Planning simulation space needs in uncertain contexts

Abstract

Introduction

King’s College London delivers simulation-based education to approximately 5500 different students who use a Simulation and Interactive Learning (SaIL) centre which was designed for significantly different needs. The management team needs to plan for current and future demand in the context of changing regulatory and curricular requirements, and changing technologies. The authors developed a model to address this challenge.

Methods

We used a structured approach involving university data and interviews with module leaders. Assumptions were needed on target student numbers, group sizes and time spent in a simulation centre. The algorithm at the core of the model applies the targets and assumptions to a profile of room use by type of room, as experienced by a typical student over the course of one year. Individual course results are summed and rounded up, to provide an overall demand for space which can be compared to capacity.

Results

The number of SaIL rooms required is expected to almost double by 2022/2023. The mix of spaces must also change to enable more simultaneous immersive simulation of diverse types, with associated debrief spaces also increasing.

Conclusions

This model helps maximise impact by creating connectivity between a vision, business targets and the physical space. Financial resources are used more effectively by avoiding potential over or underprovision. Flexibility of build will remain important, as no-one can predict the future; however, a detailed, quantified model can help raise the quality of discussion about future aspirations and how closely we might meet them.

Introduction

King’s College London (KCL) and its partner Guy’s and St Thomas’ NHS Foundation Trust (GSTT) are developing a new education centre at St Thomas’ Hospital. The issue facing both organisations is how to plan for current and future demand in the context of changing regulatory, educational and curricular requirements while taking into account capital budgets and the constriction of refurbishing a listed building.

In its own right, KCL delivers simulation-based education (SBE) to a wide range of healthcare students, with around 5500 different students using our two Simulation and Interactive Learning (SaIL) centres. To facilitate the spatial brief, the authors were tasked with defining the requirements for the mid 2020’s for the simulation and clinical skills element of the new education centre within an uncertain context. Changes to nursing curricula following new Nursing and Midwifery Council (NMC) requirements and the need to forecast student numbers show the level of uncertainty faced when defining the requirements of a £45 m building project planned for several years ahead.

Our work was to define the spatial requirements to allow the design team to design the new centre, and, concurrently, define the expansion of the current centre on Guy’s campus to allow for increases pending the building of the new centre. The model helped underpin the spatial needs with good quality supporting information, and through sensitivity analysis understand the range of possible outcomes to consider.

The General Medical Council requires medical students to be competent in 32 practical procedures by the time they graduate.¹ To become a registered nurse, the NMC requires student nurses to have completed 2300 hours of clinical practice. Until recently, up to 300 of these hours could be delivered by the university in simulation and contribute to the practice learning component of the programme.² Recently the NMC guidelines have removed this limit to recommend that ‘simulation-based learning opportunities are used effectively and proportionately to support learning and assessment’.³

While the use of simulation in healthcare education is well established, another level of uncertainty going forward is the rapid development of technology, such as virtual, augmented and mixed reality simulation and how they can be used in the future. These have the potential not just to provide students with state-of-the-art technological developments for skills acquisition, but also to provide them with exposure to situations where there are multiple options on how to respond, often as one of a team, moving away from a learning pedagogy that centres on knowledge appropriation, towards developing receptiveness to complex and emerging situations with inherent uncertainty: Simpkin and Schwartzstein, 2016.⁴ It is proposed by Campbell et al ⁵ that future learning spaces will blend traditional classroom and innovative virtual learning environments and that developing technologies offer ‘amazing possibilities’ as a learning tool ‘to bridge the knowledge and skills gap to the next learning goal’.⁵ However, the level of unknown makes it difficult to incorporate the effect of this on spatial requirements for a simulation centre. The approach taken by the authors was to recognise the university strategy to promote a blended learning approach, and focus on expectations about the need for physical space, which we do not expect to be significantly reduced by the introduction of virtual reality (VR) technology under development within the time frame of this study. The training of future health professionals cannot be fully replaced by VR, because physical interaction is required. Rather, we envisage that the use of augmented or mixed reality approaches may enable the university to make fuller and more flexible use of the physical space in the centres and other resources that are geographically disparate.

VR technologies also have the potential to improve assessment strategies, and could help to meet the university’s strategy to ‘shift our assessment focus to assessment for learning, away from assessment of learning’.⁶ However, Glowatz et al ⁷ suggest that there is still work required to develop effective assessment using virtual environments, and is outside the purview of the authors’ remit.⁷ For spatial requirements, our principal concern was to understand how much time overall a student may be expected to attend a SaIL centre, be it provided by KCL or by GSTT, to achieve the requisite training and assessment.

The authors have developed a model using a structured approach involving data analysis, interviews with module leaders and university targets. The premise of the algorithm at the core of the model applies targets and operating assumptions to a profile of room use by type of room, as experienced by a typical student over one academic year on a given course.

Methods

The methods described in this paper cover the specific scope of simulation and clinical skills education in medicine, nursing (adult, child and mental health), midwifery and the allied health professions as taught at KCL and Guy’s and St Thomas’s Hospital. The scope excludes requirements for simulation in dentistry for which separate facilities are provided, and for in situ simulations as undertaken in a real hospital setting.

Data gathering

We held interviews and meetings with senior management, education leaders, clinical tutors, module leaders, teaching and technical staff within both organisations to elicit ambitions, targets, descriptions of preferred ways of teaching their material and where possible, quantified data. In total 22 individuals contributed, in addition to the authors.

Interviews were unstructured, with questions typically covering the following:

• A request for quantified data, if available, covering time-tabled hours per student per year that requires SaIL space, including any planned increases over the next 5 years.

• Description of any new courses planned that are not currently taught, with intended start date, target intake, course duration.

• Local strategy for SBE and the impact of any proposed curriculum change, for example, on the scale and volume of what must be taught or learnt in a specialised SaIL centre.

• Description of specific modules and how each is delivered—the class size, number of classes learning simultaneously, the hours of SaIL each student receives.

• In-depth look at example sessions within the modules to understand to how space is being used, the types of space required and how groups rotate between or within rooms.

• Elicit from the conversation whether this is optimal or would ideally be different if space was not constrained.

• In the future, will the module be delivered differently and if so how?

Data analysis

The interview process provided rich descriptive information on the variety of ways that SBE is delivered at module level. However, quantified information for modelling was incomplete owing to a combination of time limitations, the devolved nature of some information, and the future curriculum for nursing being at an early stage of development. We were asking novel questions that could not always be answered, requesting interviewees to consider the simulation aspect of their teaching in fresh ways that differed from the normal approach.

Therefore, we supplemented the interviews with a historic analysis of a full 12 months in the Chantler SaIL Centre at Guy’s campus. During the year 2017–2018, the centre handled 5045 room bookings with an average booking duration of 4.5 hours. Every booking was reviewed and based on the interviews and information from centre technicians, was assigned a room type, assuming the ideal type of room for the session, rather than it being dictated by limited availability.

The types of spaces, and a description of each, is in table 1.

Table 1.

Types of rooms required, identified during interviews and analysis

Type of room	Functional description
Single bay simulation and control room	Used for full-patient simulation. Requires use of control room for technician, manikin voice. High-quality audio-visual capture, live viewing and recording.
2 bay ward	Used for full-patient simulations and clinical skills practice. Access to a control room. High-quality audio-visual capture, live viewing and recording.
4/6/8 bay ward	Used for simulations and clinical skills practice. Regular ward fit out including suctioning, oxygen and nurse’s station. Range of equipment to accommodate adult, maternity, child and neonatal scenarios. Some cameras over beds required for demonstrations and in simulation use. Not appropriate for seminar-style teaching.
Home environment	A simulated small dwelling with living room, bedroom, kitchen and bathroom, reflecting typical living conditions. Alternatively, a simple sitting room or a flexible space that could represent another type community setting.
Simulation debrief	A non-clinical space with tables and chairs in formal or informal layout. Used for the debriefing part of the simulation. Observation of the live simulation takes place here by other students in the group.
Mandatory skills	Large clinical teaching space, with hand washing and infection control equipment. Hoists, beds, couches. Good circulation space. Floor work required. Chairs and audio-visual equipment also needed. Camera and screen for demonstration purposes. Not used for simulations. The model confined ‘mandatory skills’ to health and safety training such as moving and handling, and basic life support, for internal practical space-planning reasons.
Consultation room/general practitioner setting	Used for simulations and clinical skills practice. Desk and computer, patient chair, patient couch, sink and infection control equipment. Comfort of room and privacy are important as consultations may involve real patients in an educational role.
Other interactive learning and skills	Also known as skills laboratories. Rooms used for clinical skills practice and simulated scenarios that do not require a ward environment because of the way they are taught. Must have look and feel of a clinical environment. Hand washing and infection control equipment. Benchtop skills, floor work, couch work, pair work. Groups can be seated watching demonstrations before relocating to other rooms to practise. May run circuits of skills in different rooms. Used for communications, for example, actor and student simulates a history taking with others watching in a semicircle. Very flexible spaces.
Outpatient environment	A simulated clinic containing at least a waiting area and consultation room. Some specialist equipment may be present. Optionally include a check-in or reception desk. High-quality audio-visual capture, live viewing and recording
Seminar room (excluded)	Standard seminars take place in standard teaching facilities. If the seminar requires demonstrations using Simulation and Interactive Learning equipment, it may take place in a skills laboratory.

The historical data from the Chantler SaIL Centre contained no information on group size, which is required to plan future space. Based on working knowledge and interview information, an average group size was assigned to each room type, which allowed the authors to uncover underlying patterns of room use. A key facet of the model was development of the generic student profile based on the following formula

$${\displaystyle {\mathrm{h}}_{\mathrm{c},\mathrm{t}}={\mathrm{B}}_{\mathrm{c},\mathrm{t}}\ast {\mathrm{g}}_{\mathrm{c},\mathrm{t}}\ast 1/{\mathrm{S}}_{\mathrm{c}}}$$

Where

• h_{c, t} is the number of hours a student spends in a room of type t on a course c over 1 year.

• B is the historical booked annualised hours of room type t on a given course c.

• g is the group size for that course in the room type.

• S is the historical number of students on the course.

This enabled a profile of use—as experienced by a typical student—to be reverse-engineered for each analysed subgroup: typically one department, one course or, for the medicine degree, one stage. The calculated profiles were simplified to half-hours to reflect the realities of university timetabling and converted into a distribution of room use adding to 100%.

We validated or revised the historic profiles using information from strategy and interviews, to represent a possible scenario for the implementation of changes. Typical changes might include, for example, a greater focus on community care, more time dedicated to full-patient, team-based simulation with associated debrief and more time spent in a simulated ward environment.

Data for figure 1	Historic, %	Future, %
Single bay simulation and control room	3.0	2.5
Ward simulation (2 bay)	0.0	0.0
Ward simulation (4 bay)	4.5	5.0
Ward skills (8 bay)	1.5	13.3
Home environment	3.0	3.3
Simulation debrief	15.2	20.8
Mandatory skills	15.2	4.2
Other interactive learning and skills	57.6	47.5
Small consultation rooms	0.0	1.7
Outpatient environment	0.0	1.7

Figure 1.

Historic and future assumed profiles of room use for BSc nursing courses. BSc, bachelor of science.

In figure 1, the future profile represents more time spent by students in a greater diversity of room types, while mandatory skills hours remain fixed in absolute terms, thereby reducing as a percentage. The higher use of ‘ward skills’ rooms in future reflects our strategic intention to deliver clinical skills training in a representative hospital environment, around the bed, replacing some use of the ‘other interactive learning and skills’ rooms. These assumptions change the profile noticeably, demonstrating the value of the approach taken.

Development of model

The assumptions required, all of which can be flexed in the model, are centre operating hours, room frequency (used hours per week), number of teaching weeks per year, group sizes and area in square metre for each room type. University space use is defined in terms of frequency (% of time a room is used compared with operational hours), occupancy (average per cent rate of occupation compared with a room’s maximum capacity) and utilisation, defined as the product of frequency and occupancy as a percentage. Evidence suggests that best practice frequency is in the range of 65%–70%.⁸

The model enables adjustments to the frequency to allow for room set-up, footfall between classes and the ‘peakiness’ of the timetable. To understand the effect of operational hours of a centre on a spatial basis, we stress tested the existing operating paradigm against an extended paradigm.

One of the concepts of the model was to base the room areas around group sizes to ensure an appropriate occupancy and enhance the student experience in the planned centre. Using work that was simultaneously being carried out on other parts of the KCL estate, the authors were able to confirm group sizes to an area with an architectural practice.

The inputs used by the model are student numbers, hours per student and profiles of use found during data capture and analysis. These reflect the business targets set by the university, and the interpretation of education and simulation strategies explored during interviews. Our objective was to determine the total demand for space, across all relevant SaIL centres, to understand the additional capacity requirement that must be incorporated into the new build. The model assumes that teaching staff will be available as required to meet the targets set by the participating faculties; we did not model a constraint on teaching workload or resources.

The algorithm in the model was used to determine space use for each course and room  type. The individual course results were summed, and rounded up to provide an overall demand for rooms of each type, which can be compared with capacity. The relevant calculations are:

$${\displaystyle {\mathrm{R}}_{\mathrm{t},\mathrm{c}}={\mathrm{H}}_{\mathrm{c},\mathrm{t}}\ast 1/(\mathrm{T}\ast \mathrm{O}\ast \mathrm{f})}$$

where

• R_{t, c} is the total number of rooms required of type t on course c.

• H is the total annual hours that course c needs rooms of type t.

• T is the number of teaching weeks available.

• O is the operational hours per week.

• f is the frequency of room use expressed as a number between 0 and 1.

The total annual hours H is determined by

$${\displaystyle {\mathrm{H}}_{\mathrm{c},\mathrm{t}}=({\mathrm{p}}_{\mathrm{t},\mathrm{c}}^{\prime }\ast {\mathrm{h}}_{\mathrm{c}}\ast {\mathrm{S}}_{\mathrm{c}}^{\prime })\ast 1/{\mathrm{g}}_{\mathrm{t}}^{\prime }}$$

where

• p_{t, c} is the revised profile element of room use for the type t of room on a course c.

• h is the number of hours spent per student per year in all centres on a given course.

• S is the input future number of students on the course.

• g is the modelled average group size expected for that type of room across all courses.

The space requirement is determined by rounding up the resulting number of rooms, for each type of room, and applying the appropriate area in square metres. In principle, rounding may be carried out at different levels of granularity within the model according to the physical or organisational separations of space. In this instance, the concept is for a shared facility between KCL and GSTT, to optimise the efficiency of space use, so rounding was carried out as the final step.

Results

To validate the accuracy of the model, the authors created a baseline scenario using existing student numbers, current profiles of use (as determined above) and operating paradigms. The results confirmed existing suppositions and experience that KCL had the correct number of rooms, although not necessarily of the correct types, and this is shown in table 2.

Table 2.

Comparison of existing space and modelled demand for the current scenario

Simulation space type	Existing King’s College London Space	Modelled rooms, rounded	Variance
Single-bay sim and control room	2	2	0
Ward simulation (2 bay)	2	1	−1
Ward simulation (4 bay)	0	2	2
Ward skills (8 bay)	2	1	−1
Home environment	1	1	0
Simulation debrief	0	3	3
Mandatory skills	2	2	0
Other interactive learning and skills	14	9	−5
Small consultation rooms	0	1	1
Total	23	22	−1

Over a period of 6 months, the authors provided different options to assess the impact of the different variables. What became clear, as revised curricula were developed, was the significant impact in the changing profiles of the students on spatial requirements.

With the development of the agreed scenario, the results are shown in table 3, which shows that the number of SaIL rooms required is expected to almost double by 2022/2023, compared with current availability. The mix of spaces must also change to enable more simultaneous immersive simulation of diverse types, with associated debrief spaces also increasing.

Table 3.

Comparison of existing space and modelled demand for the agreed scenario

Simulation space type	Existing King’s College London Space	Modelled rooms, rounded	Variance	Modelled m2
Single-bay sim and control room	2	3	1	25
Ward simulation (2 bay)	2	1	-1	−50
Ward simulation (4 bay)	0	2	2	140
Ward skills (8 bay)	2	3	1	125
Home environment	1	2	1	20
Simulation debrief	0	7	7	315
Mandatory skills	2	2	0	0
Other interactive learning and skills	14	12	-2	−90
Small consultation rooms	0	9	9	225
Outpatient environment	0	1	1	60
Surgical sim	0	0	0	0
Total	23	42	19	770

The sensitivity analysis shows the fundamental importance of the profile of use in determining overall requirements (table 4). The base case profile represents our central scenario, after the expected curriculum change has taken place. Changes in curriculum provision are modelled by adjusting two inputs: the hours per student and profile of use. Full patient (high fidelity) simulation with debrief afterwards, and small group work, is demanding in terms of time, space requirement and staff workload: sensitivity case 1 shows the effect of limiting this kind of simulation. In case 2, the overall hours per student were increased by 10%–20% which for some room types can be absorbed in the same provision (since the base case requirements must be rounded up to whole rooms) leading to a relatively modest increase in space. Cases 3 and 4 show how potential diversions from expectation could, in principle, be managed by changing the operating protocol. Future changes in technology are likely to require consideration of the fit-out of space, rather than the demand for space itself, in the timescale of this study. It is possible that students may undertake some virtual forms of simulation off-site, reducing pressure on the centres’ physical resources by limiting the time spent per student in the physical centres. However that would be a secondary effect overlaid on the basic finding that levels of simulation are increasing and physical space for that is very much in demand.

Table 4.

Sensitivity test results for total modelled rooms in different scenarios

Sensitivity case	Student numbers	Base case	+25%	−25%
0	Base case profile	42	48	32
1	Lower student hours and less sim+debrief	26	32	20
2	Higher student hours with new profile	44	55	34
3	Room frequency reduced by 10% points	47	58	38
4	Extended operating hours	36	43	27

Discussion

The model has enabled more informed decision-making by delivering initial results through a reliable approach that can be easily updated for future use as information is refined, and by exploring ‘what-if’ scenarios and sensitivity analysis. The principal finding is that the university’s simulation space requirement, in terms of room numbers, is expected to nearly double over the next four to 5 years. The forecasts assume a relatively high frequency of room use, with smooth demand across the year, on the premise that building to the maximum peak demand would not be cost-effective. However, there is variability in terms of peaks of demand within the base case demand that will need to be managed. The model does not include a technique to allow for the day-to-day variation in requested roommix, which must, therefore, be managed flexibly through the design response of the new build, and through a robust timetabling process. While the model provides an understanding of types of space and area, it does not look at adjacencies and other factors to allow the seamless operation of a centre, and this would require more analysis as part of the detailed design of the new centre.

The work to date helps maximise impact by creating connectivity between a vision, business targets and the physical space. The model has helped to develop a clear understanding of the current situation and its limitations, and helped the leadership team plan the impact on college infrastructure. Having been through the process of developing the model, it has allowed the university senior management team to have more confidence in prioritising capital investment, with financial resources being used more effectively by avoiding potential over or underprovision.

It is beneficial to separate the planning and delivery functions to ensure the requirement, and project response to it, are clearly differentiated. The model enables teaching staff to understand how their strategy and business targets drive space demand, and how the planned, preferred styles of SBE delivery determine the mix of rooms that will be required. However, while the model has defined the number of rooms and the types of space to inform the spatial brief, it is important for the design team to ensure that the space is built with flexibility to allow future proofing as there will no doubt be unpredicted demands from changes in technology, curricula and the pedagogical delivery of clinical skills and simulation training.

The process of developing this model has enhanced the quality and depth of discussion between teaching staff, centre management and the estates department over future requirements. It has allowed a quantified brief to be developed that is clearly connected to future direction, maximising the impact of the role of SBE in the future.