through the whole chain of acute care.

5.2 Information sharing practices between chain partners

patient is provided with a referral note, this note is a hard-copy letter which is retrieved from the GP’s information system. This note consists of background information of the patient and the current complaint or care need. The patients then provide the referral note to the front-office of the ED or GP out-of-hours and another triage is conducted. This triage is based on the Dutch Nationale Triage Standaard-protocol (NTS) and is conducted in a digital system, which transfers the assessment information to the ED physicians. The general patient information is being stored in X-Care which is the hospital wide information system. Based on this information the treatment is executed. The referral note is manually scanned by the ED and then allocated into the Electronic Health Records (EHR) of the hospital. When a patient has an acute care need in the out-of-hours then the patient goes to the GP Out-of-Hours organisation. This is a different GP then the fixed daily GP of the patient and this GP is not able to retrieve information from the patient files due to IT compatibility problems and information ownership. This means that the temporary GP is not able to retrieve information about the patient on the same level as the daily GP. This already provides tensions in the consult due to a lack of patient information and not knowing each other.

out-of-hour services, then the temporary GP does not get this consultation form, only the ED will receive it.

Once a patient has been treated at the ED and the patient will be discharged, a discharge letter is generated. This discharge letter is send by means of secured email to the fixed GP of the patient. The discharge letter consists of information about the consult and treatment that has been provided by the ED and will only become available to the fixed GP, not the temporary GP in the out-of-hours services. This is because of the treatment relationship as mentioned earlier. If the patient needs to go to the pharmacists after being discharged at the ED, then the pharmacist is able to retrieve information about the patient through the EHR of the hospital and also calls with the ED physician or the GP to retrieve additional information about the patient situation. If a patient needs care at home then the district nurse will also be involved, a GP or ED physician will only share limited medical information with the district nurses they find necessary to share, there are no agreements on this.

In practice, all the medical information about the patient is being generated over and over again by every care provider in the acute care chain. The patient has to tell his story multiple times (if the patient is able too) and the information that is being generated is stored decentralized at the specific care provider. None of the care providers is able to retrieve information about the patient from each other digital systems or in a centralized way.

"The patient has to tell his story over and over again, this is not favorable for the quality of care. If the GP had an overarching system where everyone can retrieve information from the

patient, this would improve the service of the chain significantly" (ED Physician).

district nurse will store this information in the information system of the district care organisation. When the GP conducts a visit to that specific patient this new information is not available for the GP and has to call with the district nurse to retrieve the updated information which can be very time consuming.

The same counts for the GP out-of-hours, if a patient comes in the afterhours then the patient files are not accessible, which means that patient specific agreements are not known. During day time services it is already a time-consuming activity to retrieve information about the patient at the different care providers involved, in the out-of-hour services it is almost impossible to retrieve historical data about the patient. Because of this, patients, especially frail elderly, are treated at the ED while the treatment was not necessary. All these barriers towards patient information sharing has influence on the quality of care delivered. The interviewees state that this way of providing care is not very successful if you look through the eyes of the patient. The patient goes through the whole chain and is being questioned by every care provider that the patient sees. The interviewees say that a patient most of the time is surprised that care providers do not have any information available since they already have been in an ambulance or the ED before.

“The patient is always surprised that we don’t have any information about them, they call 112 and then they say “yes but you know this information because I also called before” they have

no idea that we can’t share anything with the chain partners. This has influence on how the patient perceives the quality of care” (Quality Advisor EMS).

Next to that, not all the chain partners are using the same protocols for handover processes which also influence the amount of information and the information quality that is provided to the next care provider. If any of the care provider is provided with full information about the patient it is easier to prepare the care services needed for that specific patient before the patient arrives at the care organisation.

6 ANALYSIS

The results from the interview data are analysed in this section in order to evaluate the influence of the variables within the three perspectives on inter-organisational information sharing within the chain of acute care. Analysing the results will provide insights on how the relationships between the several constructs and variables are present and how these are interdependent from one another. Findings in the analysis are underpinned with so called “illustrate of events” and “quotes” to strengthen the outcome. This section will end with an integral analysis of the different variables resulting in a revised empirical framework.

6.1 The Organisational Perspective

From the start of this research, the organisational perspective consisted of four variables which were pre-defined within the research framework. These constructs were all measured in the interviews conducted with the stakeholders involved. Next to the pre-defined variables, inductive variables arose whilst analysing the interview data. Within an acute healthcare setting, information sharing also seems to be influenced by the stakeholders’ interest. All the variables that were found in the organisational perspective are outlined and elaborated below.

Trust

information sharing practices, everyone who has a treatment relation with the patient, must have insight in the patient information, but only when there is permission of the patient explicitly”. This implies that the policy perspective is also influencing the information sharing practices between the chain partners and thus interdependent of each other. The EMS operator stated: “Yes I trust the chain partners, this is essential for collaboration. This trust does not necessary mean I am willing to share more information, I am bounded to our protocols and agreements, such as NEN7510 and the new AVG-legislation”. To summarize, the variable trust has a strong influence on the information sharing practices amongst chain partners but can also be negatively moderated by the variables policy & legislation.

Commitment

relationships. Since our presence at the GP out-of-hours and the ED of the hospital, I also have to work with different GP’s that I don’t know that well, and these relationships are not very developed yet”. This implies that process design is also a crucial factor in establishing relationships and commitment, which in turn strongly affects the information sharing practices between organisations in a positive way. This is confirmed by the manager of organisation B which said: “Because of the proximity and intensive collaboration with the ED, we established a very intense relationship with this chain partner rather than with people from the ambulance services for instance. The reason behind this is that you meet and ran into each other every single day, next to that everybody sees each other as equal and as a colleague”. This event illustrates the importance of commitment to the chain but also how integrating services and proximity of chain partners can help breaking down barriers in the collaboration. Getting to know each other and have daily talks enhance patient information sharing practices. The quality advisor of organisation C confirmed that frequent meetings and establishing agreements really help to build commitment in the chain. The quality advisor stated: “The chain partners that are the closest to our organisation are the ambulance services and the ED of the hospital, because these are the partners we work together with the most. The relationship with the GP Out-of-Services organisation is getting better and better, also because we recently established new covenants about how we share information and collaborate”. Commitment to the chain is definitely present in the chain of acute care. Process design and policy turns out to be the variables that positively moderates between commitment and inter-organisational information sharing.

Shared Vision

and the integration of the GP Out-of-Hour services, I still think we have this island mentality". This statement is contradicting to what the Manager of organisation B stated about seeing each other as equals and colleagues, but once the researcher validated this statement at the manager the following response was provided: “There is no shared vision yet, but that’s a thing that we as a chain of acute care providers need to work on in the future. I hope this is coming soon”. It also came forward that the district care organisation feels less involved in the care chain then other chain partners are. The manager and the district nurse of organisation E felt less adopted by chain partners since the collaboration with this organisation is still in a beginning stadium. “We need to have more eye for each other, every organisation is still acting from a stand-alone position. People are still acting in way that is good from themselves and not for the whole chain, we need to develop an overarching vision, so the chain can work towards mutual goals and provide patient-focused care”. To summarize, despite all the collaboration efforts of the chain partners there is still no shared vision established yet. In this case, it means that the lack of shared vision will lead to less information sharing practices. Establishing a shared vision means that collaboration in the chain can reach a higher level only when chain partners share a common view and goal for the whole chain of acute care.

Stakeholders’ interest

insures still ensures competitive attitudes and different interests”. This statement can be confirmed with a good example of the integrated GP Out-of-Hours at the ED of the hospital. The Manager of organisation A stated: "Of course, there are different stakeholder interests, a good example is integrating the X-ray function into the GP Out-of-Hour services. A GP can now redirect a patient straight to the X-ray department, which is nice because this filters’ out unnecessary patients at the ED, but it costs the hospital a Diagnosis Treatment Combination, in simpler terms "money". These contradicting stakes are acknowledged by every interviewee in this research. The chain partners are transparent and honest to each other if it comes to this kind of issues, because they also recognize the benefits that a patient has from this set of services. The GP of Organisation B said: "So more intense integration and information sharing is subject to different stakes, that's a typical thing for acute care, but if you look from a patient perspective, the quality of care improves". This suggests that integration and information sharing not only comes with different interests of the several stakeholders but seems to be beneficial for its purpose at first, delivering better quality care to the patients.

Process Design

partners because of the established trust, commitment and process design, even when it is behind legal boundaries. Also, recently a pilot-project started with the district care institution who’s providing district nurses to the ED. “Currently we are executing a pilot-project in which the district nurses are present at the GP out-of-hours services, this establishes trust between the GP’s, the ED and the district nurses and it’s also good for getting to know each other”. The proximity of organisations and improving the process design not only results in better relationships and more intense collaboration, but also in more standardized processes and shared protocols. The manager and the ED physician of organisation A confirmed this with the following statement: "At the ED we integrated the GP Out-of-Hours services, this results in better information sharing and triage of a patient is only executed once, with standardized protocols, we speak the same language now". All these arrangements result in enhanced patient information sharing and improve the care delivered to the patient.

6.2 The Policy & Legislation Perspective

Regarding the policy and legislation perspective, the participants could not provide that much information as expected. But analysing the data demonstrates that this theme is one of the greatest barriers in information sharing. The Dutch government has established a new privacy legislation which impedes patient information sharing even more. With this new privacy legislation, some systems have to be redesigned in terms of information sharing.

Policy

it is urgent to share information, we just ask the patient for permission. We don't have any chain agreements, we just act trough common sense and legislation". This illustrates that the established policies are not familiar with the employees and they just act based on the rules they know by heart. Policies regarding information sharing seems to be very fragmented as well, every chain partner have certain agreements but only with the chain partner they collaborate with the most, the rest of the chain seems to be underspecified in this context. The Manager of organisation E stated: “We have established policies about sharing patient information, but this is only with the GP’s of the patients we are treating at that moment and not with all the other chain partners, so this differs as well”. The policies are of dyadic nature and only active between district nurses and the fixed GP’s of the patients, not between other chain partners. Every interviewee stated that new policy agreements were made in the overarching organisation for acute care, which is established based on the initiative of the government. The government pressurizes the acute care organisations to collaborate more and to improve the whole care chain. Within this policy there are also agreements about the information sharing practices: “A new policy is established called “kwaliteitskader spoedzorg”, in this policy, chain partners made agreements which says that we have to intensify our information sharing practices instead of reinventing the wheel over and over, which is a bad form of providing care”. But all the interviewees state that this policy is contradicting with the new privacy legislation. All the interviewees recognized the following problem: “There is a paradox present in these policies, on the one hand we have to be extremely careful due to privacy legislation, on the other hand we want the best possible chain collaboration, and this asks for transparency and intensified information sharing”. This outcome implies that policy and legislation are two contradicting issues and that enhanced information sharing is being blocked by legislation in general. The willingness to intensify information sharing is absolutely present with in the acute care chain but is blocked by barriers established by the government.

Legislation

hand the government restricts us by law and legislation". The legislation part seems to have a very strong influence on collaboration within the chain and it also influences the quality of care provided by the whole chain in a negative way. The GP said: "In our occupation there are a lot of rules and procedures which are established by the government, this can be a big barrier in information sharing and deliver quality care". Legislation also seems to have its positive side, because it makes the chain partners more aware of which information they can share and which they cannot. all the interviewees acknowledged that current legislation made them careful with sharing information and commonly check if it can be shared or not. They also stated that if the outcome of the patient will improve if certain information is shared, they are prepared to share it with a chain partner to improve the outcome of the patient and the care delivered.

6.3 The Technical Perspective

The technical perspective in healthcare context forms a substantial barrier and does not enable information sharing at all. The IT-systems within the chain of acute care are all fragmentised and are stand-alone systems. Every chain partner has his own software programs which is not accessible for an external party. The IT systems are currently bounded to internal usage and therefore not foster information sharing. The data sharing standards are sub-optimized because certain protocols and standards are only used by a part of the chain. Information is being shared hard-copy, by fax or just by phone which results in a high amount of paper work and low information quality.

Compatibility

that systems are not connected. This is especially the case in the out-of-hours and with frail elderly patients. These patients have made certain treatment agreements with their own GP and these agreements are unknown for the rest of the chain partners. For instance, "The biggest compatibility problem at the moment is the one with the frail elderly, because we can't look into each other systems, we don't know which agreements the patient made with the GP and then a lot of unnecessary treatment takes place". The Manager of organisation B stated that some of the daily GP’s are connected to the Landelijk SchakelPunt (LSP) which provides the GP out-of-hours the possibility to look into the patient files during the out-of-hour shifts. As being stated: “If there is a treatment relationship, then the GP Out-of-Hours has insight in all the patient information that a patient has at its own daily GP. The GP is able to communicate through LSP, which gives a summary of the patient to the GP Out-of-Hours, only when the patient gave permission to do so”. The pharmacist of organisation A also works with the LSP and is able to retrieve information about the patient at the specific GP or pharmacists’ organisation the patient is involved with. The pharmacists arranged this individually with the patients by asking them for permission and a signature to retrieve patient information, to bypass the legislation and also to be able to work more efficiently.

All the systems seem to be fragmentised and sub-optimised and it is only possible for a small part of the chain to share information digitally. There is no chain wide system that enhances information sharing. Another outcome of the interviews is that internal digitisation of information sharing is influencing the external information sharing practices. As stated by the district nurse of organisation E: “We are working with client files, in the past we hand wrote these files which means that if a GP visited a patient at home, he could look into the files and knew what we had done. We made this patient files digital and now this information is not accessible for the GP anymore, he has to call us, and we have to provide him with information in an oral way, this is not very efficient and is also a downside of digitization”. This implies that more internal digitization can negatively influence the information sharing between chain partners.

Data Sharing Standards

would use the SBAR, they will get the right information, with the right amount and of the right quality. But eight of the interviewees also said that it is really a hard job to re-educate GP’s that are in the field already for quite some time. Because not all the chain partners use the same protocol this influences the whole chain. For example, the EMS operator said: “The EMS, the ambulance services and the ED uses the SBAR-protocol, this gives a comprehensive and systematic handover process. The GP’s don’t use this protocol which means that we don’t have enough patient information sometimes”. The findings also prove that chain partners who are using the same triage system, improved their patient information sharing practices as well. The ED and the GP out-of-hours are using one triage system which means they speak with the same language, the same colour codes and the same urgency. For the rest of the chain it can still be quite different such as the Quality advisor of the EMS said: “We all use different protocols, the EMS uses ProQ&A, the GP’s uses NTS or NHG and the hospitals are using the ABDCE standards, it would be great if everyone in the chain would use the same standards regarding patient information and handover processes”. this is also recognized by the district nurse who just use their own simple handover forms instead of using the SBAR-protocol similar to the chain partners.

Information sharing technology

The information sharing technology within the chain of acute care is still conservative. Because of the fragmentised IT-systems which are only developed internally, communicating with chain partners is still executed by phone or fax. "Because our systems are not connected we use a lot of fax, phone traffic and secured email communication to share patient information, it’s not organised very efficiently”. This is widely recognized by all the interviewees in which they state that the IT development in healthcare is very underdeveloped. They all want a centralized system in which they can retrieve patient information to improve the quality of care and also to eliminate all the waste they have regarding patient information sharing.

Information quality

characteristics must be present to fully realize the potential of the information shared between chain partners. For example, the EMS operator provided the following evidence: “Because not every care provider is using the SBAR-protocol, there is less information available than you want, which influences the information quality but also the quality of care that we can give. You can imagine if you provide a hospital full information about a patient, they can prepare their services in a much better way than with half of the information”. The information quality is also strongly influenced by the compatibility problems within the chain of acute care. Because there are no connections between IT-systems the care provider is not able to get full information. The Manager of organisation D stated: “The patient information is generated by the one who has responsibility for the patient, this is not shared through information systems and thus doesn’t provide us with good quality information”. This again, illustrates how the different variables are interdependent of one another and they influence each other in either a positive or negative way. Table 6.1 presents a list of enablers and barriers that occur within the different variables towards information sharing.

TABLE 6.1

Overview of enablers and barriers towards information sharing

Category Enablers Barriers

Trust Informal talks, meetings, proximity,

transparency Legislation, privacy concerns, intensity of collaboration Commitment Regular meetings, mutual agreements,

transparency Involvement in the chain, intensity of collaboration Shared vision Mutual goals Island mentality, absence of shared vision Stakeholders’

Interest Quality of care, integrating services, outcome of the patient Financial issues, capacity, autonomy, consequences of information sharing Process Design Integrated services, proximity and close

communication Intensity of collaboration, mutual efforts in collaboration Policy Covenants with chain partners, mutual

agreements of providing care, patient permission

Legislation, privacy, patient rights

Legislation Awareness, carefulness Bounded framework, privacy legislation Compatibility Couplings between systems, secured email,

LSP Fragmentised IT, internal digitization, conservative technology use Data Sharing

Standards

Shared protocols & procedures Fragmentised protocols & procedures

Information