Costs of medical specialist training in the Center for Women and Child Health of VUmc

(1)

Costs of medical specialist training in the Center for

Women and Child Health of VUmc

“Tijd voor opleiding”

Amsterdam, January 24th, 2016

Prof.dr. Reinoud J.B.J. Gemke, consultant pediatrician VUmc University Medical Center

Amsterdam rjbj.gemke@vumc.nl

Thesis MBA Health Care Amsterdam Business School

Studentnumber 10733671

MBA – HC supervisor: Prof.dr. J. de Mast

In company supervisor: Prof.dr. J.B. van Goudoever

(2)

MBA – HC thesis Reinoud J.B.J. Gemke

Executive summary

In this project the training costs of residents in 2 medical specialties (Pediatrics and Gynecology) in the Center for Women & Child Health Care of VUmc were assessed. The goal of the project is to enhance the accountability of very expensive training programs for medical specialists that are fully funded by public means. In recent years the accountability of medical specialty training has been improved by collaborative efforts of large general teaching hospitals (Berenschot). Despite efforts, in University Medical Centers (UMC’s) insight in the costs and efficiency (i.e. cost-effectiveness) of residents training programs is lacking.

The methodological principle underlying the project was a Lean Six Sigma approach for improvement of processes in health care that has 5 subsequent phases: Define (D), Measure (M), Analyze (A), Improve (I) and Control (C). Being the first in-depth study of the costs for medical specialist training in UMC’s, the focus of this project will be on the first 3 phases. Improvement and control are activities for a subsequent project.

When costs are clarified, efficiency and opportunities to reduce (potential) waste may be uncovered, using a value chain analysis. For the current study three main determinants of costs were defined, also denoted as CTQ’s (Critical To Quality) in the DMAIC approach:

1. Costs of time of specialists/consultants allocated for residents’ training, 2. Costs of employing specialty residents

3. Other (e.g. indirect and/or overhead) costs of residents’ training.

We used a cost calculation model that was developed in 2012 for a study among large general teaching hospitals in the Netherlands. Fixed and variable costs of residents training were calculated. Fixed costs were derived from financial statements. The variation in costs are particularly determined by variations in variable costs due to specialists’ time for residents training, which is highly

intertwined with (in)direct patient care. Therefore questionnaires were sent to all staff members and residents of the department of pediatrics and gynecology of VUmc for an in-depth assessment of variable costs. Time for categories of activities was asked including (in)direct patient care and non-patient care related teaching and training activities.

As another method to assess variable costs, observations of daily activities of couples of one resident and one staff member (supervisor) on a sample of different work (“shop”) floors of both

departments were performed. The results of the questionnaires and of the observations enable comprehensive assessment of variable costs.

The total costs per resident per year that we calculated in the current study (€ 177,994 for pediatrics and € 194,185 for gynecology), were substantially higher than in Berenschot (€ 140,525) and

somewhat higher that in Robijn (€ 166,000). As we used a well-accepted model for cost calculation and also included an in depth study of variable costs based on broad response from clinicians (staff and specialty residents), we infer that our results are robust. This indicates that the current

reimbursement for specialty residents in the Center for Women and Child Health of VUmc is insufficient. So far this under-compensation has been unnoticed and probably cross-financing from other sources occurs. It is likely that this is also the case in other specialty training programs. The most likely explanation is that the initial assumption of “production-neutrality” on which the reimbursement of specialty residents was based, no longer applies.

The results of this project enable evidence based steps to improve the efficiency of residents’ training programs in the departments of Pediatrics and Gynecology VUmc. The anticipated merging of the Centers of Women and Child Health Care of VUmc with the AMC counterpart offers significant opportunities to optimize the efficiency of residents training programs. The results of the current

(3)

project enable a critical appraisal of how the programs may be merged and what improvements of cost-effectiveness of residents training programs can be realized while maintaining or improving the quality of training for future specialists in the UMCA. Discussions and brainstorming sessions of stakeholders from both VUmc and AMC are likely to provide improvement actions that can be implemented and controlled in the merged future training programs. From a broader perspective, the methods and results of the current study provide a template for the assessment of costs and efficiency of resident’s training programs of other specialties in VUmc and AMC. Moreover the results of this study may contribute to enhanced accountability and transparency of the public funded statutory tasks of UMC’s.

Acknowledgements:

This study would not have been possible without the consent of the staff and residents of the departments of pediatrics and gynecology Vumc. Particularly the support of prof. dr. Johanneke I.P. de Vries, director of the residents training program gynecology is gratefully acknowledged. The author is also thankful to Danii Suyk, Jorine Geertsema and Marlijn Melief, bachelor students in health and life sciences, who completed the observations of daily activities.

The support of prof.dr. Johannes B. van Goudoever, chair of the Center for Women and Child Health VUmc and AMC, Margreeth van der Meijde, MBA, director of the VUmc School for Medical Sciences and Koen Meijssen of the financial department of VUmc are gratefully acknowledged.

Last but not least the project would not have been possible without the coaching of prof.dr. Jeroen de Mast and dr. Marit Schoonhoven from the Amsterdam Business School MBA Health Care program.

(4)

Table of contents page

Executive summary 2

Acknowledgements 3

Table of contents 4

I. Introduction 5

II. Framing and methods used 6

III. Case description 10

A. Define 10

B. Measure 20

C. Analyze 23

IV. Results and reflection 30

A. Results and discussion 30

B. Reflection 32

V. Conclusions and recommendations 34

C. Conclusions 34

D. Recommendations 35

List of abbreviations 36

List of references 37

Appendix 1 Direct activity observations staff pediatrics Appendix 2 Direct activity observations staff gynecology Appendix 3 General costs of residents’ training

(5)

Chapter I Introduction

This project was accomplished in VUmc as thesis of the first Amsterdam Business School MBA Health Care program that was initiated in connection with the alliance of VUmc and AMC. A generic lean six sigma approach for improvement of processes in health care was used as theoretical framework. Following a well described and quantitative problem definition and data-supported diagnosis, a basis for solving the problem is presented. The results are discussed within the context of the literature and also within the context of the management of the project in achieving its deliverables.

In this project the 2 medical specialty training programs of the Center for Women’s and Child Health (CWCH) in VUmc, Amsterdam are assessed as a service process. The goal of the current project is to assess the training costs of residents in the 2 medical specialties of the Center for Women and Child Health of Vumc (Pediatrics and Gynecology) in VUmc. It provides a foundation for the improvement of accountability regarding medical specialty training, which is increasingly required by both internal and external stakeholders. While in recent years the accountability of medical specialty training has been improved in large general teaching hospitals, the University Medical Centers (UMC’s) are chasing behind. Ultimately, for sustained funding of medical specialty training, transparency is mandatory as it is fully provided by public means.

The main question to be answered in this project is: What are the costs of specialty residents training (pediatrics and gynecology) in the Center for Woman and Child Health of Vumc. This can be divided in the following sub questions:

a. Costs of time of specialists/consultants allocated for residents’ training b. Costs of employing specialty residents

c. Other (e.g. indirect and/or overhead) costs of residents’ training.

The complex intertwined accomplishment of providing health care and resident’s training, mostly at the same time, means that it is complicated to perform an adequate costs analysis. Consequently there have been a very limited number of previous studies assessing the costs of residents training.

Benefit analysis

Benefit for the customer (primarily the specialty resident)

Better accountability is mandatory for sustained public funding of their training. In this perspective, cost-effectiveness is a prerequisite for expensive residents’ training programs. Transparency and accountability may also improve the allocation of (currently unprotected) time for specialist training

(6)

that has long been considered as an inherent part of daily work and fully integrated in patient care. Increasing requirements and the more formal structure of training programs with increased numbers of residents, increased requirements from training staff and increased facilities for teaching hospitals should stimulate better accountability. Moreover it may also support the satisfaction of staff and residents and the quality of teaching hospitals regarding their teaching and training obligations. Benefit for the organization (VUmc).

In this perspective it should be realized that yearly about € 44 mln (i.e. 6,5% of the total yearly budget) is funded by reimbursement for residents’ training. This includes not only the training programs pediatrics and gynecology but over 20 other residents’ training programs, totaling about 350 specialty residents in VUmc. Hence the analysis of this project may become a template for analysis of the cost and effectiveness of all other specialty residents training programs. Benefits for the alliance of VUmc and AMC.

In the strategic alliance of VUmc and AMC, both Center’s for Women and Child Health are

frontrunners. The collaboration of the residents training programs for pediatrics and gynecology are part of the frontrunner projects. Experience from assessment of cost-effectiveness and quality of residents’ training programs of these two lead specialties may support the merging of other specialist training programs.

Potential benefits for other UMC’s and Nederlandse Federatie Universitaire Medische Centra (NFU): From a broader, national perspective, the current project may contribute to the UMC’s catching up regarding financial transparency of residents training programs as compared with the STZ hospitals.

Chapter II Framing and methods used

While medicine and health science have provided an enormous array of new diagnostic, therapeutic, preventive and prognostic possibilities, the process of provision of health care is hampered by high costs, low efficiency and many incidents adversely affecting its quality and safety. Increased realization of this has sparked numerous initiatives to address these important shortcomings. The landmark report “To err is human” by the Institute of Medicine (USA) has put this issue in the spotlights and has given an enormous boost to initiatives improving the quality of health care.1 As a result, in the past 25 years, improvement in health care quality has grown from demonstration projects into a worldwide movement. Dominant in this movement has been an improvement approach, founded by the work of William Edwards Deming on the Plan-Do-Check-Act cycle,2 which was based on quality circles introduced by Walter Shewhart. Subsequently Joseph Juran addressed

(7)

the costs of poor quality and offered a further perspective by introducing the quality trilogy, encompassing quality planning, quality control and quality improvement.3

More recently managers and professionals in health care recognize the potentials of Lean and the Toyota Production System (TPS) as a powerful method for improvement of quality and cost reduction in manufacturing.4 Lean synchronization aims at achieving a flow of products or services that exactly meets what the customer wants at the lowest possible cost. The 3 main topics addressed in Lean are: (1) Elimination of waste (comprising the categories transportation, motion, redundant or duplicative work, over- processing, complexity, defects or rework and inventory), (2) Involvement and

empowerment of employees in shared tasks to smooth the process and eliminate waste and (3) Adopt principles and strategies for continuous improvement.5 Examples of successful

implementation of lean in health care by influential organizations such as Virginia Mason Medical Center are well appreciated.6 Nevertheless health care professionals remain reluctant in accepting that important lessons about optimal delivery of the health care process could be learned from the production and service industry, despite the differences in content. Experts in the area of operational effectiveness in health care increasingly share their conviction that there are opportunities to

improve the process of delivery of health care substantially.7 Indeed the enormous possibilities of modern medical care in combination with increasing financial constraints demands substantial improvement of many sub-efficient health care processes and the reduction of waste therein. The successful adoption of lean principles in hospitals is predominantly affected by (1) the sense of urgency (often sparked by quality and/or financial issues), (2) initial projects demonstrating

important improvements of the quality and the financial position and (3) prolonged involvement and unequivocal support of the board and leading professionals.8

In the Netherlands there are initiatives where business school affiliated research institutes have adapted Lean Six sigma models for operational excellence from the production and service industry for application in health care.9 In a number of collaborations of large hospitals with these research groups, lean six sigma has been implemented in these hospitals. So far the results have particularly been in reducing costs and improving cost-effectiveness. For examples of specific applications in health care see these references.10111213

In this project the costs of specialty residents training are assessed, with the perspective of subsequent improvement of cost-effectiveness. Therefore some theoretical background on the (potential) financial strategies of this process is appropriate. Historically the funding of specialty residents training comes from public means, provided by the Ministry of Health. Until a decade ago this funding was included in the all-inclusive budget that (university) teaching hospitals received from the Ministry of Education and Ministry of Health Care for their public tasks. The lack of accountability

(8)

in the spending of this part of the budget and lack of control of the number of future specialists have induced major changes. From 2005 on, the reimbursement for each residents in training has become a yearly grant, to be obtained from the Ministry of Health, based on rigorous criteria including a strict control of the total number of residents per specialty and the distribution of the numbers of

residents per specialty to each (university) teaching hospital.

There are 3 ways possibilities to allocate this reimbursement: (1) Backpack model in which each selected and appointed specialty resident is provided with a scholarship to complete his/her training within strict regulations, (2) OOR model in which each cluster of collaborating teaching hospitals, headed by one UMC, is yearly provided with the grants for a strictly allocated number of specialty residents and (3) Tender model in which OOR’s (clusters of UMC with teaching hospitals) make a bid (“quote”) for an all-in training of a specific number of specialty residents.14 Currently model 2 is adopted and for the last (half) year of resident’s training model 1 is possible. Model 3 has been discussed a decade ago but not been adopted. Recently the report of a multi-ministry policy group proposed a yearly self-investment by specialty residents for their training.15 Following a vigorous debate, so far this proposition was not adopted by the Government.

In the current project a generic lean six sigma approach for the improvement of processes in health care is adopted. This approach comprises a well described and quantitative problem definition, as well as a data-supported diagnosis, preceding attempts of problem solving. Improvement actions are tested before they are accepted and implemented.1617 The methodological principles underlying Lean Six Sigma have been integrated in 5 subsequent phases: Define (D), Measure (M), Analyze (A), Improve (I) and Control (C). This cycle may be regarded as more intuitively obvious than the PDCA cycle insomuch that it follows a more “experimental” approach. The major definitions of the steps within of these phases can be described as follows:

Define: Define the scope of the analysis (which business unit, which process, which variables) Measure: Establish the key performance indicators: CTQ’s (Critical To Quality) and their target

Validate the measurement procedures

Analyze: Diagnose the current process and compare to target Identify potential influence factors

Improve: Establish the effect of influence factors

Design improvement actions for the great influences Control: Improve the process control

The emphasis in current project will be on the Define, Measure and Analyze phases of the costs of the CWCH specialty residents’ training program. The Improve and Control phases are beyond the

(9)

scope of the current project, but will be discussed from the perspective of subsequent steps that may be taken in future projects in the alliance and/or merging of VUmc and AMC.

A process analysis that is derived from industrial and service organizations may be used for a theoretical description of the subsequent steps a resident goes through during his/her training to become a medical specialist. Depiction of this so called SIPOC (supply, input, process, output to the customer) provides a high level overview of process that is studied (see figure 1 below).

After graduation from medical school (including various rotations called internships) junior medical doctors may start a training period to become a medical specialist. This period as specialty resident lasts 5-6 years in which a gradual increase of independent professional activities of a consultant specialist is pursued. A (very) large proportion of this training is combined with clinical work (patient care). In this process of gradual increase of independency, initial observation of professional activities is followed by directly supervised care, indirectly supervised care and finally independent patient care, which resembles working as a full qualified specialist.

Figure 1. SIPOC (supply, input, process, output to the customer) chart of process that is studied

Suppliers Inputs Process Outputs Customers

Medical school Specialty residents Specialist training Consultant/specialist Patients

Step 1 : observation of patient care

Step 2 directly supervised patient care

Step 3 indirectly supervised patient care

Step 4 independent patient care

(10)

Chapter III Case description

A. Define

In University Medical Centers (UMC’s) there is an intensive cooperation between the faculty of medicine and the affiliated university hospital. UMC's provide a wide range of medical care including top clinical care (specific categories of advanced care, the provision of which is restricted to a limited number of large hospitals). In addition they fulfill three public functions: (1) provision of top referral (last resort) patient care, (2) research and (3) teaching/training of health care professionals. The latter are particularly medical students to become doctor (MD) and residents in training to become a consultant medical specialist. The latter are called specialty residents (in Dutch: AIOS [Assistent In Opleiding tot Specialist], not to be confused with a PhD student formerly called AIO, currently OIO). These public functions of the UMC’s are financed by several different funds. The goal of each of these funds is to finance a certain public function. However, in reality it is not only difficult to discern what funding is used for which purpose, it is also unclear how the different funds contribute to the different results of the UMC. In order to protect and to avoid further explanation of this situation. The complexity and intertwined relations are often (mis?)used as reason to maintain the status quo. Some stakeholders use the “pancake” metaphor: When you get a good pancake you don’t want (have) to know all the ingredients and their cost but you will just pay the bill.18 Due to financial constraints, the Ministry of Health increasingly demands accountability of the different funds transferred to the UMC’s.

In figure 2 below the income of VUmc in 2014 is depicted, split by the different funds from which it is obtained and totaling € 675 mln. For training of residents, yearly 44 mln (6,5%) is available in Vumc.

(11)

MBA – HC thesis Reinoud J.B.J. Gemke Financing of specialty residents

Teaching and training of specialty residents to become consultant specialists takes 5-6 years after completion of medical school, is very expensive and is funded by public means. Therefore in the last decade the Dutch government has strictly limited the number of residents training positions for each medical specialty. For the whole country there are currently about 10,000 residents in training to become a consultant-specialist, the total funding of these training positions amounts € 1,5 billion per year. After a governmental policy of slow increase in the last 5 years, the number of specialty

residents is expected to remain stable in the coming years. Nevertheless, the distribution of the number of specialty residents positions every year is fiercely negotiated by 8 groups of regionally collaborating teaching hospitals (clusters, each comprising of one UMC and a number of general teaching hospitals).19 Not in the least, these negotiations are driven by the financial reimbursement (grant) of € 130,000 per resident per year. Apart from the residents salary (about € 70,000, including additional employers costs) this reimbursement is based on largely unverified assumptions of actual costs and expenses. When this funding was instituted in 2006, an important assumption was that specialty residents, during their 5-6 years of training, on balance their productivity would be neutral (“productie-neutraal”), meaning that the loss due to a resident’s limited production capacity and the consultant’s delay-time, incurred in the junior training phase, is compensated by the gain in

production capacity and the resemblance of consultant’s work in the senior phase of the resident’s training period.

With around 340 specialty residents in training, for VUmc the total funding amounts to about € 44 mln/ys. Financial constraints in health care as well as the demand for improved accountability of this expensive, publicly funded, training programs has driven attempts to assess the costs for these specialist training programs. It also has sparked discussions on the way specialty training should be funded in the future. Even in an affluent country like the Netherlands, scenario’s that (post-graduate) specialty residents, similar to (pre-graduate e.g. medical) students, have to pay a substantial

proportion of their training themselves are debated and may only be alienated by efficient training programs and accountability of funding.

Results in large general teaching hospitals

In 2009 the large general teaching hospitals, collaborating as the Stichting Topklinische Ziekenhuizen (STZ), have commissioned the Dutch consultancy firm Berenschot to assess the costs of medical specialty training. 20 Not surprisingly, but so far undetermined, it was noted that in the junior phase of residents’ training, (in)direct patient care takes substantially longer than in the senior phase, where residents resemble working as a consultant. It was also recognized that for medical staff, the

(12)

training of specialty registrars meant that they cannot provide patient care as efficient as they would be able to do without a resident/trainee. This was referred to as delay-time (“vertragingstijd”). It was noted that modern residents training not only comprises (in)direct patient care but also requires a substantial period of time for other training related activities. Despite the uncovering of these major qualitative insights, the first attempt to clarify costs of residents training was not successful. This was primarily because the study did not include an in-depth and quantitative analysis of the intertwined activities of residents and staff in combining patient care with training/teaching.212223

Based on these preliminary conclusions, a second study by Berenschot was performed in 2011. This comprised an in-depth study of 5 medical specialty training programs, two of which in 2 different large general teaching (STZ) hospitals. The results can be found in the report "Cost Study on the training of medical specialists in four STZ hospitals".2425 The main conclusions of the Berenschot study are:

1. Training of residents to become medical specialists involves benefits but also (high) costs for hospitals (and for specialists if they have a private practice within the hospital).

2. Variation in costs are particularly determined by variations in variable costs, namely: a. Variation in delay time: extra (hospital) costs due to longer usage of expensive

resources (e.g. OR, cath lab)

b. Variation in revenues due to variation of residents’ contribution to care production c. Variation in costs and revenues for medical specialists with hospital based private

practice

i. Costs: supervision and/or attendance of (in)direct patient-care

ii. Revenues: independent patient care by residents (e.g. rounds, on-call duties) saving other costs

3. Most important determinants of the total costs of residents training program are: a. Utilization of expensive hospital resources (e.g. OR, cath lab)

b. Attendance of staff at training related indirect patient care activities (e.g. shift hand over, grand rounds)

c. Number of residents in training (more residents corresponds with higher efficiency) 4. Total costs comprise:

a. Fixed costs: on average approximately € 100,000 (see point 5) b. Balance of variable costs and revenues (see point 6)

5. Fixed costs comprise for about 2/3 residents’ salary and for about 1/3 overhead, including residents’ training related capital expenses (e.g. clinical training center)

(13)

6. Balance of variable costs and revenues is determined by the (extra) hospital costs due to residents delay time, specialists costs and revenues (e.g. saved costs for other personnel) 7. Both national and local regulations and conventions of residents’ training programs have

major impact on costs

The average costs of the resident training programs was € 140,000, comprising about € 100,000 for fixed costs and about € 40.000 for the balance of variable costs and revenues. There were substantial differences between the 3 main categories of medical specialties (surgical, medical, and supportive: “snijdend”, “beschouwend”, “ondersteunend”). The total costs of a resident training for a surgical specialty (surgery, gynecology) were € 172,400/yr, for a supportive specialty (radiology) was € 148,250/yr and for a medical specialty (internal medicine, neurology) were € 104,800/yr. Also large differences between hospitals were found in variable costs, revenues and the net balance of variable costs and revenues. Hence specialty category and hospital appear to be (independent) determinants of variable costs and revenues.

This was the first study to provide insight in the costs for medical specialist training, yet it was restricted to the large general teaching hospitals (Stichting Topklinische Ziekenhuizen, STZ). It should be realized that, due to its complexity, a small sample of specialties (5) in (only) 2 hospitals were studied. Nevertheless the relevance of the subject, the high amount of public funding and the fact that for the first time an in-depth study with appropriate methodology that provided adequate results has led to broad acceptance of the results, despite the fact that the generalizability of the results may well be disputed.

The NFU project Robijn

In 2014 the collaborating Dutch University Medical Centers completed a broader program to improve accountability of public funds that were not related to reimbursement of insured patient care. This project, called Robijn, included an assessment of the accountability of (1) residents training, (2) teaching of medical students, (3) academic work place function and (4) government based funding of research. All these funds are highly significant for the total income of UMC’s, see also figure 2 for the income components of VUmc in the perspective. In Robijn the costs of residents training programs were assessed and compared to international references. The report described that the

reimbursement percentage that Dutch UMC’s receive for residents training on average is 8,6% of the total care budget (income from health insurance + Ministry of Health). However substantial

variations among UMC’s were found, ranging from 6,5% - 10,9%. For the Vumc the reimbursement percentage is 44/(370+57+44) *100% = 9,3%. For AMC the reimbursement percentage is

(14)

(47/(450+101+47)=7,8%. In 2 other European countries where these figures are available for comparison, in Finland 8,3% was found and in Spain 9%.2627

Because of its transparency and acceptance, Robijn also used the Berenschot methodology as the basic model for calculation of costs. To obtain an additional insight in the variable costs of teaching and training, a questionnaire was sent to directors of UMC resident’s training programs to obtain an estimate of time involved in residents training. Although the Berenschot study had concluded that in-depth field study is mandatory for adequate assessment of costs, this was not included in the Robijn project.

Like the STZ hospitals, also in UMC’s large variations were found in the numbers of residents, the structure and deployment of the residents training programs. In a provisional analyses, Robijn used an average 11-12 hrs of specialists time per resident per week spent on non-patient care related training activities (i.e. 20-25% of total full time equivalent specialists’ time). Based on residents working 43 weeks per year effectively (excluding holidays and other absence) € 64,000 per resident per year was calculated for average indirect (non-patient care related) training costs. In Berenschot € 54,000 was found for this category of activities. This difference was largely attributed to activities of directors of UMC’s residents training programs for their role in local, regional and national

coordinating activities. In total the average costs for resident’s training in UMC’s was € 166,000 per resident per year. Although in Berenschot large differences among the 3 main categories of

specialties (medical, surgical and supportive) were found, no subgroup analysis was performed in Robijn.

Table I. Summary of NFU Robijn results regarding costs of residents’ training

Estimated costs of medical specialty training in UMC’s €/resident/yr

- Resident’s salary (incl employers costs) + training related expenses* 71,000 - Overhead (43,7%) incl. UMC’s training infrastructure and related personnel** + 31,000

Fixed costs (in Berenschot ca € 99,000) 102,000

- Variable teaching and training costs*** + 64,000

Average estimated total costs of medical specialist training in UMC 166,000

*e.g. external obligatory courses (on average € 5000/resident/year)

**This overhead percentage is based on investment (depreciation costs) and exploitation (incl. personnel and inventory costs) of teaching and training infrastructure. The percentage is used by NFU and is based on UMC’s accounting information.

(15)

An important deviation from the Berenschot model was that no assessment of delay time (to obtain the balanced result of variable costs and revenues) was included in the Robijn study (Berenschot conclusion points 4 and 6). In this perspective it should be realized that in the Netherlands, all specialists are employees and do not pursue private practice within the UMC. Hence, although there are no delay-time related costs or revenues for the specialists, this may well be the case for UMC’s but this remains uncovered.

In summary Robijn provided a rough estimate of the costs of medical specialist training, largely based on broad assumptions and general UMC accounting information.28 The main conclusion from the report is that current reimbursement of € 130,000 per resident per year is not overfunded but provides insufficient compensation as the average costs were estimated to be around € 166,000. At least because, as stated earlier, estimates of specialists’ delay time were not included. It should be realized that revenues from residents were not included, nor was the net balance of residents’ variable costs and revenues.

Consequently, comprehensive insight of the costs of medical specialist training at UMC level, let alone at specialty level, is currently unavailable. This despite the fact that in the near future better accountability is expected to be required by the Ministry of Health. Failing to provide this, there is a substantial risk of discontinuation of the current level of public funding for high quality medical specialist training.

Residents training programs in pediatrics and gynecology

In the Netherlands about a decade ago, Pediatrics and Gynecology were the first to initiate the restructuring and modernization of their specialty training programs according to the nationally adopted CANMEDS competencies for all health care professions.2930 Before this modernization, residents were merely making high numbers of working hours to become a consultant more or less automatically after 5-6 years. Following the revision, the residents’ training programs still comprised long periods of training on the job but these became much more structured, including better

described rotations, frequent feedback, regular clinical training center sessions (e.g. skills lab), monthly courses, periodic exams etc. Also the obligations of staff became more structured and increased substantially, both as trainers and in taking over residents’ work during their mandatory absence for off-site training activities. Moreover professional training of staff became mandatory (post graduate Teach-the-Teacher programs). With the implementation of these revised training programs, further clarification of the time and content of activities spent by staff

(trainer/teacher/coach) and by residents (trainee) became increasingly relevant. From the same starting points and with the same perspective all other medical specialties have revised their training programs in the last 5 years.

(16)

The aim of this MBA thesis is to provide a more comprehensive insight in the costs of residents training by a more in-depth study of the two residents training programs of the Center for Women and Child Health Care in VUmc. The costs of indirect (extra) time of specialists/consultants (trainers) and specialty residents (trainees) as well as other costs for medical specialist training will be

assessed. We used an accepted cost calculation model that was developed in 2012 for a study among large general teaching hospitals in the Netherlands. The basics of this model were also applied in the project Robijn of the NFU (Federation of collaborating Dutch UMC’s), a more cursory study of the costs of residents training that was part of a broader project to enhance the accountability of UMCs’ public funding, published in early 2015.

The main question to be answered in this project is: What are the costs of specialty residents training in pediatrics and gynecology, respectively, as provided by the Center for Woman and Child Health of VUmc. This can be divided in the following sub questions:

a. Costs due to time of specialists/consultants allocated for residents’ training b. Costs of employing specialty residents

c. Other (e.g. indirect and/or overhead) costs of residents’ training (e.g. availability of a clinical training center)

Stakeholder analysis

Before its conception the potential benefits and drawbacks of the project were discussed with Margreeth van der Meijde, MBA, vice-dean for medical education and director of the School for Medical Sciences at VUmc. Despite the fact that she recognized that that not everyone might be happy with the findings of this project, she supported the proposition of the author that sustained continuation of public funding was best served by transparency and accountability. Therefore she unequivocally supported the project. Subsequently Prof. dr. Johannes van Goudoever, head of the division of Women and Child Health at VUmc, was consulted. With similar arguments and with the potential to include the benefits of more efficient residents training in the alliance of AMC and VUmc he became a sponsor of the project. Both Mrs. van der Meijde and Mr. van Goudoever consented to become in-company coach of the project. Consultation with prof.dr. Saskia Peerdeman, chairperson of the Centrale OpleidingsCommissie (COC, central commission for resident’s training), Werner Koster, senior policy officer supporting this commission, also resulted in their support. Upon request, Prof.dr. Johanna de Vries, director of the Gynecology residents training program, offered

(17)

director of the financial department of VUmc, Koen Meijssen became involved providing the financial expertise to the project.

The vulnerability of the current project lies in the possibility that internal stakeholders may discontinue their support if teaching and training time cannot be adequately accounted for in the current budget structure. Subsequently inappropriate attempts to restructure internal budgets items to suit individual stakeholders might occur. In absence of accountable spending, external

stakeholders (e.g. the Ministry of Health), are likely to advice reduction of the budget and funding for residents’ training programs.

If transparency, obtained by the current study, would reveal that too little or too much time and/or money is spent on residents’ training programs the adaptation and/or scaling up the methodology and conclusions of this project to other residents’ training programs would be hampered. In that case the Center for Women and Child Health might literally be(come) the only division in VUmc to “pay the price” for transparency while potential over-financing remains concealed in other divisions. In a broader perspective, if more money is spent on specialist training than received, cost cutting without increasing efficiency would seriously impede the quality of the specialists’ training program.

Project management

The project was prepared in May-June, 2015. Before the start of the measurements, the Departments of Pediatrics and Gynecology were familiarized with the background, purpose and content of the project called “Tijd voor Opleiding”. The author presented the project to the residents and staff of both departments on a regular time slot following the morning shift hand-over. This was on October 13th for Pediatrics and October 28th for Gynecology.

The deliverables of the project include an in-depth assessment of the yearly costs of resident’s training, with specific attention for an analysis of variable costs, by the end of 2015. Analysis of variable costs is primarily based on the input from a comprehensive questionnaire, to be completed by a large representative sample of staff and residents. This first deliverable will enable robust internal and external accountability of CWCH residents’ training programs. As a complementary analysis, the time (hence costs) of coaching of residents on the “shop floor” by direct observations of activities of staff and residents will be performed. Also a deliverable is a clear and readable report, explaining the cost-model in a way that it can be used as a template for subsequent projects for cost analysis in other medical departments of VUmc and AMC. In subsequent studies, initiated by other stakeholders in VUmc and/or AMC, the results from the project may be used for improvement of cost-effectiveness and imminent pilots on performance based resource allocation of residents’ training.

(18)

To achieve these deliverables, the planning and control of the project was based on the PRINCE2 principles for good project management.31 This acronym means PRojects IN Controlled Environments in which 7 principles and 7 themes are involved in 7 project processes that were also used in the management of the current project:

- Starting up by conceiving the idea, writing the project/business plan, forming a project board and project team, taking up project management, communication with stakeholders.

- Initiating the project by planning and monitoring the start-up phase, subsequently refining the business case and including assessment of quality and control of the project.

- Directing the project by authorizing different phases of execution, direction of (un)foreseen events, project closure, setting up project files.

- Controlling more specific stages (e.g. alterations during planned observation of activities) - Managing product delivery: Accept, execute and deliver results of a work package (e.g.

distribution of questionnaires, calculation of direct costs, delivery of concept report) - Managing stage boundaries: This was informal but strict e.g. regarding the time period in

which questionnaires could be returned, defining floors and periods for activity observations - Closing the project: With the completion of this thesis the project as such will be closed but

subsequent actions, based on recommendations are proposed. These actions may be initiated by stakeholders of VUmc and/or AMC (see section Recommendations).

Problem definition with Critical To Quality (CTQ) flowdown

CTQ: Costs of teaching (staff time), training (residents time) and other costs for residents training. - Staff (teaching) time: Time/costs for resident teaching (e.g. supervision, feedback, classes) - Residents’ (training) time: Time/costs for training: salary and additional expenses (e.g. for

courses, exams etc)

- Other costs: indirect (non)personnel overhead costs e.g. clinical training center, administration etc.

The target of the costs is equal to the current reimbursement rate (grant) i.e. € 130,000 per specialty resident (regardless of specialty category) per year.

(19)

Figure 3. CTQ flow down of costs of specialty residents’ training

Strategic focal point Project objective CTQs Operational costs

Assessment of specialist training costs (pediatrics & gynecology)

Assessment of direct teaching costs

Assessment of direct training costs

Staff time for teaching Time of all residents' activities

Assessment of indirect training costs

Overhead costs of residents' training Staff costs for teaching Costs of all residents'

activities

Figure 4. Process flow chart of specialty training (Pediatrics and Gynecology)

1. Direct patient care

year 1-2 year 2-3 year 3-4 year 4-5

Ward/delivery suite Specific rotations (Outpatient) consultation Choice (sub)specialty rotation

2. Indirect patient care

- Rounds Daily

- Shift hand over Daily - Multidisciplinary consultation Daily 3. Non patient training activities

- With own specialty at least one per week - With other specialties at least one per week 4. Cursory training activities

- specialist resident’s class 6-12 days/year - interdisciplinary specialist residents class 1-4 days/year

Observation of care

Directly supervised care

Indirectly supervised care

On call ward & ER

Observation of care

On call IC/perinatology

On call for all activities

Independent patiënt care as specialist

(20)

B. Measure

We measured the time (hence costs) of activities of staff and residents with questionnaires and direct observations. We measured other costs of residents’ training programs by assessing the allocation of relevant budget components, obtained from the administration of VUmc.

Questionnaires

Based on the methods of Berenschot and Robijn, a comprehensive questionnaire was designed. Following pilot-testing, it was distributed to all specialty residents and medical staff of the

Departments of Pediatrics and Gynecology. The questionnaire comprised of 2 versions: one for the residents and one for the staff of both specialties.

We formulated questions addressing similar categories as the Berenschot report. Whereas in Berenschot and Robijn a limited number of stakeholders were used as respondents (usually the director of the residents training program), for the current in-depth project we decided to include all staff members and residents as respondents. Therefore we expect to find valid and more reliable results of residents and staff activities from our study than the Berenschot and/or the Robijn study. For review of the questionnaire these links to Surveymonkey may be used. Staff questionnaires for both specialties were identical, residents questionnaires for both specialties were also identical. We defined a response rate > 50% as appropriate for a representative sample.

Staff pediatrics https://nl.surveymonkey.com/r/L96FPHQ Residents pediatrics https://nl.surveymonkey.com/r/L8XMXV5 Staff gynecology https://nl.surveymonkey.com/r/MMH63RD Residents gynecology https://nl.surveymonkey.com/r/MN3BVLY Data files and survey summaries are available for review.

Direct observations

Direct observations of daily activities were performed among a sample of residents and staff with specific attention for different categories of training and teaching activities on the work floor. Direct observations of activities were included for a more detailed assessment of routine days on the “shop-floor” where working and training are highly intertwined. This was unequivocally the most relentless issue in the (first) Berenschot report and an acknowledged drawback of the Robijn project. Direct observations were expected to provide a more realistic and comprehensive picture of the close cooperation of residents with supervising staff, incurring a substantial amount of (indirect) time. But it also constitutes a highly significant contribution to the training of residents in its combination with

(21)

complex patient care. Supervision of residents lies at the heart of their daily training. Therefore couples of staff/trainer with resident/trainee were assigned for registration of their daily activities. Results from the observation study are supplementary to, but not directly comparable to, the results from the questionnaires. Whereas the results from the questionnaires reflect the training efforts of the average staff, the direct observations provide an insight of the over-average training activities of a limited number of assigned (and rotating) supervising staff-members.

Validity of measures:

Although the interconnection of care and training in the daily work of residents with staff remains an unresolvable issue, the methods we used were derived from previous studies (Berenschot, Robijn) that were well accepted. The different categories in the questionnaires and observation studies (direct patient care, indirect patient care, non-patient care in-hospital training activities, other training activities) and the activities of staff and residents that were attributed to these categories were first discussed in brainstorming sessions and subsequently tested in pilot study in the department of pediatrics in May-June, 2015. The 2 residents training programs of the Center for Women and Child Health of VUmc comprise both a medical (“beschouwend”) and a surgical

(“snijdend”) specialty: Pediatrics and Gynecology, respectively. As substantial cost differences were found in Berenschot between these 2 categories, the inclusion of both main specialty categories in the current study is a relevant advantage. Lack of this differentiation is an acknowledged drawback of the Robijn study.

Definitions of activities

Staff activities. Firstly each staff member was asked to estimate the weekly hours and percentage of their time allocated to the 3 primary tasks of an academic medical specialist comprising (a) patient care, (b) research activities, (c) teaching & training activities and (d) other activities (e.g.

administration). Subsequently questions were asked to obtain a more specific answer to the time allocated for residents training.

1. Direct patient care, being one-on-one activity of resident with specialist in the presence of a patient (e.g. consultations, procedures, operations).

2. Indirect patient care, being one-on-one patient related activity of resident with specialist in the absence of a patient (e.g. shift hand-over, (grand)rounds, patient-related interaction with other health care professionals, patient-related administration).

3. Non-patient care related activities, comprising in-hospital training related activities for residents (e.g. mono- of multidisciplinary teaching sessions, specific residents’ classes, feedback, coaching).

(22)

4. Other activities for teaching and training of residents (e.g. teaching contributions to local, regional and/or national residents classes, postgraduate education for residents training [Teach-the-Teacher]).

Particularly regarding frequent and regular (in)direct patient care activities (e.g. morning shift hand-over), it may be difficult to disentangle the staff’s contribution to resident’s training and to patient care, respectively. Therefore the average percentage that consultants assigned to the proportion of each of these activities that contributes to the residents training was included in the questionnaire. This percentage was used as a correction factor (factor 2) in the subsequent analysis.

Resident’s activities. All residents’ activities and associated time were regarded as training costs, primarily because of external (public) funding. However to enable potential future improvement of efficiency (cost-effectiveness) of residents’ training, the different categories of residents’ activities were questioned. These comprised the following categories:

1. Direct patient care, being one-on-one activity of resident in the presence of a patient, regardless of the presence of a specialist (e.g. consultations, procedures, operations). 2. Indirect patient care, being care for (a) patient(s) without his/her (their) presence (e.g. shift

hand-over, (grand)rounds, patient-related interaction with other health care professionals, patient-related administration).

3. Non-patient related activities, being in-hospital training related activities (e.g. learning sessions, residents’ classes, feedback, coaching).

4. Other training related activities, regarded as work activities out of hospital (e.g. scheduled (mandatory) regional of national residents classes, exams). Hence not including home based self-study.

As the salary and employment of residents are fixed costs (in this project as well as in Berenschot and Robijn) that are included in the publicly funded reimbursement rates, the activities of residents were not analyzed in the current project. Rearrangement of residents’ activities would not affect the (total) costs of their training. Residents activities that were registered in the study might be used in subsequent studies assessing the efficiency and quality of their training activities.

Project organization

Following the introduction on a regular morning shift hand-over session the presentation and extra information of the project was mailed to all staff members and all residents of both departments in order to address those who could not attend the oral presentation. This mail was distributed by the secretaries of the departments of pediatrics and gynecology. The mail also included a link to the web

(23)

based questionnaire for staff (42 pediatricians, 30 gynecologists) and residents (24 pediatrics, 17 gynecology), respectively. One and two weeks later reminders were sent by the secretaries, again including the links. The questionnaires were closed after 3 weeks. The web based questionnaire was completed by 22 of 41 pediatric staff members (54%) and 19 of 24 pediatric residents (79%). In the departments of gynecology 15 of 30 staff members (50%) and 14 of 17 residents (82%) completed the questionnaire.

Observations of staff and residents were scheduled in the 2 weeks following the introduction of the project and the distribution of the questionnaires. The “shop-floor”, where the primary process of patient care and most of the residents training takes place, of the department of pediatrics comprises 2 medium care pediatric (sub)specialty wards with 20 beds each, 2 intensive care units (one neonatal ICU with 14 beds and one pediatric ICU with 6 beds) and outpatient clinics which also providing accident and emergency (A&E) services. The “shop-floor” of the department of gynecology comprises wards (gynecology, obstetric), delivery suite, OR and outpatient clinics.

Observations of daily activities were scheduled to include a representative sample of the shop-floors of both departments. In total 13 working days were included in the department of pediatrics, each comprising the time for activities of a resident and a staff member (supervisor). In the department of gynecology 6 working days were included, also comprising the time for activities of couples of one resident and one supervising staff member. Activity categories were the same as described for the questionnaires before. Times were rounded to the nearest 5 minutes of observed activities. All observations were done by 3 trained bachelor students in Health Care Sciences and were noted on printed activity registration forms.

C. Analyze

Analyzes of information from questionnaires Overall distribution of core activities

The staff of the Department of Pediatrics reported that on average patient care takes 54,4% of their time, teaching and training of residents/students 19,2%, research 14,7% and other activities 11,7%. The staff of the Department of Gynecology reported that on average patient care takes 61,0%, teaching and training 14,3%, research 10,7% and other activities 17,5%.

For calculation of costs the Berenschot model was used. This model was adapted with the financial standards and conventions for UMC’s, as derived from the Robijn project.

(24)

Relation of time with (personnel) costs

Most costs are due to time of personnel, particularly residents and staff. To convert physicians’ time to costs, information from the financial department of VUmc was used to obtain the average costs of staff and residents from Pediatrics and Gynecology. Tables with GPL (Gross -annual- Personnel Costs) are presented below. From annual costs, the personnel costs per hour were calculated for each of the 4 physician categories. The tables also provide the total number of staff and residents in full time equivalents (fte’s) in the Departments of Pediatrics and Gynecology.

Table II. Conversion of time to costs of staff and residents, respectively.

GPL = Gross Personnel Cost (mean costs per fte per year)

Based on the adopted model, cost items include fixed and variable costs.

Fixed costs of residents training comprise

a. Resident’s salary including additional employers costs b. Overhead costs based on NFU standard (as in Robijn: 43,7%),

c. Costs of hospital’s residents clinical training center (in VUmc this budget for residents is allocated to the VUmc Medical School (Instituut voor Onderwijs en Opleiding, IOO), see appendix 3): In total this amounts € 965,000 for about 340 residents in VUmc, i.e. €2838 per resident per year

d. Costs of (vice) directors of residents training program (0,5 fte GPL per year pediatrics and 0,3 fte GPL per year gynecology). Although it is agreed that about these amounts of protected time should be available for these functions/activities, in reality this appears impossible to realize in busy clinical practice.

e. Costs of local and regional residents training commission meetings (ca 30 hrs/yr for 5 other staff members) based on GPL per hr, divided by the number of residents for each department.

f. Cost of secretary/coordinator per resident per year

g. Other internal invoices of training facilities (rent of facilities e.g. simulation lab, catering) per resident per year

GPL staff/yr GPL staff/hr fte staff GPL resid/yr GPL resid/hr fte resid.

Pediatrics 194318 106 42 73749 35 21,7

(25)

Table III. Fixed costs of specialty residents training (per resident per year)

Total fixed costs for residents in pediatrics amount € 116,279 per year, Total fixed costs for residents in gynecology amount € 120,038 per year.

Variable costs of residents training comprise

Costs of residents training include time for teaching and training activities of staff during fluently alternating collaborative and independent work with residents while providing (in)direct patient care. Particularly in these activities of (in)direct patient care, the contribution of staff to resident’s training and patient care, respectively, are hard to disentangle. Therefore in the questionnaire the average proportion that consultants assigned to each of these activities was corrected for (factor 2).

Table IV. Correction factor for attribution of staff time to training and patient care (factor 2).

Fixed costs Pediatrics Gynecology

Salary (1,0 fte) incl other expenses 73749 77348

Overhead costs UMC's (43,7%) 32228 33801

Clinical Training Center (1) 2838 2838

(Vice) director residents training (2) 4477 3350

Residents training committee (3) 732 914

Secretary (4) 1467 1090

Internal invoices (rents, catering) 788 697

Total fixed costs 116279 120038

1. In total € 965000 divided by 340 residents in VUmc

2. Pediatrics 0,5 fte (21,7 residents), gynecology 0,3 fte (14,6 residents) costs based on GPL/yr. 3. Costs of 30 hrs/yr for 5 staff members, divided by # residents. Costs based on GPL/hr

4. GPL secretary € 39,780/yr. Pediatrics 0,8 fte and Gynecology 0,4 fte, divived by # residents

(In)direct patient care activity pediatrics gynecology a. Attendance of staff at shift hand-over 0,39 0,42 b. Attendance of staff at grand rounds 0,48 0,34 c. Attendance of staff at multidisciplinary discussion 0,24 0,41

d. Bed side teaching 1 1

e. Feedback and coaching of residents 1 1

f. Resident related administration 1 1

g. Post graduate education of teaching competencies for staff 1 1 h. Teaching local, regional or national residents' classes 1 1

(26)

To extrapolate the answers of the respondents to all staff members activities for residents training, the total staff fte’s was divided by the fte’s represented by the staff respondents of the questionnaire (factor 1). Subsequently the result was multiplied with the number of workable weeks per year (52 minus 6 weeks holiday minus 2 weeks post graduate education for staff = 44 weeks) providing the number of hours per year for resident teaching/training. This was multiplied with the hourly GPL and the result was divided by the total fte’s of residents to obtain the variable costs of staff provided teaching and training activity per resident per year.

The tables V and VI show the aggregated results from the questionnaires assessing pediatric and gynecology staff activities for residents’ training, respectively.

Table V. Costs of pediatric staff activities for residents’ training (per resident per year).

Legend

# resp = number of respondents hrs/wk = hours per week for this activity

factor 1 = correction factor to extrapolate results from respondents to total staff activity

factor 2= correction factor adjusting for allocating activity to both patient care and residents training wks/yr = number of workable weeks per year

GPL = Gross Personnel Cost (average cost per hour of staff activity)

/21,7 = dividing by number of residents in pediatrics to obtain costs per resident per year (2015)

Teaching related activity #resp hrs/wk factor1 factor2 *wks/yr *GPL (€/hr) /21,7 Category

Direct patient care/bed side teaching 19 18 2,1 1 44 106 8124 8124

Shift hand over 21 54 2 0,39 44 106 9053

(grand)rounds 21 64 2 0,48 44 106 13205

Multi disciplinary consultations 21 41 2 0,24 44 106 4230

Subtotal indirect patient care 26488

Resident feedback and coaching 21 28 2 1 44 106 12036

Postgrade education teaching skills 17 17 2,2 1 44 106 8038

Training related administration 22 13 1,9 1 44 106 5309

Residents' classes teaching sessions 21 4 2 1 44 106 1719

Subtotal non patient related 27103

(27)

Table VI. Costs of gynecology staff activities for residents’ training (per resident per year).

Legend

As in preceding table

/14,7 = dividing by number of residents in gynecology to obtain costs per resident per year (2015)

Subsequently the total yearly costs per resident in training can be calculated by adding the fixed and variable costs, which is presented in table VII. Also the mean costs for residents training in the Center for Women and Child Health Care of VUmc have been calculated.

Table VII. Total costs of specialty residents training (per resident per year).

Direct observations

The observation of teaching/coaching activities of supervising staff and the daily activities residents are shown in the tables in Appendix 1 (Pediatrics) and Appendix 2 (Gynecology). Underlying data recordings are available for review. The activity categories were similar as in the questionnaires. Results from the observation study are supplementary to but not directly comparable to the results from the questionnaires. Whereas the questionnaires reflect the training efforts of the average staff member, the direct observations provide an in-depth view of the above-average training activities by a limited number of assigned and rotating supervising staff.

Teaching related activity #resp hrs/wk factor1 factor2 *wks/yr *GPL (€/hr) /14,6 Category

Direct patient care/bed side teaching 14 16 2,14 1 44 89 9184 9184

Shift hand over 14 51 2,14 0,42 44 89 12295

(grand)rounds 14 23 2,14 0,34 44 89 4489

Multi disciplinary consultations 14 27 2,14 0,41 44 89 6354

Subtotal indirect patient care 23137

Resident feedback and coaching 14 26 2,14 1 44 89 14924

Postgrade education teaching skills 14 22 2,14 1 44 89 12628

Training related administration 14 23 2,14 1 44 89 13202

Residents' classes teaching sessions 15 4 1 1 44 89 1073

Subtotal non patient related 41826

Costs of teaching related per resident per year 74147 74147

Pediatrics Gynecology

Fixed costs 116279 120038

Variable costs 61715 74147

(28)

Work time of couples of one resident and one supervisor on 5 work floors of the Department of pediatrics activities were registered on representative days including samples from medium care wards, Neonatal Intensive Care Unit (NICU) and the outpatient clinic (including A&E). In total 13 working days of couples of resident and supervisor in pediatrics were observed. Because in the current project, the focus is on costs of resident’s training, only the time of teaching and training activities by the supervisor/staff are reported. In a subsequent project, where efficiency of residents’ training may be addressed, the results of residents activities become relevant.

On an average working day, the supervising staff in the department of pediatrics spent 5h18min on resident training related activities including (in)direct patient care and non-patient care related training activities. The range varied from 3h45min on the ward to 8h15min in the NICU. See also Appendix 1.

Work time of couples of a resident and a supervisor on 3 work floors of the Department of

Gynecology (wards, delivery suite, outpatient clinics) were registered on 6 representative days. On an average working day, the supervising staff in the Department of Gynecology spent 4h26min on resident training related activities including direct, indirect patient care and non-patient care related training activities. The range varied from 2h30min to 7h15min. See also Appendix 2.

The results regarding indirect patient care, based on the observation study, are shown in table VIII.

Table VIII. Supervisor’s total time and indirect patient care time for residents’ training

Staff time for Indirect patient care Total time residents training

mean range mean range

Pediatrics: 2:47 1:45 – 4:05 5:18 3:45 – 8:15

Gynecology: 2:40 1:00 – 4:25 4:26 2:30 – 7:15

Process analysis, value stream map

From the perspective of value stream analysis, particularly indirect patient care takes relatively much time, as can be concluded from these tables. This means that qualified supervision of 1 resident to provide high quality patient care in combination with residents training in the Departments of Pediatrics and Gynecology requires 4-5 hrs. of consultant’s time per day. This equals about half a normal specialist’s working day (9-10h) for each work (“shop”) floor where residents are trained in both departments. Expressed in financial terms, 0,5 fte staff per floor means the following:

Pediatrics: Calculation based on 5 floors, Pediatric staff GPL and 21,7 residents:

0,5 x 5 x 194,318 / 21,7 = 22,387 per specialty resident per year for staff providing supervision only. Gynecology: Calculation based on 4 floors, Gynecology GPL and 14,7 residents:

(29)

Hence on average roughly half the time that supervisors spend on resident’s training on the shop floor comprises indirect patient care and wide ranges were found for this category of activities. Hence particularly this category of activities should be addressed in subsequent steps aiming at improving the efficiency of residents training programs. Optimization by reduction of staff time used for indirect patients care, while maintaining the quality of the training, may offer opportunities to reduce costs.

Probably the most important factor for the allocation of staff time to resident’s training are the regulations and conventions regarding the (mandatory) attendance of staff at specific activities for resident’s training in the category indirect patients care. National resident’s training regulations mandate the attendance of (all) staff members e.g. at shift hand overs. With about 40 staff members in the department of pediatrics and about 30 in the department of gynecology, in reality this is often not the case. Although not a deliberate policy, this limited attendance at mandatory sessions is a hidden contribution of staff reducing time for indirect patient care. Particularly in smaller teaching hospitals where there may only a few residents and the number of staff is a multiple of the number of residents, the efficiency of attendance of the plenary staff at primary teaching activities may be disputed.

In addition to national requirements for all medical specialties, also specialty specific and/or hospital specific conventions regarding the staff participation in residents’ training activities (often in the category indirect patient care) vary and in some respect may be regarded as arbitrary. Although the time (and costs) connected with these regulations and conventions are often taken for granted and remain obscure, the current project helps to uncover these issues, and their quantity appears to be substantial. Critical discussion of these activities on local and national level may offer opportunities to reduce costs while preserving or improving the quality of residents’ training programs.

(30)

Chapter IV. Results and reflection

In this chapter, firstly the results of the content of the project will be discussed and secondly a reflection on the process (including project management and deliverables) of the project will be presented.

A. Results and discussion

In this project an in depth assessment of costs for residents training in the Center for Women and Child Health of VUmc was performed and compared with 2 similar analyses. For direct comparison, the most important results are summarized in table IX. In the last column the results of a recently obtained unpublished internal study from the department of pediatrics of the UMC Radboud, Nijmegen have been added.32 Of note is that in this department, where patient care and residents training have been thoroughly separated, the costs of specialty residents training are much higher.

Table IX. Summary of fixed, variable and total costs of current study, Berenschot, Robijn, Peds UMCN.

The costs that were found in our project are substantially higher than in Berenschot and somewhat higher than in Robijn, the latter is mostly due to higher fixed costs. Using an accepted model for cost calculation, we derived our results from an in-depth study of variable costs, based on input from a broadly distributed questionnaire. Moreover the results were obtained recently, i.e. at a point in time when all (time consuming) revisions due to the modernization of residents training programs during recent years have been implemented. We therefore infer that the results from our study are a more robust basis for further analysis and improvement actions than Robijn and Berenschot.

The key common conclusions from the three studies (Current project, Berenschot and Robijn) are: 1. Costs and accountability are highly affected by hospital specific factors, e.g. conventions for

allocation of funds to internal budgets for residents training related costs and expenses. 2. The highest costs of residents training are variable costs, incurred by

Pediatrics Gynecology Berenschot Robijn Ped UMCN

Fixed costs 116279 120038 98763 102000 123606

Variable costs 61715 74147 41762 64000 115815