An institutional view on the acceptance of medical managers
to use a benchmark, in order to enhance management
control in a top clinical hospital
Master’s thesis
MSc. Business Administration Organizational and Management control
University of Groningen June 2013
Name L.A. de Kroon
Student number 1355473
Email address l.a.de.kroon@student.rug.nl
First supervisor drs. M.M. Bergervoet
Second supervisor drs. A. van Beilen
1
Management summary
Due to external pressures, the board of directors of a top clinical hospital is implementing a benchmark. However, to make this implementation successful, key stakeholders need to adopt this benchmark. The board of directors aims for the benchmark to be a tool to get the conversation started about costs of different treatments. It will provide the hospital with information about its own performance in comparison with other hospitals, and could therefore strengthen its position in negotiations with the healthcare insurer. According to preliminary interviews, the medical managers do not seem to see the additional value of the benchmark. Physicians can control the costs of treatments by their ‘request behaviour’. However, when these key figures for enabling cost reduction do not seem to believe this can be achieved, it reduces management control.
From a managerial perspective, this study aims to provide practical tools for the management team of a top clinical hospital to encourage medical managers to use a benchmark and therefore facilitate its implementation. The following research question will be addressed:
‘How to encourage medical managers to accept using a benchmark, in order to enhance management control in a top clinical hospital?’
This study shows that according to the medical managers, the main causes for a reduced acceptance of the benchmark are: no optimal balance between top-management pressure and autonomy of medical managers, little communication with the BE department, sub-optimal presentation of data, little information provision about opportunities of the system and comparability of the data, not enough emphasis on the gains that could result from benchmark information, too much focus on costs instead of quality, and not enough sense of urgency.
Recommendations to enhance the acceptence of medical managers to adopt the benchmark are: 1. The role of the board of directors. The board of directors should show their interest in
improving processes during the quarterly meetings. They should hold medical managers accountable for their costs. However, this should be focused on improving the underlying processes by learning from other hospitals instead of letting the medical managers justify their exceedances and therefore affect their sense of autonomy.
2. Increase curiosity and added value. There should be a positive communication concerning gains, financial or non financial, in comparison to the benchmark. The board of directors and the BE department should share success stories with medical managers and compliment medical managers when either achieving cost reductions or process improvements.
3. Increase collaboration with advisors. The meetings of the management team of the RRUs together with a BE advisor should be performed more frequently. This could be once every 2 months, so it will create a more familiar feeling. This would contribute to the view of BE advisors as advisors and not as controllers.
4. Not only focus on costs. The BE advisors should explain to a higher extent which information could be drawn from the benchmark to show the medical managers contributory opportunities of the system. The focus should not be on saving costs, but on improving processes, because quality is more important for medical managers than costs.
5. Clear and structured presentation of the data. The benchmark data should not be delivered in large Excel files. An overview of the ten DTC’s that are most exceeding in costs compared to the benchmark should be worked out in more detail. This could be done with tables and graphs, which have to provide a neutral image instead of an offensive or accusing picture. Therefore, arrows pointing downwards, minus signs and red colors should be avoided. 6. Create a crisis. For medical managers to accept using the benchmark a sense of urgency is
2
Table of content
1. Introduction p.3
1.1. Problem statement 4
1.2. Purpose of the study 4
1.3. Significance of the study 4
1.4. Research question 5 1.5. Sub-questions 5 2. Literature review 6 2.1. Organizational control 6 2.2. Benchmarking 6 2.3. Institutional theory 7
2.4. Power relations in a hospital setting 8
2.5. Case study 9 2.6. Conceptual model 9 2.7. Sub-questions 10 3. Research Methods 11 3.1. Research Design 11 3.2. Data collection 11 3.3. Research population 12 3.4. Data analysis 12 3.5. Reliability 13 3.6. Validity 13 4. Results 14 4.1. Relationship stakeholders 14 4.2. Benchmark 15
4.3. New rules and routines 16
4.4. Acceptance of the system 18
5. Discussion 20
5.1. Relationship stakeholders 20
5.2. Benchmark 21
5.3. New rules and routines 23
5.4. Acceptance of the system 23
5.5. Limitations and further research 24
6. Redesign 25
6.1. Problem definition 25
6.2. Problem analysis and diagnosis 25
6.3. Plan of action 26
7. Conclusion 27
8. References 27
3
1. Introduction
Over the past years the costs of the healthcare system in the Netherlands have increased1. To obtain
better insight in the causes of these rising costs, changes in the healthcare system have to be investigated. In the Netherlands, a system of regulated competition exists. This system tries to
control the health care expenses and aims to improve quality and expedience2. With innovative
contract forms between healthcare insurers and providers, the care purchasing market can be shaped. Before 2011, when hospitals exceeded their budgets, the insurance companies covered these additional costs with state financing. In 2012, the financing structure of hospitals changed from
traditional budgeting practices to performance financing3. A transition model was implemented to
facilitate this change. According to the ‘Hoofdlijnenakkoord’, an agreement between the ministry of health, care providers and insurers, hospitals have to transform their financing from DTC’s (Diagnosis
Treatment Combinations) to performance based DOT’s (DTC’s On the way to Transparency)4. Also,
the B-segment, which includes frequent non-urgent hospital treatments for which a free market
exists, was planned to increase from 34% in 2011 to 70% in 20124,5. Furthermore, a ceiling for
healthcare funding was set. Healthcare insurers aim to reach an average spending growth of 2,5% in their contracting policy, excluding wage- and price adjustments. They provide hospitals with as much money as the previous year with no possibility of extra funding from the state when exceeding this
amount4. This does not mean that hospitals will be penalized for an increased number of patients
due to improved quality of care for example, but they will not be compensated for loosing patients to
another hospital anymore3. In this view of the current political and economic climate, the
negotiations with insurers are becoming stricter. An increasing number of conflicts between health
care providers and insurers arise6. Insurers will not purchase healthcare packages from hospitals that
are too expensive compared to others anymore. Costs, processes and activities have to be made more transparent and public sector organizations are to a larger extent held accountable for their actions (Dey, Hariharan & Despic, 2008).
As these external pressures are increasing, the need for hospitals to control their organizations and to be more cost effective becomes more important. Hospitals require tools to collect data about costs, to analyze this information and to compare it with other hospitals. One way to control the organization is to implement a performance measurement system (Dey, Hariharan & Despic, 2008). To achieve this, the Martini Hospital in Groningen decided to collaborate with five other top clinical hospitals across the Netherlands and developed a benchmark. This hospital group,
named Santeon7, exchanges information about costs and treatment methods. By comparing these
data it is possible to improve quality of care and at the same time to make treatment processes more cost effective7.
In this study the implications of institutional theory for the successful implementation of a benchmark in a top clinical hospital are explored. To increase management control, enhanced by a performance measurement system, key stakeholders need to participate and inter-professional rivalries have to be resolved (Major & Hopper, 2005). In this process, pressures from institutions and power relations play a role. According to Brignal and Modell (2000), institutional theory adds interests and power of different stakeholders to organizational analysis. However, institutional theory is also criticised for neglecting power and interest issues (Abernethy & Chua, 1996). Despite these contradictory beliefs, institutional theory is used to explore the change that the implementation of a benchmark has on the rules and routines of medical managers. In addition, the concept of power in a hospital setting is described further. By integrating the influence of institutional pressures and power relations in a hospital, this study aims to explore the attitudes of medical managers towards adopting a benchmark. Furthermore, a redesign will be developed to encourage medical managers to accept using a benchmark in order to enhance management control.
4 investigated in this study, but are seen as independent variables. Furthermore, this benchmark will be part of the broader existing set of accounting and non-accounting control systems within the hospital’s planning and control cycle (P&C cycle). However, this integration process in the P&C cyle, as well as the fit with the existing control systems are beyond the scope of this research.
In the remainder of this section the business problem, purpose and significance of the study and the research question will be explained. In the next section relevant topics for the business problem provided by academic literature, namely, benchmarking, institutional theory and power relations in the hospital setting will be discussed. The methodology outlines the setup of the study and the result section presents the information obtained from interviews with medical managers. The redesign section gives an overview of the possible solutions of the existing business problem of the hospital. Then, the discussion and subsequently the conclusion of this study are provided. Finally, the last sections of this paper will include the reference list and appendices.
1.1.Problem statement
Six preliminary interviews were conducted to reveal the underlying assumptions of the business problem that the hospital is coping with. As a result of the information obtained from these interviews, this study will focus on the following business problem: due to external pressures, the top management team of a top clinical hospital is implementing a benchmark. However, to make this implementation successful key stakeholders need to adopt this benchmark. At this point, the medical managers in particular do not seem to see the additional value of the benchmark, which reduces management control. Therefore, factors that encourage the medical managers to accept using the benchmark need to be explored.
1.2. Purpose of the study
The goal of this study is to understand what the necessary conditions are, for a successful adoption of a benchmark by medical managers, in order to enhance management control in a top clinical hospital.
1.3. Significance of the Study
Most research in the field of management accounting systems within a hospital setting focuses on the institutional perspectives of control packages (Abernethy & Chua, 1996), of benchmarking on different levels in a hospital (Guven-Uslu, 2005), on the bureaucratic orientation and budgeting behavior of physicians (Abernethy & Stoelwinder, 1990) or on the development of frameworks for performance management (Dey, Hariharan, & Despic, 2008; Conrad & Guven-Uslu, 2012). However, there is little research that has attempted to integrate institutional perspectives and the adoption of a management performance system in a hospital setting focused particularly on the medical managers.
5 professionals are willing to use the benchmark in order to enhance management control. In this study, factors that encourage medical managers to accept using the benchmark will be explored.
1.4. Research question
The following research question will be addressed is this study:
How to encourage medical managers to accept using a benchmark, in order to enhance management control in a top clinical hospital?
1.5. Sub-questions
In order to answer the research question, the main concepts in this question need to be explored through literature study. These main topics are management control and benchmarking in a hospital setting. Furthermore, more theoretical knowledge about institutional theory and the power relations in hospitals is needed to answer the research question. Therefore, the following sub-questions need to be answerd through theory:
1. What is organizational control?
2. What is the role of benchmarking in a top clinical hospital? And what are the key success factors for implementing a benchmark?
3. What is the role of institutional theory in a hospital setting?
6
2. Literature review
2.1. Organizational control
Traditionally, the concept of organization control was perceived as an administrative process designed to regulate the activities of organization’s participants (Mills, 1983). Management accounting systems (MAS) are often implemented to facilitate managerial decision making and to control behavior of subordinates (Abernethy & Bouwens, 2005). From this conventional view, management accounting practices are seen as information providing for management planning and control. Therefore, most management control systems are developed from the perspective of senior management to control work (Chenhall, 2003). These control mechanisms are often called administrative or bureaucratic controls, which include rules, standard operation procedures, budgets or reward systems.
Traditional accounting systems, like budgeting, focused only on financial systems or used even narrower financial measures, for instance return on investment. Over the past years the use of accounting systems has changed (Miller & O’Leary, 1993). The change of management accounting practices is seen as a process that is evolving over time (Burns & Scapens, 2000). A number of more recent developed systems, like activity-based costing and the balanced scorecard (Kaplan & Norton, 1992) have tried to overcome the shortcomings of these financial based control systems, by including more factors related to the performance of the company.
In earlier years, hospitals were mainly dominated by the interests and values of physicians, and informal and qualitative controls existed (Abernethy & Chua, 1996). Many tasks require teamwork en therefore it is difficult to measure individual performance. An extended socialization process of skill and value training is required. A group of people in different organizations but with similar values are called clans or professions, in case of a hospital setting (Ouchi, 1979). According to Orlikowski (1991) ‘professional control’ can be seen as an external form of control, in which specialist employees are trained in outside institutions, such as professional schools. These controls are used in organizations with complex production processes and where highly specialized skills and knowledge are needed (Orlikowski, 1991).
However, the public sector was not as efficient and innovative as private enterprises. Over the past decades, controls shifted towards more formal, bureaucratic controls, which serve management instead of public interests (Abernethy & Chua, 1996).
2.2. Benchmarking
The information derived from a management accounting system (MAS) is often used to ‘assess and reward subordinate performance’ (Abernethy & Bouwens, 2005). Performance measurement, planning, implementing and evaluating improvement measures are all part of performance management (Dey, Hariharan & Despic, 2008). Performance assessment in the public sector is more complex compared to the private sector, as these organizations have to respond to social demands and their primary goal is not to increase profitability (Wynn-Williams, 2005). Furthermore, it is difficult to develop relevant performance indicators, as there is a lot of variability, diversity and complexity in not-for profit organizations (Meyer & Gupta, 1994), compared to the more homogeneous products in factories for example. To overcome these problems there is an increased need for performance measurement systems, like benchmarking.
7 The adoption of a benchmark cannot be an immediate success. A number of primary conditions which enhance the implementation of a benchmark are described in the literature. According to Elmuti, Kathawala, and Lloyed (1997), who investigated the role of benchmarking in a number of private companies, these key success factors were: senior management support, understanding of the organization’s improvements, culture of openness to change, willingness to share information, and dedication to ongoing benchmarking efforts. The application of benchmarking is influenced by changes in the environment, for instance the dynamics in the population, health economics and political climate (Wynn-Williams, 2005).
Guven-Uslu (2005), adopted the ‘receptive context of change model’ developed by Pettigrew, Ferlie, and McKee (1992), to investigate the implementation of benchmarking in three hospital trusts in the United Kingdom (Appendix 1). This model includes 8 factors involved in change, divided into 3 categories: The first category, external factors of benchmarking (BM), includes ‘environmental pressures for benchmarking’, the pressures arise from patients and society, and ‘benchmarking agenda and its locale’, this includes external business factors, like for instance healthcare insurers. The second category, organizational factors of benchmarking, includes ‘policy and strategy of BM’, so the internal business factors, ‘simplicity and clarity of BM’, for instance the performance measures and processes, and ‘Leading benchmarking’, which addresses the type of measure that should be benchmarked. The third category, individual factors of benchmarking, includes ‘benchmarking in professional groups’, for instance clinicians and accountants, ‘benchmarking culture’, which addresses the fit between professional groups, and ‘benchmarking networks’, for instance the communication and co-operation between groups.
The way of thinking involved in benchmarking needs to be adopted by managers on the strategic level, but also by the medical professionals on the operational level. Although this way of thinking is in line with the professional codes of conduct of improving performance to achieve best practices, Evidence Based Medicine (Finkler, 2004), benchmarking needs more formal recognition. Therefore, according to Jones (2002), management support and resources are needed for medical professionals to adopt benchmarking. Individual managerial orientation also influences the willingness of individuals to adopt administrative forms of control (Hall, 1967). Conrad and Guven-Uslu (2011) mention the importance of the context in which performance measurement systems (PMS) are developed and the key role of agency. The study of Abernethy and Stoelwinder (1990) shows that physician managers are willing to adopt an orientation towards organizational goals, however this will be enhanced by management development programs, like training and conflict management. According to Lémieux-Charles et al. (2003), the development and use of performance measurement systems is a search for legitimacy and rationality.
2.3. Institutional theory
Studies of control systems not only focus on the integration of financial and non-financial measures, but also on the link between the organization and its environment. From the structural-functionalist approach, for instance contingency theory, organizations are viewed as technical systems that need input from and give output to their environment (Hopper & Powell, 1985). In contrast to these technical interdependencies, organizational theorists increasingly focus on social and cultural interdependencies between the organization and its environment, in for instance institutional theories (DiMaggio & Powell, 1983). In order to survive, organizations need to reach both technical efficiency and social legitimacy (Abernethy & Chua, 1996).
8 organizations are trying to imitate other organizations which are perceived to be more succesful or legitimate, mimetic isomorphic change occurs. Normative isomorphism results from pressures from professional groups, reinforced by education, training and professional standards. The definition of an institution according to Burns and Scapens (2000) is ‘ the shared taken-for-granted assumptions which identify categories of human actors and their appropriate activities and relationships’. These taken-for-granted rules can also lead to organizational structures becoming similar, isomorph, to conform to society to obtain legitimacy (Jensen, Kjærgaard & Svejvig, 2009). The framework, ‘the process of institutionalization’, developed by Burns and Scapens (2000), shows the interaction between institutions, rules and routines, and action (Appendix 2). They viewed management accounting as organizational routines and rules. These rules and routines can provide stability as well as change. Rules are seen as the ‘things that should be done’. Routines are developed from repeatedly following rules, which makes behaviour programmatic, and are therefore seen as the ‘things that are actually done’. At the one hand, institutional principles can be ‘encoded’ into rules and routines, and rules and routines can also become institutionalized again. At the other, hand rules and routines can be enacted, which can be subjected to resistance if they challenge existing values for example. The other way around is also possible in which repeated behaviour leads to change in rules and routines (Burns & Scapens, 2000). A study of Abernethy and Chua (1996) shows how forms of control emerge and become institutionalized in a hospital setting. However, institutional theory is criticised for neglecting power and interest issues (Abernethy & Chua, 1996). It lacks explanation about why actors behave the way they do and what their interests are. Furthermore, it does not address the influence of human agency on the social practices which create the institutions (Jensen, Kjærgaard & Svejvig, 2009).
2.4. Power relations in a hospital setting
Traditionally, the dominant group in hospitals have been the physicians. The organization is dependent on the physicians, as their specialized knowledge and skills are critical for the existence of the hospital (Abernethy & Vagoni, 2004). Furthermore, they have the power to control costs. The relationship between the organization and medical professionals is seen as conflicting, as their norms and values are incompatible, according to Abernethy &Stoelwinder (1990). Power of medical professionals can result in conflicts, which can be problematic for the implementation of effective management control systems (Abernethy & Stoelwinder, 1991). The professional and bureaucratic organization seem to have different principles of control. In this dual authority structure, in which professionals and administrators are equal in power, conflicts arise through differences in professional and managerial priorities. To overcome resistance of professionals toward bureaucratic rules and formal control strategies, attempts have been made to integrate clinicians in hospital management structures. By involving professionals in management related issues, their orientation and attitude toward management control strategies changes, which is essential for hospital effectiveness (Abernethy & Stoelwinder, 1990). In the current structure the physicians have same rights, but not same responsibilities as organizational members. Medical professionals use resources for clinical decision making, however they are often not accountable for the resulting financial consequences (Abernethy, 1996). If bureaucratic controls, like supervision and standardization, do not threaten the autonomy of professionals, they can be very effective. Supervision of junior medical professionals by senior clinicians can be seen as part of the socialization process instead of a bureaucratic control. Also standardization in the form of patient care protocols is not seen as a form of bureaucratic control (Abernethy & Stoelwinder, 1990).
9 authority at strategic level, as they have the power of decision rights derived by their expertise and their ability to control critical resources (Abernethy &Vagoni, 2004). Power can be perceived as enabling and conflicting (Collier, 2001; Modell, 2002). As described by Tsamenyi et al. (2006), top management can mobilize power to implement a new control system (coercive pressures), however power of professionals (normative pressure) can resist changes. They address the interplay between institutional forces and intraorganizational power relations in the implementation of a new accounting and financial system.
According to Guven-Uslu (2005), openness, understanding and co-operation between managers and clinicians are needed to overcome resistance. It should be clear who is responsible and accountable for different processes. Medical professionals have to consider the needs of their patients, the organization and their own financial needs (Shortell et al., 1998). Waring and Currie (2009) try to understand how ‘managerial strategies for controlling professional work can converge with the autonomous working practice of professional groups’. Participation of clinicians is essential for successful benchmarking, and therefore a bottom up approach is assumed to be the best approach. Also medical professionals need help with cost related matters, and should therefore be provided with training and a more business oriented culture should be developed (Guven-Uslu, 2005).
2.5. Case study
The Martini hospital is a top clinical teaching hospital located in Groningen, the Netherlands. The hospital was established in 1991 through the merger of the Deaconesses Hospital and the Catholic Hospital. Nowadays, the hospital has over 2.556 employees, 31 medical specialties and about 140 medical specialists. The hospital is controlled by a board of directors, which is assisted by staff functions (organization chart is added in Appendix 3). Some responsibilities and decision making are decentralized to Result Responsible Units (RRU’s). A RRU is an organizational unit organized around a type of patient care and service. There are 19 RRU’s on the basis of medical specialties and 10 RRU’s based on the support function of the medical specialties. The management of an RRU consists of an organizational manager and a medical manager. The organizational manager is responsible for costs and quality of personnel and other resources needed for the care pathways within the RRU. The medical manager is a physician, who is responsible for costs and quality of the medical treatments. The medical managers are chosen by their departments. Furthermore, the department of business
and economics gives advice and support to RRU’s8.
Together with 5 other top clinical hospitals in the Netherlands, the Martini hospital participates in the Santeon group. This hospital group cooperates in different areas, for instance information technology, human resources and purchasing decisions. It exchanges knowledge, talent and skills. One area focuses on expedience on the basis of production and resources which lead to the development of the financial benchmark. With this benchmark, cost prices of different treatments are compared between the 6 Santeon hospitals, in order to learn and to increase quality of care7.
2.6. Conceptual model
10 are seen as independent variables and the box in the conceptual model is colored grey. This study will adopt institutional theory to investigate the effects of the benchmark on the rules and routines of medical managers and how changes in these rules and routines can result in the acceptance of usage of this benchmark by medical managers (the dependent variables). Furthermore, when looking at the main stakeholders, the relationship between management control of the hospital board and the acceptance of the benchmark by medical managers, is explored.
Independent variables Dependent variables
a1 a2 b d1 c d2 2.7. Sub-questions
In order to answer the research question, the answers to the following sub-questions need to be explored by this empirical research:
a). Relationship main stakeholders:
- a1).What is the influence of the acceptance of the benchmark by medical managers on management control of the board of directors of the hospital?
- a2).What is the influence of the management control of the hospital board on the acceptance of the benchmark by the medical managers?
b). Benchmark:
- Which factors of the benchmark influence the acceptance of the benchmark by medical managers?
c). New rules and routines:
- In what way does the benchmark influence the rules and routines of medical managers?
d). Acceptance of the system:
- d1). What is the influence of the change in rules and routines, caused by the benchmark on
the acceptance of medical managers to use the benchmark?
- d2). What is the influence of the acceptance of the benchmark on the rules and routines of the medical managers?
External pressures - Macro level
- Meso level
Management control of the Hospital Board
Directors Acceptance of the Benchmark by medical managers Benchmark - Content - Purpose - Added value - Urgency
11
3. Research Methods
3.1. Research Design
This research adopts a business problem solving approach. This is due to the fact that the organization of interest is experiencing a business performance problem. For this approach, the regulative cycle of Van Strien (1997) is used as described by Van Aken, Berends, and Van der Bij (2007). This cycle has five basic process steps (Appendix 4). The business solving project has 3 parts which cover the five steps of the regulative cycle. The first part, the design part, provides a redesign of the business problem and covers the first three steps of the cycle: problem definition, analysis and diagnosis, and a plan of action. This is the phase that is addressed in this study. The next 2 parts include the change part and the learning part. In the change part, the redesign is realized though changes in the organization and also covers the fourth step of the cycle, intervention. The learning
part, in which the organization learns to deal with the change, covers the fifth step of the cycle,
evaluation. Unfortunately these very important phases are beyond the scope of this research . A case study has been conducted at the Martini hospital in Groningen. By means of theoretical and empirical information this study will investigate the causes of the problem regarding the acceptance of the Santeon benchmark by medical managers in order to enhance management control.
3.2. Data collection
For the theoretical analysis, papers useful for this study were obtained from academic journals through an extensive search in the online databases, Business Source Premier and Pubmed. These two databases were used, because these were sources of validated knowledge. Furthermore, a combination of a business database and a medical database could provide papers about the research topic from different viewpoints. Key search terms were: organizational control, organizational control AND public sector, benchmarking AND hospital, benchmarking AND medical managers, benchmarking AND physicians, institutional theory, institutional theory AND hospital, and power relations AND hospital. Only publications derived from this literature search that were most recent and highly cited were used as theoretical basis for the conceptual model.
Furthermore, archival data and internal documents of the hospital were used. These included annual reports, policy letters, documents about administrative and financing arrangements of hospitals, and information about the Santeon group. These documents were analyzed to develop a view about the institutional factors that could influence behavior of top management and medical managers. Furthermore, these documents were used to create a wider view of the healthcare landscape in the Netherlands.
For the empirical analysis, first, preliminary interviews were conducted with six main stakeholders of the business problem. This stakeholders’ group consisted of a member of the board of directors of the hospital, the director of finance and care administration, the head of the department of business and economics (BE), an advisor of the department of BE, an organizational manager of a result responsible unit (RRU) and a medical manager of an RRU. The preliminary interviews were conducted to see whether the problem really existed and to get a more organization wide view, which helped to formulate a more accurate description of the business problem.
12 benchmark. Fourth, factors that could enhance the acceptance of the system. (Appendix 5: Format semi-structured interviews medical managers).
The interviews were performed at the offices of the medical managers at the different outpatient clinics. The average duration of the interviews was aimed at 50 minutes, however, due to lack of time of the physicians, the duration of the interviews varied from 30 to 50 minutes and the average duration was approximately 40 minutes. The interviews were audio recorded to increase reliability.
3.3. Research population
Twelve medical managers were initially approached by email. Eight medical managers responded and were willing to give an interview. The other 4 medical managers were emailed again, but unfortunately did not respond.
The 8 different medical managers, who are physicians as well, are chosen on the basis of the differences of their RRU’s. The medical managers, MMs, are addressed by the letters A to H. Two medical managers of RRU’s of the A-segment, namely, of the intensive care unit (ICU), MMA, and of the pediatrics department (PED), MMB, were interviewed. The A-segment includes the medical specialties which are regulated, and therefore not competitive in the market. These physicians are employed by the hospital. Only a few medical specialties still belong in the A-segment. The B-segment includes the medical specialties which are free and competitive in the market. These physicians are not hospital employees, but independent sellers or contractors. Most RRU’s of the hospital belong to this segment. For this study the medical manager of dermatology (DER), MMC , and of orthopedics (ORT), MMD, were interviewed as these specialties are perceived as being most competitive in the current market, because of the rising number of private clinics. The division between the A- and B segment, specialties that are regulated versus specialties that are competitive in the market, is made to explore different views towards the acceptance of the benchmark. It is expected that the acceptance of the medical managers to use the benchmark will be higher in the B-segment, as there is a free market perspective behind it and therefore increases the need for these specialists to provide a more cost effective treatment. The other 4 medical managers, which all belong to the B-segment, are categorized by 2 characteristics: ‘surgical (snijdend) versus internal (beschouwend) medical specialties’ and ‘small versus large diversity in patients’. These characteristics are chosen to cover the broad heterogeneity of medical specialties. For a small diversity in patients and internal type specialty, the medical manager of gastroenterology (GAS), MME, is interviewed and for a small diversity in patients and a surgical type of specialty the medical manager of gynecology (GYN), MMH, is chosen. For a large diversity in patients and a surgical type of specialty is chosen for the medical manager of surgery (SUR), MMF. For a large diversity in patients and internal type specialty, the medical manager of internal medicine (INT), MMG, is interviewed.
3.4. Data analysis
After the interviews were transcribed in Microsoft Word 2010, the data could be analyzed. As it was spoken language, the sentences and grammar were not always correct. Due to the large size, these interviews are not added as appendices to this thesis. However, they are available for inspection.
13 Appendix 7. To answer the problem statement, step 6, the codes needed to be viewed in a hierarchical way. The codes that were seen more often, are believed to be more important for answering the research question.
3.5. Reliability
To increase reliability, research methods need to be precisely performed and documented, and circumstances need to be similar. In that case, when another researcher would perform the same study, the results will be similar (Baarda, De Goede & Teunissen, 2001). The format for semi-structured interviews in Appendix 5 was used in every interview. The interviews were tape recorded to provide an accurate data collection. The tapes were played slowly, so transcribing went as precise as possible. The interviews were held in the office of the medical managers, so in a familiar context for them. The time of the day the interviews were performed varied, 2 interviews were performed in the morning, 3 at lunch time and 3 at the end of the afternoon. The interviewees were unknown to the interviewer. A repetition of this study is definitely possible, however, subjectivity in semi-structured interviews is inevitable. The same topics can be addressed by a different researcher, but the formulation of questions and answers could differ. The consequence is that this can change the fragments used for the analysis, but not ideas of the interviewees about the main topics, so it will still provide nearly similar results.
3.6. Validity
To explore if the derived data are a good reflection of reality, a matter of internal validity is addressed (Baarda, De Goede & Teunissen, 2001). The fact that the interviews were semi-structured gave the interviewees the opportunity to express their ideas and feelings in their own words, which increases the realistic view of this study. To determine the inter subjectivity, it is important to look at the extent in which the results are dependent from the researcher. As mentioned before, semi-structured interviews are always accompanied with a bit of subjectivity. Furthermore, I am a medical doctor myself and I have my own opinion about benchmarking in healthcare, however, I have tried to adopt an objective view when analyzing the data. It can be seen as both an advantage and bias to speak ‘their’ language. However, the developed labeling system is based on the answers given by the medical mangers, and I believe that other researchers would develop a similar system out of this information.
14
4. Results
In this section the most important findings from the interviews with the medical managers are presented. The interviews consisted of four main topics. These topics cover the sub-questions of the conceptual model, namely: the relationship between the main stakeholders, benchmark characteristics, new rules and routines, and the acceptance of the system.These parts together will contribute to the development of a redesign solution for encouraging medical managers to accept using the benchmark.
As starting point, every medical manager knew that benchmarking was about making comparisons. Also, all eight medical managers (MMs), which are addressed A to H, experienced making comparisons as a positive action. However, the views about the content and purpose of the benchmark differed. All eight medical managers knew that the content included costs. However, seven out of eight (7/8) MMs rather would like to see a quality aspect attached to it, rather than it to be only focused on costs. MMC of the B-segment said, ‘We focus on quality, but I am wondering how
you can see that aspect in the benchmark data. Quality is measured on the basis of perception.’ As
purposes of the benchmark were mentioned (Appendix 8): comparison with other hospitals (7/8), to explain your results (7/8), to provide insight into your processes (3/8), for learning (1/4), as reflection (1/8), evaluation (1/8) or even as punishment, as it could be used against you (1/8).
4.1. Relationship stakeholders
For this study, the main stakeholders of the benchmark are interviewed. As main stakeholders are seen, the board of directors and the medical managers. The preliminary interviews already showed the view of a member of the board of directors. He mentioned that the benchmark should be a tool to get the conversation started with professionals about why their care profiles deviate from other hospitals and to make arrangements about expediency of care. For the board of directors it can be an instrument to control the organization, as a fixed element of the planning and control cycle, to reach the goals and objectives of the internal organization. To accomplish this there has to be some urgency, it should be implemented in an informal way, and it should be important enough to make managers and medical specialists think about it and act according to it, but it should not be too strict that it evokes a lot of resistance, according to a member of the top management team.
Seven out of eight medical managers (7/8), perceived the role of the board of directors regarding the financial performance of the different medical specialties, as a critical (table in Appendix 8). However, the extent of this pressure varies between latent and explicit. MMC of the B-segment says for instance: ‘They (the board of directors) don’t tell it explicitly, they don’t really
demand things, but I do experience pressure when the quarterly figures are presented.’ While the
medical manager of the B-segment, which has a specialty with small diversity and internal medicine (MME) perceives a more explicit role, ‘They (board of directors) try to penetrate the intended
performance into the pores of the organization. They try to be very critical in the bigger picture and I think that is excellent.’ Only one medical manager, MMH (B-segment, small diversity/surgery), would
like to see a more critical role, ‘They should to a higher extent demand explanations about the figures
of the hospital in comparison with the benchmark.’
All MMs perceived the critical role of the board of directors as being normal or even as something positive. The medical manager of the large diversity/surgery specialty, MMF even thinks this is the only way that the financial performance of the medical departments is addressed, he said, ‘[…] it is a ‘conditio sine qua non’, because if they don’t do it, I don’t have to do it either, that’s the
way it works.’ However, the role of the board should be critical to a certain extent. They should be
motivating, but not too pushy as this could result in resistance or even conflict. Although the medical managers think that cooperation with the board of directors is very important, they often have different interests. MMD (B-segment) said, ‘It is dangerous when they are going to play the role of
15 Furthermore, seven out of eight (7/8) medical managers addressed feelings of autonomy, authority or accountability. Three (3/8) medical managers addressed the fact that they worked in a non-academic hospital, because they want to be in charge. All (8/8) medical managers wanted to have control over their own processes. MMC of the B-segment, said
‘I think it is very good that we do a lot of things on our own. That keeps the doctor in the lead, to so to speak. That is very important. They control things, but part of the fun of working at a department in a non academic hospital is to have your own shop and that you have a lot of influence on the policy of the department…I think that fits the personality of a lot of doctors.’
Regarding the role of the board of directors, no differences between the A- and B-segment are perceived. Medical managers of both segments think that the board should have a critical role towards the hospital’s performance and costs, but to a certain extent, because of different interests and feelings of autonomy of the medical managers.
For the successful implementation of the benchmark not only the medical managers, but also their fellow medical doctors need to be willing to work with the benchmark. For other medical specialists to be willing to use the benchmark there need to be an open culture. All (8/8) medical managers perceived that their colleagues were open for their recommendations and innovations, as long as they would actually lead to improvements. The degree of active involvement of the fellow medical doctors in benchmark data varied. In both internal medical specialties, the physicians were perceived to be most actively involved and open to innovations, as they also came up with new things themselves. Change takes time and some people have to grow into it, ‘but the conversations
are getting easier’ (MME). Three medical managers (3/8) addressed the fact that they already saw a
cultural change amongst medical doctors. MMA (A-segment) said, ‘Some old colleagues left and I see
an evident change in culture.’ 4.2. Benchmark
In the interviews, the medical managers pointed out a lot of factors that could encourage them to work with the benchmark. The factor that was most often mentioned, was the presentation of the benchmark data. All but one medical manager (7/8) agreed that the amount of data they get is very large. Large Excel files with cost prices are perceived by medical managers as difficult to interpret, which could result in ignorance of the data or even resistance. MMA of the A-segment said, ‘[…] on
those cost prices, they came with miserable long documents. I don’t want huge files anymore.’ The
medical managers agree that a more structured presentation of the data would be more manageable. They would like to see for instance only the 10 DTC’s were their cost prices are much higher compared to the benchmark worked out in graphs and figures. One medical manager, MMC of the B-segment already received the data worked out, but this was made by an external party which they consulted. Two MMs said that even when the data are presented clearly, they could be perceived as offensive. MMG (B-segment, large diversity/internal) said, ‘If you only get to see lists
with arrows up and down, that is not stimulating’. Also the colors in which the data are presented
seem to matter. When a red color is used to point out a negative thing, for instance an exceeding in costs, that could immediately trigger a defense reflex. This could negatively influence the role of the advisor of the BE department that shows the data to the medical managers. The advisor is in most cases seen as the business or financial man, the man of the figures, not mainly as an advisor.
To the question whether training should be organized to make the system easier to use, the
medical managers were not very enthusiastic, mainly because of time reasons. However, they would like more information about what they could do with the benchmark.
MME (B-segment, small diversity/internal):‘I sat with our BE agent for 15 minutes and he
16 More communication with the BE department is seen as desirable. In this way the medical managers could learn more features of the benchmark system, so they could use it more efficiently, without giving them extensive training. MMA (A-segment): ‘You have to do it together. Very rarely I sit behind
the computer with a BE advisor, but perhaps this should be more frequent. […]’. Also the
organizational manager of the RRU has to be involved with the conversations. The BE advisor and the medical manager have separate meetings with the organizational manger, but rarely with the three of them together. A physician looks at the same data differently than a BE advisor. However, the two different professions need each other for an adequate interpretation. Two medical managers (2/8) think that the organizational manager and the BE advisor do not have enough time to focus purely on their RRU, which diminishes the quality of the meetings. ‘Customized information provision’, which is preferred by the medical managers, could be created by planning meetings with the management of the RRU and the BE advisor, and more time for adequate information provision.
Furthermore, the benchmark should not be focused on costs only. MME (B-segment, small diversity/internal) said, ‘It (benchmark) gives us some control. It (processes) can be done more
efficiently, I don’t care if it saves money.’ Seven out of eight managers (7/8) would also like to see a
quality aspect attached to it. Quality is more important to them than money. The fact that the benchmark is about cost prices, and no patient satisfaction or consumer quality index is added, does not contribute to the acceptance of medical managers to use the benchmark. Only one medical manager thinks it is good that the benchmark is about costs:
MMF (B-segment, large diversity/surgery): ‘I think hard numbers are more workable than soft
ones…Patient satisfaction, no, I don’t have a lot with that. It is important, for window-dressing. But the quarterly meetings are mainly focused on numbers.’
However, the other medical managers agreed with what MMA (A-segment) said, ‘If you want to have
the medical managers against you, you have to tell them that it is all about money.’
Also issues about the comparison of the data are addressed. All medical managers think that the benchmark data are comparable with other hospitals. However they remain a bit skeptic. They believe the data are comparable only generally, not on detail level. The medical managers do see differences in population or registration system.
MME (B-segment, small diversity/internal):‘The thing is…well, is it really comparable? I know
that our hospital is very accurate in the registration of things, you don’t know how this is done in the other hospitals. […] If we have a nursing day price that is more reliable than that of others, you have a huge difference.’
The medical managers also realize that this is the only thing they have got to compare data, so that is at least something. Furthermore, although there are differences, more important is that these differences can be explained.
There are no evident differences seen between the A- and B-segment according to factors that could contribute to the acceptance of the benchmark by medical managers. Medical managers of both segments agreed that the data should be presented more simple and structured, that the data are generally comparable and that communication with BE advisors is important.
4.3. New rules and routines
17 manager A believes that the benchmark could provide more insight in the request behavior and this alone would already lead to changes.
MMA (A-segment):’We have seen more patients and requested less laboratory tests, only by
providing more insight. […]. A lot of things are done on routine basis, but there is so much to gain in there. Only when you provide insight, which the Santeon benchmark obviously can do.’
There are, however three medical managers (3/8), who think that the implementation of the benchmark would not change that much. At the ICU, for example, a lot of routine laboratory tests are performed. Another medical manager believes that because every patient is different, you should not want to have protocols that state how often which laboratory test should be performed. Therefore he does not see changes in request behavior that could arise though the implementation of the benchmark. However, three medical managers (3/8) addressed the importance of evidence based medicine (EBM). This means they only try to work according to medical protocols and standards developed through scientific research. The medical managers do not believe that the benchmark would change these national EBM principles, but it can change processes within this particular hospital. Also control and decision making could be effected by the benchmark. MMD of the B-segment said, ‘[…] for the shoulder we have made a standardized work-up […], so we try to
conform to that.’ Before those changes in rules become changes in routines, time will pass, and old
routines have to be unlearned. Seven medical managers (7/8) believe that the benchmark could make everybody more aware of their rules and routines. Three medical managers (3/8) are actively involved in questioning old routines.
MMG (B-segment, large diversity/internal): ‘ [..] I notice that my residents have to get used to
the fact that I am so critical about their requests. Why did you do that? Because we always do it like this, they answer. Well, that is the wrong answer.’
Furthermore, costs seem to get more attention of the medical managers and they try to communicate this to increase the common feeling of cost consciousness. Four medical managers (1/2) address the fact that cost consciousness is very important, and should be a social competence of doctors. They also believe that their residents should already be trained in this.
MMF (B-segment, large diversity/surgery): ‘I inculcate the residents regularly that they
shouldn’t do too much laboratory tests and that an X-ray is not necessary for a particular treatment, because it costs money!’
Two medical managers (1/4), however, don’t teach their residents about costs. One of them believes that residents already have to learn so much, that learning about costs would be too much. The other medical manager believes that quality comes first, and then it does not matter how expensive a particular treatment is.
The implementation of the benchmark could, according to three medical managers (3/8), also have an effect on the interaction with other medical specialists. These three medical managers talk or would like to talk to colleagues in other hospitals about benchmark information. However most of the times, everybody is too busy and these conversations do not take place anymore.
18
4.4. Acceptance of the system
For the medical managers to be willing to accept to use the benchmark, the previous topics, the relationship between the main stakeholders, the attributes of the benchmark and the change in rules and routines, could all contribute. Next to this, other factors can play a role in the acceptance of the benchmark. First, most often mentioned in the interviews is the added value of the benchmark (Appendix 8). Medical managers need to see an added value, to be willing to use it. However, this added value lies in different areas according to the medical managers. Five medical managers (5/8) see the additional value of the benchmark in improving their processes.
MMB (A-segment): ‘Their (healthcare insurers) aim is to economize. That is the wrong
approach! The aim is to analyze if you are doing well at this moment and to see if things can be done more efficiently.’
Another added value of the benchmark lies in increasing the performance of the medical departments. For the B-segment, in which the prices are competitive in the market, the medical managers were expected to be more competitive than the medical managers of the A-segment. However, both medical managers of the B-segment specialties that are most competitive in the market, MMC and MMD, did not make such statements. MMC said, ’[…] well actually, we think we
are doing quite well. So, we feel a limited need to change things even if we notice it.’ The fact that she
used the word ‘think’ indicates that they are not certain, but perceive to be doing well and apparently there is no need to increase their performance. In contrary, in the A-segment, which is regulated and the prices are not free in the market, the medical managers were more competitive towards other hospitals .
MMA(A-segment): ‘A lot of people don’t think that the healthcare sector is competitive,
however, I do not totally agree. I want to be better than the others. With this kind of thinking you push the level of performance upwards I think.’
Furthermore, according to three medical managers (3/8), the benchmark could provide insight which can help to control their processes. An example, provided by a medical manager of the B-segment, is that now they can see the number of first outpatient clinic visits in the first months of the year compared to the other hospitals. If this is too little, they can control these processes. Not only the medical managers of the B-segment, but also of the A-segment are concerned about controlling their processes:
MMB (A-segment): ‘At the moment you have insight in those matters (number of laboratory
tests and urine cultures), you can use it to control your processes. We never had any insight into that before, and now we can start controlling.’
One medical manager, MMG, thought that sharing success stories would increase the acceptance of MMs to work with the benchmark. She believed that when you tell stories about processes that improved or became more efficient, it would become more tangible for MMs.
19 compared to other hospitals. Furthermore, MMD also did not perceive some form of urgency, because concerning his own salary he does not dependent on the hospital.
MMD (B-segment): ‘Honestly, at this moment I don’t have a lot of interest in our
expenditures, because our honorarium is not part of it, so I don’t care. […] And if they say, this is not working anymore, than we will all quit and we will continue our practice outside the hospital. It is that simple.’
However, three medical managers (3/8) perceived some kind of urgency. MME of the B-segment could image a threat of the board of directors, that when his cost prices are low compared to the benchmark, the board would say, ‘very good that you try so hard, but you are running a loss for the
hospital, so maybe you need to have less medical specialists.’ MMC of the B-segment does not feel
much urgency, but she thinks that working at the same hospital for years ‘can lead to blind spots’, which could be reduced by benchmarking. Another medical manager of the B-segment, MMH, thinks very business-like and emphasizes the importance of looking out for other hospitals, ‘I consider
ourselves as a company and within a company you should look around you very carefully. What is the market doing? How can I increase my adherence?’ He also could image that a healthcare insurer
would stop ‘doing business with some hospitals anymore, because they are scoring bad on the
benchmark’. However, he says, ‘due to political considerations, these hospitals stay open anyway.’ So
this diminishes the sense of urgency he perceives. Also a medical manager of the A-segment could imagine a treat from healthcare insurers.
MMA (A-segment): ‘If somewhere a good product is delivered for less money, than I would
know what to do if I was a healthcare insurer. There are lot of people that agitate against it, but be prepared, or otherwise you’ll be eaten, I always think.’
Furthermore, the people involved could also influence the acceptance of the benchmark by medical managers. Four medical managers (1/2) addressed the importance of communication between the medical manager, the organizational manager and the BE advisor. Regular meetings (3/8), close contact (1/4), adequate information provision (3/8), right people for the job (1/8), and a proactive mentality (1/4) are mentioned as factors that could contribute to better communication (Appendix 8). It is important that the views are aligned and that everybody acts according to the same interests. MME believes that for the acceptance of the benchmark, ‘curiosity has to be aroused’.
20
5. Discussion
In this section the main findings of the study are summarized and discussed in the context of academic literature. The section is divided according to the questions that followed the conceptual model and were to be answered by conducting this study. Furthermore, this part addresses the limitations of this study and further research.
5.1. Relationship stakeholders
Question a1. ‘What is the influence of the acceptance of the benchmark by medical managers on
management control of the board of directors of the hospital?’ In case the medical managers accept
to use the benchmark, the board of directors will get much more insight in the processes and associated costs of the different medical specialities. When comparing these data with other hospitals this can be very valuable for the board of directors to increase management control in their own organization, to increase their negotiating position with respect to the insurers if necessary or to increase the accountability and transparency towards the public.
To answer question a2. ‘What is the influence of the management control of the hospital
board on the acceptance of the benchmark by the medical managers?’ this study shows that the
hospital board should be critical towards the performance of medical specialists, but to a certain extent. The board of directors is asking the medical managers to explain their performance in costs and numbers in comparison with the benchmark in quarterly meetings. The fact that the medical managers are held accountable for their results, also in comparison with other hospitals, gives a kind of pressure which amongst other factors, results in a higher acceptance of the benchmark.
In hospitals, professionals are performing the core production activities. These activities include complex tasks, therefore administrative controls poorly fit (Zucker, 1991). However, still administrative controls, like a benchmark are implemented in hospitals, as seen in this case.
Kohli and Kettinger (2004) studied the attempts of a hospital to exercise cost and outcome control over physicians with a system, which informed physicians’ practice decisions with performance information. This increased transparency, which is also a result of the benchmark in this study. However, the overall result in their study was a failure, because a significant change in the clinical practice of the physicians was not achieved. A lack of management legitimacy resulted in the absence of control benefits over the physicians. The results of this study show that every medical manager perceived some kind of pressure of the hospital board, regarding performance and costs. For most medical managers in this study the pressure on performance and costs was a normal, accepted thing or was perceived as something positive. Therefore, the hospital board in this case had enough legitimacy, which can be seen as a prerequisite for a change in behavior of physicians.
According to Abernethy and Stoelwinder (1995) there can be a ‘clash of cultures’, when the autonomy of professionals demands so much control that their special rights are in conflict with management goals and bureaucratic principles of efficiency and accountability. However, the results of this study indicate that efficiency is not only a bureaucratic principle of the management team, also the medical managers strive for efficiency. They are very much open to learning new things, to increasing their performance and improving the efficiency of their processes. On the other hand, all medical managers have a relatively strong feeling of autonomy, they want to be in the lead. Most of them work in a non-academic hospital, because they want to run their own shop and to control their own processes. So a self-governing structure of the profession exists. Therefore, too much pressure of the board of directors results in resistance or even conflicts. This confirms the findings of Abernethy and Stoelwinder (1995), who state that professionals with a high professional orientation feel offended by targets and performance measures imposed by superiors.
21 managers towards output controls, found in this study, are not as bad as expected. When the output measures are well defined and it is not all about costs, but also quality aspects are taken into account, the medical managers would perceive these controls as useful tools to control their processes. According to Chua & Degeling (1991), the implicit objective of the implementation of new management control systems is to increase control over the behaviour of professionals. This study shows that when medical managers accept to use the benchmark, next to the other existing control tools, it can add to the management control of the board of directors.
5.2. Benchmark
To answer question b. ‘Which factors of the benchmark influence the acceptance of the benchmark
by medical managers?’ this study addresses: a clear and structured presentation of the data,
communication and collaboration with staff functions, a critical role of the board of directors to a certain extent, open culture to change, focus not only on costs but also quality and efficiency, increase curiosity and added value by sharing success stories, increase the sense of urgency, emphasize cost consciousness as social competence, and address comparability.
A number of attributes of the benchmark that could enhance or decrease the acceptance by medical managers are addressed in this study. Here, they will be discussed by using the research performed by Guven-Uslu (2005), who investigated the implementation of benchmarking in three hospital trusts in the United Kingdom. That study adopted the ‘receptive context of change model’ developed by Pettigrew, Ferlie, and McKee (1992). This model includes 8 factors involved in change, divided into 3 categories:
a). External factors of benchmarking (BM)
1. Environmental pressures for benchmarking: these pressures arise from patients and society. Only one medical manager in this study addressed the fact that the implementation of the benchmark is also needed for patients, as they become more demanding and critical about the care they consume. However, this topic is beyond the scope of this research.
2. Benchmarking agenda and its locale: this includes the external business factors, like for instance healthcare insurers. Guven-Uslu (2005) found that the external stimulus for change is only present when the benchmark becomes a decision-making agent for contracting. According to the preliminary interviews, the member of the board of directors and members of the BE department think that this scenario can become reality in the near future. However, none of the medical managers was afraid of this. More than half of them did not believe this would actually happen and the other half thought that in case this happens, a large capacity problem would arise and waiting lists will be that long, that insurers still needed all doctors for the total production, not only the cheapest ones. This means that the medical managers in this study did not saw the urgency for implementing the benchmark, which according to the study of Guven-Uslu (2005) is necessary to generate change.
b). Organizational factors of benchmarking
3. Policy and strategy of BM: this includes the internal business factors. Guven-Uslu (2005) concluded that a bottom-up approach for BM policy was preferred as it provided flexibility and inter-disciplinary involvement. However, although the medical managers wanted to have insight in their own processes and to increase efficiency, the benchmark was mainly seen as top-down. One medical manager even said: ‘if they don’t do it, we don’t have to do it either.’ 4. Simplicity and clarity of BM: with this, the performance measures and processes are
