Strange Observations – Part 6

Door with Conflicting SIgnage Web 2Continuing in a series of strange things that I have seen while consulting at hospitals….

This series of pictures is just plain sad… and wrong. The top picture shows a door in the path of egress with an ‘Exit’ sign over it. If you look closely, there is a magnetic lock on this door and there is a sign posted on the door as well. This door is in an exit access corridor and lead from one physician’s office area to another physician’s office area and for some reason was equipped with a magnetic lock. But the magnetic lock did not qualify for the exceptions to of the 2000 LSC for delayed egress locks or access-control locks.

Take a look at the bottom picture… The picture is a close-up of the sign on the door. The sign says “This is not an exit…” but the ‘Exit’ sign over the door says “This is an exit”. That’s a conflict. To be sure, the door is an exit as this door was located in an exit access corridor and without the designated exit it would have been over a 100 foot dead-end corridor (which is not permitted). So, the exit is necessary.

But the conflicting signage is just wrong. Just because one physician’s office did not want people exiting into his area, you cannot lock the door and say the door is not an exit.

Close up of Door With Conflicting Signage Web 2I remember we tried to find out what actually releases the magnetic lock and the staff that I was with could not say. They said they were not sure if it released on a fire alarm and they checked with security and it did not release on any signal that security could transmit.

This is the type of deficiency that will lead to an Immediate Jeopardy decision and cause all sorts of headaches for a hospital.

Strange Observations – Part 4

Deadbolt on Exit Door Web 2Continuing in a series of strange things that I have seen while consulting at hospitals….

This is a picture of a door to an exit stairwell. Note that there is a deadbolt lock on the door above the fire exit hardware. That is so not allowed, according to of the 2000 LSC. You cannot lock a door in the path of egress that requires the use of a tool or a key. (There are some exceptions but this door does not qualify).

Also, the sign on the door indicates it is a delayed egress lock but it says the door can be opened in 30 seconds. Section allows for a 30 second delay but only when it is approved by the authority having jurisdiction (AHJ).  Most national AHJs for hospitals do not approve of the 30 second delay so this is not permitted. (Some hospitals mistakenly believe that the local or state fire marshal is the only AHJ they need to ask permission, but they don’t realize that the accreditation organizations and CMS are AHJs as well.) If you want to utilize an exception in the LSC whereby it is permitted with the AHJ’s permission, then you need to have permission from all the AHJs who have authority over hospitals:


  • CMS (Federal agency who controls Medicare/Medicaid reimbursements)
  • Accreditation organization
  • State agency who license hospitals
  • State fire marshal
  • Local fire inspector
  • Insurance company

The likelihood of getting all of them to agree on an exception to the LSC is slim, if not impossible.

Controlled Access Locks

Q: While conducting fire drills in the hospital, one of the questions on our drill evaluation sheet is, “Did the security doors in the fire zone release properly?” We have controlled areas where the doors are locked to control access into the unit. To exit the unit only requires the push of a button and the doors release. So are we in compliant with this controlled access not releasing during the fire alarm activation since the exit is not controlled? Or should the doors release to allow free entry and exits?

A: Doors in the path of egress in a healthcare occupancy are not permitted to be locked. However, there are three (3) exceptions to this requirement:

  • Delayed egress locks complying with section, 2000 LSC
  • Access-control locks complying with section
  • Clinical needs locks complying with section

By the sound of your situation, it appears to me that you do not have delayed egress locks and you do not have clinical needs locks, which leaves access-control locks. However, it also appears that your description of the security door locks may not be in compliance with section Here is a summary of the requirements for access-control locks:

  1. A motion sensor must be mounted on the egress side to detect occupants approaching the door, and automatically unlock the door in the direction of egress
  2. A loss of power to the control system automatically unlocks the door in the direction of egress
  3. A manual release button must be mounted 40 to 48 inches above the floor, and within 5 feet of the door, that when operated will directly interrupt the power to the lock, independent of the control system, for a minimum of 30 seconds. The button must be labeled with the words “PUSH TO EXIT”.
  4. The door must unlock in the direction of egress upon activation of the building fire alarm system or the building sprinkler system.

So, it appears to me that you are missing the motion sensor on the egress side of the door that would automatically unlock the door when someone approaches. Also, it sounds like your locks are not interconnected to the building fire alarm system to automatically unlock on an alarm. According to section, these are required. Also, check the ‘PUSH TO EXIT” button to make sure it interrupts power to the locks for a minimum of 30 seconds, when depressed.

Expert Witness

imagesXLH1EQVPI received a telephone call recently from a lawyer that asked me if I would be willing to serve as an expert witness in a pending litigation suit. The situation of the lawsuit is a patient who was admitted to the inpatient psychiatric unit of a hospital busted the protective cover over the fire alarm manual pull station, and activated the fire alarm system. The entrance doors to the psychiatric unit unlocked on the fire alarm, and the patient ran out, climbed to the roof of the facility and jumped off.

Now the family of the patient is suing the hospital and the fire alarm company who installed the interface between the door locks and the fire alarm system. The plaintiff’s lawyer is arguing that the interface should never have been installed and the doors should not have unlocked upon activation of the fire alarm system. I agree with the plaintiff’s lawyer, but the problem is, it was the defense’s lawyer who contacted me.

The two major codes and standards that apply in this situation are the NFPA 72 National Fire Alarm Code (1999 edition), and the NFPA 101 Life Safety Code (2000 edition). The Life Safety Code (LSC) allows clinical needs locks on doors in the path of egress, but only in healthcare occupancies (hospitals). They are not permitted in any other occupancy. These types of locks are permitted where the “clinical needs of the patients require specialized security measures for their safety, provided that staff can readily unlock such doors at all times”. Clinical needs locks are permitted for Behavioral Health units, such as psychiatric and Alzheimer units, to prevent patients from leaving the unit unauthorized, and potentially harming themselves or others.  Clinical needs locks requires all staff who work on that unit (which includes physicians, nurses, aids, clerical, maintenance, foodservice, housekeeping, etc.) to have a key or device to unlock the door on their person at all times. [Code reference:, Exception No. 1, 2000 edition of the LSC].

But when it comes to connecting those clinical needs locks to the fire alarm system so they automatically unlock on a fire alarm activation, the codes are not so clear. Other permissible locks, such as delayed egress locks and access control locks, are required by code to be connected to the fire alarm system and automatically unlock the door whenever there is an alarm. However, there is no such similar language for clinical needs locks to unlock on an alarm. In addition, NFPA 72 section 3-9.7.1 and 3-9.7.2 says if you have a lock on an exit door it must be connected to the fire alarm system, and it must unlock the door on a fire alarm. However, the exception to 3-9.7.2 essentially says this is not required if an AHJ says so, or if another code says so. So, we go back to the Life Safety Code, and we see that delayed egress locks and access control locks are required to unlock on an alarm, but the section on clinical needs locks is very silent on the subject. When a code is silent on a subject that means it is open for interpretation. Therefore, section 3-9.7 of NFPA 72 does not apply to clinical needs locks in a hospital, because the interpretation Life Safety Code permits it. It is documented that psychiatric patients are astute enough to actuate a fire alarm to unlock the exit doors. Therefore, the code is available to be interpreted in such a way to not require clinical needs locks to unlock the doors in the path of egress on activation of the fire alarm system provided all staff carry a key to unlock the door in case of an emergency.

But who makes that interpretation? Usually the authorities having jurisdiction (AHJ) makes that interpretation. Section of the 2000 LSC says the AHJ shall determine whether the provisions of the LSC are met. But when the AHJ does not issue an interpretation, then it is up to the individual organization’s to decide. As far as I know, there is no national AHJ for healthcare that has it written in their standards that clinical needs locks should not be connected to the fire alarm system. Therefore, the hospital needs to ask the question: Based on the potential risk to patient safety, is it best to connect the door locks on the psychiatric unit to the fire alarm system so they unlock upon activation of the fire alarm system?

I declined the offer to be an expert witness for the defense, and suggested they settle the lawsuit out of court.

Main Lobby Locked Doors

Q: We have a separate building on our campus that includes our behavioral health unit, along with related offices and meeting rooms for the support staff. The behavioral health unit is segregated from the offices and meeting rooms and patients would never be in these offices or meeting rooms. There are secured doors preventing the patients from entering the main lobby area. The path of egress for the offices and meeting rooms is through the main lobby. This is a secured facility, so much so that the egress doors from the main lobby of the facility are locked with the use of electronic mag-locks. The only way the doors will release is through one of the three methods: Swipe an employee badge to release the mag-locks; a person in the cubicle to push a button to release the mag-locks; or when the fire alarm system is activated. Are we in compliance with the Life Safety Code with our mag-lock doors for egress concerns in our lobby?

A: It does not appear that you are. As you state, the behavioral health unit is segregated from the rest of the facility by secured doors and the path of egress for the offices and meeting rooms is through the main lobby, so locking those egress doors would not be permitted. Section of the 2000 Life Safety Code does not allow doors in the path of egress to be locked. The exceptions to allow delayed egress locks and access-control locks, but in this case, it does not appear that you could use clinical needs locks on doors in the path of egress that are shared by the offices and meeting rooms.

Typically, authorities having jurisdiction do not allow clinical needs locks on more than one set of doors in the path of egress for behavioral health units. You could install delayed egress locks on the main lobby egress doors as long as the entire facility is either sprinklered or protected with detectors. A card-swipe reader could be installed to deactivate the delayed egress function so people could exit without activating the delayed egress alarm. Or, you could install access-control locks on the doors, although they really are not locks for people egressing. Follow the requirements for delayed egress and access-control locks found in sections and of the 2000 Life Safety Code.

Infant Abduction Locks

Hospitals want to keep their nurseries, mother/baby units, and pediatric units secure, so they lock the doors. This causes a problem with the Life Safety Code because you can’t lock the doors in the path of egress in a hospital, other than three exceptions: 1) Clinical needs locks, which nurseries, mother/baby units, and pediatric units do not qualify; 2) Delayed egress locks; and 3) Access-control locks. Access-control locks really do not lock the door in the path of egress because a motion sensor will automatically unlock the door as a person approaches. So, in this situation the doors can only be locked using the delayed egress provision (found in section of the 2000 Life Safety Code).

But hospitals want the infant security systems used on the babies. These systems have a bracelet that is attached to the baby, and some have bracelets to attach to the mother as well. If the bracelet gets too close to the exit door, an alarm will sound and the door will lock. The problem is, these infant security systems do not comply with any of the three exceptions for locking the doors in the path of egress, listed above. Even if the doors will unlock on a fire alarm the hospital says, that is still not enough to qualify for the any of the three exceptions.

But then the hospital says their accreditation organization approved this door locking arrangement. Why should it be considered non-compliant if the accreditor allows it?  Sorry… just because the accreditation organization says it is okay, still does not make it compliant with the requirements of the Life Safety Code. When the state agency who surveys on behalf of CMS takes a look at it, they will not be as benevolent as the accreditor, and they will cite it as a deficiency.

So, to be compliant with the Life Safety Code, when the doors lock because the bracelet gets too close to the door sensor, the doors should lock into a delayed egress mode (again… see section in the 2000 Life Safety Code). Then it would be legal. But the 2012 LSC has made a change in this area and will allow locks on doors for the specialized protective measures for the safety of the occupants (see section 18/ in the 2012 LSC). This will allow you to lock the doors without delayed egress, provided you meet the requirements listed in that section. CMS has already approved categorical waivers to allow hospitals to begin using this new section of the 2012 LSC before they adopt it.

Take a look at your locks that are used on the nurseries, pediatric, mother/baby units, and even the ICUs and the ERs. If they are not delayed egress, then take a look at the CMS categorical waivers and consider modifying the doors to meet those requirements.

Door Locks

Q: We have been asked to install a lock on a door in the path of egress through an office. For security reasons they would like to lock the doors to and from this area. We are thinking of using an electric strike fail safe connected to fire alarm on both doors. This is not in a patient care area, and the doors would only be used by staff.

A: In a hospital, there are only three permissible methods to lock a door in the path of egress: 1) Delayed egress; 2) Access-control; and 3) Clinical needs. Let’s eliminate clinical needs right off the bat, as that refers to a behavioral health unit or an Alzheimer’s unit. Delayed egress may be a possibility, but the hospital needs to be fully sprinklered or fully smoke detected, and you cannot have more than one delayed egress lock in the path of egress to the public way. Delayed egress does not provide true security for the doors, just a 15 second delay which if activated can be very annoying to the staff. I don’t see this as a suitable arrangement.  The more logical approach is the access-control locks, which allows you to provide security to prevent unauthorized individuals from entering the space, but it does not prevent anyone from exiting the space. Section of the 2000 edition of the Life Safety Code describes the requirements for access-controlled egress:

  • A sensor must be installed on the egress side of the door to detect an occupant approaching the door and automatically unlock the door. This sensor must also be wired where a loss of power to the sensor unlocks the door.
  • A loss of power to the access-control system must unlock the door.
  • A manual release ‘Push to Exit’ button must be installed on the egress side of the door, 40 – 48 inches above the floor, and within 5 feet of the door. The manual release button must be labeled with a sign that reads ‘Push to Exit’. When operated, the manual release button must directly interrupt power to the lock independent of the access-control system, and the door must remain unlocked for a minimum of 30 seconds.
  • Activation of the building’s fire alarm system and/or sprinkler system must unlock the door, and remain unlocked until the fire alarm system has been manually reset.

Access-control locks do not provide any security in the path of egress. In your question, you stated that the door in question is in the path of egress. If that is truly the case, then there is no way you can legally lock this door. I would advise the hospital to re-configure their walls and path of egress to allow the office space they desire without locking a door in the path of egress. Another issue to consider: Is the door in question required to latch? If so, then access-control locks cannot be used in lieu of latching. Even though the door may be locked by a mag-lock, it still needs to latch (if required). The phrase “fail safe” means different things to different people. Typically, for locksmiths ‘fail safe’ means when power is removed, the locks remain locked, but for fire safety people the phrase ‘fail safe’ means the lock remains unlocked.

Card Readers on Door Locks

Q: During a recent survey, the surveyor said a card reader on the stairwell door cannot be located on an adjacent wall or door frame, but it must be an integral part of the lockset itself. Is this true? They were talking about card readers on the stairwell side for re-entry to a floor.

A: Did the surveyor cite you for non-compliance? If not, surveyors sometimes say things that are misunderstood, especially if there is no citation. Doors not in the path of egress are permitted to be locked, and a re-entry door from a stairwell usually is not in the path of egress. As long as the re-entry door in the stairwell is not in the path of egress then I do not see any reason that what you describe would be a problem. The Life Safety Code would allow for a card reader device to unlock a stairwell re-entry door as long as the door is not in the path of egress.  If the card reader is mounted in the stairwell on the door leading to a floor of the building (not a discharge door), then the card reader is not on the egress side of the door. There is nothing in the Life Safety Code, or in NFPA 80 Standard for Fire Doors and Fire Windows, (1999 edition) that would require the card reader to be mounted on the door leaf, rather than the on the wall near the door. Therefore, it is clear that the LSC permits card-access readers to be mounted on the wall near the door, since it is not a device or motion to operate the door. I cannot think of any situation that would require the card-access reader to be mounted on the door leaf, itself.

Locked Doors for Utility Rooms?

Q: Where is the reference in the Life Safety Code that requires the doors to housekeeping or soiled utility rooms to be locked? I have a Risk Management director that tells me the code requires these doors to be locked.

A: There is no Life Safety Code requirement to lock housekeeping or soiled utility room doors. There is no Joint Commission, CMS or any other national authority that requires housekeeping or soiled utility room doors to be locked. Where hospitals get into trouble with CMS and the accreditation organizations on this issue is the failure to assess the risk to safety for patients and staff, when these doors are left unlocked. Each of the national authorities has a standard that requires hospitals to either identify safety and security risks in the environment, or their standard requires the hospital to maintain a safe environment for their patients.  An unlocked utility room that contains a risk to the patients would certainly be suspicious to a surveyor that the environment may not be safe for the patients. A housekeeping room may contain cleaning supplies that could be considered dangerous to unauthorized individuals (such as children). If the door to the housekeeping room was left unlocked, then people could gain access to the hazardous items and hurt themselves or others. Likewise for soiled utility rooms, which by definition would have soiled linens which may be bio-hazardous. This does not mean all soiled utility rooms or housekeeping rooms need to be locked. They just have to be assessed for the safety or security risks associated with the contents of the rooms. In my encounters, most of the soiled utility rooms that I see in hospitals are unlocked. Only soiled utility rooms where children are prevalent are the ones that are typically locked. Now, on the other hand, most (if not all) housekeeping janitor’s closets that I see are locked, partly due to the hazardous cleaning chemicals stored in them, but also because Housekeeping doesn’t want their other supplies stolen. But, to be sure, there is no direct requirement in the LSC or in the accreditation organization standards to keep these doors locked.

Delayed Egress on Doors in Ambulatory Health Care Occupancies


A friend of mine was working on a project involving ambulatory health care occupancy, and they wanted to use a delayed egress lock on an interior door.  The 2000 Life Safety Code (LSC) limits special locking arrangements to exterior doors in ambulatory health care occupancy, but the 2012 LSC does not include that limitation.  They wanted to know what I would recommend.

Well, they are correct: The 2012 edition of the LSC did away with the limitations of the special locking arrangements found in section of the LSC. Section 20/ of the 2012 edition of the LSC now permits delayed egress locks on any door in the path of egress, where the 2000 edition of the LSC limited them to the exterior door.

I assumed that the individual who asked me the question was bound by the 2000 edition of the LSC, such as a Joint Commission accredited organization, or perhaps a CMS provider for Medicare. My initial thought is the organization would have to comply with the conditions of the 2000 edition of the LSC, and cannot take advantage of the more lenient 2012 edition until such time that edition is adopted by CMS and/or Joint Commission.

However, CMS did issue a categorical waiver to healthcare organizations to allow them to use many of the provisions of the 2012 edition of the LSC now, before the 2012 edition is actually adopted (which may be at least another 12 months away). I reviewed the CMS S&C memo 13-58 once again, and while CMS did state in one of their opening paragraphs that they have the authority to grant waivers for ambulatory surgical centers, they failed to do so in the body of their memo. They have a categorical waiver on doors to allow healthcare occupancies to use the more liberal 2012 LSC position on delayed egress locks, but that categorical waiver only applies to healthcare occupancies, and not ambulatory surgical centers, assuming the ambulatory health care occupancy my friend was referring to was an ambulatory surgical center.

Therefore, I concluded that the organization should (or must) comply with the 2000 edition of the LSC and only install special locking arrangements on exterior doors. Once the 2012 edition of the LSC is adopted, they can then install delayed egress locks on interior doors.

I also mentioned that if the facility in question is only accredited by The Joint Commission, and does not receive any funds from CMS as a Medicare or Medicaid provider, then they could contact the Standards Interpretation Group (SIG) at Joint Commission and ask them if they would accept a Traditional Equivalency to allow them to use special locking arrangements on interior doors of their ambulatory care occupancy. (The telephone number for SIG is:  630-792-5900, select option 6.) My guess is they will, provided the organization meets all the requirements for a traditional equivalency.

If the facility in question actually is part of a larger organization that does have a CMS control number (CCN), then I advised my friend that they have no choice by to comply with the conditions of the 2000 edition of the LSC.

My friend replied asking if they could request a standard CMS waiver to allow the organization to install the delayed egress lock on an interior door now, before the 2012 edition of the LSC is adopted.

My reply was yes, there is always the possibility for a waiver, but CMS will not accept a waiver request unless it is in response to a survey deficiency. In other words, the waiver process is not valid until someone representing CMS cites a deficiency.

As a safety professional, I would never recommend or advise a client to knowingly violate the current edition of the LSC, even though we know that issue will be viewed differently in a more recent edition. There is always the chance that CMS may not adopt the 2012 LSC, although I would be very surprised if they did not.

However, any organization may do what they want, and often times they disregard the advice of a safety professional, and violate the LSC, taking the risk that they will not get caught. In this case, it is understandable as the presumption is the 2012 edition will be effective within the next 12 months or so, and they may feel they will not have any surveys or inspections before then.

So… in summary: The waiver process is not available to them since they have not been cited for non-compliance with the LSC. And, as a safety professional I cannot advise them to violate the LSC. What they do after that is their own business, and risk.

I have some clients who ignore my advice, preferring to ask other safety experts until they find someone who agrees with the answer they want. Hey… it’s their hospital, not mine. I always advise clients to follow the current rules, regulations, codes and standards, but if they decide otherwise, then that’s on them.