USP 800 Compliance – A Cautionary Tale

The newest guideline, with an enforcement date of July 1, 2018 (real-time countdown to USP 800 is on my home page), has the intention of protecting the operators who compound hazardous drugs. While there are negative pressure environments already utilized in the preparation of medications in commercial manufacturing (Fentanyl comes most readily to mind), if particular precautions and forethought isn’t put into the planning stage of building clean rooms that comply with USP 800 there will be a major price to pay. The cost I speak of is not only in real dollars but if the operators don’t understand the basic principles of a negative pressure environment they could also be contaminating the products they’re compounding.

While I agree with safely and properly compounding drugs that may be hazardous to those preparing them, I also know firsthand all of the stumbles and falls (and extra unplanned $ that was spent) that I went through in figuring out how to make it all work correctly. My biggest fear is that the impending deadline will cause many hospitals and compounding pharmacies to rush into trying to comply. Without knowing just a few key points those attempting to comply will end up perhaps doubling their original budget.

For the purposes of this article I’m merely going to state the facts. I’m going to give the timeline of our build out and what we subsequently did to remediate a very serious issue we had with our clean room. Full disclosure: I’m a pharmacist, not an engineer. I had MANY consultants and a lot of help to get to the point of complete remediation.

When we contacted our clean room designer we asked them to design the rooms with the goal of trying to also come in line with what was known about USP 800 at the time (find out why we started this whole project here). They presented us a design that included a positive pressure ante room, a positive pressure buffer room, a negative pressure hazardous buffer room and a separate non-sterile hazardous area – pictured below.

The clean room engineers had given us a complete plan of what to install for the the air handling for the room and a complete air flow diagram (complete plan will not be published); the plans were actually very detailed. Everything was followed by our subcontractors per the plans given to them by the engineers. Our clean room company came and built the room and our HVAC contractor connected the clean room to the air handling unit and exhaust system.

Something to keep in mind about our build in particular: We’re based in Kentucky, in the summer it is hot and HUMID. This is a key point to our build because another goal was to maintain humidity levels below 45% degrees and sub 68 degree temps.

The ceiling to the rooms were caulked twice and the room was triple cleaned (twice with vestacyde followed by sporklenz). We felt like we were ready for the room to be certified. Our initial certification was 10/27/15. During this initial cert both total particle counts and viable air sampling was performed and in all 4 locations of the hazardous buffer area there were multiple colonies of growth observed when incubated; cladosporium, penicillium, rhodotorula, yeast and arthrospore-former. We also noticed similar growth (of the same species) in two of the locations in our ante room (near sink and entrance to the negative pressure room).

Given the picture above my initial thought was that we had bacterial growth in the sink and since the negative pressure room was directly adjacent to the ante room, this MUST be the source. OH, I wish it were that easy. Sit tight.

(NOTE: I’ll stick to my day job of NOT being a microbiologist, thank you)

From my notes documenting this QA event:

AeroMetric Air Sampling results from Agape instruments concluded there was bacterial growth in both the Ante and Hazardous Sterile clean rooms. See attached results for details. An investigation into the possible contamination has been launched. The compounding procedures will be resumed after a thorough cleaning and recertification has been completed.

There are a few possibilities that could’ve led to the contamination.

  1. The construction of the clean room is such that our positive pressure ante room is located directly adjacent to our negative pressure hazardous ISO 7 room. Therefore, if there is any contamination in the ante room it will be pulled into the ISO 7 room with the negative pressure.
  2. It has been noted that our ante room positive pressure averages on the lower side and it is possible that there is not enough positive air flow to the outside of the clean room. This is a possible area for adjustment.
  3. Our cleaning procedures will be reviewed with the technicians who perform the cleaning. One area that will be noted is the amount of contact time for our specific cleansers. It is possible that when the triple clean (2x clean with disinfectant followed by sporicide) was performed the contact time for the cleanser and sporicide were not reached and therefore not maximally effective.
  4. We have also reviewed proper dilution of quat as this was another possible area that we were not correctly performing.

On 11/9/15 another triple clean will be performed on our entire clean room and (certifying company) will be scheduled to re-sample our clean room areas with issues to ensure that the problem has been resolved.


We scheduled another cert and on 11/18/15 they came again to sample the locations in the ante and hazardous buffer room. Again, we saw growth in both rooms. Growth was found by the entrance to the ante room (NOTE: the growth was not found here the last time but in the other two locations previously mentioned) and in every location in the negative pressure room. At this point I’d like to say that while this testing was done under dynamic working conditions, there isn’t much IN these rooms. I’m not one for storing really anything in my rooms in terms of supplies. I take what I need into the rooms only as I need it and bring everything out after compounding is finished. That means I do not even have sterile syringes, needles, drug packaging…nothing in my rooms except except excipient bottles and a small amount of equipment (hot plate/stirrer, peristaltic pump and homogenizer – and hoods of course).

To quantify what we saw this second sampling:

  • 4 CFU of Epicoccum in 1 location of the ante room
  • Location 1 of Hazardous buffer: 2 CFU bacillus, 17 CFU micrococci, 1 CFU streptomyces (bacteria), 1 CFU Alternaria species, 1 CFU Aspergillus, 25 CFU Cladosporium, 4 CFU Penicillium, 2 CFU Non-sporulating colonies
  • Location 2: 4 CFU bacillus, 1 CFU Coag. Negative Staph, 9 CFU Micococcus, 1 Streptomyces (bacteria), 23 CFU Cladosporium, 1 CFU Epicoccum, 3 CFU non-sporulating, 9 CFU Penicillium
  • Location 3: 6 CFU bacillus, 14 CFU Micrococcus, 11 CFU Cladosporium, 1 CFU Epicoccum, 4 CFU non-sporulating, 2 CFU Penicillium
  • Location 4: 4 CFU bacillus, 25 CFU Micrococcus, 3 CFUAcremonium, 6 CFU Cladosporium, 3 CFU non-sporulating, 6 CFU Penicillium
  • and a partridge in a pear tree.



Aerometric results still show quite a bit of growth in the ante room but even more disturbing in the hazardous drug buffer room (see attached documents from certifier). Further investigation is required to finally understand the root cause of this concerning contamination. After phone calls to both redacted from certifier and the owners of clean room builder (the clean room engineer and builder).

certifying tech’s suggestions:

  1. Review cleaning procedures again
  2. May need to tweak air pressures within rooms as he is concerned ante room is at threshold of passing; 0.02 inches of water.
  3. Air balancer should possibly be called to accomplish

Clean room company suggestions:

  1. Asked if there by chance if there are any leaks in the room, ceiling, walls; go over with fine tooth comb.
  2. They have built multiple rooms with this configuration and the only time this has been an issue is when the room isn’t 100% sealed.
  3. Review cleaning procedures and sent a way to dilute bleach and alter pH to make a good cleaning solution.
  4. They said that since the hazardous buffer room was by far the worst contamination between the two rooms with growth that they believed the buffer room was more likely the cause of the contamination.

On Tuesday 12/8/15

pharmacy technician noted when cleaning that she noticed “hair and blue stringy stuff” in the hazardous buffer. She said the mop is by far the dirtiest when cleaning that buffer room by far over any other room.

Discussion: does this mean that there may be something coming down from above the ceiling? The first time X certified the room, one ceiling tile in each room was unsealed and opened. Could this be the source of contamination?

Action: 12/9/15 each of the ceiling tiles that were opened by X have been resealed with 100% silicone caulk.

Air Co X has been called to assess our situation. They will be coming 12/14/15 to look at our room and see if they see any other possibilities into our contamination issue.


Weird, “hair and blue stringy stuff?” Hmm. Ok…

More QA notes:


Viable air sampling was performed of the ante room and our hazardous buffer room by Air Co X (NOTE: a different company from our original certifier), inc. Results were reported on 12/29/2015 and it was found that we still had multiple bacterial counts in both areas (see report dated 12/29/15). We performed a nine time clean of the clean room (triple clean 3 days in a row) and scheduled another air sampling to be performed on 1/8/16

On 1/8/16 our report showed that we still had micrococci, staph coag (+) and cladosporium growth in multiple areas. (see report dated 1/19/16 from Air Co X). BF of Air Co X wanted to further investigate our clean room to see possible root causes of what was going on. He stated that he would like to have the room’s pressures tested and the number of air changes per hour in each room. In an email on 1/11/16 BF hypothesized that the room was exhausting more air than was being supplied back to the room, creating a situation that made the room try to draw from whatever air source possible (potentially the air space above the clean room – which is far from clean).

BF’s hypothesis was confirmed when readings were taken of the pressure differentials of our clean room including measuring the amount of air that was leaving through the hood. It was discovered that there was a major difference in the volume of air being exhausted and what the HVAC system could supply.

The following was written by BF in an email dated 1/18/ is an outline of the plan of action to remedy the clean room’s viable air growth:

Hello, Seth

To follow-up and to confirm our conversation and give you an action plan to get you guys producing….

First off, thanks for getting the non-viable reports to me so quickly after our conversation.  Having reviewed them, I am not surprised at results to-date.  “Passing” non-viable criteria does not mean all is OK.  Chris has had negative viable aerosol findings in pharmacies with much lower particle counts.

1 – DO this Now – Open hole above “ante-ante room” door to provide unrestricted path (sized to accommodate 2 16X25 X 4-5inch media filters ) for ~approximately 1500 cfm of make-up air into cleanroom ass’y.  Right now, our differential pressure measurements from the other day show large amounts of make-up air is coming across the ceiling tile from dirty space above.  This is causing humidity control issues during cooling season, plus a source of dirty air.  With such a pathway, you still have a negative building with an increased humidity load during cooling season, but the make-up air is now primarily from conditioned space versus the current unconditioned (and unfiltered) space above the ceiling.  This will help control humidity and should reduce particulate load in make-up air.  (Alternatively, leave the door open will have a similar effect, but that just fails the cosmetic test in multiple ways.)

This cleanroom system requires at least 1500 cfm of make-up air to replace the air being exhausted out of the 2 lab hoods.  Reducing the lab hood exhaust rate will not solve the problem, because it must be above a certain minimum to maintain face velocity.

SIDEBAR:  Clean air is created in the controlled environment primarily by one of two ways: 1 – Make-up air is filtered prior to entry or 2 – Air change rate is high enough to clean dirty air sufficient to meet standards.

2 – Now armed with baseline particulate data from original certifier report, I show up with a particle counter to measure improvement.  If things look markedly better, we show up Thursday for formal certification run.

3 – If things could be better, rent a portable HEPA filter and measure particulates.  If better, then Thursday is a go.

4 – If things are still not markedly better from particle count perspective, the next option is the positively pressurize the ante-ante room by inserting filter media in the “hole above the door”  (recall “hole” is large enough to accommodate 2 16X25 X 4-5inch media filters) and adding a booster fan to pre-filter make-up air.

Thursday, we make non-viable and environmental measurements first, if they are not satisfactory, there is no sense in proceeding until Option 4 is installed.

Thanks again for your help, Seth

Before I forget, I still need to see report to read blacked areas, but I found what I needed for today.)



There are some VERY key things to note from this consultant.

  1. When dealing with a negative pressure room, it almost doesn’t need to be said but I’m going to say it, YOU NEED TO SUPPLY CLEAN AIR! This is essentially a vacuum you’re dealing with.
  2. Negative pressure rooms will pull air from wherever it can…and will. If you’re not supplying sufficient amounts of clean air it will draw from wherever it can. What goes out must equal what comes in (well, it’s going to be slightly negative since it’s negative pressure).
  3. Regardless of us having an ISO 7 ante room, our negative pressure room was still drawing air either from outside of the ante room or from above the ceiling. You would have to be a NASA engineer to 100% seal a negative pressure room. Don’t try to caulk your way out of this situation.

Our first step was to dial down the clean room so we weren’t exhausting as much air. We also made an easier path for the clean room to draw air from OUTSIDE the ante room rather than from above the ceiling (which is where we think air might’ve been coming from). We literally made holes above the door coming into the area where the clean room resides (see drawing below – this is a walled area outside of our clean room – with a door – noted by the angled line) and put MERV 14 filters into these holes. This area outside the clean room contains our depyrogenation oven, an autoclave, gowning materials, cleaning and compounding supplies and glassware.



After executing the plan of action outlined in the email from BF, we have performed another set of viable air samples. Our viable air samples showed marked improvement (see Air Co X report dated 1/29/16). We had 1 non-sporulating fungi, 4 cfu of gram positive cocci, micrococci, staph coagulase (+), & bacillus in our ante room. We Also had micrococci and staph coagulase (+) in our hazardous buffer room. After nasal swabbing each of the operators of the clean room it was discovered that all 3 operators have staph coagulase (+) bacteria


Reached out to Jim Polarine from Steris, who is a consultant on clean rooms for the pharmaceutical industry. After discussing our situation and reading all of our viable air sampling reports he discussed the idea of setting our own action/alert limits for our clean room and that our current growth is not uncommon. He also gave us ideas for a environmental monitoring plan that includes weekly fingertip testing (following compounding), settling plates (while compounding set in high traffic areas), and using an active air sampler (suggested MAS-100, SAS, or Biotest Air sampler).

After making the engineering changes that were suggested by BF we will also test one final time to confirm that our efforts are showing progress.


It should also be noted that typically air sampling resulting in the growth that we were seeing doesn’t necessarily have to do with cleaning. While you can clean surfaces all day every day, this is usually a problem with the environment itself or poor gowning technique. Each of the operators that had clearance for going into the clean room had previously passed a fingertip glove sampling test and we didn’t feel that it had to do with gowning technique. Moving along…here’s what we did:

  • keeping the filters in the “holes” we created above the door
  • ordered 2 more 2×2 MAC 10 envirco HEPA fan filter units and installed these in the ceiling outside the clean room
  • the ceiling tiles were replaced with the same type of ceiling tiles you’d typically see in a clean (they can be cleaned and caulked into place)
  • This area was tested to be an ISO 8 area only in theory (NOTE: we purposely did NOT classify this area, we only want it to perform as such but do not want to classify).
  • An air balancer was able to dial down the exhaust for the room so that only the amount necessary for the negative pressure was being exhausted while still maintaining all other specs for the environment. (see picture below)

Subsequently, we had an inspection from the NABP and their surveyor noted that since the non-sterile hazardous room has a direct pass-through into the hazardous buffer room it should be gasketed also go the extra mile in keeping the air in that room as clean as possible thereby classifying it to ISO 7 standards. So we did.

In the end here’s what we have:



Results received on 4/7/16  (see certifier report dated 4/7/16) for final area of negative pressure hazardous buffer room had no growth; indicating that our changes have worked in minimizing growth in our clean room. To summarize the engineering changes we’ve made:

  1. Make a pathway for air to come into the clean room area by making “holes” in the door and above the door (these are 16×25” cutouts that we have mounted filters in).
  2. Add 2 hepa filter fan units to the area just outside the clean room so that we are cleaning up the make-up air being supplied to the clean room.
  3. Add door sweep to door going into prep area (this is the area just before entering the clean room).

By making these changes we have mitigated growth of all bacteria and fungi in the areas that had previously tested positive.


In the fall of 2016 we also took on an additional project to once an for all put our issues with humidity and temperature to bed. We contracted with a local company here in Lexington to design a system to supply dry, temperature controlled air to the clean room. An engineering firm from Louisville assessed our needs with the qualifications of:

  • continuously maintaining temperatures below 68 degrees and relative humidity below 45%

This is somewhat of an engineering feat but CAN indeed be accomplished. The solution for us was a desiccant dehumidifier. Without boring you with all of the details essentially what this does is take outside air, cools it down to a fairly low temperature (30-45 degrees) passes it through a desiccant wheel which dries it (and I mean DRY) and then reheats the air to be pushed into the desired location. This machine is now what supplies all of the air for the clean room and the prep area outside of the clean room. We’ve not only accomplished complete remediation from our bacterial and fungal issues but now have an environment that is cool, dry and even more resistant to having any issues in the first place.

Again, this wasn’t to warn people that USP 800 is a bad idea, rather to educate those who are (and will be) going forward with becoming compliant to not make the same mistakes we did…and pay the price (in literal dollars).

Some final take away notes (this is just what I’ve learned from this experience):

  • Negative pressure sucks…literally. When you’re creating a negative pressure using a biological safety cabinet you are potentially drawing any room contamination in through what is supposed to be the cleanest area of your room (SEE SECOND AND THIRD BULLET POINT).
  • DO NOT store anything that isn’t absolutely necessary in your buffer room for THAT day. In other words, keep your hazardous room essentially empty 99% of the time (in fact, this should go for every room in your clean room – even the ante room).
  • When creating a negative pressure environment, this will take air from wherever it wants. It will take the air from the path of least resistance. What YOU need to do is create that path of least resistance. It also needs to be fairly clean (ISO 8) air that you’re feeding the area (even the area beyond the ante room – make an ante ante area – or what I call my prep area).
  • Not classifying area outside of clean room was done on purpose. Why? Simple: under promise and over deliver. If we classified, we would have to maintain that certification EVERY time. We’re not interested in paying that cost OR keeping that certified; we just simply want that air to be clean.
  • If you’re a pharmacist that works in a clean room environment (or responsible for one) and you’ve never looked at your certification documents, crack the book. It’s your duty to know what these numbers mean. Your certifier is not going to be there when you need to interpret these for an inspector NOR should they. You NEED to know what this stuff means…have high particulate load near a cart? What’s on the cart? Cardboard!? Oh no! – if you don’t know what this stuff means: or better yet…reach out to someone who does. I’ve found some real experts in the field (i.e. Jim Polarine just to name one – thank you Jim) to be EXTREMELY helpful…and more than happy to!

Soap box dismount…

Thanks for reading!

Good luck.

Related Articles


Your email address will not be published.

  1. Unbelievably accurate account of how it is. We have such a similar experience and layout I have chills and at the same time a level of relief that someone else had the same exact issues, Thank you Seth for publishing this article. All of you who are about to encounter USP 800 take this man’s advice to heart.

    1. Thanks Phil!

      I’ve heard this from a couple of people and feel like the more people that come online with USP 800 standards, the more we’ll hear of this. I unfortunately did not submit this for comment to USP but feel like some kind of asterisk or something should be added with details about this.

      Thanks so much for reading!