Building a Clinic Prototype You Can Use Again
Write your clinic prototype as a document set with one named owner, a version number, and a written exception process. Fix the room module (dimensions, casework, power and data locations) and let the arrangement of rooms adapt to each building. Publish v1.0 when site 1 opens, revise it after the 90-day review, and require a priced cost delta before you approve any exception.
By site 3 you have a prototype whether or not you wrote one. It lives in the head of the person who ran sites 1 and 2, in the last architect's CAD file, and in what your clinical director remembers wanting to change. That version holds until that person leaves, you hire a different architect, or you design two sites at once. Then the newest permit set becomes your standard by default.
I have watched this on multi-site builds where the design was good and nobody owned it. Site 1 got three years of operating refinement. The site 2 architect received a copy of the site 1 permit set and no record of which dimensions were deliberate. Site 3 opened with a room layout divorced from decisions made on Site 1 and Site 2. Each site inherited the previous one's compromises and added a few because nobody owned a change log. Fix the process by deciding who owns the standard, what is allowed to change, and what a change costs. Better design will not fix it. Here is how to set that up.
What is a clinic prototype, mechanically?
A package of documents. The floor plan is one output of the package and the least portable one. A given plan is shaped by one column grid, one window line, and one lease outline. When you hand next architect the plan, you've given them a suggestion. Hand them the package with context on how the prior sites came together and you give them constraints they can design around.
Produce the package from site 1. Test it on site 2. Do not make the mistake of "finalizing" (its never really done) your prototype package before you have operated a clinic. You will fill it with assumptions you cannot defend, and you'll end up with a new site that doesn't account for what you've learned before.
- A room data sheet per room type. Exam, consult, treatment, lab draw, group or activity room, provider workspace. For each: dimensions and clearances, door swing, power and data locations, plumbing, medical gas if any, casework, the equipment list, and finishes.
- A typical-room drawing set. Plan, reflected ceiling, elevations, and details at a scale the architect can drop into any plate. This is the document that travels between firms.
- An equipment and FF&E schedule with model numbers. Not categories. The model determines the rough-in, the clearance, and the lead time.
- A finish and materials standard. Flooring by room type, wall protection, base, ceiling, paint, and the cleaning protocol each finish has to survive.
- An MEP basis of design. Air changes and exhaust for the rooms that need them, power and cooling for imaging or lab equipment, dedicated circuits, and the load assumptions the next engineer starts from. On metabolic and longevity projects, body composition scanners, cold plunge and sauna equipment, and IV suites drive real infrastructure. On a PACE center it is exhaust, hot water, and the kitchen and therapy loads.
- A low-voltage and IT standard. Drop counts and locations per room type, wireless coverage, check-in kiosk power and data, and where the rack lives.
- A signage and wayfinding standard. Type, mounting height, language, and which decisions are corporate and which are site-specific.
- A change log. One page. Version, date, what changed, who approved it, and why.
Do not skip the change log. Without it, version 3 of your standard is a document nobody can explain, and your GC will win the first argument about why a dimension is what it is.
What do you standardize and what do you let the building decide?
Standardize what repeats and what your staff depend on. Start with the exam or consult room module, and go past square footage to the internal layout: where the provider sits relative to the patient and the door, which side the sink is on, where the computer goes, whether the exam table runs parallel or perpendicular to the wall. Your staff build muscle memory around that arrangement. Flip it at the next site and you add a half second to every reach, dozens of times a day, for the life of the lease.
Standardize the casework and equipment package, the low-voltage layout, the finish palette, and signage. Repetition pays you in each of those: the GC prices a known assembly, your architect's fee on repeat work typically drops 20 to 30 percent because the typical-room set already exists, and your equipment vendor quotes from a schedule instead of a wish list.
Let the building decide almost everything above the room. The arrangement of rooms, the size and shape of the waiting area, staff support and break space, back of house, the exterior and entry sequence, and anything the column grid, the window line, or the existing restroom core dictates. If your standard fixes the arrangement, you will reject good buildings. I have seen an operator pass on a well-priced second-generation medical suite because it would not accept the site 1 plan, when the module fit fine and only the order of the rooms had to change.
Write a rule for the third category: items that need a decision, not a fixed answer. Room count is the clearest example. Set exam room count from the market model, the panel size, and the staffing ratio, not from what site 1 happened to build. Same for specialty rooms. Write the rule (clinical ops sets room count from the volume model and approves it before the test fit) so the right person answers the question instead of whoever is drawing that week.
Treat the prototype as a governance document that contains drawings. You are paying for one thing: a decision made once, by the people who had the operating data, still applies at site 4 when those people are busy and the architect is new. The drawings carry the design. The change log and the exception process carry the reasons.
Who owns the standard and who approves exceptions?
Name one owner. In a two to six site operator, make it your COO or clinical operations lead, because the standard exists for operational reasons and the tradeoffs are clinical. Make your real estate lead the keeper of the documents: version control, distribution to each architect and GC, and the change log. That split puts the person who cares about throughput in the decision seat and the person who cares about schedule in charge of the record.
Version it on a rule anyone can apply. Publish v1.0 when site 1 opens. Issue v1.1 after the 90-day review, which is usually small corrections. Move to v2.0 when a change alters cost per square foot or throughput, because your CFO and your clinical leadership both need to see that one.
Write the exception process in four lines. Who can request one: the architect, the GC, or the site clinical lead. What they have to bring: the building condition that forces it, the cost delta in dollars, and the operational effect. Who approves: the owner, in writing, inside a stated number of business days, because a slow approval becomes a field decision. And the closing rule: an approved exception either becomes a change to the standard or stays a documented one-off. Do not let it sit in between. That is how standards erode: the site 2 architect gets a verbal okay for a room 6 inches narrower because of a structural bay, nobody records it, and the site 3 architect measures the site 2 drawings and treats the narrow room as the standard.
No owner, no version
- The newest permit set becomes the standard by default.
- Whoever is on the call grants exceptions verbally in the field.
- Each new architect re-solves the room and bills for it.
- The GC prices every site as a first-time build.
- Your clinical staff learn a different room at every location.
Named owner, versioned, governed
- One person approves changes; the real estate lead controls distribution.
- Exceptions require a building reason, a cost delta, and a written decision.
- The typical-room set drops into any plate, so fees fall on repeat work.
- Subcontractors bid a known assembly, which tightens pricing.
- Operating feedback has a place to go and a date you decide it.
How do you get what staff learned at site 1 into the site 2 drawings?
Understand why this is hard before you fix it. The people who know exactly what is wrong with site 1 are your clinical staff, and they are not in the room when site 2 is designed. Site 2 design happens six to fourteen months later, often with a different architect or at least a different project manager, working from a permit set that records what was built and none of the reasons. Nobody is withholding anything. You have not built the channel.
Make the 90-day post-opening review that channel. Treat its output as input to the standard, not as a punch list. Ask one narrow question: where does what staff actually do differ from what the drawing assumed? Expect small, specific answers. The supply cart living in the corridor because the room came up 6 inches short of the clearance the cart needs. The check-in desk nobody uses because patients walk past it to the kiosk, which means you built and staffed a position the workflow does not want. The second sink nobody touches. The provider workroom that seats four and holds seven at 11 a.m.
Then do four things in order. Capture it. Price it: ask your architect and your GC what each change costs on the next set, and expect some of the loudest complaints to be cheap and some of the quiet ones to be expensive. Decide it, with the owner. Version it. Do the arithmetic while you are at it: a room dimension corrected at v1.1 costs a few thousand dollars of drawing time before site 2 goes to permit. The same correction discovered in the field at site 3 is a change order against framing that is already up, plus schedule, and on the projects I have run that gap is routinely an order of magnitude.
If you are building site 2 while site 1 is still stabilizing, you are designing site 2 before the feedback exists. Write the standard anyway. Mark every item you have not tested in operation, and put a date on when you will revisit it.
What do you give up by freezing the standard?
Freeze early and you buy repeatability: predictable pricing, faster permit sets, lower design fees, staff who can move between sites. You also lock a first-site mistake into the platform. If your exam room is 6 inches tight, freezing at v1.0 replicates that at every location until you pay to change it. Freeze late and you keep flexibility, and you pay for it in fees and schedule at every site, plus a team that re-decides settled questions.
The exception process carries the same tension. Approve fast and the standard drifts, because the easiest exception to grant is the one that keeps the GC moving this week. Approve slowly and the field decides without you, which is the same drift with worse documentation. I set a 3 to 5 business day approval window with a required cost delta. The cost delta does most of the work: a fair number of requested exceptions disappear once someone has to price them.
Look at what each end of the market does. A large system or a well-capitalized national brand runs a rigid kit of parts, buys consistency and speed, and loses deals, because a rigid standard rejects buildings that would have worked. A founder-led operator with no documentation keeps every building in play and gives up cost predictability, staff portability, and design fees at every site. For a 2 to 10 site operator, run a fixed module inside a governed exception process. That position also protects a differentiated, non-institutional design, because the things that make your space feel unlike the healthcare system people expect (light, materials, the entry sequence, how the waiting experience works) mostly live in the layer you let the building decide.
What does version 2.0 need to know that version 1.0 did not?
You have a prototype by site 3, written down or not. Decide now who owns it, what it takes to change it, and where you record the change so the next architect finds it.
Then ask your team one question: what did operating your clinics teach you that your current drawings do not know? Room count against actual panel size. Storage against actual supply volume. Provider workspace against the staffing model you moved to after go-live. Write the answers down, price them, and put a version number on the result. That document is how the people who ran site 1 reach the building you have not leased yet.
Key takeaways
- Write the prototype as a document set: room data sheets, a typical-room drawing set, an equipment schedule with model numbers, finish and MEP standards, a low-voltage standard, and a change log. The floor plan is the least portable part.
- Fix the room module and its internal layout. Let the building decide the arrangement, the waiting area, back of house, and the entry. Set room count from the volume model, not from site 1.
- Name one owner (usually the COO or clinical ops lead) and make the real estate lead keeper of the documents. Publish v1.0 at site 1 opening, v1.1 after the 90-day review, v2.0 when cost or throughput changes.
- Require a building reason, a priced cost delta, and a written decision for every exception, inside a 3 to 5 business day window. Every approved exception becomes a standard change or a documented one-off.
- Correct a room dimension at v1.1 and you pay drawing time. Find it in the field at site 3 and you pay a change order against framing already up, routinely an order of magnitude more.
Frequently asked questions
What should a clinic prototype or design standard include?
A room data sheet for each room type, a typical-room drawing set the architect can drop into any floor plate, an equipment and FF&E schedule with model numbers, a finish and materials standard, an MEP basis of design, a low-voltage standard, a signage standard, and a one-page change log. The floor plan itself is the least portable item, because it is shaped by one building's column grid and window line.
When should a healthcare operator create a clinic prototype?
Write it from site 1 after it opens, and test it on site 2. Publish v1.0 at opening, v1.1 after the 90-day operating review, and v2.0 when a change alters cost per square foot or throughput. If you are designing site 2 before site 1 has operated, write the standard anyway and mark every item you have not tested.
How do you keep a clinic design standard from eroding across multiple sites?
Name one owner, version the document, and write an exception process: who can request a change, what they must show (the building condition, the cost delta, the operational effect), who approves and how fast, and the rule that an approved exception either changes the standard or is recorded as a one-off. In Retained CRE's projects, requiring a priced cost delta removes a fair share of exception requests on its own.
Should every clinic location look the same?
No. Fix the room module and the things operations depends on: the internal exam room layout, casework and equipment, low-voltage, finishes, and signage. Let each building decide the arrangement of rooms, the waiting area, back of house, and the entry sequence. The approach I use at Retained CRE is a fixed module inside a governed exception process, which is also where a non-institutional design has room to respond to the building.
Documenting Site 1 Before Site 2?
I turn a first clinic into a prototype package you can hand to the next architect and GC, and set up the exception process that keeps it intact.
Schedule a ConversationYour standard also needs the lease to allow it: use, alterations, and delivery condition. What You're Actually Signing grades those clauses and 23 others.
