A case-mix system in mental healthcare – influences on empowerment of clinicians

(1)

A case-mix system in mental healthcare –

influences on empowerment of clinicians

Master Thesis

MSc Business Administration

Organizational and Management Control

University of Groningen

Faculty of Economics and Business

Bart van der Wielen

Bedumerweg 19A

9716 AB Groningen

student number: s2216396

b.van.der.wielen@student.rug.nl

Supervisor: dr. B. Crom

Co-assessor: dr. M.P. van der Steen

January 2017

(2)

Abstract

Since the introduction of a case-mix system in Dutch healthcare, many authors have asserted that this system is experienced negatively by clinicians, especially in mental healthcare. This implies a potential threat for empowerment of clinicians and thereby a potential source of management control issues. Since academic knowledge on this topic was lacking, this study has investigated in depth how the case-mix system influences psychological empowerment of clinicians in mental healthcare. A case-study at a major Dutch mental healthcare provider was performed, including ten semi-structured interviews with clinicians. Six mechanisms were identified through which the case-mix system negatively influences one of the four the cognitions of psychological empowerment. Influences related to the cognition meaning were found to be most important, followed by influences related to the cognitions autonomy and impact. The cognition competence was found to be unaffected by the case-mix system. Three moderating variables were identified that decrease the negative effect of one or more mechanisms; one moderator was found to have a positive effect on a mechanism. The findings provide managers in mental healthcare with insights that form a knowledge basis for improving empowerment of clinicians. Future research should further test the conceptual model, clarify if the findings also hold for somatic healthcare organizations, and investigate if the relative importance of cognitions as found in this study holds outside healthcare.

(3)

1. Introduction

Dutch healthcare expenses as a percentage of GDP have known a long period of continuous growth since 1972, up to a percentage of 14.4 in 2012. In the decade from 1998 to 2008 these expenses have even doubled (Centraal Bureau voor de Statistiek, 2016). This growth can be explained by several factors, including an ageing population, a demand for better quality care by citizens as a result of increased welfare, and technological innovations that enable better but also more expensive treatments (Van der Horst et al., 2011). The Dutch government has put effort into stopping this trend, resulting in a gradual decrease of this percentage since 2012 to 14.1 in 2015 (Centraal Bureau voor de Statistiek, 2016). One of the means of doing this was establishing market forces in the healthcare sector by introducing a new health insurance law in 2006. This resulted in a regulated healthcare market with a triangular relationship between patients (or clients), healthcare providers and health insurers, in which patients pay monthly premiums to their health insurers and insurers pay the healthcare providers for the care provided to their clients (figure 1).

Figure 1. The regulated healthcare market (based on Janssen & Soeters, 2010)

Like any market, a healthcare market requires the definition of prices and products. For this purpose, a case-mix system is used. Such a system is based on the idea of Diagnosis-Related Groups (DRGs) which is a patient classification system used to generate homogeneous groups of patients with the same diagnosis. The assumption is that the patients within these groups require a similar treatment and therefore employ a similar set of resources (Preston et al., 1997). Such a classification therefore improves planning of medical activity and makes the subsequent allocation of resources more manageable (Bloomfield, 1991). The relative mix and volume of these homogeneous patient groups can be described as the case-mix (Fetter et al., 1980).

(6)

reimbursement of healthcare services (Oostenbrink & Rutten, 2006; Jansen & Soeters, 2010). Because of the applications of the case-mix system for activity planning and costing purposes, it can be seen as a management control system (Chenhall, 2003). The use of DBCs for reimbursement purposes should lead to more transparency and thereby enable competition between healthcare providers in a market for healthcare provision. In the Netherlands, health insurers have been granted a central role in this market. Insurers negotiate on behalf of their clients with healthcare providers for higher quality treatments and lower prices. In the end, these reforms should lead to a situation of higher quality of care and decreasing healthcare expenses (Hoogervorst, 2004). After the introduction of the new system in somatic healthcare in 2006, the DBC system was also gradually introduced in mental healthcare, starting in 2008 (Nederlandse Zorgautoriteit, 2011).

Ever since its introduction, however, the Dutch case-mix system has received a substantial amount of criticism from the mental healthcare sector. One of the major issues is that the system is said to inhibit the autonomy or discretion of qualified professionals (Hees et al., 2006; Van den Berg, 2006; Palm et al., 2008; Van Sambeek et al., 2011). Also clinicians experience that the focus on efficiency interferes with their desire to provide high quality care and the ability to pay attention to complexities (Palm et al., 2008; Van Sambeek et al., 2011). All in all, the case-mix system in the Netherlands seems to receive little support from practitioners which is a threat for intrinsic motivation of professionals (Van Sambeek et al, 2011). As management control is not only concerned with planning and costing, but also with motivating employees to ensure that they act in the best interest of the organization (Malmi and Brown, 2008), this threat to employees’ motivation is a relevant issue of concern from a management control perspective. Research on the effects of similar DRG-based case-mix systems in healthcare (both mental and somatic) in different countries is available. However, these studies generally focus on the effects on quality of care or production planning (e.g. Rosenheck et al., 1990; Rosenheck & Massari, 1991). The effects on mental health professionals working with the system – the clinicians – remain underinvestigated (Lehtonen, 2007).

(7)

healthcare. Second, the clinical pathway (the sequencing of all interventions by staff members based on a particular diagnosis) in mental healthcare is less predictable than in somatic healthcare (Emmerson et al., 2006; Evans-Lacko et al., 2008) and it is more difficult for clinicians to demonstrate which treatment will be most effective (Frank et al., 2000). Therefore, it is more difficult for financiers (such as health insurers) to understand and be in control of the care that will be provided. This is a reason for them to take more control over the primary process of healthcare provision itself, more than is the case for somatic healthcare, and thereby impair the clinicians’ autonomy. The unpredictability of the care pathway also creates an extra tension for clinicians: the case-mix system forces them to establish diagnoses in the beginning of the care process, whereas this is more difficult in medical healthcare than in somatic healthcare (Palm, 2008). The tension created by this might therefore be larger and hence more influential than in hospitals. Finally, a growing number of news articles about failures within the Dutch mental healthcare sector, fueled by a general concern about the added value of the sector, have pointed more attention towards mental healthcare providers and their performance. The societal pressure on these organizations to perform well is therefore high (Gupta Strategists, 2016).

(8)

criticizing the system that organizations do not have any influence on. Finally, Van Sambeek et al. (2011) did not apply a conceptual framework to their study but provided a descriptive presentation of their findings. It is therefore difficult to synthesize the findings into a theoretical insight that is applicable and useful for a broader, perhaps international, audience.

To tackle the latter issue, this study makes use of the concept of psychological empowerment. Psychological empowerment is a concept from psychology and can be most easily defined as intrinsic motivation (Thomas & Velthouse, 1990), a psychological state of employees that results from their working conditions (Laschinger et al., 2001). It consists of four cognitions: meaning, competence, impact and autonomy. Many of the implications of the case-mix system that have previously been mentioned by Van Sambeek et al. (2011) and others can be related to one or more of these four cognitions, which makes psychological empowerment an appropriate framework to synthesize these findings. Furthermore, empowerment has a strong link with management control. Malmi and Brown (2008) state that an organization with empowered employees requires less extensive management control systems to control the behavior of their employees. Simons (1995) also relates control to empowerment by outlining the importance of a balance between control (here exerted by the DBC-system through control of costs control of behavior) and empowerment. However, the amount of academic research available on empowerment related to case-mix management is limited.

Besides synthesizing the findings into a theoretical framework, this study also aims to enhance our understanding of the way the case-mix management system influences clinician empowerment. So far, research concerning psychological empowerment has been mostly quantitative and aimed at revealing antecedents or consequences of psychological empowerment (e.g. Koberg et al., 1999; Laschinger et al., 2001; Drake et al., 2007; Hall, 2008; Mahama & Cheng, 2012; Kong et al., 2016), answering to the call of Spreitzer (1995) to further examine the effects of organizational manipulations on empowerment. However, Zimmerman (1990) already argued that the emphasis on quantitative research methods limits our understanding of the concept. He states that it is necessary to “develop

research strategies that incorporate qualitative procedures and the voices of the research participants.” (Zimmerman, 1990: p. 176). Also he emphasizes the importance of considering “environmental influences, organizational factors, or social, cultural or political contexts” (p. 173) in

(9)

the non-weighted average of the Likert-scale based scores on the four cognitions). However, it is important for managers to know the relative importance of the influences in determining perceived psychological empowerment, so that they know which ones are most worth paying attention to. Thereby this knowledge gives direction towards the managerial agenda for improving psychological empowerment in the organization. The third contribution of this study is that – again following Zimmerman (1990) – it takes a first step in identifying moderating variables that influence the prevalence or importance of the effects found. Existing literature on empowerment has paid attention to variables that influence empowerment. However, moderating variables have not yet been discussed. Knowing under which circumstances certain influences play a more important or less important role is also important step in drawing the managerial agenda for coping with empowerment issues.

There are several reasons why managers in mental healthcare should have this knowledge about the effects on the case-mix system on the professionals in their organization. Understanding how professionals perceive the system enables a more effective implementation of policy if this is adjusted to these perceptions. This understanding also enables managers to cope with potential dysfunctional behavior that might results from diminished intrinsic motivation, which is traditionally one of the main objectives of management control efforts (Merchant & Van der Stede, 2007; Abernethy & Chua, 1996). Eliminating negative effects or enforcing positive effects for employees not only contributes to a higher employee satisfaction but also enables these employees to use their energy for their patients and thereby to increase their quality of care, which should be one of the main objectives for managers in healthcare. Moreover, in the light of the negotiations between healthcare providers and health insurers, a clearer picture of how the case-mix system affects the professionals providing the care adds clarity to the conversation. As both parties should be concerned with the quality of care provided, both should be interested in satisfying the mental health professionals providing the care. Finally, also the political field could benefit from this research as, according to Lipsky (2010), the effects of government policy are most directly reflected by the people working with it. Also Tummers et al. (2012) indicates in his study about DBC implementation that it is important for the debate about such public policies that the experiences of the public professionals implementing these policies are addressed. Apart from the applicability to the Dutch healthcare sector, this research could also be valuable for other countries that use DRG-based systems, which nowadays includes the U.S., Australia and a large range of European and Asian countries (Tan et al., 2014; France, 2013).

The Research question used in this study is the following:

(10)

This research question is divided into three sub-questions:

1. How does working with a case-mix system influence clinicians’ perceived meaning, competence, impact and autonomy?

2. What is the relative importance of the influences of the case-mix system on the four cognitions of empowerment?

3. How are the reported influences of the case-mix system affected by moderating variables?

Given the explorative nature of the research questions, the study is a qualitative one, aimed at developing theoretical insights and building a conceptual model. Ten semi-structured interviews with clinicians, collected within one organization providing mental healthcare, form the main data source of the research. As an addition to the information provided about the consequences of using the case-mix system, clinicians have be asked how they think their level of empowerment could be improved, in order to be able to provide better recommendations for management.

(11)

2. Literature review

In this section, an overview of the existing literature on de subject will be presented. First, the Dutch mental healthcare sector and the associated case-mix system will be introduced. Then the main theoretical concepts related to case-mix will be discussed, resulting in a clear overview of the case-mix terminology. Also the concept of empowerment will be defined and explained in conceptual terms and related to management control. After that, an overview will be given about the research previously conducted concerning the influence of case-mix systems on empowerment. This results in a conceptual model presenting that forms the basis for the empirical part of the research.

2.1. The Dutch mental healthcare system

Mental healthcare is aimed at prevention, treatment and cure of psychological diseases, enabling people with chronic psychological diseases to participate in society, and helping people who are confused or addicted and do not seek help themselves. The care provided in the Netherlands is divided into basic mental healthcare, provided by GPs, social workers and first-line psychologists, and specialist mental healthcare, offered by specialized institutions (GGZ Nederland, n.d.). Mental healthcare expenses constituted € 6,6 billion in 2014, equal to 8% of total healthcare expenses (excluding welfare, youth and childcare) (GGZ Nederland, 2015).

Since 2006, registration of DBCs in Dutch mental healthcare is mandatory. A DBC is defined by the national DBC-workgroup as “the whole of activities and operations of a hospital and medical

specialist that results from the demand for care a patient is consulting a specialist in the hospital for.”

(12)

job types that are allowed by law to be a head clinician (e.g. psychiatrists and nursing specialists), but health insurers can have their own additional requirements for head-clinicianship.

2.2. Case-mix terminology

The concept case-mix is used in a wide array of contexts and therefore sometimes represents different meanings. Therefore it is good to pay some attention to defining the terms related to case-mix and to make clear how the term case-mix system will be defined in this study.

In healthcare, case-mix is defined as the relative mix and volume of homogeneous patient groups (Fetter et al., 1980). Averill et al. (1998) distinguish a clinical perspective and an administrative perspective on the concept of case-mix: “From a clinical perspective, casemix refers to the condition

of the patients treated and the treatment difficulty associated with providing care. (...) From an administrative or regulatory perspective, casemix refers to the resource intensity demands that patients place on an institution.” (Averill et al., 1998, p. 2). This distinction already indicates that

case-mix information is used for various purposes. The information about these homogeneous patient groups is usually captured into a system, which is referred to as a case-mix system. Reid (2013) describes a case-mix system as a system that gives information about the healthcare system and makes it more effective and efficient. A case-mix system can also use case-mix information for registration and reimbursement (Oostenbrink & Rutten, 2006) or funding (Madden et al., 2013). The exact meaning, attributes and applications of a case-mix system largely depend on the country in which the system is used, since case-mix systems across countries are designed and used for different purposes (Steinbusch et al., 2007). Case-mix information can also be used to adjust for case-mix differences or changes, which enables a better comparison between healthcare providers and a more valid internal analysis as well in terms of quality and cost analysis (Averill et al., 1998). When case-mix information is used for accounting purposes, it is referred to as case-mix accounting (Fetter et al., 1980). It is then aimed at providing a complete financial picture of the costs of treating specific types of patients, and is thereby an addition to the management accounting and financial accounting systems, which provide accounting information about departments and the entire organization respectively. When this case-mix accounting information is used for budgeting purposes, it is sometimes referred to as case-case-mix

budgeting (Lowe, 2000). Case-mix management, to conclude, is defined as the management of clinical

activities based on patient categories and the resource allocation based on those activities (Bloomfield, 1991), for example by prioritizing particular groups of patients. Because the aim of this study is to discover how the mix system influences mental healthcare clinicians, in this study the term

case-mix system will be used.

(13)

case-mix system (Oyebode et al., 1990; Sutherland & Botz, 2006). Hence, a system that is in literature referred to as ‘DRG-based’ can also be interpreted as a case-mix system. Therefore, the following sections will not only review previous findings concerning case-mix systems but also review literature about DRGs and DRG-based systems.

2.3. Empowerment

The concept of empowerment will be discussed in two parts. First, the different definitions and uses of the concept will be discussed. It is followed by a section that explains the link with management control in more detail.

2.3.1. Conceptualization of empowerment

The aim of this study is to discover the impact of a case-mix system on clinicians from a management control perspective. The concept of empowerment is very appropriate to assess this influence for several reasons. First, it is a perceptual concept so it can be applied very directly to clinicians. Second, the conceptualization following next will show that it can be linked quite well to many of the issues raised in the debate around case-mix systems. Finally, empowerment has been found to be a mediating factor in the relationship between management accounting and control practices, and individual-level performance of employees (Hall, 2008; Mahama & Cheng, 2012).

Although empowerment has been defined in a variety of ways, its definitions remain centered around ‘giving power’, a notion that is central in much of the management control literature (Simons, 1995; Spreitzer & Mishra, 1999). However, the conceptualization of empowerment has developed over time. Burke (1986) describes this granting of power as a delegation of authority. Conger & Kanungo (1988) call this idea the relational construct of empowerment, but also distinguish a motivational construct which means ‘to enable’ rather than only ‘to delegate’. They define the motivational construct as a process of enhancing feelings of self-efficacy among organizational members by removing conditions that foster powerlessness. Thomas & Velthouse (1990) define empowerment as ‘increased intrinsic task motivation’ and add to previous conceptualizations that the motivational construct can better be described as ‘to energize’. Koberg et al. (1999) define the relational construct as related to an individual’s power and authority relative to others, and the motivational construct to individual cognitions and perceptions that constitute feelings of behavioral and psychological investment in work. Laschinger et al. (2001) also makes a distinction between two perspectives but uses the terms

structural empowerment, which describes the conditions of the work environment, and psychological

(14)

self-determination). Meaning involves the fit between the requirement of the work role and personal beliefs and values. Competence is the individual’s belief or confidence in the capability to perform the job with skill and compares to the self-efficacy concept of Conger & Kanungo (1988). Impact can be described as the influence people have on outcomes at work (Spreitzer, 1995). Autonomy, or self-determination, is the sense of having a choice in initiating and regulating actions, or the autonomy in the initiation and continuation of work behaviors. Following most recent researchers on the topic of psychological empowerment (Drake et al., 2007; Hall, 2008; Mahama & Cheng, 2012), in this study the conceptualization of psychological empowerment consisting of meaning, competence, impact and autonomy will be used.

It should be noted that, despite the fact that both Koberg et al. (1999) and Laschinger et al. (2001) conceptualize psychological or motivational empowerment using the same four cognitions, Koberg et al. describe it as a process whereas Laschinger et al. describe it as a psychological state, hence an outcome rather than a process. It can easily be argued that the process of empowerment leads to a psychological state of perceived empowerment. Since it is assumed to be easier for clinicians to describe how their current perceived state of empowerment differs from the situation before the case-mix system, than to reflect on a process of changed empowerment over a period of ten years, the notion of a psychological state of empowerment will be used in this study.

2.3.2. Empowerment and management control

Research about the effects of empowerment of healthcare employees in healthcare shows that it has numerous beneficial effects. Psychological empowerment was found to be a predictor of increased productivity or effectivity, decreased propensity to leave the organization (Koberg et al., 1999), decreased job strain (Laschinger et al., 2001), increased job satisfaction (Koberg et al., 1999; Laschinger et al., 2001), innovative behavior (Knol & Van Linge, 2009) and it is beneficial for the prevention of burn-outs (Boudrias et al., 2012). The last three effects mentioned have been validated once more by a systematic review of Wagner et al. (2010). This empirical evidence shows that it is highly relevant for healthcare organizations to pay attention to maintaining or improving the levels of psychological empowerment of their employees.

(15)

therefore interesting to see how this balance between control and empowerment is influenced by the implementation of a case-mix system in mental healthcare. Lowe (2000) also addresses this potential imbalance, although he does not use the concept of empowerment. He does highlight the potential problem of decoupling that may arise if the case-mix system is little related to the backstage operations (Covaleski et al., 1993). In terms of empowerment this could encompass a conflict of meaning (if the case-mix system emphasizes values that are different from existing professional values) or a lack of impact (if the case-mix system promotes certain behaviors that professionals do not see as benefiting patients). Decoupling was already recognized by Van Sambeek et al. (2011) in Dutch mental healthcare as a result of the implementation of the case-mix system. Lowe (2000) suggests that the low level of decoupling he found in New Zealand may be explained by the fact that the case-mix system here had an internal focus. When the external focus is prevalent (for example the system is implemented for reimbursement), such as in the US, higher levels of decoupling may be found. This is in line with the observation that in the Netherlands the case-mix system also has an external focus and the subsequent decoupling found by Van Sambeek et al. (2011). In the study of Doolin (1999) the case-mix system of the hospital was actually used as a means to empower clinicians. The system provided clinicians with information that enables them to better manage themselves. This study shows that there can also be a positive effect of the case-mix system on empowerment, rather than a negative effect.

2.4. Influences of a case-mix system on empowerment

Previous research on the influences of case-mix systems or DRGs in mental healthcare is diverse. The earliest research stream in this field is mainly concerned with resource planning and focuses on the effects on certain performance measures such as length-of-stay and occupation of beds (Rosenheck et al., 1990; Rosenheck & Massari, 1991). Also DRGs are evaluated as a means to predict resource use (Oyebode et al., 1990; Burgmer & Freyberger, 2002). The applicability and accuracy of DRG in mental healthcare receives attention as well (Ben-Tovim & Elzinga, 1994; Hunter & McFarlane, 1994; Hay & Pearce, 1996; Lien, 2003). However, international studies on the experience of clinicians in mental healthcare working with a case-mix system or DRGs are lacking.

(16)

case-mix system as suggested by literature. In the empirical part of this study, these influences will be verified and I will investigate under which circumstances they apply.

2.4.1. Meaning

Multiple sources emphasize the new values that have come to play a role in mental healthcare. Palm et al. (2008) state that the desire for efficiency interferes with the opportunity for complexity, depth and quality. Van Sambeek et al. (2011) also suggest that professional values conflict with economic and bureaucratic values that are made very important in the case-mix system. Also transparency is an essential new value, according to them. Hence, a lack of meaning could be caused by a conflict between values of quality, complexity and depth that conflict with efficiency, transparency or bureaucracy.

2.4.2. Competence

According to Hees et al. (2006) and Van Sambeek et al. (2011), the case-mix system leads to a considerable increase of bureaucratic and administrative activities. Some of these are performed by supportive personnel, but also clinicians themselves have to perform a part of these tasks. This means that the whole of tasks and activities of clinicians has changed. It is therefore valuable to evaluate whether clinicians feel skilled enough to perform these newly obtained activities, which tells us if there is a potential lack of competence.

2.4.3 Impact

(17)

2.4.4. Autonomy

Hees et al. (2006) argues that clinicians have to disclose and report what happens in their consulting rooms to governmental regulators and (commercially driven) health insurers. They are however not explicit about how clinicians are actually inhibited by this and how this might affect clinicians’ autonomy. Palm et al. (2008) conclude from their survey research that healthcare insurers heavily interfere with the content of the care-provision process. According to Van den Berg (2006) and Palm et al. (2008), the Dutch DBC-system forces clinicians to establish one diagnosis that is based on the DSM-classification (the generally used handbook that classifies (mental) diseases and symptoms), which limits the discretion in the further treatment of this patient. Van Sambeek et al. (2011) also acknowledge the resistance to work with a DSM-based system because of its constraining implications for the treatment.

2.4.5. Conceptual model

The conceptual model summarizes the potential ways in which the case-mix system could influence psychological empowerment, based on the previously discussed findings. This model should not be interpreted as an attempt to predict all influences up forehand, but rather to serve as a starting point for the empirical part of the study, in which these possible mechanisms will be verified and the model will be developed further. As indicated, moderating variables that moderate the effect of the influences are to be added to the framework.

(18)

3. Methodology

This chapter describes the method of this study and its justification. Following from the research questions, the general research design will be discussed, followed by the data collection method. This section is further specified into a description of the case organization, the selection of interviewees and the structure and measures used for the interviews. Finally, the approach of the data analysis will be discussed. Validity and reliability will be addressed throughout the chapter.

3.1. Research design

The aim of this research is to develop a better understanding of how the use of a case-mix system influences empowerment of clinicians in mental healthcare. There is already (both qualitative and quantitative) research available that suggests that there exists a relationship between working with a case-mix system and the cognitions of empowerment, as discussed in the literature chapter. However, this study is the first one to unite these findings in the concept of empowerment. In addition, knowledge about the relative importance of the different influences and the circumstances under which they play a more important or less important role is also still lacking. A qualitative approach is the best way to develop such an understanding, as it is focused on understanding human beings’ experiences and reflections on those experiences (Lincoln & Guba, 1985). As a basis for the research process, the theory development process as described by Van Aken et al. (2012) was used (figure 3).

Figure 3. Theory development process (Van Aken et al., 2012)

The theory development process starts with defining a business phenomenon that has not yet been explained in academic literature. The phenomenon of interest - the influence of a case-mix system on empowerment of clinicians in mental healthcare - has been discussed in the first two chapters of this paper, based on an extensive literature review. Second, following the theory development process, the phenomenon was observed during a case-study at a major Dutch mental healthcare provider, for which the main data source consisted of ten semi-structured interviews with clinicians. After a structured analysis of the results by coding the transcribed interviews, explanations were developed and the insights were compared to additional literature. The analysis led to additions to existing theories and enriched our knowledge of the way in which a case-mix system influences empowerment of clinicians in mental healthcare. Business phenomenon not explained in academic literature Observation of phenomenon in one or more case

(19)

3.2. Data collection

Semi-structured interviews with ten clinicians from a case organizations form the main data source for this study. A case study focuses on understanding the dynamics in a single setting and is an appropriate method to enhance the understanding of a relationship (Eisenhardt, 1989), which is indeed the objective in this study. After ten interviews, theoretical saturation was reached. That is, new insights from additional interviews were only marginal since the phenomena mentioned have already been brought up in previous interviews (Eisenhardt, 1989). The interviews were semi-structured in order to ensure that the most important issues were addressed in every interview, which enhances the reliability of the findings (Van Aken et al., 2012). On the other hand semi-structured interviews allow discretion to discover which aspects are found to be relevant by respondents (Blumberg et al., 2011; p. 265). The interviews have been conducted in the period between 17 November 2016 and 10 January 2017 and had an average duration of 44 minutes.

In addition to the interviews, other data sources were used to broaden the understanding of the phenomenon at hand and its context. This triangulation enabled the researcher to draw better conclusions and thereby increase validity of the study (Yin, 2003). These additional data sources include: observations of four project group meetings concerned with implementing a new control system for DBC-registration; documentation of a survey among team leaders regarding implementation issues of and opinions on this control system; two meetings with the director of finance, control and healthcare sales; a meeting with the project leader responsible for implementing the DBC-system at the case-organization at the time; a briefing and documentation by the Board of Directors about the quarterly results, state of the organization and future policy actions, in which the case-mix system and the financial implications for the organization were discussed; numerous formal and informal talks with various employees from the departments of controlling and healthcare sales; a briefing and documentation about the current financing system of healthcare in The Netherlands. Discussing the findings with members of the organization helped preventing misinterpretations and therefore increased the validity of the results.

3.2.1. Description of the case organization

(20)

and care is financed and managed in a different way than mental healthcare. The mental healthcare part of the case organization counts over 2,700 employees and treated over 28,000 patients in 2014. Its size makes it a major provider of mental healthcare for a large region in the Netherlands.

3.2.2. Interviewee selection

The interviews have been held with ten clinicians with various jobs and working in different departments. Their specifics are listed in table 2. Since there are many different types of employees within mental healthcare, a conscious selection of interviewees was required. This was done by quota sampling, a purposive sampling method which improves representativeness by selecting respondents based on relevant characteristics (Blumberg et al., 2011: 195). To omit respondent bias and thereby to increase reliability (Van Aken et al., 2012), respondents were selected to ensure a fair representation of different jobs, organizational units, care types and gender. For this purpose, a list was made with the jobs that are relevant for his study. These jobs have been selected on the basis of two criteria which they both need to meet: 1) the job concerns employees who work as a clinician, and 2) the job concerns employees who work with the case-mix system (in this case DBC registration). This resulted in table 1, which also shows the number of employees per job within the case organization.

Table 1. List of relevant jobs

Job Employees Job Employees

Healthcare psychologist 1 142 Physician in training to psychiatrist 20 Senior differentiated nurse 115 Physician not in training 15

Psychologist 2 87 Social worker 12

Psychiatrist 63 System therapist 8

Nursing specialist 40 Psychotherapist 7

Remedial educationalist 3 29 Senior psychotherapist 3 Clinical psychologist 28 Clinical neuropsychologist 3

1) includes healthcare psychologists in training to become a clinical psychologist, clinical neuropsychologist or psychotherapist.

2) includes psychologists in training to become a healthcare psychologist.

3) includes remedial educationalist generalists and remedial educationalists in training to become a healthcare psychologist

(21)

representation of the organization’s diversity. Table 2 presents an overview of the participating respondents.

Table 2. Interviewee specification

Job Gender Age Unit Care type

1 Nursing specialist M 53 Youth Outpatient

2 Clinical psychologist F 55 Adults Inpatient

3 Psychiatrist M 43 Youth Outpatient

4 Psychiatrist F 39 Forensic Inpatient

5 Nursing specialist M 52 Youth Outpatient

6 Psychiatrist F 53 Adults Inpatient

7 Clinical neuropsychologist F 50 Adults1 Inpatient

8 Psychiatrist M 58 Elderly Inpatient

9 Senior differentiated nurse F 40 Adults Outpatient

10 Healthcare psychologist F 47 Adults Outpatient

1) this unit is officially part of the Youth unit but in fact mostly treats adults

The participating interviewees all work in different departments, dispersed over five different cities in the North of the Netherlands. All interviewees received an information letter after the appointment for the interview was settled, which has been included in appendix A.

3.2.3. Interview structure and measures

(22)

and explanations of the scores improved the reliability of results. The scores on the measurement items have been included in appendix C. Clinician number 6 did not score the list as this clinician, despite careful selection of respondents, lacked the experience of working in mental healthcare before the implementation of the case-mix system.

Table 3. Measurement items for empowerment by Spreitzer (1995) Meaning

The work I do is very important to me.

My job activities are personally meaningful to me. The work I do is meaningful to me.

Competence

I am confident about my ability to do my job

I am self-assured about my capabilities to perform my work activities I have mastered the skills necessary for my job

Impact

My impact on what happens in my department is large

I have a great deal of control over what happens in my department I have significant influence over what happens in my department

Autonomy

I have significant autonomy in determining how I do my job I can decide on my own how to go about doing my job

I have considerable opportunity for independence and freedom in how I do my job

Some of the items are about the job or the work of the clinician. Other items ask about job activities. It should be clear that some of the clinicians’ activities, such as DBC-registration, are directly related to the case-mix system, whereas other activities, such as having a conversation with a client, are not. These activities together make up the job. Therefore, when talking about the ‘job’ or ‘work’, if necessary a distinction between both types of activities was made.

In the second part of the interview, specific questions were asked about the influences of the case-mix system on empowerment as suggested in the conceptual model. This part therefore contributes to answering sub-questions one, but also contributes to an answer on sub-question three, as the difference between experiences of clinicians pointed out under which circumstances certain influences play a more important or less important role.

(23)

motivation. They therefore provide results for sub-question two. Finally, as this study also aims to provide recommendations for management, clinicians were asked how they think the implementation of the DBC-system could be altered to best meet the interests of clinicians.

The interview guide as used for the interviews has been included in appendix B.

3.3. Data analysis

(24)

4. Results

In this chapter, the collected data from the interviews will be presented. Following the sub-questions of the research, the influence of the case-mix system on each of the four cognitions of empowerment will be discussed first. Then the insights about the relative importance of the four cognitions will be presented. Finally, based on a comparison of contrasting findings, relevant moderating factors will be discussed.

4.1. Influences on cognitions of empowerment

The influences on empowerment will now be discussed for each of the cognitions separately. 4.1.1. Meaning

The expected influence of the case-mix system on the cognition of meaning was that the values efficiency and transparency are being made more important by the system whereas these are not values that are intrinsically very important to clinicians. Both values indeed manifest themselves in different ways. Regarding efficiency, clinicians confirm that the importance of efficiency is very prominent now. As clinician 5 explains: “I don’t think it is a bad thing that you have to show that you are

working, but now it seems to become more important that I am earning money than that I am making patients feel better.” The importance of production is mainly experienced as it has become a part of

the performance measurement of clinicians. Clinician 9: “Your productivity is actually discussed with

you during performance reviews.” However, as clinician 10 clarifies, this might be more harmful for

motivation than beneficial: “I think intrinsic motivation of clinicians is sufficient to achieve a good

production and that there is no need to apply production norms. In fact, if that norm is emphasized, it influences motivation in a negative way.”Also, clinicians feel that the focus on efficiency and cutting costs comes at the cost of quality of care for the patients. Clinician 9: “It is very hasty and hurried.

People don’t take their time to really discuss patients, because everything has to be done so fast. This comes at the expense of quality of care.” When they cannot offer the quality they want to offer, the

value of quality and helping people conflicts with the value of efficiency. For some clinicians, changing values in their work makes their work less satisfying. Others however are able to rationalize the need to comply with today’s requirements when reflecting their intrinsic motivation. That is, they disconnect the administrative part from what they see as ‘work’. In some cases their intrinsic values, such as acting in the best interest of the patients, even outweigh the pressures from health insurers:

(25)

This illustrates that clinicians’ intrinsic, professional values are so strong that they cannot easily be replaced or changed. It shows that clinicians can go against the system as long as they can justify to themselves that it is for the sake of the patient’s wellbeing. It is also evidence of decoupling as suggested in the literature review: clinicians are complying with the system’s rules but still find ways to do whatever they think is good. Clinician 5 illustrates how strong the intrinsic motivation to treat patients can be: “By dictating what is reimbursed and what is not, the insurer forces me to establish a

different diagnosis.” Stopping the treatment is not considered an option by this clinician. In addition to

decoupling as an influence on impact (see 2.4.3), decoupling can therefore also be seen as a mechanism that diminishes the negative influence of conflicting values as a moderating variable. It is yet unclear why some clinicians decouple more than others.

Having to work more transparent also influences the cognition of meaning, but in a different way. For the sake of transparency, clinicians have to register their activities very extensively. This is a time-consuming activity that is nonetheless not personally meaningful to them at all. Clinician 10: “Even

the shortest phone calls I have to register. Sometimes the time to register it is longer that the activity itself. (…) You have to justify everything you do. Even if you just think about someone for two minutes, so to speak.”. Another example of unmeaningful activities is the required direct contact that head

clinicians periodically need to have with their clients, regardless of how necessary they find this. Clinician 1: “Out of a case load of 30, there are around 25 that only shake hands with the psychiatrist

and that’s it. (…) Then the system says it is correct and the treatment can be invoiced, but the psychiatrist has added nothing to the content.”

The negative influences of the case-mix system on empowerment through the emphasis of efficiency and transparency are confirmed by the empirical data. To this should be added that decoupling can diminish these negative effects.

4.1.2. Competence

There were no clues in literature that suggest a specific influence of the case-mix system on perceived competence. Therefore, this study has tried to find out how the whole of activities and tasks of clinicians has changed since the implementation of the case-mix system and if this affects their perceived competence. Only one clinician reported that the registration itself was difficult. According to others, the challenge that the case-mix system brings is not the registration itself, but the coordination of all the administrative requirements, especially for head clinicians. Clinician 3 explains:

“The head clinician is responsible for the correct execution of the DBC. (…) You can as a psychiatrist make sure that everything is checked and divide tasks, but you are still responsible. That brings a certain perceived pressure.” Also clinicians that meet many different patients in a day experience

(26)

You have to write down everything and that is inconvenient. Especially when I worked at the department where I met thirty different people on a day, because people come to you with questions, they call, patients visit, I had to write letters, and so on. Then you are not going to register how much time you spend on which activity. (Clinician 6)

However, both influences described by the clinicians are characterized more as an inconvenience than as a lack of competence. It also does not lead to a low perception of competence. In fact, this cognition received the highest score of all four cognitions: an 8.4 on average (appendix C). It can therefore be concluded that the case-mix system has no effect on perceived competence.

4.1.3. Impact

When discussing impact, a distinction is made between three types of impact: impact on patient outcomes, impact on efficiency and impact on transparency. From the theoretical framework, it was expected that impact on patient outcomes has decreased due to less available time, and that impact on transparency is low because of decoupling.

Impact on patient outcomes is indeed diminished through the decreased availability of time for patients. This happens in two ways. First, the extensive registration activities take up time that cannot be spent on seeing patients. Clinician 2 explains: “You can only do one thing at a time. In my opinion,

I do not see patients often enough and that is caused by all the administrative activities. For that matter, patients do not benefit from this.” Second, treatments sometimes have to be terminated

whereas, according to clinicians, patients do need further treatment. “You are dealing with financing

streams that are so confined that sometimes funding is stopped whereas there is still a need for help or treatment.” (Clinician 4).

Regarding efficiency, almost all clinicians share the opinion that their work has become more inefficient since the implementation of the case-mix system. “I think the goal was to minimize the time

spent on a patient, but I think that it has accidentally become more time, because you have to register so much.” (Clinician 7). Clinician 6 adds: “Registering and checking the registration has become an entire industry next to the primary process.” Also the meetings with head clinicians that are perceived

unnecessary (see 4.1.1.) are seen as an inefficient use of time. Clinician 8 does support these statements, but simultaneously argues:

(27)

It seems that this clinician takes into account his awareness of the efficiency implications on the organizational level and experiences a positive influence on efficiency. However, for most clinicians, when asked about their own impact on efficiency, this was not the case.

The impact on transparency is perceived differently among clinicians. Regarding transparency towards patients, the process has become more structured. However, “the desire to be transparent towards

patients and health insurers is disproportionate. There are so many processes involved to monitor everything that it becomes a job in itself for a clinician.” (Clinician 7). There is also no consensus on

the transparency of the system as a whole. Some say it is now more clear what is done during a treatment, which makes it more transparent. However, others confirm the suggestion of decoupling that was made in the theoretical section and thereby argue that transparency is not achieved as much:

Many people, like me, do their registration afterwards. They make an estimate: for one I did about this, for the other I did about that, the one you don’t register and for the other a bit more so in the end it all works out. This way you create the impression that you have a good overview of what happens in practice, whereas this is just not the case. It is a bogus transparency. But on these flawed data, policy is actually based. And that, I think, is a very dangerous development.

(Clinician 6)

The empirical data show that there is no consensus on the perceived impact on efficiency and transparency. However, an important point is that, just as is the case with meaning, clinicians admitted that they think of impact as in their impact on patient outcomes. They do not reflect on impact on efficiency or transparency unless asked for, which means that these are not influences that affect perceived empowerment. The fact that clinicians have less time available for patients is therefore the only confirmed influence on impact.

4.1.4. Autonomy

Based on existing literature, it was proposed that autonomy of clinicians is affected by the case-mix system through constrained diagnostication and interference of health insurers. Both proposed effects are supported by the empirical evidence. The process of establishing a diagnosis is constrained because less time is available to establish the diagnosis, since it must be established before treatment can be started. Also, the case-mix system forces clinicians to give patients a DSM-based classification that allows little discretion but is determinative for the further treatment. That is, based on the classification, the clinician has to follow certain guidelines concerning the treatment that affect their perceived autonomy. “If you omit many things from the guideline, you have to explain why you didn’t

do it.” (Clinician 3).

(28)

parts of the treatment that they deem unnecessary. For example, treatments may have to be terminated preliminary (as described in 4.1.2) or treatments are not compensated at all: “Sometimes you cannot

treat a patient that you would like to treat, simply because the patient does not meet all criteria of the classification.” (Clinician 10). The same goes for prescribing medication, as clinicians 5 explains: “Some children benefit from long-acting medication, but that is very expensive. The health insurer only compensates from a part of those costs. If insurers would compensate it entirely, you would prescribe such medication more easily.” He also summarizes the feeling that is held by many

respondents in one statement: “The health insurers have in that sense become in charge of how you

treat and why you treat.”

There is a third, quite specific way in which autonomy is affected since the implementation of the case-mix system, which comes from the distinction that is made between head clinicians and non-head clinicians. First, because non-head clinicians have lost certain authorities. Clinician 9: “I now have to

discuss a diagnosis with a head clinician, whereas before I could establish this myself. It feels like I am accountable to my colleague. That part was taken from us and I regret that.” Second, because

there has been a shift of activities towards head clinicians, who now have many obligations that are scheduled for them. “Autonomy has declined for me. That has changed from an 8 to a 6 or a 7. I see

that in my consultation hours. Those are fully scheduled for the coming weeks.” (Clinician 3). An

example of such obligations is the required direct contact with a head clinician as discussed in section 4.1.1. This influence can be described as a redistribution of activities.

This leaves u with three ways in which autonomy is affected by the case-mix system: constrained diagnostication, interference of health insurers, and redistribution of activities. All effects are negative.

4.2. Relative importance of influences

(29)

The insight that stands out is that the cognition meaning is the most determining one for clinicians. As clinician 10 states: “Essentially it is all about values. The other categories play a role, but in the end it

is good to keep in mind those values. I deliberately chose this profession and this is what I like.”

Whether this also holds for employees in different industries is not sure. Despite the several negative influences on meaning, clinicians still score this cognition with an average of 8.2. This might be explained by the fact that clinicians interpreted the concept of ‘work’ as working with patients, despite the fact that they admitted that a substantial part of their activities now exists of registration and other activities that they do not feel committed to (see 4.1.1). Two clinicians explain this paradoxical observation: “Meaningfulness has not changed for me. But if you also count in the registration, sure it

has changed. Still, if you ask me about the meaningfulness of my work, I see that separate from the side issues.”(Clinician 5).

When I think about the meaningfulness of my work, I think about the work I do with patients. I understand the core of the work and I also understand that it requires processes. But if you look at all the processes, including the administration, I somehow manage to suppress that. (Clinician 7)

Hence, a mental distinction is being made between core activities concerning patients and other activities, which from a psychological perspective can be seen as a coping mechanism. More precisely, it can be seen as a form of cognitive dissonance: clinicians try to avoid the uncomfortable feeling that their work now involves more unmeaningful activities by rationalizing that these activities are actually not part of their work (Gray, 2011). Therefore, although some of the activities in their work are not meaningful at all to clinicians, their perceived cognition of meaning remains high.

Autonomy is affected in many different ways, as discussed in section 4.1.4. It has also the lowest average score (6.6) of all four cognitions. However, most clinicians nevertheless find ways to still do what they think is necessary. We can call this decoupling, as Van Sambeek et al. (2011) already recognized. The clinicians themselves give numerous examples of it. Clinician 5: “The health insurer

does not find an adjustment disorder severe enough to be treated by a secondary care provider. So then we establish a different diagnosis, if we are convinced that this patient does need care from us.”

Clinician 6 explained the same (4.1.1). Clinician 2: “I just keep doing what I do. I just call everything

cognitive therapy.” Another way of retaining autonomy is to fight for it. Clinician 3: “There is still room to treat patients very individually, but you have to justify it.” Clinician 7: “We try to explain to the health insurers again and again why we need time to do proper diagnostication. That is the reason that I still perceive some autonomy.” These statements make it very clear that one should be cautious

(30)

for the negative influences on autonomy. This is especially the case for non-head clinicians. This difference will be further discussed in section 4.3.2.

When clinicians reflected on their impact, they were again referring to impact on patient outcomes, as was the case for meaning. This cognition is perceived relatively low, 6.7 on average. Clinician 7 explains why:

I am only part of someone’s life for no more than an hour per week. For the remainder, there are so many things happening in someone’s body and mind that it makes me think: even if I can only cause a tiny movement that makes that person feel a little bit better, than I am satisfied. But I don’t have the idea that if I try something with a patient and it doesn’t work out, that I didn’t do my job well.

The profession of a clinician in mental healthcare can be identified as one in which achieving results is not a measure for satisfaction, which makes impact as a cognition relatively unimportant for perceived empowerment.

No major work changes were reported that had a significant influence on competence. That is, only one clinician reported difficulties with registration that influenced his perception of competence. Together with the fact that competence is scored an average of 8.4, this cognition is assumed to be the least important one in determining changes in empowerment as a result of the case-mix system.

4.3. Moderating variables

The way in which the case-mix system has influenced clinicians’ empowerment differs among the interviewees. Based on the interview data, four variables that moderate the extent to which influences affect perceived empowerment have been identified. By taking these factors into account, the insights generated from this research can be applied more consciously to specific organizational units or individual clinicians and it also increases the generalizability of the study for other organizations. 4.3.1. Severity of psychiatric problems treated

Since the case-mix system is meant to increase efficiency, there is a tendency of health insurers to scrutinize care processes of which the necessity is doubted. Especially clinicians treating less severe psychiatric problems encounter situations in which the care that they deem appropriate is not reimbursed, which results in a negative influence on autonomy (clinician 10, section 4.1.4) and impact (clinician 5, section 4.2). However, for more severe psychiatric patients, for example in forensic care, the importance of treatment is rarely questioned. Clinician 4 argues: “I deal with more severe

psychiatric disorders so I don’t experience that as much.” Clinicians that treat less severe psychiatric

(31)

The severity of the patients’ illnesses especially becomes visible in the distinction between inpatient and outpatient care. More severely ill patients receive inpatient care (care that involves overnight stay in a clinic). There are two ways in which the difference between inpatient and outpatient care affects the degree to which certain influences of the case-mix system on empowerment are experienced. First, most patients start and finish their care process in outpatient care and only receive inpatient care when their condition deteriorates. Since the majority of administrative activity and responsibility lies in the start and termination of the total care program, the administrative burden is mostly encountered by outpatient clinicians. Second, in contrast to outpatient clinicians, inpatient clinicians do not see their patients on a regular, scheduled basis. This alleviates the pressure by the organization for clinicians in an inpatient setting to be productive. Clinician 2 from an inpatient department clarifies: “They

[outpatient departments] have to see patients 85% of their time. You cannot require that in a clinical setting because I cannot do something with these patients every day.”

Clinicians that treat more severely ill patients, especially clinicians who work in an inpatient setting, experience less negative effects of unmeaningful activities and less interference of health insurers. 4.3.2. Head-clinicianship

Head clinicians are usually psychiatrists or clinical psychologists who have the responsibility of the entire care process, including the administrative justification. They therefore suffer the most from the increased registration requirements: “As a head clinician I experience that I am coordinating many

processes that are not so much related to providing care but with justifying to others what we are doing.” (Clinician 7). This suggests that head clinicians deal more with unmeaningful activities than

non-head clinicians. On the other hand, those non-head clinicians suffer from stronger feelings of decreased autonomy that head clinicians do. Clinician 9 explains how she was affected by the fact that clinicians of her profession (senior psychiatric nurse) are no longer allowed to be head clinician: “I

deliberately became an SPV because I wanted more responsibility and to take control myself. Now I cannot do intakes myself anymore. That was taken over by physician assistants and psychologists. That is actually painful.” Hence, the negative effect on autonomy is stronger and also more influential

on eventual perceived empowerment for non-head clinicians than for head clinicians. 4.3.3. Size and complexity of the organization

Clinician 3 explains the several ways in which the relevance of the organization’s size and complexity plays a role:

(32)

psychiatrists have moved to smaller providers. That collectively happened as psychiatrists in smaller organizations receive more support in their administrative obligations.

Despite the fact that this issue was brought up by only one clinician, the statement that clinicians in larger organizations experience more interference of health insurers seems plausible. However, it should be noted that this effect interferes with the severity of the case-mix as described in 4.3.3. Clinician 1 explains: “Patients sometimes complain that everything goes a lot faster at other, smaller

providers. But I understand why, those organizations treat less severe problems.” Even within large

healthcare providers, the degree of experienced interference of health insurers therefore still depends on the severity of patients’ problems. The influence of this moderating variable therefore deserves more attention in future research.

4.3.4. Decoupling

(33)

5. Conclusion and discussion

This research has studied the implications of the implementation of a case-mix system on the psychological empowerment of clinicians in mental healthcare. It thereby contributes to our knowledge about the way in which management control systems have an impact on empowerment of employees. This final chapter discusses the most important findings of this study. These are compared with the theoretical framework and presented in a revised conceptual framework. Finally, the limitations of the study will be discussed and suggested directions for further research are presented, as are managerial recommendations.

5.1. Conclusion

The results in chapter 4 clarify how psychological empowerment of clinicians is influenced by the case-mix system. For meaning, this occurs through an emphasis of productivity in performance measurement, and the increase of unmeaningful activities. For autonomy, it occurs through constrained diagnostication, the interference of health insurers, and the redistribution of activities. Impact is affected because there is less time available to treat patients. No influences on perceived competence were identified. All influences on the cognitions of psychological empowerment that were found are negative. Positive influences were also identified, for perceived impact on transparency and efficiency. However, transparency and efficiency were found not to be determinative for perceived impact, which makes these positive influences not relevant for psychological empowerment.

The relative importance of the identified influences is largely dependent on the cognition they are related to. Meaning is the most important cognition for clinicians, followed by autonomy and impact. Competence was found to be the least important cognition for perceived empowerment in this context. Four moderating variables were identified that moderate the negative effects of the case-mix system on empowerment. Two of them are organizational variables: the severity of the patients treated and the size and complexity of the organizations. Two are personal: head-clinicianship and the degree of decoupling by the clinician.

A case-mix system in mental healthcare – influences on empowerment of clinicians