Almost every fire department, EMS agency and police department in the United States has something on a wall about mental health, and a large share of them have very little behind it. This piece deals with the parts that decide whether a behavioral health program works at all: exposure that accumulates across a career rather than arriving in one catastrophic call, what the research on single session debriefing actually tested and what it did not, how peer teams are selected and what records they should never create, why confidentiality governs the decision to call, and the scheduling and supervision choices that outrank any resilience class on the training calendar.

Cumulative exposure across a career

The programs most agencies built in the 1980s and 1990s were designed around the rare catastrophic event, the line of duty death or the school bus crash, because those were the incidents that visibly knocked a department sideways. The exposure that actually accumulates in fire, EMS, law enforcement and dispatch is more ordinary than that, and it arrives in small repeated doses: the pediatric arrest that goes badly, the frequent caller you have run on for four years who finally dies, the hanging in a house you have been in before, the overdose where the family is in the doorway. A paramedic working a busy urban system will be present at more deaths in five years than most physicians attend in a decade, and nobody counts that anywhere.

The published prevalence numbers should be handled with care, because they vary widely by discipline, by country, by instrument and by whether the sample was a general workforce or people who responded to one named disaster. A systematic review and meta-analysis by Berger and colleagues, published in Social Psychiatry and Psychiatric Epidemiology in 2012, pooled studies of rescue workers worldwide and reported a current post traumatic stress disorder prevalence of roughly ten percent, while individual studies in that literature range considerably higher and lower. The honest reading is that a clear majority of responders will never meet criteria for a trauma related disorder, and that a minority large enough to matter in any department of size will.

What shows up first in a firehouse is usually not a diagnosis. It is sleep that stops coming, drinking that moves from Saturday to every shift change, a short fuse at home, avoidance of one particular type of call, and cynicism that a crew reads as normal because everyone around the person has some of it too. Suicide is the outcome that gets the attention, and the counts remain poor: the Firefighter Behavioral Health Alliance and, for law enforcement, the organizations that tracked officer suicides before the FBI began its statutory Law Enforcement Suicide Data Collection, have all said plainly that their numbers are undercounts because reporting is voluntary. Treat any total you are quoted as a floor and ask who collected it and how.

What the debriefing research actually tested

Critical Incident Stress Debriefing came out of a 1983 article by Jeffrey Mitchell in the Journal of Emergency Medical Services, describing a structured group process for responders after a bad call, and it spread through the American fire and EMS services faster than almost any other single idea in the field. Mitchell and George Everly later positioned it as one component of a larger Critical Incident Stress Management system that also included pre-incident education, on-scene support, defusing, family support and referral. By the mid 1990s, a great many agencies had reduced all of that to a single practice, which was to convene a session a day or two after the incident and consider the matter handled.

The research that unsettled this arrived mostly from outside the fire service. A Cochrane systematic review by Rose, Bisson, Churchill and Wessely, published in 2002 and later withdrawn from the Cochrane Library as no longer maintained, examined randomized trials of single session individual psychological debriefing and concluded that there was no evidence it prevented post traumatic stress disorder, that two trials found worse outcomes at follow up, and that compulsory debriefing of trauma victims should stop. The National Institute for Health and Care Excellence in the United Kingdom has, in its post traumatic stress disorder guidance including the 2018 update, recommended against offering psychologically focused debriefing for the prevention of PTSD, and the VA and Department of Defense clinical practice guideline has taken a similar position, so check the current edition of either before you quote it.

The defenders of the model make a fair objection that readers should understand rather than dismiss. The trials in that review largely studied one-on-one sessions with primary victims, people who had been burned or injured in road crashes, and not group sessions with an intact crew inside a multi-component program, so the intervention tested was not the intervention most fire departments believe they are running. The trouble for that argument is that the controlled evidence for group CISM in occupational populations is thin and methodologically weak, so the fair summary is that the practice has been shown not to work in the form that was tested, and has not been shown to work in the form its proponents advocate. Sources here genuinely disagree, and no trial published to date settles it.

What replaces the single session

The absence of good evidence for one-and-done debriefing does not mean doing nothing, and the practices with better support are dull and cheap. Psychological first aid, as set out in the field operations guide developed by the National Child Traumatic Stress Network and the National Center for PTSD, is essentially structured practical help: safety, basic needs, accurate information, connection to people who matter, and linkage to services, delivered without pressing anyone to describe what they saw. Hobfoll and a large group of co-authors set out five elements in the journal Psychiatry in 2007 that most immediate response models now hang on, which are promoting a sense of safety, calming, a sense of self and collective efficacy, connectedness and hope.

Applied in a station, that looks like the officer taking the unit out of service for an hour, getting food into people, sending a relief crew if one exists, telling the crew what is known about the patient or the investigation, and making sure nobody drives home after a 26 hour shift that ended with a child. It also means a contact three or four days later from someone the member trusts, and another at three or four weeks, because the natural course of acute distress is improvement for most people, and the ones who are worse at a month rather than better are the ones who need a clinician. That interval, sometimes called active monitoring or watchful waiting, is the part agencies skip.

Mandatory attendance is worth singling out. Requiring a member to sit in a group and speak about a call imposes disclosure in front of supervisors and peers who will be riding with them next shift, and it exposes people who were fine to the graphic detail carried by people who were not. A mandatory contact is a different and defensible thing, where a peer supporter or officer reaches the member, confirms they are functioning and tells them how to get more, and the member is free to say that they are fine and end the conversation there.

Confidentiality decides whether anyone calls

Everything else in a behavioral health program is secondary to the question every member asks before dialing, which is who will find out. The legal ground here is narrower than most agencies think. Peer supporters are not licensed clinicians, so the privilege that attaches to a therapist does not attach to them by default, and anything a peer supporter is told can in principle be sought in litigation or in an internal investigation unless a statute says otherwise. Many states have enacted peer support confidentiality statutes covering law enforcement, fire and EMS, and they differ substantially in who is covered, what training the peer must have completed, and what the exceptions are, which usually include threats of harm to self or others, admissions of criminal conduct and mandatory abuse reporting. Have your agency attorney tell you in writing what your state’s statute says and does not say.

Public records law is the exposure that gets missed. A peer support coordinator who keeps a spreadsheet of who was contacted, or who emails a chief from an agency account to say that a named firefighter is struggling, has created a record on a government system that may be subject to your state’s open records act and to discovery. The protection that works is to not create the record: no names in logs, no clinical notes by peers, no case files, and aggregate counts only. Peer supporters should also be told, in training and in writing, that they use personal contact methods for peer conversations and that they do not write anything down.

HIPAA is routinely misunderstood in this context. It binds covered entities and their business associates, meaning health plans, clearinghouses and health care providers who bill electronically, and it does not convert a fire department peer team into a protected channel; employment records held by an employer in its role as employer are not protected health information. Your department may still hold real obligations under the Americans with Disabilities Act to keep employee medical information in a separate confidential file, and under your state’s peer support statute if one exists, but the mechanism matters because members eventually learn whether what they were told was true.

Building a peer team that gets used

Selection determines everything that follows, and rank is the wrong criterion. The people who work are the ones crews already talk to, which you find by asking the floor for nominations rather than by asking battalion chiefs who they think should do it. Keep anyone who holds the discipline function off the team, spread membership across shifts, stations and ranks so that no member has to drive across the county to find one, and include the disciplines that always get left out. Telecommunicators hear the entire incident, often in the caller’s voice, and frequently never learn whether the patient lived, yet many agency trigger lists still require a member to have been on scene.

Training has to be an actual curriculum with a certificate and a refresher interval, not an afternoon. Published options exist from the International Critical Incident Stress Foundation and from the International Association of Fire Fighters among others, and the choice matters less than the requirement that every team member has completed one, that the training date is tracked by your training officer, and that the team meets regularly with a licensed clinician who provides supervision and takes the referrals the peers cannot handle. Peers also need a rotation off the team and their own support, because the people who volunteer for this work are frequently the same people carrying the heaviest exposure.

Scope needs to be written down and enforced. Peer support is contact, listening, practical help and referral, and it is not therapy, diagnosis, advice about medication, or an investigation. Activation should not require permission: publish a policy stating that any member may contact a peer supporter directly for any reason, that the team may self-activate on a defined list of incident types, and that no chief officer approval is needed. The trigger list should name the events at minimum, including line of duty death or serious injury, the death or suicide of a member, any pediatric death, mass casualty incidents, incidents involving a victim known to a responder, prolonged rescues with a conscious patient, and violence directed at responders.

Getting people to clinicians who can help

Peer support is the front door, and it fails if there is nothing behind it. The first thing to test is whether your health plan’s behavioral health network exists in practice: have someone in human resources call ten in-network providers from the plan’s own directory and record how many answer, how many take new patients, how long the wait is and whether they offer telehealth. Directories in every line of business are notoriously stale, and a member who calls four numbers and gets voicemail four times has concluded something about the program that no poster will undo.

Fit matters more here than in most referrals. A clinician who has never heard a scene described will visibly react to one, and members do not go back after that. Agencies that get this right maintain a short vetted list of clinicians who have worked with responders, who understand shift work and duty status implications, and who can say which treatments they actually deliver. The trauma focused treatments that current clinical practice guidelines from the VA and Department of Defense and from NICE recommend include cognitive processing therapy, prolonged exposure and eye movement desensitization and reprocessing, and you should verify the current editions rather than rely on this paragraph, but the practical use of knowing the names is that a member can ask a prospective therapist whether they are trained in any of them.

Two administrative facts shape whether people seek care, and leadership should know both. A number of states have enacted workers compensation coverage or presumption laws for post traumatic stress in first responders since roughly 2017, and they vary enormously in which disciplines are covered, what has to be shown, and what benefits attach, so ask your workers compensation administrator for a plain summary of your own state’s law and its filing consequences. Separately, licensure and certification applications have historically asked about mental health history, and the Federation of State Medical Boards adopted recommendations in 2018 urging boards to limit questions to current impairment rather than diagnosis or treatment history; whether your state EMS office, nursing board or POST commission has followed suit is a question you can answer by reading the actual renewal form.

Schedules, staffing and the discipline process

Ask responders what wears them out and a large fraction of the answer will be administrative. Research on police stress, including the long running Buffalo Cardio-Metabolic Occupational Police Stress work led by John Violanti and colleagues at the University at Buffalo, has consistently found that officers rate organizational and administrative stressors at or above critical incident exposure, and the fire and EMS literature points the same direction. Held-over shifts, cancelled leave, unpredictable schedules, promotional processes nobody trusts, and internal investigations that run for months with no communication all produce chronic strain that no coping skills module addresses, because the problem is not located in the member.

The concept that best describes a particular kind of damage is moral injury, which Jonathan Shay developed in Achilles in Vietnam in 1994 as the betrayal of what is right by a person in legitimate authority in a high stakes situation, and which Litz and colleagues defined in Clinical Psychology Review in 2009 in terms of perpetrating, failing to prevent or witnessing acts that violate deeply held moral beliefs. In a response agency that looks like being held at staging within sight of a patient, working an arrest with no second unit available because the county would not fund one, being ordered to do something the member believed was wrong, or watching an agency publicly blame a crew for an outcome the agency’s own staffing decisions produced. Distress arising that way does not respond to being told to breathe.

The same problem runs through quality assurance. Albert Wu described the clinician involved in a serious error as the second victim in a 2000 article in the BMJ, and the term has since been criticized for how it sits alongside actual patient harm, but the operational point survives: a member who has made a mistake that hurt someone needs both a clinical review and support, and the same person should not deliver both. Separate the functions by name in your QA policy, so that the person conducting the chart review is not the person checking on the medic, and so that a referral to peer support is never recorded as a disciplinary step.

Measuring the program without collecting names

You can tell whether a program is alive without keeping a single identifiable record. Count activations by trigger type, count self-initiated contacts separately from team-initiated ones, count referrals made to the clinician list, and track how many trained peer supporters you have by shift and station with their training dates. Ask your employee assistance provider for the aggregate utilization report they already produce, which gives you a rate without names. Run an anonymous two-question survey with the annual physical or the SCBA fit test asking whether the member knows two ways to reach a peer supporter and whether they believe a peer conversation would stay private, because the second answer is the one that predicts use.

The failure modes are consistent enough to list. A team exists on paper with nobody currently trained, or all five members work the same shift at headquarters. The clinician list was built in 2019 and two of the four have retired. The program is announced at new hire orientation and never again. A chief mentions someone’s counseling in a promotional interview, or a member who disclosed a problem quietly lost a specialty assignment, and every member of that department now knows exactly what the policy is worth regardless of what the policy says. Peer support gets used as a compliance step attached to discipline, at which point the team becomes an arm of the administration in the eyes of the floor.

What leadership does after the bad call is the real communication. Taking the unit out of service for the rest of the shift, telling the crew what happened to the patient when you find out, and sitting with the probationary firefighter at the kitchen table for twenty minutes all register in ways that campaign materials do not, and members compare those actions against the policy language when they decide whether to call anyone. Budget the staffing that allows an officer to make that call without leaving a district uncovered, because an officer who cannot take a unit out of service has no way to act on any of it.

What to do at your agency

• Ask your agency attorney this month for a one page written memo answering two questions: what your state’s peer support confidentiality statute protects and what it excepts, and whether peer support contact logs kept on agency systems are subject to your state’s open records act.
• Have the peer support coordinator stop keeping any record that contains a member’s name, replace it with a tally of contacts by trigger type and by month, and destroy existing identifiable logs after the attorney confirms the retention rules.
• Direct your human resources or benefits coordinator to call ten behavioral health providers from your own health plan’s directory and report in writing how many answered, how many are accepting new patients, the average wait to first appointment, and how many offer telehealth.
• Add telecommunicators, mechanics and other support staff to the trigger and eligibility language in your existing behavioral health policy, and delete any requirement that a member must have been on scene to receive contact.
• Put one item on the agenda of your next officers’ meeting: each company officer states the two specific ways a member of their crew can reach a peer supporter without going through a chief officer, and anyone who cannot answer gets the card before they leave the room.
• Have your training officer produce a current roster of peer team members with training dates, shift and station, and schedule refresher training for anyone past their refresher interval before the end of the quarter.
• Have your operations chief pull the last twelve months of call volume between 2200 and 0600 by unit, and bring the three busiest units to the next budget discussion as a staffing and sleep question rather than a wellness question.

Takeaways

• The exposure that damages responders accumulates over years of ordinary bad calls rather than arriving in one catastrophic incident, which means programs built only around rare events miss most of the problem.
• Published prevalence estimates vary widely by discipline and instrument, with a 2012 meta-analysis by Berger and colleagues reporting a pooled current PTSD prevalence around ten percent in rescue workers, and suicide counts from voluntary reporting programs should be treated as floors rather than totals.
• The Cochrane review by Rose, Bisson, Churchill and Wessely found no evidence that single session individual psychological debriefing prevents PTSD and recommended that compulsory debriefing stop, while proponents of group CISM argue the trials tested a different intervention, and the controlled evidence for the group model remains thin.
• Practical help, accurate information, a contact at several days and another at several weeks, and referral for anyone who is worse rather than better at a month, are the elements that current mass trauma guidance supports.
• Peer supporters hold no clinical privilege by default, state peer support statutes vary in scope and exceptions, HIPAA does not cover a department peer team, and public records law can reach any log the team creates, so the safest protection is to write nothing down.
• A peer team works when members are nominated by the floor, kept out of the discipline chain, trained on a real curriculum with a clinician providing supervision, and reachable without any chief officer’s permission.
• A referral path is only as good as the clinician list behind it, which means testing your health plan’s behavioral network by telephone and maintaining a short list of clinicians who have actually worked with responders.
• Sleep, staffing, predictable schedules, timely internal investigations and the separation of quality assurance from support do more for behavioral health than any resilience course, because they change the exposure rather than the individual’s coping.