Adapted Lean Thinking for Healthcare Services: An Empirical Study in the Traditional Chinese Hospital

Abstract

This chapter looks at how Lean Thinking can be adapted using a model derived from a case study of a large Traditional Chinese Hospital. After a restructuring in divisions and the implementation of the care programmers and clinical pathways, hospital management found that they had no tools to evaluate if these changes were resulting in a Lean Thinking approach on the work-floor. In agreement with hospital management, an existing tool of Business Process Re-engineering measurement was adopted and adapted to the specific context of health care. This chapter reports on how the quantitative model was changed and validated in order to come up with a useful instrument to measure the Lean Thinking of the employees in the hospital. The Hospital Lean Thinking (HLT) tool can be useful to measure the effects of changes that are assumed to lead to more Lean Thinking or even patient focus. In this way the pay-off of these investments can be made more tangible. The HLT tool offers hospitals a way to evaluate how they are evolving towards more Lean Thinking.

1 Introduction

Issues of health care in Chinese hospitals include crowding and long waiting times; a visit to a hospital in China is rarely a smooth and satisfying experience. Nearly 400 outpatient departments (ODs) operate in Beijing and provide care for patients around the clock. There are annually 24 million OD visits, which result in 4 million hospitalisations (accounting for 60 % of hospital admissions overall). Patients in Beijing rely more heavily (40 %) on OD services than other big cities (22 % other cities).

Causes for OD overcrowding are well known (Jones and Mitchell 2006) and include hospital bed shortage, high medical acuity of patients, increasing patient volume, shortage of examination space and shortage of RN (Registered Nurse) staff. Even though these issues are well recognised, alleviation of these problems in OD is not trivial and requires addressing complex systems issues. This is not the focus of the paper which instead concentrates on exploration of improving effectiveness of OD operations.

It has been shown that despite various management, methods and techniques have been developed over the last years to approach the need for constant change and improvement in OD management. There are still many difficulties and obstacles between doctors and patients in the aspects of expression, transmission, understanding and communication of information. These obstacles affect the improvement of perioperative efficiency, such as reducing waste time or overtime of cases, adding more cases during regular diagnosing hours and increasing profitability of the hospitals. Thus, these impediments need to be reduced to shorten the overall outpatient time. The Lean Thinking approach potentially can play a vital role in improving outpatient performance and raising the level of hospital services (Womack and Jones 1996).

The Lean Thinking strategy is considered to have the capacity to enable delivery of better health care at the lower overall cost (Jones and Mitchell 2006). Lean thinking originated in the Toyota Production System in the 1950s and has been further developed by Womack and Jones (1996). The aim of Lean Thinking is to provide what the customer wants, quickly, efficiently and with little waste (Jones and Mitchell 2006; Womack and Jones 1996; Young et al. 2004). It can provide the following benefits (Jones and Mitchell 2006): improved quality and safety, improved delivery and improved throughput, that is, using the same resources with higher efficiency and accelerating momentum. A stable working environment with clear, standardised procedures can create the foundations for constant improvement.

From a perspective of manufacturing, Lean Thinking is a strategy to achieve competitiveness through identification and elimination of wasteful steps in products, services or processes (Womack and Jones 1996; Shinohara 2006). It aims to substantially smooth the flow and drastically reduce waste and process variations (Womack and Jones 1996; Shinohara 2006; Taj and Berro 2006; Reichhart 2007). Waste is defined as the activity or activities that a customer would not want to pay for and that do not add value to the product or service from the customer’s perspective (Shinohara 2006). Once waste has been identified in the current or existing state, a plan is formulated to eliminate this to attain a desired future state in as effective and efficient a manner as possible. These activities belong to one of three sets of operations (Moden 1993):

• Non-value-added activities

• Necessary, but non-value-added activities

• Value-added activities

Similarly, in a healthcare service organisation, wasted time leads to high cost and affects the quality of patient care, and thus it should be reduced. To achieve the leanness target, the activities that add little or no value, or that adversely affect the smooth flow of the process, are considered for elimination. Because the main mission of health care is to treat and cure patients who are the end consumers in the care process, it is argued that the patient should have an input into defining what creates value in health care (Young et al. 2004).

Some management professionals argue that lean manufacturing does not translate well to service industries. Bowen and Youngdahl (1998) show how it can apply to health care by providing theory, case studies and context for lean applications. Flinders Medical Centre, a medium-sized public sector teaching hospital in Adelaide, South Australia, has for some time been implementing lean strategies (King et al. 2006) and has been able to operate below its budgeted costs (Jones and Mitchell 2006). Lean thinking has also been advocated in the healthcare setting of the USA through the use of the Six Sigma methodology, which in many ways resembles lean production techniques (Young et al. 2004; Tolga Taner et al. 2007; Dahlgaard and Dahlgaard 2006). Other related literature also reveals that the implementation of Lean Thinking brings benefit to health care (Jones and Mitchell 2006; King et al. 2006; De Koning et al. 2006; Jimmerson et al. 2005; Ahluwalia and Offredy 2005).

2 Literature Review

2.1 Lean Thinking in Health Care

The strategies of Lean Thinking are applicable to health care (Reichhart 2007; Moden 1993; Bowen and Youngdahl 1998). An application of Lean Thinking to health care lies in minimising or eliminating delay, repeated encounters, errors and inappropriate procedures (Moden 1993 p, 162). Hospitals may apply Lean Thinking to provide better services to their patients, especially in the use of operating theatres. One of the key principles of Lean Thinking is respect for the customer. In a healthcare setting, the patient is the primary customer to the healthcare services since the patient justifies the existence of such services. Applications of the Lean Thinking approach in health care need to consider how to engage patients who are the end consumers in the care process (Moden 1993; King et al. 2006). Literature indicates that the implementation of Lean Thinking has the potential to benefit health care (Tolga Taner et al. 2007). Some healthcare services around the world are considering applying the Lean Thinking approach. Applications of lean in health care have been published in academic journals and other media (see Table 8.1).

Table 8.1 Summary of Lean Thinking literature for healthcare services

2.2 Performance Measurement Systems

Traditional performance measurement systems (PMSs) have often failed to measure and integrate all the factors critical to success of a business (Yurdakul and Ic 2005; Wegelius-Lehtonen 2001). To deal with the new environment, new PMSs have been proposed, such as the Activity-Based Costing System (Koota and Takala 1998), the Balanced Scorecard (Kaplan and Norton 2005), the SMART System (Hudson et al. 2001) and the Performance Measurement Questionnaire (Park et al. 1998). There are also approaches for proposing criteria for the design of PMSs (Neely et al. 2005).

Despite the availability of the various approaches to develop PMSs, there are a few performance systems that are exclusively designed and applied to service businesses (Bititci 1995). Atkinson et al. describe a stakeholder-based PMS and while they do not draw attention to the service measurement aspect, nevertheless they do apply their system to measure a bank’s various outputs, one of which is service innovation (Atkinson et al. 1997).

The literature on Lean Thinking and performance measurement reveals that there is a lack of quantified and unique indigenous management models (Yu et al. 2000). Existing management models can be separated into three distinct types: the static, (Harding and Popplewell 2000; Candido and Morris 2000), the dynamic (Dobni and Luffman 2003; Dye 2004) and the mixed models (Zhang and Prybutok 2004; Candido 2005; De Toni and Tonchia 2005). Existing static models offer a representation of an organisation, dynamic models offer a generic process of strategy formulation and implementation and mixed models show what dimensions can be changed at each stage. The mixed model can be further developed in the form of quantification.

2.3 AHP Approach for Performance Assessment

The analytic hierarchy process (AHP) developed by Saaty (1980) provides a suitable and appropriate way of analysing a performance measurement model because AHP is a multiple criteria decision-making technique that allows subjective as well as objective factors to be considered in a decision-making process (Rangone 1996; Dey et al. 2006). Performance measurement is usually a team effort and AHP is one available method for forming a systematic framework for group interaction and group decision-making (Saaty and Vargas 2001).

Although AHP has been used before to measure performance in service industries, Dey et al. (2006) applied AHP to performance measurement of intensive care units in hospitals, and Chow et al. (2005) adapted the AHP methodology to the service innovation of restaurant industry; it appears not to have been applied as a performance measurement model specifically for service innovation strategy. The objective of this study is to develop a quantitative performance measurement model that can be used for service process performance and strategy selection.

3 Research Objectives and Methodology

The objective of the research adopted under the heading of Quantitative Models for Service innovation Strategy (QMSQS) was to identify tools and techniques that would facilitate:

• Identification of factors affecting service innovation

• Identification of the relationship between factors affecting service innovation

• Quantification of the different factors affecting service innovation and on the overall performance of the service processes

• ‘What if?’ analysis on process performance and strategy selection

The six steps of the approach were developed as a result of the QMSQS methodology implementation as depicted in Fig. 8.1. The details of this approach have been explained and discussed through a case study in Sect. 8.5.

Fig. 8.1

A methodology for service innovation strategy

Stage 1: Establishing Strategic Perspectives

At the start of a QMSQS approach, the strategic perspectives should be established within the strategic plan. These perspectives should be used for all the service processes of a business. Perspectives are connected together with the help of the coupling operation. It is not necessary to establish a methodical sequence. However, if the perspectives structure (i.e. the sequence of individual perspectives) is logically established, the automatic generation of a model is considerably simplified and the logical structure of the cause-and-effect chain is made clear.

Stage 2: Defining Strategic Performance Measures

With the QMSQS method, strategic objectives and critical factors taken from different points of view are allocated to the company (the so-called perspectives). These include internal perspectives (e.g. learning and development perspective, process perspective) and external performance perspectives (e.g. patient perspective, hospital and economic perspective). This arrangement of the key performance indicators achieves a certain balance between short-term and long-term goals, hospital and non-hospital key performance indicators, leading and lagging indicators and internal and external points of view. The introduction of specific key performance indicators for different sectors adds a further benchmarking component to the concept.

In general, the QMSQS approach classifies the relationships between critical factors which affect service innovation as follows:

• Direct (vertical) effect

• Indirect (horizontal) effect

• Self-interaction effect

Stage 3: Describing Service Processes

The quantitative model for service innovation strategy (QMSQS) developed in Sect. 8.4 uses system dynamics models (SDMs) to describe service processes.

The components necessary to provide a full description of a service process are thus procedures, events, products/statuses, processors, organisational units and information technology resources.

Considering all the effects on all the elements of the procedure for every event would severely complicate the model and lead to redundancies in the description. In order to reduce this complexity, the general context is divided into individual models that represent separate modelling and design aspects. These can be processed largely independently of each other. The models are divided in such a way that relationships between the components within a model are very high while those between the models are only relatively loosely linked.

The AHP approaches to the measurement of service innovation are presented in the Evaluation Models.

Stage 4: Identifying Factors to Measure Performance

The identification of the factors follows on from the service process model. Having created a service process model for the performance measure in question, in this step the activities within each process were analysed in order to establish factors which may contribute towards the particular performance measure. This was achieved through the use of cause-and-effect analysis. The cause-and-effect analysis shows how each service process (shown as a major cause) may have an impact on the performance measure.

These factors provide the structure view for that particular result model. In addition, they also lead to the identification of performance measures for use at the tactical and operational levels. The measures identified against each process are considered to be tactical performance measures corresponding to each process. The causes listed under each service process provide the basis for operational measures for that process.

The use of the cause-and-effect analysis technique as described above provides a useful approach to analysis and provides some form of guidance for modelling of the structure of performance measures. It is possible to transpose the information contained in the cause-and-effect model into a more visualised data structure commonly used in information systems analysis.

Stage 5: Linkage of Resource Class to Data Model

The project team should consider linking the PMSs to competency development. This enables employees to focus not only on their strategic goals but also on competencies that may be critical for strategy execution, such as teamwork and communication. A company’s core competencies (including management competencies) should be selected based on the company’s business strategy, core values and culture.

The PMSs should be linked to variable pay to motivate all employees to work together to achieve the company’s strategic goals. When designing the variable pay component, it is important to consider the relative importance and priority of objectives in each perspective and at each level of the organisational structure. The decisions made about relative importance and priorities communicate clear messages to all managers and employees.

Stage 6: Development of Control Models

The control model is a model of the interactions between states and events of a real-time system. The system responds to outside events and passes through a series of modes or states. As events occur, they initiate changes in the system’s state.

A possibility for creating the control model, which is a means to display how the measures are influencing each other, is the use of a system dynamics (SD)-oriented approach, using a matrix. In the vertical column the influencing measures are written and perpendicular to them the influenced measures. The levels are chosen according to the levels in the structure view.

In the matrix the level of relationship between individual measures is illustrated in the normal SD fashion. That is:

• ++ Strong positive impact

• + Positive impact

• Blank no impact

• − Negative impact

• −− Strong negative impact

This SD-oriented matrix approach provides a valuable technique to describe the dynamic view of a PMS, because it is simple and straightforward to understand.

The use of the SD approach also promotes deployment of priorities between various levels of the PMS. Once the business objectives and strategy are established this will allow prioritisation of the strategic performance measures according to the objectives and the strategy. Once the top-level priorities are in place these could be deployed to lower levels through the relationship matrix using the relationship level as a deployment aid. This type of deployment approach is well described in the widely available SD material. However, in using this approach one must practice particular attention to ensure that priority deployment is carried out between two distinct levels.

4 Quantitative Model for Service Innovation Strategy (QMSQS)

This chapter introduces four types of quantitative models, shown in Fig. 8.2, which can support the aforementioned methodology to service innovation indigenous management. Each model is a collection of operation formulae, which can capture the information necessary to describe a service process’s state and behaviour. These models are considered essential: a business model, control model, result model and an evaluation model.

Fig. 8.2

Descriptive views of QMSQS

The business model also includes four classes of objects: an organisation object, an information object, a function object and an operator object as shown in Fig. 8.2. These classes are proposed based on the object-oriented approach (OOA).

The OOA makes the basic assumption that the world is made of an organised collection of objects. According to this hypothesis, anything within a service firm is also considered an object characterised by its unique and invariant identifier, its object class and its state, defined by the values of its attributes. A business object might be a concrete thing (e.g. a piece of equipment, an employee or a product), an abstract thing (e.g. an enterprise goal, a business process, an enterprise activity, a performance measure or a service) or a relationship between things (e.g. a logical link between two objects).

Individual models are not described in detail in this chapter, since the purpose of each model has been identified and a detailed explanation of its information content and functionality is given in Sect. 8.3 and an explanation of their application is given in the case study (see Sect. 8.5). The purpose of this section is to identify and describe the content and functionality of the QMSQS, since these quantitative models must satisfy the methodology requirements of all the steps.

4.1 Business Model

Business models are the most important building blocks of our quantitative models. From the OOA point of view, they are a description of a set of abstract business objects that share the same attributes. Our business models are divided from four classes:

4.1.1 Organisational Class

The organisational class is a typical form of representing organisational structures. This kind of class reflects the organisational units (as task performers) and their interrelationships, depending on the selected structuring criteria. In order to show the individual positions in the company that have, for example, job descriptions, the separate position object type is available. Organisational units and persons can also be assigned a type.

4.1.2 Function Class

The functions to be performed (processes) and their interrelationships with each other form a second view, the function class. It contains the description of the function, the enumeration of the individual sub-functions that belong to the overall relationship and the positional relationships that exist between the functions.

Modelling methods often display functions in connection with objects from the other descriptive views of QMSQS. The relationship between data and functions is displayed, for example, to specify the transformation process of a function via the input/output data of that function.

In the QMSQS architecture, however, the various areas of analysis are kept strictly separate. Within the function only those representational forms are used which illustrate the connections between the functions. One example is the relationship between functions and data displayed in the control model of QMSQS.

4.1.3 Information Class

Events such as ‘patient order received’ or ‘invoice produced’ define the point at which a change in the state of information objects (data) occurs. They are described in the information class of the QMSQS architecture.

An information class includes a description of the semantic data model of the field which is to be examined. According to the QMSQS division principle, this description contains both the objects which specify the start and end events of a process chain and the status descriptions of a service process chain’s relevant environment.

An entity-relationship model (ERM) is the most widely used designing method for semantic data models. This modelling method uses a number of specialised terms such as entity type, relationship type and attribute. The relationships which exist between these objects and information technology (IT) are linked via coupling operators which will be introduced in the next section.

4.1.4 Flow Class

The flow class is structured as an integrated view in which the relationships between the other object classes are described by the operator object.

The operator object consists of a set of operations that take one or two objects as the input and produce a new object as their result. The fundamental operations in the flow class are select, project, Cartesian product and associative.

The select operation selects tuples that satisfy a given predicate. We use the lowercase Greek letter sigma (σ) to denote selection. The predicate appears as a subscript to σ. The argument relation is in parentheses after the σ.

The project operation allows us to produce this relation. The project operation is a unary operation that returns its argument relation, with certain attributes left out. Since a relation is a set, any duplicate rows are eliminated. Projection is denoted by the uppercase Greek letter pi (Π). We list those attributes that we wish to appear in the result as a subscript to Π.

The Cartesian-product operation, denoted by a cross (×), allows us to combine information from any two objects.

The associative operation is a binary operation that allows us to combine certain selections and a Cartesian product into one operation. It is denoted by the ‘join’ symbol ‘’. The associative operation forms a Cartesian product of its two arguments, performs a selection forcing equality on those attributes that appear in both relation objects and finally removes the duplicate attributes.

4.2 Control Model

Indigenous management is mainly concerned with detailed action planning and the preparations to take these action steps. Thus, all general decisions have already been made and—sometimes—a model might already exist. The purpose of system dynamics modelling in service innovation indigenous management is to communicate the decisions that have been made, involve stakeholders in the implementation and provide means for ‘optimisation’ within the given decision frame. The usage of system dynamics in indigenous management differs from system dynamics modelling in strategy formulation insofar that:

• The scope of changes that can be made to the actual strategy is very limited.

• Frequently, a high number of people from all organisational hierarchical levels are involved either in the modelling itself or in connected activities.

• The main purpose of modelling is for understanding and refining a decision that has already been made.

This section is a detailed discussion of the system dynamics modelling, which allows for simple representation of complex cause-and-effect relationships. For the discussion that follows, it is important to understand that it is the levels (or state variables) that define the dynamics of a system. For the mathematically inclined, we can introduce this in a more formal way. The following equations show the basic mathematical form of the QMSQS:

$$\begin{array}{ll}measures[i]_{t}&={\displaystyle{\int }_{0}^{T}levels{[j]}_{t}}dt;\\ \frac{d}{dt}measures[i]_{t}&=levels[j]_{t}\end{array}$$

(8.1)

or

$$ rate{s}_{t}=level{s}_{t}dt={\displaystyle {\int }_{0}^{T}rate{s}_{t}dt\frac{d}{dt}level{s}_{t}}$$

$$ rate{s}_{t}=\mathbf{g}(level{s}_{t},au{x}_{t},dat{a}_{t},const)$$

(8.2)

$$ au{x}_{t}=\mathbf{f}(level{s}_{t},au{x}_{t},dat{a}_{t},const)$$

(8.3)

$$ level{s}_{o}=\mathbf{h}(level{s}_{o},au{x}_{o},dat{a}_{o},const)$$

(8.4)

In these equations g, h and f are arbitrary, nonlinear, potentially time varying, vector-valued functions. Equation (8.1) represents the evolution of the system over time, (8.2) the computation of the rates determining that evolution, (8.3) the intermediate results necessary to compute the rates and (8.4) the initialisation of the system.

The symbols aux, const, data, levels and rates represent different types of variables:

• aux _t Auxiliary. These are computed (see (8.3)) from Levels, Constants, Data and other Auxiliaries. Auxiliary variables have no memory and their current values are independent of the values of variables at previous times.

• const Constants. These do not change with time.

• data _t Data (also called exogenous). These have values that change over time but are independent of anything that happens to other variables.

• levels _t Levels (also called accumulations, stocks and states). These change only over time and the values they take on at any time depend on the value they (and other variables) took on at previous times. Equation (8.1) shows how the Levels integrate or ‘accumulate’ based on the values themselves and other variables in the system. The Level variables ultimately determine the dynamic behaviour of a system.

• rates _t Rates (also called flows). These are the variables that directly change the Levels. Rates are essentially the same as Auxiliaries and differ only in the way they are used in a model.

Rates are implicitly determined based on Auxiliaries and other variables and are not broken out as a separate variable type. Put another way, an Auxiliary that is used to change a Level can also be thought of as a Rate.

In the following section, we underline our thesis by a case study from a traditional Chinese hospital trying to implement a strategy. After that, we discuss some general issues of system dynamics modelling for indigenous management of service innovation.

4.3 Result Model

The result model proposed here is capable of constructing a PMS, using a set of metrics generated by a control model.

Static models of service innovation gaps (SQGs) are, thus, representations of the organisation, at a given moment, which identify, define and interrelate the fundamental organisational dimensions for successful indigenous management.

By listing 20 essential dimensions—represented as ellipses—and by overlapping each ellipse with every other, Candido and Morris (2001) introduced a static model which emphasised the diversity of dimensions that can be involved in indigenous management and the intricacy of their relationships.

Essentially, the model aims to provide a list of all basic dimensions that can constitute important areas for management intervention during strategy formulation and implementation. The model, however, does not imply that managers must intervene on all 20 variables. The specific group of dimensions that a manager will choose to manipulate depends on his/her personal experience and knowledge. But, more importantly, the choice should depend on the current internal and external situation of an organisation, particularly on the SQGs that have been identified before and during implementation (Candido and Morris 2001).

4.4 Evaluation Model

Evaluation model uses simulation technology which enables experiments of ‘what if?’ scenarios to be carried out, giving the designer a better insight into how the proposed enterprise will work. This chapter focuses on the use of simulation for dynamic evaluation. Following the simulation experiments, the design can be refined further, possibly by revisiting the control model. Case studies in the next section will demonstrate the creation and population of evaluation models.

5 Illustrative Case

In order to test the concept of the QMSQS model described above, two alternative strategies have been carried out and detailed steps have been given in Sect. 8.3. In this section an application of the QMSQS to model the strategy to ‘improve quality of service via e-business’ at ‘TCM’ is presented. The case will show how the QMSQS was used to identify factors affecting performance and their relationships and quantify the effects of the factors on this indigenous management.

‘TCM’ is a leader in the non-prime automobile financing industry. For 10 years ‘TCM’ has fulfilled the auto financing needs of consumers across China. Today the company serves patients who may not qualify for loans for their new or used vehicle based on conventional criteria. With a reputation for quality control, an efficient processing system and sturdy capital to fund loans, ‘TCM’, delivers responsive service to its national network of dealers.

5.1 Establishing Strategic Perspectives

The QMSQS methodology’s use of different perspectives for organising company objectives enables a significant improvement in the development of the company’s strategic management infrastructure. We discuss five perspectives in this company: hospital, patient, process, employee and partner and innovation.

Customers’ concerns tend to fall into four categories: time, quality, performance and service and cost. Lead time measures the time required for the company to meet its patients’ needs. For existing products, lead time can be measured from the time the company receives an order to the time it actually delivers the product or service to the patient. For new products, lead time represents the time to market, or how long it takes to bring a new product from the product definition stage to the start of shipments. Quality measures the defect level of incoming products as perceived and measured by the patient. Quality could also measure on-time delivery or the accuracy of the company’s delivery forecasts. The combination of performance and service measures how the company’s products or services contribute to creating value for its patients. Senior managers at ‘TCM’ established general goals for patient performance: to get standard processes to make credit decisions quicker and to improve patients’ time to wait. The managers translated these general goals into four specific goals and identified an appropriate measure for each, as shown in Fig. 8.3.

Fig. 8.3

Organisation structure graph of TCM

Using business objects in Information Class (see Sect. 8.4.1.3), the patients’ perspective can be described as follows:

5.2 Creating Service Processes

The modelling of the company’s service process is the starting point for indigenous management. The organisational units, positions, person and locations are shown in Fig. 8.4.

Fig. 8.4

Establish patients’ perspective

From organisation class (Sect. 8.4.1.1) the person-type and organisation-type objects can be instantiated as follows:

The organisational object is built by the operator object via links between the person type and organisational units. In this context, a link can have one of the three meanings: ‘is technically superior to’, ‘is disciplinarily superior to’ or ‘is a component of’.

In this case the process model represents the underlying process which leads to achievement of the strategic objective. It was decided to use business models technique (see Sect. 8.4.1) to model the underlying process. Although it was possible to create a single-process model to encapsulate all activities which affect all the strategic performance measures, it was simpler to create a model of the process focusing on one measure at a time. This resulted in considerable duplication between the models, but it also simplified the model, making the process model more visible. Therefore, in this case study it has been necessary to create a process model for each one of the strategic performance measures identified. These duplication models can be eliminated via flow class objects. In this chapter, for illustration purposes, the strategic measure ‘patient satisfaction’ has been selected for illustration purposes. Figure 8.5 shows the process model developed for this particular strategic performance measure.

Fig. 8.5

Online services process of TCM

5.3 Establishing Cause-and-Effect Relationships

Having created a process model for the performance measure in question, in this step the activities within each process were analysed in order to establish factors which may contribute towards the particular performance measure. This was achieved through the use of cause-and-effect analysis as shown in Fig. 8.6.

Fig. 8.6

Cause-and-effect views for ‘TCM’

Using organisation class, information class, function class and flow class, the cause-and-effect relationships views for TCM can be analysed to show how each process (shown as a major cause) may have an impact on the performance measure. For example, the Online Private Customer Transactions process can affect the performance of the organisation with respect to ‘optimal e-processes’. In turn the factors listed on the upward arrow, such as broad range of products/services, high patient satisfaction, low process costs and low product costs, can constrain the process’s ability to fulfil the plan.

5.4 Linkage of IT Systems to Monitor Objectives

IT systems play an invaluable role in helping managers disaggregate the summary measures. When an unexpected signal appears on the performance measures, executives can query their information system to find the source of the trouble. If the aggregate measure for on-time delivery is poor, for example, executives with a good information system can quickly look behind the aggregate measure until they can identify late deliveries, day by day, by a particular plant, to an individual patient.

If the information system is unresponsive, however, it can be the Achilles’ heel of performance measurement. Managers at ‘TCM’ are currently limited by the absence of such an operational information system. Their greatest concern is that the service innovation information is not timely; reports are generally a week behind the company’s routine management meetings and the measures have yet to be linked to measures for managers and employees at lower levels of the organisation. The company is in the process of developing a more responsive information system to eliminate these constraints which are shown in Fig. 8.7.

Fig. 8.7

Key performance indicator allocation model

5.5 Evaluating the Performance of Alternative Strategies

In the following, we discuss two alternative strategies that can be adopted by ‘TCM’ depending on the line of loan. If the line is high, the company should adopt the ‘customised service’ strategy. If line is low, the company should adopt the ‘indirect face to patient’ strategy. These two strategies require different competitive characteristics as illustrated in Table 8.2.

Table 8.2 Competitive characteristics of two alternative service innovation strategies

The differences between the two groups lie in the elements of the characteristics and their relative positions in the group (ranking). The customised strategy primarily stresses service innovation and the ability to perform the service dependably and accurately. To have the ability to be reliable, the employees must be able to convey trust and confidence. In other words, they should have personal skills and knowledge. Referring to the banking performance of service innovation, the customised service must be excellent in reliability, responsiveness and empathy. Although it is necessary for TCM to control cost, it is not the principal characteristic for competitiveness.

The indirect-patient strategy prioritises quick credit decisions and dependable loans. To have the ability to make credit decisions rapidly, the service system must be able to shift from one type of loan to another very quickly. In other words, the service system should be flexible and automatic. The next three competitive characteristics are tangibles, consistent personal skills and responsiveness. The ability of the indirect-patient strategy to win competition is affected much more by performance in speed and tangibles than performance in reliability and cost.

Using the QMSQS approach, the hierarchical structure of the evaluation of the performance of customised and indirect-patient service strategies can be constructed as indicated in Fig. 8.8. The level 0 of the structure is the overall performance of the service innovation strategies. The performance of the service innovation strategies depends on the line of the loan (the scenario) as indicated by level 1 of Fig. 8.8. There are three possibilities of the line of loan: low (pessimistic scenario), average (normal scenario) and high (optimistic scenario). Level 2 of the structure shows performance criteria. Based on the generic performance of service innovation strategy, the performance of the alternatives can be evaluated based on the criteria of reliability, responsiveness, tangibles, assurance and empathy. Finally, level 3 of the structure shows the alternative service innovation strategies which could be adopted.

Fig. 8.8

Key performance indicator model

We conducted a presentation to explain the concept of the QMSQS model to the management team of the company. It is very critical in this step to make clear to the management the concept of the pairwise comparison questionnaires used by the model, which ask, ‘Comparing factor A to B, which one has a stronger effect on performance?’ and ‘How strong is that effect?’ At the end of the presentation the research team asked the management team to start thinking about the problem that might be selected as the case study.

Evaluation of these alternative strategies is carried out level by level starting from the top level down to the lower levels. The first evaluation assesses the possibilities of scenarios occurring in the planning period. The second evaluation assesses the relative effects of each criterion on performance under a particular scenario.

For example, what are the relative effects of reliability, responsiveness, tangibles, assurance and empathy on performance if the line is low? The relative effects of each criterion on performance are not necessarily the same under different scenarios. The third evaluation assesses the performance of each alternative on each performance criterion. Finally, the overall performance of each alternative can be computed through the composition process as explained earlier.

Using the control model, the performance of customised and indirect-patient strategies can be evaluated as indicated in Fig. 8.8. From the evaluation, it can be seen that the performance of customised patient strategy (0.514) is better than the performance of indirect-patient strategy (0.486), given that the probability of line for low, average and high are 26.3 %, 26.6 % and 44.1 %, respectively.

5.6 System Dynamics Analysis of the Service Innovation

Management uses a set of hospital and non-hospital performance measures to monitor and control the operation of companies through a set of performances. As external environments change rapidly, the set of performance measures employed by companies should also change to reflect changes in the environment. That is, performance measures reported to the management should change as a result of changes in patients, competitors, internal improvement and so on. We proposed the use of SDMs to simulate both the macro level of the entire company with interrelationships to patients—which are shown in Fig. 8.9—and the micro level of a credit department—shown in Fig. 8.10—to see the individual tasks performed.

Fig. 8.9

System dynamic analysis of macro level

Fig. 8.10

System dynamic analysis of micro level

Changes in performance measures can be in the form of deleting, adding or replacing some performance measures with other performance measures or just changing the priority of some performance measures. A performance measure classified as high priority may move to other classes because of changes in the internal or external environments of the business. The QMSQS can cope with the dynamism through the what-if simulation analysis.

For example, for the service innovation strategy evaluation explained earlier, the actual line of loan cannot be known in advance. The judgement of the probability of an occurrence of low, average and/or high loan is based on the information available at the time of evaluation. The judgement may change some time later if more information is available. Based on the current judgement, the priority of customised strategy is better than the priority of indirect strategy. However, it is important to analyse further how the priority will change if the probability of demand level changes. Again the SDMs could be used to evaluate the sensitivity analysis. The results of such analysis, based on the model presented earlier, are illustrated in Table 8.3. If the probability of low line is 100 %, the performance of the indirect strategy will be better than the performance of the customised strategy. While, if the probability of high line is 100 %, the customised strategy will perform better than the indirect strategy.

Table 8.3 Sensitivity analysis of service innovation indigenous management

Finally, if the probability of the occurrence of average line is 100 %, the performance of indirect strategy will be better than the performance of customised strategy. In general, if the probability of the occurrence of low line is greater than 43.3 %, the performance of indirect strategy will be better than the performance of customised strategy as indicated by Table 8.3. The sensitivity analysis can also be carried out on changes of the impacts of performance criteria on performance under different scenarios.

6 Discussion and Conclusion

6.1 Achievements and Benefits

An approach for quantifying the relationships between various factors affecting performance has been developed and demonstrated. The benefits of the QMSQS approach may be summarised as follows:

• Factors affecting performance can be identified, and then their effects can be quantified.

• Effects of multidimensional factors on performance can be aggregated into a single dimensionless unit (priority).

• Managers can be helped to quantify the level of impact of each factor on overall performance and therefore assisted in focusing improvement activities.

• The relationships between factors can be clearly identified and expressed in quantitative terms.

• Models can be easily altered to assist understanding the dynamic behaviour of factors affecting performance.

• A reduction in the number of performance measurement reports is facilitated.

An important benefit gained from the QMSQS approach is that the interaction of the factors can be clearly identified and expressed in quantitative terms. This identification will bring us one step forward in understanding the dynamic behaviour of factors affecting performance.

6.2 Subjectivity vs. Objectivity of Approach

People may feel that the technique used in the SDMs for quantifying the effects of factors on performance is very intuitive, subjective and very difficult to use in practice. However, through careful explanation of the concept of the approach, the authors have found that people can understand and implement it with little difficulty.

In a PMS a large number of multidimensional factors can affect performance. Integrating those multidimensional effects into a single unit can only be done through subjective, individual or group judgement. It is impossible to have objective measurements and scale systems for each different dimension of measurement that can facilitate objective value trade-off between different measures. Since the quality of service uses subjective measurement, the results may not be very accurate. However, this problem can be overcome by using group judgement rather than individual judgement. This will reduce the subjectivity of the judgement. The accuracy of the QMSQS can also be improved through experience.

6.3 Practical Issues

The example presented in the paper is highly simplified. In practice the evaluation will be more complicated as all important factors affecting the performance of service innovation strategy will need to be included in the model and the interactions among factors should be considered and agreed. Some potential problems might be encountered in applying the QMSQS method. The first relates to managers’ hesitation in filling in the pairwise comparison questionnaires, particularly if the model is applied to model performance improvement. Performance improvement usually involves identification and quantification of a large number of factors affecting performance. Consequently, the number of pairwise comparison questionnaires will be enormous. Filling in all the questionnaires will be exhausting and time-consuming. However, this problem can be minimised through three approaches:

• Firstly the users, i.e. the management team, must be involved in the whole process. In the case study example presented in this chapter, the researchers helped the management team to build a model of their PMS using the cognitive mapping technique. This in turn heightened the team’s awareness of the interaction between various factors affecting performance in their company.

• Secondly the model should be decomposed into several smaller models which are then distributed to groups of people who complete only a subset of the overall questionnaire.

• The use of the interactive software makes the implementation of the model much easier. In fact the QMSQS model is now being implemented at ‘TCM’ to prioritise 100 performance measures.

The second problem of the QMSQS application relates to getting a single judgement in pairwise comparison if more than one person is involved in filling in the questionnaires. Several discussions may be required to elaborate the real situation before a general consensus of the judgement of a particular problem can be achieved. Dynamic modelling is also an effective tool which could be used to elaborate the problem.

In summary, this chapter demonstrates the theoretical feasibility of using the QMSQS approach to implement suitable service innovation strategy through quantification of the relationships between performance measures and factors affecting quality of service.