Controlling acute medical care

(1)

Controlling acute medical care

A multiple case study research investigating the

management control of Acute Medical Unit performance

Name: Ilse Mulder

Student number: s1689045

Address: Kortenaerstraat 11a, Groningen Telephone number: 06-41664799

E-mail adress: ijbmulder@gmail.com University: Rijksuniversiteit Groningen Faculty: Faculty of Economics and Business

(2)

Preface

The thesis lying in front of you is my final piece of the Master Business Administration,

Organizational & Management Control. It means my graduation and the closure of a wonderful time as being a student in Groningen.

Although writing a thesis is an individual task, I have not been all alone in the process of writing this thesis. First of all I would like to thank my supervisors for their time and effort, especially

Marcel Bergervoet for his useful feedback and support during the process of writing this thesis. Further I would like to thank Pieter Fennema, my supervisor at the UMCG, for his support during this year. I have learned a lot thanks to you, and these lessons learned were not only related to my thesis. The period at the UMCG certainly contributed to my personal development.

Also a word of gratitude to all managers and employees of AMUs incorporated in this research. I have experienced friendliness and willingness to help from all parties. All in all I can say that the cooperation was good and the information shared has been valuable and essential for this research. And last, but certainly not least, I would like to thank my family and friends for all the support I have received in all ways possible.

Hopefully, this thesis will be interesting to read for all of you and I wish you all pleasure in reading it.

Ilse Mulder

(3)

Summary

The healthcare sector is changing rapidly and a refinement of models of acute medical care to improve efficiency and patient outcomes are needed. An Acute Medical Unit (AMU) is such a new service model and can be described as a designated hospital ward specifically staffed and equipped to receive acute medical patients from the emergency department or policlinic for up to a

designated period (between 24 and 72h) prior to discharge or transfer to medical wards.

The management control of AMUs is underexposed in literature so far. However, management control is a critical function in organizations. Therefore, the aim of this thesis is to provide insight into the connection between the characteristics of an AMU and the characteristics of a Management Control System, such that AMU performance is controllable. The key question is: What should the

design of a MCS be in an AMU setting, such that the performance of the AMU can be controlled?

The conceptual model consists of three concepts; AMU characteristics (including; vision and mission, key success factors, acute medical patients, organization structure, location, and predetermined designated period), MCS characteristics (including; key performance measures, target setting, performance evaluation, reward systems, and information flows, systems and networks) and AMU

performance measures (including; length of stay, mortality rates, ED waiting times, discharge

disposition, readmission rates, bed cost and resource utilization, and patient and staff satisfaction). A multiple case study is conducted at three AMUs in hospitals within the Netherlands.

Semi-structured focused interviews have been held with people in charge of these AMUs, and further documentation is collected. An interpretation scheme is developed to interpret the results.

Three conclusions are made to answer the key question; (1) Key performance indicators that should be measured are: LOS, mortality rate, ED waiting times, discharge disposition, bed cost and resource utilization, and patient and staff satisfaction. Further, for all of these KPIs targets should be set. (2) Evaluation processes should be in place at different aggregation levels. Many operational

evaluations exist, more evaluations at policy level should be in place. And when evaluating, more use should be made of available performance measures’ information. (3) A good working hospital information system is essential. Key is that AMU managers and employees should be able to derive relevant information from the system themselves and not be dependent on staff. Many data are already available in the database but much more can be gained from this data if the access to these data is easier.

Recommendations for further research are to pay attention to the following topics; the relationship between the role of professionals and AMU performance, the role of power of the different

(4)

1. Introduction

The healthcare sector has changed rapidly since the 90s. Competition increased, influence of patients grew and there is a stronger necessity to deliver health services in a more efficient and effective way (de Vries, 2011). The pressure on healthcare expenditures has been growing. Since the 1970s, government healthcare expenditures have been rising more rapidly than GDP in OECD countries. Hagist and Kotlikoff (2005) show that this growth in expenditures is due to demographic change and growth in real benefit levels (i.e., healthcare expenditures per beneficiary at a given age). It is remarkable that 75% of overall healthcare expenditure growth and almost all of growth in healthcare expenditure per capita reflect growth in benefit levels. Since OECD countries are

projected to age dramatically and patient admissions will rise, the rapid growth in benefit levels will be unsustainable. No country can spend an ever rising share of its output on healthcare. Therefore, in many OECD countries market-oriented healthcare reforms have taken place (Harrison & Calltorp, 2000; Fotaki, 1999; Schut, 1995; Bamford & Porter-O’Grady, 2000; Benoit et al., 2010). The main goal of market-oriented healthcare reforms is to increase the healthcare system’s efficiency and its responsiveness to patient’s needs, while maintaining equal access to healthcare (Schut & van de Ven, 2005). The introduction of these reforms has the advantage that it allows to break through the vicious circle of ever-increasing healthcare expenditures (van de Ven, 1996).

These developments of rising patients admissions while constraining costs are also visible in acute healthcare. Most acute hospitals have seen a tremendous rise in emergency admissions in the last decade along with a reduction in numbers of hospital beds (Scott et al., 2009; Hanlon et al., 1997). Admissions are expected to climb further in the next decades due to increasing numbers of elderly patients with multiple chronic diseases (Kendrick & Conway, 2003). Lack of spare bed capacity leads to overcrowding and congestion in emergency departments resulting in inefficiencies in service delivery, length of stays that may be longer than necessary (Liew et al., 2003) and greater risk to patients of medical error, avoidable death and complications (Sprivulis et al., 2006). Although early assessment by medical specialists and allied health professionals is the expected standard, it is rarely the norm (Scott et al., 2009). These factors provided the impulse for the refinement of models of acute medical care to improve efficiency and patient outcomes (Blatchford, 1997; Steward, 2002). Hospitals worldwide have considered structural reforms to optimize care (Scott et al., 2009). The keystone of new service models is usually a designated unit, run by specially trained staff whose core function is to provide timely, targeted, multidisciplinary team care to acute medical patients

(Armitage, 2002). Such a unit is called an acute medical unit (AMU) and establishing an AMU is a solution that is growing in popularity (Scott et al., 2009).

Many hospitals, especially in the United Kingdom and Australia, have established AMUs. AMUs are defined by Scott et al. (2009) as “designated hospital wards specifically staffed and equipped to

receive medical inpatient presenting with acute medical illness from emergency departments and/or the community for expedited multidisciplinary and medical specialist assessment, care and treatment for up to a designated period (typically between 24 and 72 h) prior to discharge or transfer to

medical wards”.

(7)

practitioner, or the policlinic (Raduma-Tomàs et al., 2012; Vreelandgroep, 2011). In a hospital without AMU acute patients are supposed to be transferred directly from the emergency

department (ED) to the nursing ward of the specialist treating them (Scott et al., 2009). However, due to capacity restrictions it is possible there are no beds available at that department

(Vreelandgroep, 2011). That situation generally has two effects; the first one is that patients have to wait at the ED for a free bed, resulting in overcrowding EDs. The second one is that patients are transferred to a nursing ward of another specialism, resulting in patients lying in ‘wrong’ beds. Both effects are undesirable because it may decrease quality of patient care (Moloney, 2006). However, an AMU changes the situation. In a situation with an AMU, acute patients are transferred from the ED to this AMU instead of directly to the nursing ward. Both the traditional model of care and the AMU model are shown in figure 1. Patients will be admitted to the AMU for up to a designated period, typically between 24 and 74 hours (Scott et al., 2009). The aim of the AMU is to provide a diagnosis, and when there is a solid treatment plan patients are triaged to the nursing ward of their specialism for continuing care, or are discharged and send home (Raduma-Tomàs et al., 2012). By implementing an AMU, acute and elective patient flows are separated the first 24-74 hours, and a buffer is created between the ED and nursing wards. This improves the patient flow to the nursing wards, resulting in higher rates of patients transferred to the ward of their specialism (Dannenberg, 2010). Furthermore, speeding up diagnostics in the acute phase leads to shorter in-hospital stay, an adequate and effective policy (leading to a quality-increasing effect), and a better capacity use (Vreelandgroep, 2011).

The focus of existing research considering AMUs lies mainly on describing and evaluating the performance or design of AMUs (Providence et al., 2010; Edmans et al., 2011; Scott et al., 2009 ; Moloney et al., 2005; 2007; Rooney et al., 2008 ; Noble et al., 2008). No research has been

conducted so far regarding the management control of an AMU. However, this is an important topic because management control is a critical function in organizations. Management control includes all the devices or systems managers use to ensure that the behaviors and decisions of their employees are consistent with the organization’s objectives and strategies. The systems themselves are referred to as management control systems (MCS). The primary function of management control is to influence behaviors in desirable ways. The benefit of management control is the increased probability that the organization’s objectives will be achieved (Merchant & Van der Stede, 2007).

Since management control of AMUs is underexposed in literature so far, there is no knowledge about controlling AMUs. According to Merchant & Van der Stede (2007) management control is a critical function. They state that management control failures can lead to large financial losses, reputation damage, and possibly even to organizational failure.

Purpose: The aim of this thesis is to provide insight into the connection between the characteristics of an Acute Medical Unit and the characteristics of a Management Control System, such that AMU performance is controllable.

Key question: What should the design of a MCS be in an AMU setting, such that the performance of

the AMU can be controlled?

(8)

2. Research framework

In this section, the concepts of research are more explained using existing scientific literature. Literature directly linked to the key question is used. The key question of this thesis is: What should

the design of a MCS be in an AMU setting, such that the performance of the AMU can be controlled?

Therefore, this section is divided in three parts. First AMUs are described, then MCSs are described, and finally AMU performance will be described.

2.1. Acute Medical Units

Synonymous names for AMUs are acute medical assessment units (AMAU), medical assessment

and planning units (MAPU), acute assessment units (AAU), acute medical wards (AMW), acute planning units (APU), rapid assessment medical units (RAMU) and early assessment medical units (EMU), (Scott et al., 2009). This thesis continues using the term Acute Medical Unit (AMU). AMUs seem to be a proper solution to last decades’ developments in healthcare. In the UK, the Royal College of Physicians of London since 2001 has repeatedly recommended the establishment of AMUs to provide hospitals with defined medical cover for acute general medicine in order to

respond more effectively and safely to the increasingly complex demands placed on the hospital with regard to acute medical care (Royal College of Physicians in Scott et al., 2009). In Australia and New Zealand, the Internal Medicine Society of Australia and New Zealand (IMSANZ), has taken the lead in promoting acute care medicine by releasing trans-Tasman guidelines for AMUs (Henley et al., 2006). Due to these recommendations and successful performances of existing AMUs, establishing an AMU is growing in popularity (Scott et al., 2009). Ninety-eight per cent of NHS1 hospitals in the UK have AMUs. Also in Australia and New Zealand the amount of AMUs is rising (McNeill et al., 2010). AMU service provision is expected to increase to 73% of the NZ population by the end of 2012 (Providence et al., 2010). Also in the Netherlands the amount of AMUs is increasing. In 2009, 6 AMUs were operative in the Netherlands. In March 2011 already 13 AMUs were operative and this number is still increasing. It is expected that in time, all hospitals in the Netherlands have implemented an AMU (Vreelandgroep, 2011). Although differences between AMUs across countries and hospitals do exist, they all share several common objectives (Table 1) and patient flow characteristics (Figure 1). Table 1. Objectives of AMUs (Scott et al., 2009)

I. Appropriate and timely multidisciplinary assessment, diagnosis and treatment of acute patients (led by appropriately trained acute care physicians) leading to reduced length of stay

II. Early consultant review of admitted patients and referral, as appropriate, to specialty teams III. Rapid turnaround in pathology, radiology and other clinical investigative services

IV. Improved access to clinical management resources and a more effective use of resources for the hospital as a whole

V. Reduction in waiting times for patients in EDs to access in-hospital beds (access block)

VI. Reduction in ED overcrowding and numbers of patients who do not wait to be seen by ED staff VII. Elimination of the need to outlie patients in non-home wards and disrupt ward environment with

after-hours admissions

VIII. Standardized care of acute patients based on agreed care protocols and guidelines and a more organized work environment with standardized admission and discharge processes

IX. Optimization of bed management using care pathways resulting in smoother patient flows and

(9)

Table 1. Objectives of AMUs (Scott et al., 2009) avoidance of unnecessary admissions

X. Facilitation of clinical and health services research into care of acute patients

XI. Increased staff job satisfaction and reduction in staff fatigue by improved rostering and use of shifts

One of the main goals of AMUs is to separate elective and acute patient flows for the first 24 to 72 hours (Dannenberg, 2010). This leads to fewer interruptions on regular nursing wards because acute patients are admitted on the AMU instead of the nursing wards. With an AMU patients are

transferred to nursing wards on predetermined times. It will be communicated which patients will be transferred and thus nursing wards can prepare for these patients. This smoothens the process of transferring patients. Furthermore work for nurses will be less stressful, especially at night, and nurses can fully concentrate their work on the patients already lying in the wards. AMUs, thus, influence the entire process of acute medical care in hospitals. They facilitate an efficient high quality emergency admissions process with a view to a shorter length of stay (LOS). Furthermore, the concentration of acute medical patients in AMUs provide a focus of clinical care for medical staff rather than have patients spread across several different wards (Moloney et al., 2007).

(10)

Below, AMUs will be described further by three dimensions. Patients, personnel, and unit design (consisting of unit location, size and designated period).

2.1.1. Patients

In general, AMU admission policies grant entry to any patient with an acute medical condition (Scott et al., 2009). However, some specific patient groups are for medical or technical reasons not suitable for AMUs. For example patients who need intensive care, pediatrics, obstetrics and psychiatry. These groups need specific facilities or expertise (Vreelandgroep, 2011). According to Scott et al. (2009) patients with the following contra-indications should be excluded from AMUS; (1) Haemodynamic instability requiring invasive monitoring and/or critical care facilities, (2) Special need patients (e.g. acute stroke, dialysis, oncology, endoscopy), (3) Presentations for respite or residential care, (4) Geriatric syndrome presentations best suited for admission to geriatric rehabilitation or dedicated elderly care units, and (5) Severely behaviorally disturbed patients best suited for mental health care. All other acute medical conditions are acceptable. According to Cooke (2002) the types of patients accepted into AMUs can be classified into high risk discharges, those requiring short-term treatment, patients with limited medical needs and clinical conditions needing only short-term observation. He states that patients who are usually not admitted include those with vital signs, or patients needing specialist or intensive care. It is shown from practice that the more specialism’s are participating in an AMU, the more valuable an AMU is. Therefore, AMUs should include all the specialism’s possible and only exclude what has to be excluded (Vreelandgroep, 2011). Medical indications not suitable for AMUs are summarized in Table 2.

Table 2. Medical indications not suitable for AMUs

I. Haemodynamic instability requiring invasive monitoring and/or critical care facilities II. Special need patients (e.g. acute stroke, dialysis, oncology, endoscopy) and patients needing

intensive care

III. Presentations for respite or residential care

IV. Geriatric syndrome presentations best suited for admission to geriatric rehabilitation or dedicated elderly care units

V. Severely behaviorally disturbed patients best suited for mental health care VI. Patients best suited for obstetrics

VII. Children best suited for pediatrics

Looking from a patient perspective, AMUs are advantageous. Waiting times in the ED decrease significantly (Moloney et al., 2004, Munday et al., 2012) and also total in-hospital LOS decreases significantly (Moloney et al., 2006). Mortality rates drop (Rooney et al., 2008). Furthermore, the diagnostic process is faster which means that treatment starts earlier, and patients can go home sooner. This is described in more detail in section 2.3; Performance of AMUs.

2.1.2. Personnel

Healthcare is a labor intensive industry and well skilled personnel is very important. As described above, AMUs deal with acute patients with diverse medical conditions. According to Moloney et al. (2004) the balance between specialist care and good general medical skills is critical in acute

(11)

supervised by acute medicine physicians (also known as consultants with an interest in acute general medicine) featuring multidisciplinary teams that comprehensively assess and manage both medical illness and functional disability (Scott et al., 2009). Beside physicians these multidisciplinary teams imply nurses. Griffiths (2009) argues that AMU nurses play a key role in the admission of the patient from their pre-admission setting to a hospital bed. The AMU nurses’ clinical role is conceptualized as being a social and cultural construction in a distinct community of practice. The AMU nurses in the study of Griffiths (2009) outlined that the skills to be an effective AMU nurse are developed by engagement in the practice of AMU nursing. They described their role in terms of management of the clinical environment and being skilled when providing care. Furthermore they said they learned by participating in practice and not just from observing practice or being reliant on propositional knowledge. According to Wenger (1998) learning in a social context takes place through ongoing practice that draws on social energy and power generated through interaction in joint enterprises. The new identity gained as an AMU nurse is a construction gained by experiencing and engaging in what was meaningful to the community of practice (Griffiths, 2009). However, the AMU adds another stage to the patient’s hospital journey and this operational constraint reduces the time that the nurses have to engage with an individual patient, especially if the patient’s needs are not as time critical as other patients. This may not reflect the nursing ethos of holistic and individualized care delivery. Due to their bed management role coupled with caring for newly admitted and often medically undiagnosed acutely ill medical patients, nurses experience job related stress, work overload and role conflict. However, the team’s cohesion to a common enterprise, over which the nurses exercised a great deal of control, together with the respect for one others’ professional capabilities lead to a positive working environment (Griffiths, 2009). Moreover, AMUs not only influence personnel working on these wards, but also have consequences for nurses working on regular nursing wards. Patients are transferred to nursing wards on predetermined times resulting in less interruptions in patient care due to the admission of acute patients. Especially at night work pressure on regular nursing wards is lowered after implementing an AMU.

Sinclair et al. (2003) also argue in their article that professional nurses are important in AMUs. They state that advanced practice nurses (APN) are coordinating AMUs. The APN role requires nurses with relevant experience and qualifications to practice at a highly skilled, autonomous and independent level (Sidani and Irvine, 1999). Their function within AMUs is called Clinical Coordinator (CC). The CC coordinates a comprehensive, multidisciplinary team approach whereby patient issues are

contextualized in terms of their effect on patient safety, functioning and well being. In an AMU the CC’s role encompasses four core functions: assessment, planning, facilitation of information exchange, and education. The introduction of the CC’s role in AMUs has been crucial in addressing many inefficiencies of traditional models of patient treatment and planning. The importance of expert nursing is increasingly recognized and appreciated, especially in the assessment and planning phases of the medical patient’s stay; phases that are emphasized in an AMU (Sinclair et al., 2003). Furthermore, Cooke (2003) states that strong management is essential in AMUs, especially to ensure that transfers to other wards are not delayed to maintain a smooth patient flow and to avoid an overcrowding ED. Daly (2009) complements this by saying that the nursing staff collaborates with other healthcare team members to determine each patient’s ultimate disposition. Heads of the department should ensure that these dispositions are not delayed.

(12)

et al. (2007) performed a retrospective analysis of 3163 medical patients admitted before and after an AMU was implemented, with a new on-call rota: ‘consultant of the day’ changing to ‘consultant of the weekend’. A ‘consultant of the day’ system means every day there can be a different consultant on duty. A ‘consultant of the weekend’ system includes the same consultant on call from Friday morning to Monday morning, with the intervening workdays covered by different daily consultants. Noble et al. (2007) showed in their study that a ‘consultant of the weekend’ system is more

preferable, this system was complemented by an alignment of junior medical staff over the weekend period in order to maximize continuity of care.

Furthermore, besides medical staff working on the AMU itself, medical specialists should walk rounds on this AMU twice a day. These specialists are responsible for patients of their specialism lying on the AMU. These rounds are important to keep the patient flow going, and during the rounds decisions about further treatment are made (Vreelandgroep, 2011). Functions involved in an AMU are outlined in table 3.

Table 3. Personnel involved in an AMU

Consultant (physician with special interest in acute general medicine)

Supervisor of AMU, determines long-term policy and connects AMU with medical specialists Clinical Coordinator (Advanced Practice Nurse) Operational head of AMU, takes care of:

assessment, planning, facilitation of information exchange, and education.

AMU Nurse Social and cultural construction, and supporting admissions and medical care

Medical specialists and assistants Walk rounds twice a day and decide about further treatment of their patients

2.1.3. Unit design: Unit location, size and designated period

Another essential factor besides patients and personnel, is AMU unit design. Related to design are unit location, size and designated period. AMU location is important, a geographically position within or very close to the ED and diagnostic and supporting facilities is essential. This co-location is logical as these diagnostic requirements are shared. Furthermore, transfers of patients between the ED and AMU will commonly occur (Cooke, 2003). Diagnostic facilities include pathology, radiology, endoscopy, laboratory services, physiotherapy, occupational therapy and social services (Moloney et al., 2007; Bell et al., 2008). It is necessary that 24h access to diagnostic facilities is available to AMUs (Cooke, 2003), for example by saving predetermined time slots for the diagnostics of acute patients. There is no consensus in literature about the optimal size of an AMU. By size, it is meant the amount of beds in an AMU (Cooke, 2003). It is reasonable that this depends on the total amount of acute patients treated in the hospital. However, it can be argued that larger hospitals profit more from an AMU than smaller hospitals do because with a large amount of patients, variation will be more limited and bed occupancy rates will be higher.

(13)

other 5 had a period of 24 to 36h. These studies are consistent with the stay rule of 24h to 48h of Cooke et al. (2003). Edmans et al. (2011) had a slightly larger time frame, they mentioned a period of 72h. Scott et al. (2009) support all these studies by mentioning that commonly the designated period of an AMU varies from 24 to 72h. An exception to this are the studies of Moloney et al. (2005;2006;2007) and Rooney et al. (2007), they report a much higher period of 5 days. However, they all studied the same AMU so this one could be seen as an outlier.

2.1.4. AMU: standard characteristics

Different AMU characteristics are mentioned above. It is derived from the literature that differences in AMUs exist between hospitals and countries. However, here it will be tried to come up with a standard version of an AMU that includes the most common characteristics.

Patients: All patients with acute medical conditions are accepted except for the medical indications mentioned in table 2.

Personnel: Consultants, clinical coordinators, AMU nurses and medical specialists are involved working on an AMU.

Unit location: An AMU should be located close to the ED and close to diagnostic and supporting facilities.

Unit size: There is no consensus about the optimal amount of beds in an AMU. This depends on the hospital size and the amount of acute patients. To determine the amount of beds hospitals can use a computer simulation model. Nevertheless, it is stated in literature that a unit with too few beds is not working (Vreelandgroep, 2007).

Designated period: This varies from 24 hours to 5 days. However, an AMU of 48 hours is the most common.

The aim of this thesis is to provide insight into the connection between the characteristics of an Acute Medical Unit and the characteristics of a Management Control System, such that AMU performance is controllable. Since AMUs are described in detail above, this thesis will continue describing MCSs.

2.2. Management Control Systems

(14)

are not consistent; individuals are self-interested. (3) Personal limitations: employees perform poorly due to a lack of intelligence, training, experience, stamina, or knowledge for the tasks at hand. Management control issues have primarily an internal focus and it involves addressing the question: are our employees likely to behave appropriately? Optimal control is achieved if the control losses are expected to be smaller than the cost of implementing more controls. Control losses are explained as the cost of not having a perfect control system.

2.2.1. Levers of control framework

Several MCS models have been developed in literature. One common known framework is the levers

of control framework of Simons (1995). He states that managers of modern organizations operating

in a dynamic and competitive market cannot spend all their time on making sure that everyone is doing what is expected, or achieve control simply by hiring good people, aligning incentives and hoping for the best. Employees must be encouraged to initiate process improvements and new ways of responding to customers’ needs in a controlled way. Simons introduced four levers of control; (1) diagnostic control systems, (2) beliefs systems, (3) boundary systems, and (4) interactive control systems.

Diagnostic control systems are used to monitor goals and profitability, and to measure progress

towards targets. Most businesses rely on diagnostic controls to track the progress, however, they are not adequate to ensure effective control. They create pressures that can lead to control failures when empowered employees are held accountable for performance goals and then are left to their own devices to achieve them. This risk underscores the need for managers to think about the other levers of control. Typically, beliefs systems are concise, value-laden and inspirational. Senior managers design beliefs systems to be broad enough to appeal to many different groups within an organization. The principal purpose of the statements is to inspire and promote commitment to an organizations core values. They can also inspire employees to create new opportunities. Without formal beliefs systems, employees (in large, decentralized organizations) often do not have a clear and consistent understanding of the core values of the business and their place within the business.

Boundary systems are based on the ‘power of negative thinking’. Telling employees what not to do

allows innovation, but within clearly defined limits. Boundary systems are stated in negative terms or as minimum standards and are especially critical in those businesses in which a reputation built on trust is a key competitive asset. Effective managers spell out the rules of the game based on the risks inherent in their strategy and enforce them clearly and unambiguously. Working together, boundary systems and beliefs systems create a dynamic tension between commitment and punishment. The fourth lever of control is the interactive control system. This is the formal information system that managers use to involve themselves regularly and personally in the

(15)

that reinforce one another. By using the control levers effectively, benefits of innovation and creativity are not achieved at the expense of control.

2.2.2. Performance management framework

Another MCS framework was developed by Otley (1999). He developed a framework for managing organizational performance, named the performance management framework. This framework highlights five central questions which need to be considered as part of the process of developing a coherent structure for performance management systems. The five central questions are:

1. What are the key objectives that are central to the organization’s overall future success, and how does it go about evaluating its achievement for each of these objectives? 2. What strategies and plans has the organization adopted and what are the processes and

activities that it has decided will be required for it to successfully implement these? How does it assess and measure the performance of these activities?

3. What level of performance does the organization need to achieve in each of the areas defined in the above two questions) and how does it go about setting appropriate performance targets for them?

4. What rewards will managers (and other employees) gain by achieving these performance targets (or, conversely, what penalties will they suffer by failing to achieve them)? 5. What are the information flows (feedback and feed-forward loops) that are necessary to

enable the organization to learn from its experience) and to adapt its current behavior in the light of that experience?

According to Otley (1999), a complete control system involves each of the five elements identified both separately and in combination. He states that the framework can provide a checklist to help ensure that a more complete picture of control systems operation is observed.

2.2.3. Performance management systems framework (PMS)

Ferreira & Otley (2009) developed a framework based on the models of Simons (1995) and Otley (1999). This model provides a broad view of the key aspects of a MCS, and is known as the

performance management systems framework (PMS). This naming reflects a shift from the

traditional compartmentalized approaches to control in organizations to a broader perspective of the role of control in managing organizational performance. Ferreira & Otley view PMSs as the evolving formal and informal mechanisms, processes, systems, and networks used by organizations for conveying the key objectives and goals elicited by management. The PMSs framework is shown in figure 2 and consists of 12 questions. Contextual factors and organizational culture are viewed as contingent variables that might explain why certain patterns of control are more or less effective. Including all 12 questions in this research would be too time consuming. Therefore, the 8 most relevant questions will be used. Since this research is about MCS design, the focus lies on the

(16)

Figure 2. Performance management systems framework (Ferreira & Otley, 2009).

1. What is the vision and mission of the organization and how is this brought to the attention of managers and employees? What mechanisms, processes, and networks are used to convey the organization’s overarching purposes and objectives to its members?

According to Collins & Porras (1996) vision and mission statements are landmarks that guide the process of deciding what to change and what to preserve in strategies and activities in the face of changing environments. The focus of this question is to elicit information on how organizational values and purposes are established and communicated as a means of influencing the behavior of employees (Ferreira & Otley, 2009). The vision and mission are part of Simons’ beliefs systems (1995).

2. What are the key factors that are believed to be central to the organization’s overall future success and how are they brought to the attention of managers and employees?

(17)

in acute care and senior members of the management team; (2) A willingness by clinicians to accept that reorganization involves a change in working practices; (3) A willingness by management to accept that some extra resources are required. In particular the appointment of a discharge planning coordinator, physiotherapist, occupational therapist, social worker and clerical support.; (4)

Cooperation of colleagues in other specialties, such as radiology, endoscopy, and laboratory services, to provide fast-track services in their specialties; and (5) Appointment of ward managers committed to making the new system work.

3. What is the organization structure and what impact does it have on the design and use of performance management systems (PMSs)? How does it influence and how is it influenced by the strategic management process?

According to Johnson et al. (2005) there are multiple forms of organizational structures and they involve choices regarding decentralization/centralization of authority,

differentiation/standardization, and the level of formalization of rules and procedures, as well as configuration.

5. What are the organization’s key performance measures deriving from its objectives, key success factors, and strategies and plans? How are these specified and communicated and what role do they play in performance evaluation? Are there significant omissions?

Key performance measures, also key performance indicators (KPIs), are the financial or non-financial measures used at different levels in organizations to evaluate success in achieving their objectives, key success factors, and strategies and plans in order to satisfy expectations of different

stakeholders (Ferreira & Otley, 2009).

6. What level of performance does the organization need to achieve for each of its key performance measures (identified in the above question), how does it go about setting appropriate performance targets for them, and how challenging are those performance targets?

This reflects the universal tension between what is desired and what is thought to be feasible in determining targets for all aspects of organizational performance (Ferreira & Otley, 2009). Also the use of (external) benchmarks provide legitimacy for targets, this is shown by their use in the healthcare sector (Northcott & Llewellyn, 2003).

7. What processes, if any, does the organization follow for evaluating individual, group, and organizational performance? Are performance evaluations primarily objective, subjective or mixed and how important are formal and informal information and controls in these processes?

Performance evaluations can be objective, subjective, or something in between. Objective evaluations are acceptable in situations where the input-output relationship is clear, the

performance is controllable, or when it is accepted as part of institutionalized practice (Ferreira & Otley, 2009). Subjective evaluations are not traceable to facts, they are about perceptions and feelings.

(18)

8. What rewards — financial and/or non-financial — will managers and other employees gain by achieving performance targets or other assessed aspects of performance (or, conversely, what penalties will they suffer by failing to achieve them)?

Rewards are typically the outcome of performance evaluations and can include both financial and non-financial elements. Financial rewards are explicit, non-financial and informal rewards however can also significantly influence the subordinates’ behavior and thus the workings of the PMSs (Ferreira & Otley, 2009).

9. What specific information flows — feedback and feedforward —, systems and networks has the organization in place to support the operation of its PMSs?

Information flows, systems and networks are essential enabling mechanisms to any performance management system (Otley, 1999). Feedback information is aimed at the correction of past

shortcomings, feedforward information is aimed at attempt to anticipate future events and respond in advance of their occurrence (Ferreira & Otley, 2009).

Since this PMSs framework provides a broad view of the key aspects of a MCS, this framework is usable in this thesis to examine whether these key aspects of a MCS connect with the characteristics of an AMU such that AMU performance is controllable.

Now both AMUs and MCSs are described. In order to answer the key question of this paper, What should the design of a MCS be in an AMU setting, such that the performance of the AMU can be controlled?, AMU performance has to be described first.

2.3. Performance of Acute Medical Units

With performance is generally meant the degree with which a system realizes the objectives it aims at. From this definition, three conclusions can be derived; (1) if objectives are not clearly defined, it is difficult to select proper indicators to measure the performance, (2) in order to reach the

objectives, performance measures should strictly be correlated to the productive factors (3) performance measurement can be conducted in different contexts, objectives and individuals (Tardivo & Viassone, 2010). Following this definition, it can be stated that objectives should be clearly defined. Objectives of AMUs are described in the first section and summarized in table 1. Dimensions determining performance should be linked to these objectives. Then, performance indicators can be developed to measure performance, this should be done properly and indicators should strictly reflect the performance of the dimensions measured. Performance indicators are strongly emerging in healthcare because of policymakers’ and patients’ pressure on healthcare organizations to account for, and improve quality and efficiency of care (Van der Geer et al., 2009). Scott et al. (2009) conducted a study to assess the effects of AMUs in hospitals. They studied nine peer-reviewed reports of before-after analyses of seven units in the UK and Ireland. These reports were selected by looking for seven performance indicators. These were: LOS (Length of stay), mortality rates, waiting times in the Emergency Department, discharge disposition, readmission rates, bed cost and resource utilization, and patient and staff satisfaction. These seven indicators are in line with the AMU objectives (table 1) and together they provide a proper view of the

(19)

2.3.1. Length of stay (LOS)

LOS can be defined as the time spent in the hospital (in days) from the moment a patient arrives in the hospital until the moment of discharge. So when talking about LOS, in-hospital LOS is meant. A distinction can be made between LOS > 30 days and LOS < 30 days (Moloney et al., 2012). Reducing LOS is of importance because hospital facilities are expensive and a hospital stay is usually the most costly episode in the patient’s experience of care. Managers and clinicians are therefore under pressure to reduce avoidable hospital admissions (Edmans et al., 2011). Reducing LOS is desirable because a shorter LOS means increased use of capacity and the recognition that earlier discharge is possible for many patients without having an adverse effect on clinical outcomes (Moloney et al., 2006). Many studies looked at (in-hospital) LOS to determine AMU performance. Moloney et al. (2007) showed that the introduction of an AMU speeded access to an acute medical service, and reduced costs by reducing LOS. LOS was significantly improved, both short term and long term, and was less varied between medical teams compared to the situation without AMU. No change in readmission rates occurred. Furthermore, in their study in 2004, Moloney et al. saw the median LOS decrease from 6 to 5 days.

2.3.2. Mortality rates

Mortality rates can be determined by dividing the amount of patients died by the total admissions. It is clear that it is preferable to have a very low mortality rate, however, many causes of death are not medically avoidable. Only a mortality rate that reflects avoidable deaths indicates real AMU

performance. A common used measure is the Hospital Standardized Mortality Ratio (HSMR). The HSMR is the ratio between the observed number of in-hospital deaths and the predicted number of deaths, determined by comparing the patient case-mix with the national average. The outcome is standardized around 100. A value above 100 indicates higher mortality than average, and a value below 100, indicates lower mortality (Van den Bosch et al., 2012). HSMR rate is already commonly used in the UK and in the Netherlands. However, this rate indicates total hospital mortality, and not the mortality of a specific unit. Nevertheless, Rooney et al. (2008) described the outcome for 19528 unselected acute medical patients admitted via the ED of a busy teaching hospital. The development of a dedicated AMU was associated with 45% relative reductions in all-cause annual hospital

mortality, and 36% relative reduction in 30-day hospital mortality. So, when not using the HSMR, mortality rates have value if measured the same way over a period of time. For example the relative reduction in all-cause hospital mortality over 5 years can be measured, like Rooney et al. (2008) did.

2.3.3. Waiting times in the Emergency Department (ED)

(20)

2.3.4. Discharge disposition

The discharge disposition is the rate at which patients are discharged and sent home directly within 24h/48h of presentation to the ED (Scott et al., 2009). This reflects the effectiveness of the unit because the faster patients are discharged, the less capacity they use meaning there is more capacity left for other patients resulting in a higher capacity use. The aim is to achieve a high discharge disposition, meaning that patients do not have to be transferred to regular nursing wards but can be send home directly.

2.3.5. Readmission rates

Readmission rates can be defined as the amount of people readmitted to the hospital within 30 days of discharge divided by the total amount of admissions. Rising readmission rates can undo the advantages of decreasing LOS. The study of Moloney et al. (2007) showed decreasing LOS with no change in readmission rates. Further, many readmissions to hospitals may not be preventable because they represent fresh events in patients with chronic illnesses and frequent co-morbidity (Moloney et al., 2006). So, using readmission rates, medical conditions of the patient population should be taken into account, and conclusions can only be derived if there is a real significant outcome.

2.3.6. Bed cost and resource utilization

Given the growing demands to manage costs and maximize efficiency in acute hospital services, information on patterns of resource utilization is valuable (Moloney et al., 2004). Managers and clinicians are under pressure to reduce avoidable hospital admissions. Hospital facilities are expensive and a hospital stay is usually the most costly episode in the patient’s experience of care (Edmans et al., 2011). According to Moloney et al. (2004) effective and appropriate bed usage is one of the essential elements in the efficient care and management of patients. In their study cost per bed-day increased with 16.9% after the implementation of an AMU, however bed-day savings were estimated at approximately 10% by an increase in episodes of 10%. Furthermore, resource and capacity utilization should be maximized to reduce variable costs per output, not only in de units itself but also the supporting devices.

2.3.7. Patient and staff satisfaction

Beside the medical and operational factors, it is important that patients and personnel are satisfied about treatment and working conditions in the unit. Griffiths (2009) already mentioned that nurses experience job related stress, work overload and conflict. However, these nurses were satisfied because of the team’s cohesion to a common enterprise over which they perceived a great deal of control. Furthermore, the respect for one others’ professional capabilities led to a positive working environment. Both patient and staff satisfaction can be measured using questionnaires.

2.4. Conceptual model

Now AMUs, MCSs, and AMU performance are described. To answer the key question, What should the design of a MCS be in an AMU setting, such that the performance of the AMU can be controlled?

it is interesting to find out what the relationships are between these three concepts. To answer the key question, a conceptual model is developed (see figure 3). The AMU characteristics are

(21)

of the unit itself and are therefore placed as AMU characteristics. Questions 5 to 9 from the PMS model are questions about; key performance measures, targets, performance evaluation and reward systems. These concepts are ways to actively control the unit and are therefore placed in the

conceptual model as MCS characteristics. Information flows, systems and networks supports this control and is therefore also a MCS characteristic. The AMU performance measures have been outlined in chapter 2.3. Five sub-questions can be subtracted from this conceptual model, these are outlined below.

Figure 3. Conceptual model.

1. What are the characteristics of an AMU in practice? 2. What are the characteristics of an AMUs MCS in practice?

3. What are the measures of AMU performance outcomes that should be controlled? 4. In what way are the AMU characteristics linked to the MCS characteristics? 5. Are the MCS characteristics capable of controlling the performance outcomes?

All sub-questions will be answered in further sections. Questions 1 and 2 will be answered in the results section. Question 3 is already answered in the theory section. Questions 4 and 5 will be answered in the analysis section.

AMU characteristics

- Vision and mission - Key success factors

- Acute medical patients (see table 2) - Organization structure (incl. professional personnel (see table 3)) - Location close to ED and diagnostics facilities

- Predetermined designated period

1

4

MCS characteristics

- Key performance measures - Target setting

- Performance evaluation - Reward systems

- Information flows, systems, and networks

2

5

AMU performance measures

- Length of stay (LOS) - Mortality rates - Waiting times in the ED - Discharge disposition - Readmission rates - Bed cost and resource utilization

- Patient and staff satisfaction

(22)

3. Research design

The aim of this thesis is to provide insight into the connection between the characteristics of an Acute Medical Unit and the characteristics of a Management Control System, such that AMU performance is controllable. To provide this insight case studies are conducted. The objective of a case study is to obtain multiple perspectives of a single organization, situation, event or process at a point in time or over a period of time. In studying multiple subjects, a deeper understanding of the subject emerges. When multiple units are chosen it is because they offer similar results for

predictable reasons (literal replication) or contrary results for predictable reasons (theoretical replication) (Cooper & Schindler, 2006).

Since multiple cases are researched, a sixth sub-question can be added to this research:

6. What are the similarities and differences between the different units? This question will be answered in section 4, the results section.

The type of case study used in this research is an holistic multiple case study (Yin, 2009). A multiple case study is suitable if it is intended to generalize the outcomes of the case study (Braster, 2000). Furthermore, the more cases studied, the more valid outcomes are (Swanborn, 1996; p.56). Therefore, three hospitals have been selected and three AMUs are studied. Both content and pragmatic selection criteria are used (Swanborn, 1996; p.59-60); hospitals had to have an

operational AMU and needed to be located in the Netherlands. Selected cases are outlined in table 4. At the moment the AMU in the Leiden University Medical Center finished its pilot-phase

positively, there was a positive evaluation in February 2012. Future plans consist of expanding this AMU to 30 beds and to incorporate more specialties. In this research, the current situation of the AMU is used. It will be mentioned if references are made to the future situation. The AMUs in the Kennemer Gasthuis Haarlem and Slingeland hospital Doetinchem did not have a pilot-phase.

(23)

Hospital Hospital type AMU size Operational since Names interviewees Functions interviewees Kennemer Gasthuis Haarlem

Top clinical 42 beds September

2010

M. de Geus Unit manager AMU

Slingeland hospital Doetinchem

Regional 35 beds April 2012 S. Zielhuis Ward manager AMU

Leiden University Medical Center (LUMC) University medical center 10 beds, goal is 30 beds February 2011 R. Parlevliet Dr. A.J. Fogteloo Drs. W.J. Dannenberg

Nursing team leader AMU Senior medical specialist (medical chief AMU) Member steering group AMU Table 4: selected cases.

According to Yin (2009) interviews are an essential source of case study evidence, however he also states that interviews should be considered verbal reports only. Therefore, interview data should be corroborated with information from other sources. Using multiple sources of evidence, also named triangulation, can help to establish the construct validity and reliability of the case study evidence. The most important advantage of triangulation in case studies is the development of converging lines of inquiry. Any case study finding or conclusion is likely to be more convincing and accurate if it is based on several different sources or information, following a corroboratory mode (Yin, 2009). Therefore, besides the interviews a documentation research is conducted. Files included AMU policy, protocols, and agreements. Furthermore, informal field observations were made throughout a field visit during which also the other evidence was being collected.

According to Cooper & Schindler (2006) a researcher often uses ad hoc abbreviations and symbols during a personal interview. They state that soon after the interview, the researcher should review the outcomes of the interview and complete the existing gaps. Therefore in this research, soon after the data had been gathered, a field editing review was conducted. Furthermore, Cooper & Schindler (2006) state that it is important to validate the field results. Therefore in this research, interview results were written down in an interview report and sent back to interviewees for validation. Interviewees had the opportunity to add comments and complementary notes. A number of comments and notes were made and these comments and notes are incorporated in the interview reports and the results section.

An interpretation scheme is developed to interpret the results. This interpretation scheme is

(24)

4. Results

In this section results will be outlined. These include the results of the interviews held at the different units complemented by collected documents of the units. First, results of AMU

characteristics will be described, this will be done separately for each case. In doing so, sub-question 1 will be answered (What are the characteristics of an AMU in practice?). Second, results of MCSs characteristics of AMUs in practice will be described. This way, sub-question 2 will be answered (What are the characteristics of an AMUs MCS in practice?). Below, key results are outlined. The complete interview reports are outlined in appendices 2 (Kennemer Gasthuis), 3 (Slingeland hospital), and 4 (Leiden University Medical Center).

4.1. AMU characteristics

4.1.1. Vision and mission

Vision and mission statements are landmarks that guide the process of deciding what to change and what to preserve in strategies and activities in the face of changing environments (Collins & Porras, 1996). In this part, reasons for setting up the AMU and AMU goals will be described. These reasons are comparable between the three AMUs. Also the majority of goals from literature correspond to goals set in the AMUs, only small differences exist (see appendix 6 to compare the three units). For all AMUs, goals have been formally specified at the AMU start-up by management or a steering group, but are not continuously communicated or visible anymore. Documents are often stored at the data warehouse system and AMU management supposes these goals are known by employees and other stakeholders. Below, goals of the specific units are outlined.

Kennemer Gasthuis

AMU goals at the Kennemer Gasthuis are formally specified and are; creating a buffer function, fast diagnostics and reduction of LOS, and rest at regular nursing wards. Furthermore, the Kennemer Gasthuis struggled with patients lying in ‘wrong’ beds. This problem is almost solved now the AMU is set in place. Another problem were access blocks. Since the AMU opened, there have been no more access blocks. A detailed description of AMU goals can be found in the interview report (appendix 2). Slingeland hospital

AMU goals at the Slingeland hospital are; improve the plan ability of acute care and generate the acute capacity at one place, avoidance of the ‘wrong-bed problem’, improve safety and quality, faster determination of work diagnosis and shorten LOS, and avoidance of access blocks. In the Slingeland hospital there used to be a weekly access block or threat to access block. After implementing the AMU, access blocks have only occurred due to extraordinary causes such as a bacteria. The specific AMU goals are defined but still need to be officially determined. The AMU vision is that it should be accessible 24/7, be safe and effective, transparent, providing hospitality, and equipped for the future. A detailed description of AMU goals can be found in the interview report (appendix 3).

Leiden University Medical Center

(25)

4.1.2. Key success factors

Key success factors (KSFs) are those activities, attributes, competencies, and capabilities that are seen as critical pre-requisites for the success of an organization (Sousa de Vasconcellos e Sá and Hambrick, 1989; Thompson and Strickland, 2003). Looking at the KSFs similarities and differences between the AMUs do exist. Similarities are that they all mention that good communication and commitment from all parties involved is essential. Because many different parties are involved in an AMU managers of the AMUs think that making clear and official agreements before starting the unit is essential, and that these agreements should be revised when necessary. Further, all AMUs do agree with the KSFs from literature, however these KSFs are found to be very general and hard not to agree with. In all AMUs, KSFs are not specified and also not specifically brought under the attention of employees, they are supposed to be commonly known. Detailed descriptions are given in the interview reports (appendices 2, 3 and 4).

According to the AMU unit manager a low bed occupation is essential, the AMU should never be fully occupied. An occupation of 50-60% is optimal, with a higher occupation the AMU loses her buffer function. The making of agreements upfront is also essential. Agreements should be clear and made with all parties involved. Furthermore, good communication and commitment of the entire organization are important. The AMU is a new concept, proper information should be given in order to inform the whole organization and to gain acceptance. The unit manager finds the other key success factors mentioned in literature also important.

Slingeland hospital

KSFs are; setting realistic goals, having the right vision for the unit, making good official agreements with all parties involved and revise them when necessary, and good communication. KSFs are not specifically brought under the attention of employees. It is important that communication between AMU, ED and other wards stays good. Units have to work together for one hospital-broad goal, not for their own unit. According to the ward manager this is very important for the AMU. “We have to

be ‘friends’ with all units” (Interview report, appendix 3). If regular wards are not willing to receive

AMU patients, the AMU loses her function. With a clear communication, wards understand each others’ position which will contribute to a good relationship.

Leiden University Medical Center

KSFs for the LUMC are a decent unit structure and good patient throughput. The unit should be organized in a way that guarantees throughput. Also regularly meetings with parties involved and building good relationships with medical specialists and other nursing wards are essential. “It is

important to build trust” (Interview report, appendix 4). This is specifically important in transferring

patients, if nursing wards do not admit patients from the AMU, the AMU loses her function. Also the intrinsic motivation of the team is very important, employees have to work towards the dismissal of patients and this focus point is different from focus points in other units.

4.1.3. Acute medical patients

(26)

is too specific to handle on a ‘general’ unit. Only small differences exist between the AMUs. Specialties admitted at the different AMUs are outlined in the summarizing table below, and more detailed in the interview reports (appendices 2, 3 and 4). Further, certain medical conditions not suitable for AMUs were outlined in the literature. Results from practice regarding these medical conditions of all three cases are compared and outlined in appendix 7.

4.1.4. Structure (including professional personnel)

Organizational structure involves choices regarding the decentralization/centralization of authority, differentiation/standardization, and the level of formalization of rules and procedures (Johnson et al., 2005). In the results below, only the hierarchical structure of the AMU within the hospital is taken into account. Tables consisting the hierarchical structures of the units researched and

functions of professional personnel compared with literature can be found in the summarizing table below and in the interview reports (appendices 2, 3 and 4). Professional personnel structure is comparable between the three cases. Their hierarchies all have four or five levels, and functions in different levels are comparable to each other and to the functions named in literature (see appendix 8 to compare the three cases). In all three AMUs employee responsibilities are not really specified other than in the function description. Organizational and medical responsibilities exist separately next to each other. Unit managers and team leaders do have the delegated responsibility for the operation of the AMU. However, this is discussed in meetings and is not specifically formalized. Also the position of the AMU within the chain is comparable between the three units. The position of the AMU within the chain is valuable, it supports the ED because patients can be transferred directly to the AMU and ED professionals do not have to search for in-hospital beds anymore. The role of the AMU can be seen as a hub.

The hierarchy in the unit is outlined in appendix 2 (interview report). Organizational and medical responsibilities do exist separately next to each other. Other than in the function description, responsibilities are not further specified. Medical, organizational and unit management together determine unit policy. Within the chain the AMU is perceived as being valuable because ED professionals do not have to search for in-hospital beds anymore because patients can be transferred to the AMU.

For the hierarchy of the AMU, see the interview report (appendix 3). A responsibility matrix is in development, this matrix will be hospital broad and is planned to be finished by the end of the year 2012. Recently, new function descriptions have been written including tasks and responsibilities. Leiden University Medical Center

(27)

4.1.5. Location close to ED and diagnostic facilities

According to literature, location close to ED and diagnostic facilities is important. All three AMUs acknowledge that location close to ED and diagnostic facilities is logistically preferable, but none of the three AMUs is located close to ED or diagnostic facilities due to construction restrictions or specific circumstances. However, in practice this is not perceived as a big problem. In the interview reports is described in detail where the different AMUs are located in practice (appendices 2, 3 and 4).

4.1.6. Predetermined designated period

For all three AMUs the predetermined designated period is 48h.

4.2. Management Control System

4.2.1. Key performance measures

Key performance indicators are the financial or non-financial measures used in organizations to evaluate success in achieving objectives, key success factors, and strategies and plans in order to satisfy expectations of different stakeholders (Ferreira & Otley, 2009). Not all AMUs measure the same performance indicators. In the interview reports (appendices 2, 3 and 4) is outlined whether KPIs named in literature are also KPIs according to the different AMUs researched. In appendix 9 the three cases are compared to each other. All AMUs agree that the indicators mentioned in literature do give a proper view of AMU performance but none of the AMUs is measuring all these indicators. Further, other AMU specific indicators are measured. It is remarkable that all AMUs measure the obligatory IGZ2_{indicators set by government but none of the AMUs is really controlling all these} measures. A lot of data is collected but this data is not directly used.

Indicators measured are determined in cooperation with the AMU network. Besides the AMU specific indicators, other indicators measured are the compulsory indicators from the government (IGZ indicators). These are indicators to determine healthcare quality performance and are measured daily. Data is automatically stored in the data warehouse system and all sorts of

information is saved. Periodically the AMU receives management information about the measured indicators, personnel at the AMU itself is not able to get the information out of the system. According to the unit manager: “Personnel in charge of the data warehouse system can measure

anything if we ask for it” (Interview report, appendix 2). Obligatory IGZ indicators are measured, but

there is no focus for these measures. IGZ indicators are not used to control performance, AMU specific measures are perceived to be much more important.

AMU management complemented the indicators named in literature with indicators about access blocks and patients at the right unit. Other indicators measured are the dismissal percentage within 48h and the occupation of regular wards. Further, issues will be measured if a problem is

Controlling acute medical care