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 19.2.2.2.4 of the 2000 Life Safety Code does not allow doors in the path of egress to be locked. The exceptions to 19.2.2.2.4 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 7.2.1.6.1 and 7.2.1.6.2 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 7.2.1.6.1 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 7.2.1.6.1 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 18/19.2.2.2.5.2 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 7.2.1.6.2 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

safeplace-ob-door-sm[1]

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 7.2.1.6 of the LSC. Section 20/21.2.2.2 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.

Locks vs Latches

Q: A surveyor cited us for not having positive latching doors to our entrance of the ICU suite. We have magnetic-locks on those doors and pointed out that they serve as positive latching. He refused our position and cited us anyway. Don’t magnetic locks qualify as positive latching hardware on doors?

A: No, they do not. Locks are not the same as positive latches. The magnetic locks that you mentioned are most likely access-control locks as described in section 7.2.1.6.2 of the 2000 edition of the Life Safety Code. Those electronic magnetic locks are required to unlock upon activation of the fire alarm system. That’s when you need the corridor door to positively latch the most: during a fire. Also, make sure there is a motion sensor mounted on the egress side of the doors served by the magnetic locks, along with a button mounted on the wall within 5 feet of the doors, when activated will unlock the magnetic-locks for a minimum of 30 seconds. The button must be labeled “Push to Exit”.

More on Dead-Bolt Locks

Dead-bolt lock on door Web 2If you search my website on door locks, you will find a few entries concerning dead-bolt locks. For decades hospitals seem to have installed dead-bolt locks on any door they wanted, without any concern for compliance with the Life Safety Code. That’s all changing now, as surveyors and inspectors are becoming better educated on where dead-bolt locks may be used.

The picture to the left shows a dead-bolt lock on a corridor door that is separated from and is not part of the latch-set. It is mounted approximately 12 inches above the latch-set handle and in order to egress through this door, an individual would have to make two releasing motions: 1). unlock the dead-bolt, and 2). turn the latch-set handle. Those two actions make this dead-bolt installation non-compliant with section 7.2.1.5.4 of the 2000 edition of the LSC, which prohibits more than one releasing actions to operate the door.

Some people are quick to point out that dead-bolt locks installed as described above are approved for use on patient room doors in psychiatric units where ‘clinical needs’ locks are permitted. My reply would be, approved by who?  I am aware that Joint Commission has made an interpretation that permits their accredited organizations to have dead-bolt locks on psychiatric units, but that does not address the improper installation of them. Just because Joint Commission says you can have dead-bolt locks, does not mean you are permitted to install them incorrectly.

Hotel Door Latch & Lock Web 2The picture to the left shows a dead-bolt lock that is integrated (not separated) with the latch-set. This arrangement allows the retraction of the dead-bolt simply by turning the latch-set handle, and thereby complying with section 7.2.1.5.4 with a single releasing motion to operate the door. [NOTE: NFPA Healthcare Interpretations Task Force has ruled that pulling or pushing is not considered a motion to operate the door.]

This would be an excellent solution for those situations where you wanted to have a dead-bolt lock installed on a door in the path of egress. This concept of latch-set & dead-bolt lock combined could be obtained in anti-ligature arrangement, and be used in a psychiatric unit.

But there are situations where a traditional dead-bolt lock could be used on a door in the path of egress in a healthcare occupancy. Where? How about in areas where the corridor door is not required to latch?

There are multiple locations in a hospital where doors that open onto a corridor are not required to latch: bathrooms, toilet rooms, shower rooms, and sink rooms that do not contain any combustibles. Also, how about the exterior doors in a Type I or Type II building? Those doors aren’t required to be fire rated, so there is no requirement for exterior doors to latch, either. [Now, most hospitals want their exterior doors to latch even if they aren’t required to, in order to keep the weather and the riff-raff out, but in some rare situations such as the main entrance consisting of sliding glass doors, latching may not be desired.]

Dead-bolt lock on corridor door Web 2The picture to the left shows a door with a dead-bolt lock, but there is no other releasing device mounted on this door, such as a latch-set. While this door would not be permitted in many locations in a hospital corridor because it does not latch, it would be permitted on doors to rooms that are not required to latch, according to 19.3.6.3.1 (toilet rooms, bathrooms, shower rooms, and sink closet not containing combustibles). So a door in the path of egress with a dead-bolt lock and no latch-set on one of these limited rooms would still be compliant with section 7.2.1.5.4 because there is only one releasing action to operate the door: turning the thumb-screw to unlock the door.

Dead-bolt locks that are installed on corridor doors that are required to latch in such a way that they are separated from the latch-set, are just big red flags to surveyors and inspectors. If the door requires two releasing actions to operate the door, then that’s a violation, and they are easy to find. Check out your facility… it may be possible that you have a deficiency that you are not aware.

Door Locks in Healthcare Occupancies

exit_mag_lock_hospital_10662932[1]Outside of the healthcare setting, locks on doors are utilized nearly everywhere we go, and we don’t give it much of a thought. But in the highly regulated industry of healthcare, we must be very vigilant and astute to what the codes and standards will and will not permit. Unfortunately, locks on doors that are in the path of egress are greatly misunderstood by hospitals, and therefore are widely abused.

 Ordinarily, you are not permitted to lock a door in the path of egress in a healthcare occupancy however, there are three exceptions:

  • Delayed egress locks
  • Access-control locks
  • Clinical needs locks

 Delayed Egress

Delayed egress locks are a lock when a person pushes on the horizontal crash bar of the locked door, a local buzzer will sound, and the door will automatically unlock within 15 seconds. This effectively allows a person to egress through the door, but just delays their egress, hence the name. The requirements for a delayed egress lock are:

  • Delayed egress locks are only permitted in buildings which are fully protected with smoke detectors or automatic sprinklers
  • The delayed egress locks must unlock upon activation of the sprinkler system or a heat detector or a smoke detector, and remain unlocked until manually reset. Notice that activation of a manual pull station is not required to unlock a delayed egress lock.
  • The delayed egress locks must unlock upon loss of power to the mechanism controlling the lock
  • Upon 3 seconds of activating the releasing device (horizontal crash bar) a local buzzer must actuate (to alert staff someone is attempting to exit), and within 15 seconds of activating the releasing device, the lock shall automatically unlock. The locks must reset manually, not automatically.
  • The code allows for the delayed egress lock to automatically unlock within 30 seconds if approved by the AHJ, but HFAP does not approve of this option.
  • A sign, in 1 inch letters, must be posted on the door which reads:

PUSH UNTIL ALARM SOUNDS

DOOR CAN BE OPENED IN 15 SECONDS

  • Only one delayed egress lock may be installed in the path of egress to the public way.

These types of locks are not uncommon in hospitals, but are not used very much. The most common deficiency observed with delayed egress locks is they are installed in a building that is not 100% protected with sprinklers or smoke detectors. [Code reference: 7.2.1.6.1, 2000 edition of the LSC]

 Access Control

Access Control locks are very common in hospitals, and most likely are misunderstood on their correct operation by the facilities staff.  An access control lock usually utilizes a magnetic lock (mag-lock) assembly, and is often integrated with the badge swipe reader to control access into a department. The requirements for an access control lock, are:

  • A motion senor must be mounted on the egress side of the door, that will detect a person approaching and will automatically unlock the door in the direction of egress
  • A loss of power to the device controlling power to the access control lock must unlock the doors in the direction of egress
  • A manual release device must be mounted within 5 feet of the door, and between 40 to 48 inches above the floor which, when depressed, will automatically unlock the door in the direction of egress, and must be identified with a  sign that reads:

PUSH TO EXIT

  • When depressed, the “Push to Exit” button must directly interrupt power to the lock, independent of the access control system electronics, and the doors must remain unlocked for a minimum of 30 seconds
  • Activation of the building sprinkler system or the building fire alarm system must unlock the door in the direction of egress

The most common deficiency with access control locks is the absence of the required motion sensor and/or the “Push to Exit’ button on the egress side of the door. When properly installed, access control locks are not a lock for people trying to get out of the building, but they serve as a lock on doors for people trying to get into the area controlled by the lock. [Code reference: 7.2.1.6.2, 2000 edition of the LSC]

 Clinical Needs

Clinical needs locks are only permitted 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: 19.2.2.2.4, Exception No. 1, 2000 edition of the LSC]

 Fire Alarm Connection

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 to the Life Safety Code, NFPA 72 National Fire Alarm Code, (1999 edition) 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 clinical needs locks is very silent on the subject. When a code is silent on a subject that means it permits it.  Therefore, section 3-9.7 does not apply to clinical needs locks in a hospital, because the Life Safety Code permits it. The code writers understand that psych patients are smart enough to actuate a fire alarm to unlock the exit doors. Therefore, they wrote the code 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. That is the trade-off: They will not require the locks to unlock on a fire alarm, as long as all staff can unlock the doors in the event of an emergency.

 Dead Bolt Locks

Generally speaking, dead bolt locks are not permitted on a door in the path of egress in a healthcare occupancy. The only exception would be a dead bolt lock which automatically retracts when the door handle is twisted but other than that, they are not allowed. The reason for this is the LSC only permits one releasing action to operate the door. So if a dead bolt lock is installed on a door in the path of egress and the lock is not part of the latch-set, then it would require two actions: one to unlock the dead bolt and another to turn the door handle, to operate the door. Please note that pulling or pushing is not considered one of the actions to operate the door. [Code reference: 7.2.1.5.4, 2000 edition of the LSC]

 

Dead-Bolt Locks on Psychiatric Rooms

Q: In a Psychiatric hospital emergency room, are patient holding rooms allowed to have dead bolts on the corridor side with no thumb latch on the inside of the room? Our Administrator’s position is that those rooms can be used as seclusion rooms at any given time. Is that acceptable?

A: Interesting question. I would say it depends if the seclusion room would only be used for the purpose of securing individuals who present a threat to themselves or others. This is commonly called “clinical needs” and is allowed under section 19.2.2.2.4, exception number 1 of the 2000 edition of the LSC. However, if the same room is used for a non-behavioral health patient then the locks would not be permitted. Section 7.2.1.5.4 does not allow more than one action to operate the door in the path of egress, and a dead-bolt and a door latch would be more than one operation. Can you guarantee this room would never be used for seclusion? If not, then I would think most authorities would have a problem with it. This is going to end up as an interpretation from an authority having jurisdiction (AHJ). The average hospital has 5 or more AHJs and even if 4 AHJs don’t have a problem with the dead-bolt locks, and the 5th AHJ does, then you have to comply, and remove the dead-bolt locks. Most hospitals deal with this issue by having sitters watch the patient. Not the best solution, but it does meet the code requirements.