An investigation of process optimisation for a highly specialist multidisciplinary fast track clinic : a lean Six Sigma project

(1)

Company Project |

An investigation of process

optimisation for a highly

specialist multidisciplinary fast

track clinic

-A Lean Six Sigma project-

Student|

dr. J.E. van Hooft

Supervisor |

Prof. dr. J. de Mast & Prof. dr. P. Fockens

Thesis

Amsterdam Business School |

University of Amsterdam

Contact details |

J.E. van Hooft, MD, PhD | Academic Medical Centre |

Department of gastroenterology C2-116 | P.O. BOX 22660 | 1100 DD Amsterdam | The Netherlands

(2)

You can design, create and build the most sophisticated hospital in the world. But it takes people to make the care you dreamt of a reality.

(3)

- 3 -

Preface

The reasons for this research are several but first and foremost they stem from my passion to be a medical doctor. The interaction with those who suffer from a disease, who are looking for answers, information, treatment, cure and care is a privilege where you can express your knowledge and skills but is also one of continuous learning. Learning from your patients but also from your professional colleagues. The options the Gastro-Intestinal Oncological Centre Amsterdam (GIOCA) offers with this respect are tremendous. Therefore I’m very glad to be one of the players in that team.

While being a player in the team I realised that there were also constraints, especially with regard to the care for patients with HepatoPancreaticoBiliary (HPB) malignancies. The number of patients presenting to the HPB-GIOCA outpatient clinic weekly held on Thursday has been high from the first day in 2009 onwards. The HPB-GIOCA outpatient clinic became a very intensive day with many new and repeat consultations, with often 18 or more patients to discuss during the multidisciplinary team meeting at lunchtime, which takes by now 90 instead of 60 minutes.

What if the growth would continue or what if the provided HPB-care at the VUmedical centre would have to be integrated as a consequence of the envisaged alliance? Would there be room to do more work with the same team on the designated Thursday while keeping the same high standards of care or would an extension to a second dedicated outpatient clinic day be inevitable? I was not only worried about the GIOCA day itself but also about the realisation of the treatment proposed to our patients. These concerns and questions were intruding me.

In the second place being pragmatic I realised that the above-mentioned challenges might be a good topic for my “company project” which I had to conduct to succeed my Healthcare MBA. By choosing this topic I would probably be able to plait the project into my daily work and as time was an absolute bottleneck, well recognized by every MBA-student, the combination would give a chance to succeed.

(4)

- 4 -

The third reason was the opportunity to be coached in using the Lean Six Sigma method, of which the methodological stepwise approach has been appealing, forcing me to analyse the problem in depth, thereby avoiding jumping to solutions. However expressing the benefits of the project in monetary terms, one of the pillars of Six Sigma, was not really the ultimate goal I wanted to achieve with this project. My goal was to get a deep insight in our GIOCA as an example of a fast track outpatient clinic from a management perspective, from an employer’s perspective and partially from a patient’s perspective with as final aim to find options to improve the care we deliver to our patients and find out if we could probably deliver it to even more patients.

The project started in February 2015, by May the project charter was accepted and over the summer the measurements were conducted after which the analysis was conducted and the final company project written, which has been handed in in January 2016.

Altogether it was an interesting experience with lots of learning points: deduct the problem, get rid of emotions, be involved in the measurements, analyse, find the big fish and forget the small fishes, reflect and give usable advice. Besides I once again realised what the impact of a disease and the uncertainty that comes with it do to our patients, how motivated and hard working the GIOCA professionals are and that we should keep our focus on improving quality not on cutting down expenses. But above all professionals involved in a process are key to improve the process: in the end it’s all about the people.

(5)

- 5 -

Abstract

Setting: The Gastro-Intestinal Oncologic Centre Amsterdam (GIOCA) is a fast track outpatient clinic that strives to deliver the best possible and most innovative patient centred care within a short time span. The growth has been quite significant over the last years with a steady 15%. This might however also be a threat especially for the highly utilised HPB-GIOCA. A second day dedicated to HPB-malignancies would probably be inevitable. The chairs of the different departments involved however not at all envisage this, as it would require a significant investment of resources, which might consequently hamper the work process on their own departments.

Aim: To investigate whether process optimisation by increasing efficiency would be possible.

Method: This is a study on the journey of patients with HPB-malignancies, from being referred to the start of the treatment. The Lean Six Sigma methodology has been used as investigational method. Two main objectives were appointed: decreasing the throughput time to less than 21 days for 95% of the patients and to reduce the processing time by 15%, so the anticipated growth of referrals could be facilitated without allocating additional resources.

Results: The throughput time was above the upper specification limit in 40.3% of the patients, the waiting time to surgery appeared to be the main bottleneck. Processing time was mainly determined by time spent on outpatient visits and diagnostics and not on rework as had been anticipated. Both objectives were handed back and accepted by the Champion, one because improvement was not within the span of control and the other was considered too perfidious.

Conclusion: To warrant the value proposition toward our patients the current process should first be better controlled. Possibilities for process optimisation have been identified and might very well give room to alleviate a second HPB-GIOCA day, they however appeared not to be within our reach.

(6)

- 6 -

1. Introduction

To be able to fight the hidden factory of the process that is at the focal point of this Company Project it is of importance to have a basic understanding of the transcending organisation, the Academic Medical Centre (AMC) Amsterdam and a deeper understanding of the Gastro-Intestinal Oncologic Centre Amsterdam (GIOCA).

1.1. The Academic Medical Centre

The AMC is one of the eight Academic centres in the Netherlands, it is affiliated to the University of Amsterdam which has been established in 1632. The three core activities of the AMC are patient care, research, training and education. The aim of the AMC is to conduct these tasks on top referees and top clinical level.

The AMC, though not the largest, is the most prominent medical centre in the Netherlands. (Beerda H. (2013)) (Kranenburg R. (2012)) It had an annual operating income in 2014 of 956 million with a final result of 23

Figure 1: Organogram of the AMC. Medical departments organised in division structure.

(9)

- 9 -

million. (Jaarverslag AMC 2014 (2015)) Last year 56 thousands patients were admitted. In total the AMC employs nearly 7000 professionals, including over 1000 nurses and medical specialists. The AMC has internally a structure in which the medical departments are organised into divisions. In addition a number of functions are housed in services, directorates and support groups, the organogram is depicted in figure 1.

1.2. Gastro-Intestinal Oncologic Centre Amsterdam

The aim of the GIOCA is to be the best centre for Gastro-Intestinal oncology in the Netherlands with the ambition to become an international key player. Patients should receive the best possible and most innovative care, and should be given the opportunity to participate in internationally recognised, renewing research. Furthermore sharing of knowledge is considered of paramount importance.

The chiefs of the department of surgery and gastroenterology started composing GIOCA in 2008. The main reason was the long waiting time for patients with gastro-intestinal malignancy that was in part caused by the division structure of hospitals. Most patients with gastro-intestinal malignancies need multiple appointments with different medical specialists from different disciplines to ensure a correct diagnosis and stage of disease. Besides those outpatient clinic visits the specialist would also order different diagnostic tests, which were performed sequentially, implying further accumulation of waiting times. Altogether waiting times, during the diagnostic phase, of up to four weeks were more rule than exception.

To reduce those waiting times it was considered to follow Porter’s suggestion for an integrated practice unit (IPU) (Porter M.E. (2008)) outside of the AMC but as one of the key factors for success is however a high volume, which is not realistic for gastro-intestinal malignancy within the Netherlands, it was decided to create a new organisational structure as described by Basta et al. (Basta Y.L. (2015))

1.2.1. Organisational structure

The GIOCA as fast track clinic is neither a division nor a service, directorate or support group. It is a co-operation between medical specialists of different

(10)

- 10 -

departments, oncology nurses, ancillary departments, para-medical staff, and other staff, coordinated by a board that consists of a gastroenterologist, a surgeon and a chief nursing officer. The co-operation GIOCA acts as one team towards the patient, the primary care physician, and the referring physician.

1.2.2. Financial structure

Within the university hospital, the authority to open a refundable healthcare product, a so-called DTC, for a patient is assigned to a limited number of medical departments. As depicted in 1.2.1 the GIOCA itself is not an accountable entity in the internal governance system and is therefore not authorised to open DTCs. When a medical specialist or nurse practitioner working at the GIOCA opens a DTC for their patient, they do this on behalf of their own department. Consequently their department will be reimbursed, not the GIOCA. In line the department is also responsible for the reimbursement of the professionals they set to work at the GIOCA.

1.2.3. Patient centred process

Since the opening in 2009 all patients with a gastro-intestinal malignancy are diverted to the GIOCA. The care is centred round patients by allocating specific days to the different gastro-intestinal malignancies. For example patients with colorectal carcinoma are seen on Tuesday while patients with oesophageal or gastric tumours are seen on Wednesday and patients with hepatopancreaticobiliary malignancies (pancreas, gallbladder, bile duct, duodenum) visit the outpatient clinic on Thursday. A different team of dedicated gastro-intestinal oncology physicians and specialised oncology nurses manages each tumour-specific day. Patients are designated to one of the tumour-specific days based on the evaluation of the referral letter and available imaging. This process is internally called “triage” whereby the radiologist subsequently verifies that the imaging is recent and of adequate quality while the specialised nurse follows pre-defined healthcare pathways to assess if additional tests need to be performed. Hereupon the administrative team schedules the additional tests and invites patients to a tumour-specific outpatient clinic day.

(11)

- 11 -

Those days at the GIOCA consist of a morning consultation with a surgeon or gastroenterologist and additional testing if indicated, with the exception of tests that need a sedative (i.e. endoscopic retrograde cholangiopancreatography (ERCP) or endoscopic ultrasound).

Between 12:00 pm and approximately 13:30 pm all patients are discussed at the multidisciplinary team meeting consisting of at least one staff member of the departments of surgery, gastroenterology, radiology, medical oncology, radiation-oncology and pathology. During the meeting individual treatment plans are formulated for each patient and documented in a shared electronic medical record. After the meeting, the physician from the morning consultation will see the patient to discuss the (final) diagnosis, eventually required additional diagnostic and/or proposed treatment plan. Thereafter, all specialists with involvement in the treatment discuss separately their part with the patient during this specific day unless the patient indicates to rather come back another day. Following the GIOCA outpatient clinic day the patient is either referred back with a treatment proposal, has to undergo additional investigations, or is scheduled for the treatment proposed. To check if the information was adequate and whether there are further questions an oncology nurse contacts the patients some days after their GIOCA outpatient clinic visit.

Figure 2: GIOCA day in summary as provided to referring physicians and patients, pag 9. GIOCA brochure (2012).

(12)

- 12 -

2. Framework and concepts

Considering the steep rise of referrals the unique proposition towards patients and referring physicians had been clearly appreciated. However this rapid, not completely anticipated, growth could have implied some threats towards the efficiency of the primary patient processes that are the mainstay of GIOCA. As this might hamper the further growth an investigation whether and with which analytic tool this could be best investigated was conducted.

2.1. Background

When considering the traditional notion of Porter’s generic strategies (Porter M.E. (1980)) the strategy for competitive advantage chosen by the GIOCA is one of product differentiation with the scope on focus: focus on a specific group of malignancies further subdivided in specific diseases for which patients can be referred to a highly scaled dedicated team that is put together a particular day per week to deliver high quality care, in conjunction with a significant reduction of waiting time. The success of this strategy is reflected in the tremendous growth: in five years the number of patients visiting the outpatient clinic rose by 170%, a steady 14% each year. (GIOCA brochure (2015)) Though no severe problem yet, a threat was anticipated for the HPB-GIOCA, the specific outpatient clinic for pancreas, gallbladder, bile duct, and duodenum malignancy which received 45% of the total number of referrals. The team considered this tumour-specific day rather strenuously due to the average of 11 new patients per day (GIOCA brochure (2015)), which is double the amount of the second busiest day. One was wondering what the implications would be if the number of referrals would continue to grow, or what if the envisaged alliance with the VUmedical centre would repose HPB-GIOCA resources. Would it be possible to

GIOCA

Figure 3: Porter's Generic Strategies. The position of GIOCA depicted within this strategy.

(13)

- 13 -

optimise the current process or would a second day dedicated to HPB-malignancies be inevitable? The latter was not at all envisaged by the chairs of the different departments involved: secondment of their highly scaled medical specialists for 2 days a week might threaten the continuation of their departments’ primary process. In addition the revenue generated might be too sparse, as the occupation rate of the extra day would, especially in the beginning, not exceed 50%.

To investigate whether or not the process of the HPB-GIOCA would be susceptible for optimisation was considered. It could well be that the process had evolved over time and had gained complexity, extended waiting times, included more obsolete and/or redundant work.

2.2. Models, methods and systems

Over the last century several models, methods and systems have been developed to improve processes. They all have their roots in scientific management (Wren D.A. (2005)) but with a different focus: some more on industry, some more on services and some specific for healthcare. Examples are total quality management, business

process reengineering (Hammer M. (1990)), business process management (Van der Aalst W.M. (2004)), theory of constraints (Goldratt E.M. (1984)) and Lean Six Sigma (De Mast J. (2006)). One of the pitfalls is that there is often no widespread agreement as to what the models or systems are and what actions it requires of

A B

Figure 4: Similarities between cycles used in different methods. A. Business process reengineering. B. Business process management.

(14)

- 14 -

organisations. As a consequence no precise project definitions are formulated which is a well-known factor for failure of projects. (Morris P.W. (1987)) (Partington D. (1996)) On the other hand there are a lot of similarities between the different methods, not only with regard to type of cycles used, figure 4, but also the search for

the bottleneck, focus on throughput time and quality improvement.

Besides choosing a model to facilitate process improvement it is also of importance to pay attention to the project management. The traditional approach distinguishes between five developmental components; initiation, planning and design, execution and construction, monitoring and controlling and as final completion. (Wysocki R.K. (2013)). These stages are often applied in the rather straightforward construction industry. Another concept is agile project management, which encompasses several iterative approaches based on the principles of human interaction management and founded on a process view of human collaboration, a method rather often used in creative industries. (Moran A. (2015)) Probably at the moment the most applied project management method is PRINCE2 (Project IN Controlled Environment). This method finds its origin in information technology projects but it has been further developed under the guidance of the British Office of Government Commerce and is

by now generically applicable. (Office of Government Commerce (2009)) The method

is process-focused and based on 7 principles, 7 processes and 7 themes, though if not relevant they can be adjusted which makes it a flexible method.

Many facets of PRINCE are incorporated in LSS, a combination of Lean Manufacturing and Six Sigma. These two methods strengthen each other to a platform for cultural and operational changes, leading to a total supply chain transformation. (Pepper M.P. (2009)) This emphasises the relation with operations management: a mature field of study that has constructed an impressive body of knowledge. (Does R.J. (2015)) LSS further provides the most detailed method regarding standard analysis tools and techniques with the intention of mapping out and removing inefficiencies in conjunction with a structured roadmap to reduce variation in organisational processes by analysing the performance metrics and uncovering root causes with the use of statistical tools. (De Mast J. (2012))

(15)

- 15 -

there are reports on the use of the theory of constraints and business process management. (Wright J. (2006)) (Vera A. (2007))

The expertise in healthcare, including several directly applicable publications (Schoonhoven M. (2011)) (Niemeijer G.C. (2011)) (Kemper B.P. (2013)), combined with the strictly described methodology, the incorporation of project management and not to forget the values that LSS embodies like continuous innovation and improvement, focus on the customer, data-driven decisions which are so well aligned with the mission of GIOCA, made it an attractive methodology to endeavour.

(16)

- 16 -

3. Case description

Now the organisational context has been depicted, the emergence of GIOCA discribed, a rough draft of the daily process given and the choice for the methodology underpinned the case description could be fully appreciated. The DMAIC roadmap (Define, Measure, Analyse, Improve and Control) used in LSS was applied.

3.1. Define

This paragraph focuses on the define phase: determining the process to be improved, the project objectives including definitions, the potential benefits, the project details and the project organisation.

3.1.1. The process to be improved

The process to be improved concerns the process of determining the treatment of patients refered to the HPB-GIOCA. This specific outpatient clinic has been chosen as the professionals had indicated that the workload was rather high and the room to house any further growth absolutely limited. Detailed deliniation of the setting in which the project was carried out are depicted in the SIPOC model showing suppliers, input, process, output and customers (figure 5).

Patients were either referred by the general practitioner or by a medical specialist. When a referral letter concerning a new patient was received at the administration of the GIOCA data gathering started. As soon as adequate data were compiled a patient would be invited for consultation at the GIOCA outpatient clinic after which he was discussed in a multidisciplinary meeting scheduled on the same day. The

Suppliers Inputs Process Outputs Customers referring physician patients determining treatment patient under treatment patient a. medical specialist for patient with

b. general practitioners malignancies

Project objectives

To reduce throughput time till treatment to < 21 days in 95 % of the patients

To reduce processing time with 15%

informa(on gathering consulta(on pa(ent discuss pa(ent mul(disciplinary addi(onal diagnos(cs start treatment

Figure 5: The macro and meso process description of the HPB-GIOCA using a SIPOC- chart.

(17)

- 17 -

outcome of this meeting would be directly shared with the patient so additional diagnostic testing or the start of the treatment could be arranged. The current project did not cover improvements of processes beyond the start of the treatment or processes involving non HPB-malignancies.

3.1.2. The project objectives

The project objectives were set after investigating available data at the outset of this project, bearing in mind the fundamental idea of decreasing waiting time as the main reason to initiate GIOCA. An internal document concerning the year 2014 revealed a total number of referrals of 1138 new patients of which 565 specificly to the HPB-GIOCA. The admission time (from minimal required data complete till first outpatient clinic visit) was 4.2 workdays and the throughput time (from first outpatient clinic visit till start treatment) was 18.5 workdays for the HPB-GIOCA. The service level agreement of 15 workdays for throughput time was achieved in 48% of the patients. Between 2009 and 2014 the number of new HPB-GIOCA patients had grown by 133%. (De Bres E.F. (2015))

Decreasing the throughput time was therefore the first project objective; before this could be embarked it was of paramount importance to establish a sound definition for throughput time. Though the definition given in the report by de Bres might seem quite straightforward there are some limitations. The delineation between admission time and time from first outpatient clinic visit till start treatment is rather artificial, from a patients’ perspective the total time passing till start treatment is regarded as waiting time. The second limitation is the wording used in the GIOCA booklet (GIOCA brochure (2012)) and on the website (AMCweb (2015)) where it is just stated that the treatment starts (most often) within three weeks. It was decided to define the throughput from the instant the referral letter has been registered by the administrative staff, being the moment from where on the GIOCA could take responsibility, till the start of the treatment. To stay in line with the GIOCA documentation the upper specification limit (USL) was set on 21 days (not working days). As this time frame would be of great importance to all patients it was decided to put the bar to achieve this as high as 95%.

(18)

- 18 -

The second objective was to reduce the processing time by 15%, so the anticipated growth of referrals, a steady 14% per year, could be facilitated without allocating additional resources.

3.1.3. Potential benefits

Potential benefits were anticipated to stem from the awareness of involved professionals of inefficiencies uncovered by the focussing on processing time and productive time; tackling these inefficiencies would create opportunities to accommodate the steady growth and it could probably obviate the necessity to initiate a second HPB-GIOCA day, at least for the near future.

Efficiency improvement was also considered to contribute to the strategy by maintaining product differentiation and focus leadership, which enables a strong proposition towards: patients, insurances, (non) profit sponsors and high potential employees. For the patients the gains would be the preserved access to a team of excellence facing a highly efficient track determining the treatment and minimal time till start of this treatment (patient’s satisfaction).

Regarding hard benefits one could either look at the savings generated by reducing processing time or the extra earning that could be generated as more patients could be seen due to the increased efficiency or the saving related to refraining from a second HPB-GIOCA outpatient clinic day. Achieving this last point would create a saving of over 100.000 euro per year based on the minimal staffing for an outpatient clinic day (one surgeon, one gastroenterologist, one radiologist, two dedicated oncology nurses).

3.1.4. Project details

The project details consisted of the description of the type of project. While filling out the project charter it was apparent that neither of the options covered the aim of the project precisely. The Champion requested more than just a DMA: design measure and analyse. Besides the diagnosis he requested suggestions for improvement and control, he was however aware of the manpower and time limitations; therefore no true implementation was envisaged.

(19)

- 19 - 3.1.5. The project organisation

The project organisation in charge of achieving the objectives consisted of the head of the department of gastroenterology (Supplier, Champion, User), a Master Black Belt from the University of Amsterdam and a project leader called Black Belt who was supported by five team members: a master of science trainee, a dedicated oncology nurse, an administrative employee, the headnurse of the GIOCA and a financial manager. The main task of the team was to collect and provide information about the proces (mainly the trainee), to brainstorm influence factors related to the objectives, and to generate ideas for improvements (all others).

3.2. Measure

The project objectives were operationalised into the following measurable characteristics: throughput time; waiting time; processing time; preparation time main process; rework/mistakes; processing time main process; available time; productive time. A schematic overview of the operational definitions and measurement plan are reflected in appendix I. A more thorough explanation for the choices made and details on the measurement plan including the method used for validation is elaborated on in the following sections.

3.2.1. Define the measurable characteristics

To be able to define the measurable characteristics the CTQ (Critical To Quality) flowdown as a conceptual model of project objective has been explored where upon in the CTQs were derived. (De Koning H. (2007)) To guide the selection of the CTQs the article by Niemeijer et al., which describes nine generic project definition templates, has been crucial. (Niemeijer G.C. (2011)) Though template 9 from Niemeijer et al. did focus on capacity issues it included many items that could be transposed to our current project with efficiency issues as the root cause and was therefore used as guide.

First the project objectives as described in the define phase were recalled and explored in depth to better articulate the rationale: the “why” and “what” of the project. (De Koning H. (2007)) It was apparent that the main reason to initiate this project was the question whether the current resources were sufficient to accommodate the anticipated growth. As the process is labour intensive the focus regarding resources

(20)

- 20 -

is on the employees. These employees experienced a high workload and questioned the possibility of further growth with the same resources. The Champion however wondered if there was room for efficiency improvement of the resources so expansion could be alleviated. The reason to investigate this was two-sided; first an additional HPB-GIOCA day would imply serious staffing constrains on the department of the Champion as additional secondment of highly specialised medical staff would be needed. Second was a financial concern as the occupation on a second day might be too sparse to generate enough revenue to cover the salary costs of the seconded employees.

The next step focuses on the influence of the project objective on the canonical layer that consists of the strategic focal points. Strategic focal points guide and focus action at the level of a business and characterise its strategy. The strategy of the GIOCA is first of all patient satisfaction, which is heavily influenced by the pace with which treatment is started. So from this perspective efficiency improvement focuses on reducing time till treatment, after all the main reason to initiate GIOCA.

Profitability is however another part of the strategic focal point considered being of paramount importance as salary costs that are not covered by revenues should be avoided.

Though the action planning, translation of the strategic focal points to project objectives, may seem a bit after-the-fact as the project objectives have already been determined in the project selection process, it clarifies the aggregate project objective and decomposes it into the one-dimensional variables: throughput time, processing time and available. Now resource efficiency has been divided in one-dimensional variables, the additive constituents can be viewed:

• Throughput time; waiting time

• Processing time; preparation time main process; rework/mistakes; processing time main process

• Available time; productive time

The role of throughput time and processing time for this process improvement project might be self-evident. Available time and productive time might need a bit more

(21)

- 21 -

clarification. As stated earlier human resources are the backbone of the labour intensive primary patient process of GIOCA. Therefore their disposal is key for throughput time as well as for the processing time as they execute the processes. The final CTQ flowdown for this project is depicted in figure 6.

3.2.2. Operational definitions and measurement plan

Throughput time and waiting time: The characteristics regarding throughput and waiting time were measured per patient. To obtain sufficiently precise estimates the sample size was set to the minimum of 30. Historical data were chosen because the patient needed to have completed the whole diagnostic process and have started the proposed treatment. As this phase might take up to 90 days, a cohort of at least three months ago had to be selected; in addition it should cover a certain time span so

(22)

- 22 -

eventual fluctuation over time could be determined. To fulfill these criteria the Black Belt retrieved historical data of 30 patients from the electronical healthcare system covering the period from January 8th until April 16th 2015.

The sampling was done systematically by selecting the first new referrals to the HPB-GIOCA outpatient clinic to respectively the surgeon or the gastroenterologist, in total two patients per week were selected. After selection the Back Belt meticulously analysed the electronic medical records of these patients and entered required data in a measurement plan.

This measurement plan contained the questions related to the CTQs in columns and the unit of study (patients) in the rows. The plan was designed in excel file to facilitate the calculations of the waiting time between the different process steps (see value steam map in figure 7) and total throughput time. For the throughput time the goal was below 21 days and for waiting times as low as reasonably possible.

Processing time: Three subdivisions had been defined under processing time: preparation time main process (sum of time to gather minimal required data, conduct triage and prepare for the GIOCA day), rework/mistakes, processing time main process (sum of time spent on the GIOCA day and additional diagnostics). Processing time was defined as the sum of the time spent on these three subdivisions. All items regarding processing time were measured per patient.

As for precise measurements patients would have to be followed on a day-to-day basis over the course of the complete diagnostic process till the start of the treatment. Taking time constraints and available manpower into consideration it was considered unrealistic to follow 30 patients required to apply the limit theorem. As the accuracy was key, better precise information of a part of the population than vague information of the whole population, the sample was therefore brought down to 10 patients being the maximum the constraints would permit.

The processes were mapped as follows: at the moment a referral letter regarding a patient with an HPB-malignancy was received the administrative staff would start to

(23)

- 23 -

fill out a patient specific measurement form and inform the Black Belt about a potential accrual. The Black Belt would ask the patient by telephone whether he or she would be willing to participate in the study. After informed consent pop-ups were installed in the patient’s electronic medical record to inform all professionals to keep track of their time “invested” in this patient, and where to specifically note this down. A paper version of this patient tracking system was also available. The trainee would follow the patient in the electronical healthcare system to remind the involved professionals to keep track of the time spent with the patient. In addition the patient who entered the study received a personal tracking form where they could note down all moments of contact (appendix II), a short manual stated the definitions and some examples were added. Patients were reminded by the trainee the day before a scheduled appointment at the AMC or at least once weekly.

The different types of travel sheets (electronical, patient’s tracking form and manually filled out forms of others involved) were collected by the trainee and converted in a patient specific travel sheet in excel where the data were ordered in preparation time, processing time main process, mistakes/rework and total processing time. The Black Belt kept, at least weekly, an eye on the logic of the data entered. Patients were included from 22th till 31st of July 2015 and followed till treatment was started.

Available time and productive time: For the measurement of available time it was a prerequisite that the time was specifically earmarked to the HPB-GIOCA. As this requirement could not be met for the administrative staff they were expelled from this measurement. Regarding the other professionals involved, their dedicated availability for HPB-GIOCA had to be visible in their work-timetable. This appeared not to be valid for the supporting medical staff like pathologist and nuclear specialist, so also they were not taken into account regarding availability. The available time was measured per week.

From the key professionals the Black Belt obtained the weekly work-timetables either from the manager or from the person responsible for drafting the tables. Historical data were selected as they could be verified in the patients’ electronic medical records, as names of those filling out these records are traceable. For logistic reasons the same period as for throughput, January 8th_{until April 16}th_{2015, had}

(24)

- 24 -

been chosen. The data from the work-timetables were transformed into a measurement plan, availability per key profession in columns and weeks in rows. The goal for the availability was as high as possible.

With regard to productive times the aim was to follow the key professionals for two weeks performing their tasks deriving from their engagement in the HPB-GIOCA, basically concentrated on Thursdays. The scope for the administrative staff and the oncology nurses was on GIOCA in general as their productivity can’t be so clearly delineated between the different types of gastro-intestinal malignancies. Although one could argue that for the oncology nurses a focus specifically on the HPB-GIOCA day could be chosen this was considered to give too narrow an impression of the different tasks they conduct. The following tasks were considered to be GIOCA-related productivity: administration, consultation, multidisciplinary meeting (MDM), outpatient clinic and reviewing. Putting breaks, interruption, none GIOCA-tasks and waiting time aside as non-GIOCA-related productivity. A GIOCA-related productivity of >85% seems realistic considering one hour (11.1%) of breaks per day and leaving some space for other tasks stemming from the work conducted during the rest of the week.

Shadowing and self-registration were applied to gather data on the different tasks executed and the productivy. Standardised measurement forms were designed and a manual with definitions compiled (see appendix III). The measurement procedures were done per full hour but preferably per daypart or full day. Tasks were defined and differentiated in accordance with the type of profession. The different tasks were registered in detail but thereafter bundled to GIOCA-related productivity or to waiting time, interruptions, non-GIOCA-related tasks (e.g. tasks for other outpatient clinics) and breaks. To obtain the best representation at least three measures were performed: one by the professional him-/herself, one by the trainee and one by the Black Belt. In addition for every measurement a different person from the key professions was approached.

(25)

- 25 - 3.2.3. Validation of measurement procedures

The trainee and the Black Belt tested the designed measurement plan for the throughput time by means of independently completing four patients in total. The patients selected were the second new referrals to respectively the surgeon or the gastroenterologist in week four and eigth of 2015. The results were compared, differences discussed and definitions further specified if needed.

Before the measurements of processing time were conducted, the team members of this project debated the validity of the measurements. Hereupon a member of the administrative staff, one oncology nurse, a gastroenterology fellow and one patient performed a test run. Based on the results of this test definitions as stated in the manual were clarified if needed, adjustments regarding job specific processes were made and simplification applied if possible. Henceforth the tracking form and the manual were sent to all involved professionals to give them a voice and to inform them about the upcoming measurements.

Work-timetables were validated by interviews with representatives from the key professionals (oncology fellow, radiation oncologist, radiologist, surgeon, dedicated oncology nurse, gastroenterology fellow) and correctness was checked in patients’ electronic medical records. The interviews with the key professionals were also used to clarify the travel sheets used to measure productive time, leading to more specific task definitions. Hereupon a test run was done during which a fellow filled out what kind of tasks he was doing, while the Black Belt observing this fellow did the same: marginal discrepancies were observed.

3.3. Analyse

Following the measurements the data were analysed and the values of the CTQs determined. These results were carefully studied to diagnose any problems in the process, to categorise them according to importance and to consider eventual influence factors. To reflect the results of the throughput times and processing times, including waste, in a structured manner a so-called value stream map was used. (Womack J.P (2003)) (Kemper B.P. (2010)) All calculations and statistics were performed using Minitab17 (Minitab Inc., State College, PA, USA).

(26)

- 26 - 3.3.1. Diagnosing the current process

Throughput time and waiting time: With regard to the throughput time the first analysis conducted regarded an investigation on behaviour of the process over time to see whether the process was constant or contained peaks, cyclical or other patterns. As the measurements were done individually the I-MR (Individual Moving Range) chart systematics were chosen. The MR-chart revealed that the process variation was in control. The I-chart showed that one out of the 30 measures disclosed a variation pattern that was too unlikely to be considered random, as it was more than three standard deviations from the centre line, see appendix IV. This isolated disturbance indicated that an irregularity had occurred: a so-called assignable cause warranting further investigation in the next phase of the analysis. Following the analysis on process behaviour a histogram and basic statistics for the total throughput time were executed (appendix IV). This showed a throughput time of 22 days on mean (IQR 7-34 days).

Because the histogram was skewed, probability distribution functions were tested: the Weibull distribution had the lowest Anderson-Darling value and was therefore selected. The Emperical Cumulative Distribution Function was applied hereupon to

Figure 7: Value stream map including details on process flow, waiting times (in days) and processing time (in minutes). No 8/30 = 27% Yes 19/30 = 63% Yes No 11/30 = 37% 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I Triage OK? minimum required data complet Queue nr: 30 QT: 3.90 D 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 27%

referral Queue Queue

I I Gioca-day addition needed ? Queue start treatment addiBon diagnosBcs Queue 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I I I I I Queue 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I nr: 30 QT: 0.73 D nr: 30 _{QT: 1.50 D} nr: 11 QT: 13.18 D nr: 19 QT: 6.27 D nr: 19 QT: 11.47 D preparaBon Gioca-day triage cycle time:

20.40 min cycle time: 13.90 min cycle time: 18.40 min cycle time: _{192.7 min}

cycle time: 289.0 min Part 1 No 8/30 = 27% Yes 19/30 = 63% Yes No 11/30 = 37% 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I Triage OK? minimum required data complet Queue nr: 30 QT: 3.90 D 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 27%

referral Queue Queue

I I Gioca-day addition needed ? Queue start treatment addiBon diagnosBcs Queue 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I I I I I Queue 765 12 11 10 8 4 2 1 9 3 765 12 11 10 8 4 2 1 9 3 I I nr: 30 QT: 0.73 D nr: 30 _{QT: 1.50 D} nr: 11 QT: 13.18 D nr: 19 QT: 6.27 D nr: 19 QT: 11.47 D preparaBon Gioca-day triage cycle time:

20.40 min cycle time: 13.90 min cycle time: 18.40 min cycle time: _{192.7 min}

cycle time: 289.0 min Part 2

(27)

- 27 -

determine which percentage of the patients fulfilled the criteria of the predefined throughput time of 21 days. This requirement appeared to be met in 59.7% of the patients.

Further exploration of the historic data of the 30 included patients disclosed data on the waiting times between all different process steps, in the value stream map depicted as queue. In addition it revealed that in 8 patients (27%) the triage was not done at once. Additional investigations were needed in 19 patients (63.3%). These data were integrated in the value stream map depicted in figure 7.

Processing time: For all processing steps the mean time values of the ten included patients were calculated and included in the value stream map (figure 7). After which they were bundled according to the predefined CTQs. The preparation time of the main process consisted of time to gather minimal required data, time spent on triage and time to prepare for the GIOCA day. Total processing time of the main process consisted of the sum of the time spent on the GIOCA day and additional diagnostics. Statistics revealed a mean total processing time of 447.8 minutes per patients, 52.7 minutes (11.8%) needed for preparation of the main process and 395 minutes (88.2%) spent on the main process.

Rework was explicitly mentioned in only two occasions, making up a total of five minutes (1.1%). More details on the processing time statistics have been included in appendix V.

Available time and productive time: The collected and verified work-timetables showed a very stable availability of professionals. On every Thursday, during the 15 weeks investigated period, a surgeon, a radiologist, a fellow gastroenterology and a nurse were specifically allocated to the HPB-GIOCA for the full GIOCA workday. A gastroenterologist and radiation oncologist had scheduled availability during the full afternoon. Control charts were not executed for those professionals with a completely even weekly availability, as this would just show a straight-line. The medical oncologists however had dedicated GIOCA slots in their afternoon outpatient clinic, warranting availability between two and four hours. The I-chart did not reveal a specific pattern, peaks or cyclical.

(28)

- 28 -

The productive time was investigated for the key professionals during a minimum of 564 minutes (administrative staff) and 1153 minutes till 1409 minutes for the other professionals (surgeon, radiologist, gastroenterology fellow, nurse). Due to time constraints and as a consequence of the current nature of their function, more advisory, not enough data could be gathered on the GIOCA-related productive time of respectively the medical oncologist, gastroenterologist and radiation oncologist.

GIOCA-related productivity ranged from 61.4% till 90.7% per profession, see figure 8, the administrative staff and the surgeons did not meet the requirement of 85%. Another remarkable observation of measuring the productive time was the fact that the allocated time for GIOCA was regularly exceeded; in addition breaks were hardly taken and certainly not at regular times. This should be considered when weighing the measured non-GIOCA-related productive hours. More details on all different tasks per profession can be found in appendix VI.

Radiologist Gastroenterology fellow Nurse Surgeon Administrative staff 90 80 70 60 50 40 30 20 10 0 Professionals Pe rc en ta ge

GIOCA-related productivity per group of professionals

(29)

- 29 - 3.3.2. Recalculating the business case

Having generated a data-based diagnosis by now the expected benefits of the project at hand could be shifted from being based on a rough assessment to a better underpinned one. The aim of the project, improving the efficiency of the HPB-GIOCA, appeared to have remained valid. The primary project objective, decreasing throughput time to less than 21 days for 95% of patients, seemed to regard a crucial item as it was only achieved in less than 60% of the patients.

Investigating processing time it became apparent that the factor rework was not a key factor, in contrary to the expectation. This implies a significant challenge for the predefined reduction of the processing time by 15%, as this reduction of 67 minutes in total has now to be attained fully from preparation and processing time of the main process. Hereby the preparation time is rather futile, so the burden is shifted to the processing time on the GIOCA day and generated by additional diagnostics. This will require a very critical appraisal as decreasing these times might be perfidious regarding best possible care. So decreasing processing time is not the expected low hanging fruit facilitating the anticipated growth.

The GIOCA-related availability and productivity however revealed to be a more promising, availability was high but not always specifically enforced to the HPB-GIOCA and the productivity indicated involvement in other outpatient clinics.

So to summarise, thought from another angle than anticipated, there seems to be enough potential to aim at refraining from a second dedicated HPB-GIOCA outpatient clinic day, thereby fulfilling the requirements regarding the hard benefits.

3.3.3. Identifying potential influence factors

Before an effort to solve the problem could be made it is of paramount importance to try to understand why CTQs behaved the way they did. Therefore factors that might have influenced the CTQs, the so-called influence factors, had to be discovered and investigated. To be able to identify these factors, an exploration in many directions, by a creative, speculative and divergent process, had to be conducted. The Black Belt together with the members of the project team applied different approaches.

(30)

- 30 -

A brainstorming was organised with colleagues from the VUmc to gather knowledge on their specific gastrointestinal malignancy outpatient clinic. In addition a meeting with a cancer institute, in this particular case the Antoni van Leeuwenhoek Hospital, was initiated to find out how they organise the care around patients. In both sessions a patient with an HPB-malignancy was fictively sent to the different outpatient clinics and followed during the process from referral to the start of the treatment. This meeting revealed many different insights and logistic suggestions. To further deepen the insight in each other’s outpatient clinics an exchange was undertaken; hereby the trainee participated in the different outpatient clinics.

Besides a GIOCA specific brainstorm was attended aiming to avoid funnel view, investigating whether the other tumour-specific groups encountered the same challenges and if exploiting their technical knowledge and implicit know-how would be of help regarding solutions.

The Black Belt conducted exploratory data analyses searching for patterns and salient features and for possible explanations. Furthermore autopsies were performed by close examination of outliers, the specific processes related to these excesses were carefully reconstructed to gain insight in the nature of the problems. The outliers were also compared to best of the best examples in attempt to identify different patterns.

The trainee and the Black Belt conducted several so-called Gemba studies: walking around and observing day-to-day routines of administrative staff, several medical specialists and oncology nurses. For proper observation they first envisioned how the process was supposed to be happening. Thereafter observations were done on a structured way. The above-described steps generated many suggestions, which were registered in a logbook; selections of potential ones are reflected on in the table on the next pages.

(31)

- 31 -

Short description Example Improvement options Estimated effect

Potential influence factors throughput and waiting time Incomplete referral at first instance Patient is referred but adequate information to register the patient is lacking as well as imaging. So the referring hospital has to be contacted to retrieve the information required and warrant imaging to be sent. This implies that the patient’s record is put aside and the administrative staff has to keep an eye on receiving the imaging. Making an easy to use digital portal facilitating referral but only after all requirements have been filled out, including uploading images. Clearly state why it is important for the patient that al data are provided. 2-5% Identify specific

situations Like patients that have to undergo additional investigations: do they have a longer throughput time? If this is true create disease-specific time slots so additional investigation can be done abutting avoiding sequential waiting time. <2% Identify disease

characteristics Patients with specific malignancy might require additional pre-treatment according to the guidelines and care path. Reserve treatment slots for these patients on the moment they are referred, not just at the moment they are being seen. <2% Type of proposed treatment Probably patient that do have to undergo surgery do have a significant longer waiting time. Optimise capacity for HPB-specific operations 10-15% Location of treatment (whether or not in the AMC) Waiting time to start chemotherapy might differ between affiliations. Make waiting times transparent to patients and providers. <2% Potential influence factors processing time Duration of sequential out patient visits If the treatment proposal is chemo radiation followed by surgical resection the patient sequentially visits the gastroenterologist he saw in the morning to be informed about this proposal and thereafter he visits sequentially the surgeon, the radiation oncologist and the medical oncologist. After being informed about the treatment proposal the patient visits a team consisting of a radiation oncologist, medical oncologist and surgeon at once. 5% Repeated

diagnostics Though elsewhere a malignancy is pathologically proven the specimen is revised in the AMC.

Standardise when and for what reason investigations are repeated.

(32)

- 32 -

Short description Example Improvement options Estimated effect

Potential influence factors processing time -continuation- Additional diagnostics As additional diagnostics a PET-CT-scan is made while in retrospect this appeared not to be indicated according to the guidelines. Make standardised care paths that indicate which diagnostics are needed for a specific disease. 5% Delineation of

responsibility On imaging multiple liver metastases are seen, to confirm this a puncture is scheduled followed by an outpatient clinic visit where after the patient is referred back to receive chemotherapy in his regional hospital. Make service level agreement with referring hospitals including which care will be provided where and who can be contacted in case a deviation is preferred. 2-5% Potential influence factors available and productive time Time spent on non-GIOCA-productivity The HPB-GIOCA outpatient clinic population of the surgeon contains a mix of patients i.e. controls after surgery and new patients. To warrant enough capacity for the growing number of new patients, controls will have to be deviated to another outpatient clinic. 5% Change of task Triage is done by an oncology nurse and radiologist, whereby the nurse visits the radiologist and reviews the imaging together. Uploaded images are reviewed and reported directly by the radiologist. 2-5% Optimising utilisation of time slots Time slots available for HPB-GIOCA patients at the medical oncologist are not fully utilised, often because patients are referred only by the end of the afternoon when patients rather like to go home. Improve planning regarding outpatient clinic visits to medical oncologist. <2% Administrative burden Fellows gastroenterology spend over 35% of their time dedicated for HPB-GIOCA on administrative tasks. Link electronic medical records so basic patient data like previous history, allergy, medication etc. are automatically uploaded. 5-10%

(33)

- 33 -

3.4. Improve

As stated by Slack “performance improvement is the ultimate objective of process management” whereby the aim is to close the gap between the current and the desired situation. (Slack N. (2012)) To do so the current performance has been assessed and compared to the target. The next step is to interrogate which factors drive these gaps and whether there are possibilities for improvement. The methods now coming into the picture should be methodical, objective and rigorous, hereby differentiating between the most valuable ideas -the vital few- and the less valuable ones -the trivial many-.

3.4.1. Establishing the effect of influence factors

The historical data gathered to determine throughput and waiting time were considered to fulfill the quality criteria and contained sufficient details to conduct analysis of variance. Before doing this analysis an estimation was done about the more and less potent factors. Examination of the value stream map for example indicated that a factor like “Incomplete referral at first instance” would most likely belong to the less potent factors as the total time to completion of the minimal required data was relatively short. Though the value stream map did indicate that waiting times were longest between establishment of the diagnosis -either the at GIOCA day or after additional diagnostics- and the start of the proposed treatment it did not indicate what might be the underlying cause. The Black Belt being an expert in the field of HPB-malignancies anticipated that type of malignancy suspected, number of additional investigations and proposed treatment might well be factors of influence. For certainty analysis of variance was conducted for a range of potential influence factors (incomplete referral at first instance; reason for referral e.g. second opinion, take over of treatment etc.; type of malignancy suspected; number of addional investigations needed; type of proposed treatment; location of treatment (whether or not in the AMC); final diagnosis benign or malignant disease).

The analysis revealed no relation with type or completeness of referral, type of malignancy suspected, number of additional investigations (including extra outpatient clinic visits) needed or a final diagnosis of benign or malignant disease. As an

(34)

- 34 -

example the negative analysis of variance of additional investigations has been depicted in appendix VII.

In contrary a strong relation with type of proposed treatment (figure 9 and appendix VII. For analysis of variance) and location of treatment was discovered, whereby surgical treatment was strongly related to waiting time and internal treatment e.g. treatment within the AMC to throughput time. However, as surgery was only performed within the AMC it was considered a potential bias regarding the negative influence of the location of treatment on the throughput time. Extended exploration revealed that the location was not of influence for the throughput time when treatment consisted of chemotherapy, palliation or trial. The strong and robust influence of surgery on waiting time indicates the importance of tackling this bottleneck.

Further exploration of the data on the waiting time till surgery revealed that the mean was just above 29 days (IQR 15-34.5 days). The effect on the total throughput time of reducing this waiting time either to 14 or 7 days was investigated. The results showed a reduction of the mean throughput time from 22 days to respectively 17.33 and 15.23 days. To interrogate the effect on the USL empirical CDFs were conducted revealing a fulfilment in an additional 9 till 14.9% of the patients (Appendix VIII).

Trial Surgical None oncological Chemotherapy 80 70 60 50 40 30 20 10 0 Proposed treatment W ai tin g tim e

Individual Value Plot of waiting time between establisched diagnosis and start proposed treatment

(35)

- 35 -

Regarding processing time the prospective data collected from 10 patients were further analysed to search for potential influence factors. As the number of observations is restricted pure descriptive statistics have been used.

As became apparent in the analyse-phase, processing time of the main process, comprising 88.2%, is the driver of the total processing time and not as expected the rework and time spent on preparation of the main process.

Therefore exploration of this part of the process was conducted. As mirrored in the pie chart in figure 10, diagnostics and outpatient clinic visits appeared to take up most of the time. Reflected per patient two hours were spent on outpatient clinic visits and two and a half hours on diagnostics, whereby the time range for diagnostics is rather broad as depicted in appendix VIII. This was a remarkable and unanticipated finding raising questions whether or not a reduction of 15%, 67 minutes, should be achieved by reducing dedicated time for the patient. The eventual consequences on quality of care could not be safeguarded by the available data, implying a dead end track.

With the exeption of the medical oncologist all key professionals had a stable scheduled availability. The fluctuation in time from the medical oncologist, between two and four hours has however never been a bottleneck. On the contrary, zooming in on their available slots it became evident that of these slots only 61.1% (33 out of 54) were used for HPB-GIOCA patients.

Diagnostics 1516; 38,4% Outpatient 1192; 30,2% Calling 76; 1,9% Preparation 62; 1,6% Planning 128; 3,2% Reporting 327; 8,3% MDM 390; 9,9% Reviewing 108; 2,7% Data entering 151; 3,8%

Processing tasks main process

Tasks in minutes and percentages for a total of 10 patients

Figure 10: Pie chart of processing tasks main proces in minutes and percentages for a total of 10 patients.

(36)

- 36 -

Regarding the productivety it was noteworthy that the administrative staff as well as the surgeons spent a significant part of their time on non-GIOCA-related productivety respectively 36.6% and 25.1% (appendix VI). The tasks conducted were either related to other outpatient clinics, breastcancer or surgical controles after HPB surgery, or to interruptions. These interruptions were mainly patients and referring phycisians asking for information.

Another revealing matter is the time spent on administration by the medical professions, ranging from 19.3% up to 44%; main components are preparation of outpatient clinic, filling out forms and looking after correspondence, the center of gravity however differs per profession.

3.4.2. Scenarios, consequences and suggestions

After this more detailed investigation of the effect of the influence factors different scenarios were considered in an attempt to substantiate the consequent steps. A comprehensive overview of the current situation, the envisaged situation, the action undertaken and the suggestions is depicted below (table 2).

CTQ As-is To-be Action Suggestions

Throughput time 40.3% > USL of 21 days 5% > USL of 21 days Handed back and accepted by Champion Decreasing waiting time for surgery Processing time μ = 447.8

min μ = 380.6 min (-15%, -67 min) Handed back and accepted by Champion Restructering outpatient clinic visits Critical appraisal of diagnostics

Preparation time main process μ = 52.7 min None, little fish

Rework/mistakes μ < 1 min None, little fish

Processing time main process μ = 395 min μ = 335.8 min (-15%) μ = 327.8 min (-67 min) Handed back and accepted by Champion Restructering outpatient clinic visits Critical appraisal of diagnostics

(37)

- 37 - Available time Constant > 95% None, little fish GIOCA-related- productive 61.4-90.7% >85% Handed back and accepted by Champion Decreasing interruptions and time spent on other outpatient clinics

Table 2: A comprehensive overview of the current situation, the envisaged situation, the action undertaken and the suggestions.

Throughput and waiting time: The predefined requirement of throughput time of less than 21 days was only met in as few as 60% of the patients.

It became apparent that the most influential factor on throughput time was the waiting time till surgery. The first scenario that was considered was to explore the consequences if the bottleneck would be left like it is. The focus would be directed on other factors whereby any gain achieved regarding throughput time most likely would further expand the devastating effect of the bottleneck. It could also mean to adjust the predefined throughput time: this would however imply a disregard of the value proposition and our patients consequently.

The other option would be to try to decrease the waiting time till surgery. From our data it became apparent that 30% of the patients visiting our HPB outpatient clinic have an indication for surgical intervention, which is in accordance with the international literature. (Jemal A. (2011)) This would imply, based on the data of 2014, that 170 patients should be operated in a 52 weeks period, which means 3.26 per week. Having a guarantied availability of four slots per week for HPB-GIOCA patients at the operation theatre might be an effective measure to decrease the waiting time, whereby unoccupied slots are handed back 5 days before they are due, to avoid underutilisation of operating rooms. To enforce this the surgeons together with the staff of the operation theatre have to take the lead.

Because of the importance of this issue it was felt that there was only one solution: handing it back to the Champion. He accepted it and brought the topic to the attention of the chairs of the involved departments and advised urgent dissolvement of this bottleneck.

An investigation of process optimisation for a highly specialist multidisciplinary fast track clinic : a lean Six Sigma project

Company Project |