My emphasis regarding drowning signs is to act on them even when they do not
comport to our expectations or to textbook notions of how they should look. My
expectations are lifeguards should be prompted to respond when there is doubt
about what it is they perceive. This is particularly important with respect to
I have investigated several incidents where the submerged victim was not recognized by the guards. How often do you hear tales about patrons approaching lifeguards with suspicions of submerged victims the guards failed to recognize? Victim recognition is most primal, and it hinges on so many factors that conspire against the lifeguard.
Case in point:
A13 year-old African-American boy drowned during public hours. The victim was tall and extremely thin and wore dark blue swim shorts on this day. The pool had black tile lane lines 12-inches across, making them unusually wide. The victim as not a good swimmer and the pool bottom sloped precariously from shallow to deep. The guards recall seeing the victim while in the shallow water but lost track of him.
Unfortunately when the victim succumbed he landed right on top of one lane line rendering him practically invisible. One lifeguard was informed by another young bather “there is a guy down there” but even so the guard was incredulous as he could not see him. It was only after scrutinizing the area that the guard made out the outline of the body by observing the straight lines of the lane had contours where the body extended beyond into the white pool bottom, and an angled, slim limb extending seemingly disconnected from nowhere.
The time the victim was on the bottom could not be accurately ascertained but suffice it to say he’d been down there long enough that paramedics were unable to revive him.
Guards were posted on towers, the pool water was clear and there was light to moderate activity. By all accounts this should not have happened but it did. This is a case where you have the convergence of factors masking a situation that defies typical drowning signs.
My point is that anomalies should trigger a response, regardless of whether they conform to typical drowning signs. In the face of doubt, it is better to act than not to act because acting on the confidence of the signs alone is an equally unreliable determinant.
I don’t want to put words in Cory’s mouth but I suspect this is Cory’s point, a point that we can ill afford to ignore.
I am sorry that that happened, but there is nothing in the scenario you described that I would call atypical of the submerged victim, as you described it. In addition, I do not recognize how running drills that those that you describe can help the lifeguard in this situation. As proponents of these drills, you seem to imply that lifeguards will not employ good surveillance without these drills, but that is simply not the case. The drills, in fact,are not required to ensure good surveillance, and I believe the drills themselves,as an intrusion,can cause a delay in recognizing true victims. (BTW, I am not saying that this happened in this case. There is nothing in your scenario that would make me think the lifeguards were doing anything but their best.)
I have never maintained that “lifeguards will not employ good surveillance without these drills” as you claim but only that drills are necessary to gauge transfer of learning. Please read any of my previous postings for verification.
You also argue drills are intrusions that can hamper “recognizing true victims.” In turn I have suggested that high-fidelity drowning signals arouse the cognitive snap for drowning recognition that generates an intrusion in technicality only (since it introduces a false positive for drowning). Furthermore, since high-fidelity drills do not re-task lifeguards away from surveillance they are “controlled intrusions,” implying a class of intrusions within a new paradigm not formerly considered.
You also suggest live drills pose a risk factor to the public and to staff (a risk other responders recognize and hence the prohibition against live audits). In response I argued that other responder industries are adverse because live audits generate identifiable risks in accordance to accepted risk management indices, but these are absent for Lifeguarding as evidenced by the analytics.
True, disagreement is not evidence, but it is based on the evidence I have gathered throughout this thread. While you dismiss Joe's "turtle's and boxes" argument (which is the best analogy to date I've read) as too simple, you similarly are over-simplifying things by saying all victims will look like victims. I highly recommend the DVD's "Disapearing Dummies," by Dr. Tom Griffiths, and the new E&A VAT Video.
Just to clarify some word-twisting you performed - I never said drowning victims would look like red balls and caps on children. I believe in the use of live GID's, and 3-dimensional mannequins, because red-balls and caps will never look like a distressed conscious, or unconscious guest. I also never said that distressed and drowning victims won't display recognizable behaviors, I simply argue that someone who suffers from, for example, a stroke in the deep end of a pool, 25 feet away while other people are splashing around, won't necessarilly look like a body underwater.
To the statement "they won't look like trash, manikins or shadows" - I'd challenge any new aquatic professional to watch the videos I mentioned above, and make that decision for yourself. (and yes, I have been a lifeguard for years, and teaching lifeguarding for years, and teaching lifeguard instructors for years, and managing facilities for years, as well)
This link describes the contents of the video, and reinforces the points Joe and I have been making:
I am sorry if you think I have been twisting your words. I thought someone said he wanted them to see every bit of trash in the pool and how can you detect a victim if you cannot see a red ball. Those are not direct quotes but that is what I have been hearing.
The problem with the crates analogy is if a real victim is a crate with a turtle underneath, a false victim is not a crate without a turtle because that implies a seamless fake which does not occur. Fake victims can only approximate actual victims; some do not even approximate human beings let alone display victim behaviors. Even 3-dimensional manikins and live humans pretending to drown only imperfectly approximate real drowning victim. So they are hardly equivalent to crates without turtles inside. (So, hopefully without offending you, I am troubled that the crates analogy is the best analogy to date that you have read.)
It is not over-simplifying to state that victims display certain characteristics that aid in recognition. This is a fundamental principle of patron surveillance that I have tested throughout my own career as a lifeguard. I am familiar with Dr. Griffiths work, and I have seen this phenomenon and other types of glare and blind spots. These are challenges to be recognized and overcome, but they do not require the intrusion of on-duty drills to overcome. In fact, on-duty drill do not adequately address the complexies of victim recognition.
What I mean is this: You must train lifeguards about the behaviors to look for that indicate a victim as well as weak swimmers and hazardous conditions and dangerous practices. New lifeguards should shadow experienced guards to further hone their recognition skills in various areas of the facility. In-service training should include simulated rescues to get lifeguards practice moving through a rescue and basic life support procedure as well as activation of the EAP. New guards on rotation should be monitored by the PIC to determine if they are scanning properly and catching behaviors in their area that they should react to and control. On occasion, when you want to test a lifeguard, pull the lifeguard out of rotation, place a victim in the water and see if the lifeguard understands how to initiate the EAP, make a proper rescue, and provide necessary care. Have the PIC monitor all lifeguards daily to make sure behaviors that are inappropriate are being recognized and controlled.
These training techniques and others, combined with the lifeguard's knowledge and experiences, makes for a well trained lifeguard. Other factors are also required: head lifeguards with proper attitudes and work ethics, the concept of the lifeguard team, etc. Please note that new lifeguards require some training and orientation before they can enter the rotation. Please also note that on-duty drills are not included in this training program.
When you add on-duty drills, you must train lifeguards to expect something else or in addition to real victims, hazards, and practices. They must also expect drills. You may show them what they are looking for or the types of "fake" victims to be deployed and how the lifeguard is to respond and how far to carry the rescue. You may inform them of a time standard that you expect and maybe even consequences for poor performance. Once you do this, the intrusion has been imposed and the lifeguard's attention has been divided.
The extent of any distraction or intrusion is difficult to measure, but it is undesirable in any degree. The only sure way to eliminate distractions and intrusions are to prevent them all together, which is what I have been proposing from the start.
Eliminating this intrusion does not eliminate the necessary training that a lifeguard should receive. Many lifeguards grow up to be excellent professional rescuers without this problematic, intrusive practice of on-duty drills. Train your lifeguards, yes. Use these on-duty drills, no!
You have misconstrued the analogy. It is intended to demonstrate signal bifurcation (attention between two goals) and how it can be avoided. If you want a specific treatment of signal integration for a drowning scenario, please refer to my three scenario analogy in a previous posting.
Yup - it all comes back to hi-fidelity vs. low-fidelity drills, which we've already hashed out ad nauseum. Thank you for the Signal Taxonomy - I will be sharing it with my staff!
I have constructed a Signal Taxonomy matrix for your review in the hopes it will clarify the working relationships among the drills we have been we have been discussing. I believe the schema is self explanatory with one exception.
There are two factors I wish to elaborate on before moving on.
One factor is the cued/non-cued distinction made in the Taxonomy matrix. The intent is to distinguish drills that are cued (those that have embellishments that identify the “victim” as part of a drill), and non-cued “victims” (those that are not embellished and rely on the merits of associated drowning signals for their identification). It is important the “victim” be identified on the merits of drowning signals by lifeguards in order to assess learning transfer.
Another factor I have not discussed but wish to address is the potential for what Ron describes as “anticipatory feelings towards expected drills.” If I am interpreting this right, this can be rightly referred to as a pretreatment signal load.
A pretreatment signal load posits that there are staging steps the lifeguard will learn to look for in anticipation to a treatment or drill in our case, and this anticipatory search competes against the drowning signal search for attention (exactly what Ron predicts with his “anticipatory feelings”). The more involved the pretreatment staging the greater the potential for a new generative signal.
The only way to avoid this is to ply low pretreatment (staging) signals that tend not to trigger an anticipatory search. Surreptitiously slipping a Timmy dummy into the pool seems easy but generates a high pretreatment signal because it demands a series of staging actions necessary to conduct the drill that lifeguards will learn to anticipate.
Using a silhouette for drills, on the other hand, may generate a lower pretreatment signal load than a Timmy (presumably because a dummy is more cumbersome) but is still substantially more unwieldy than a physiognomic, biological, non-cued, passive “victim,” which carries no significant pretreatment signal load (as outlined in the attached Signal Taxonomy matrix).
A low pretreatment signal load minimizes signal search bifurcation because the lifeguard will not be incentivized to search for pretreatment signals while remaining focused on the search for drowning signals.
Your feedback, as usual, is greatly appreciated.
I like the taxonomy. it outlines fairly well the reason I do not use cued drills. Do you think the use of shadow dolls/silhouettes placed prior to a rotation, when the guard is not present, to test a bottom clear, would fall in the cued or non-cued category?
By the way, there is a typo under the pre-treatment signal load explanation; I believe you meant to write anticipatory cues, not cures.
Lisa these are very good questions.
I reserve “cued” and “non-cued” distinctions for physiognomic, biological drowning models to distinguish embellished from unembellished signals within a drill. For example a swimmer “victim” wearing a yellow tee-shirt or red cap is a signal embellishment for “drill in progress.” This operant needlessly generates non-drowning signals that the lifeguard will respond to, yet it is the measure to drowning response.
A nondescript swimmer “victim” is non-cued when the lifeguard must rely on drowning signals alone to identify him, and therefore is said to be “non-cued.”
A silhouette does not necessarily fall into the modality of cued or non-cued. This is particularly true when silhouettes convey high fidelity drowning signals for a bottom drowning. When silhouettes are used for vigilance testing, the potential problem lies with the “pretreatment signal” generated during the staging steps to execute the drill. Pretreatment staging potentially prompts signals lifeguards anticipate and search for; and these anticipatory signals, in turn, will tend to compete for attention with drowning signals (creating signal bifurcation).
In other words, the fewer the pretreatment signals generated, the less incentivized lifeguards are to look for them and the less likely to produce a needless signal split. A passive swimmer “victim” possesses the highest fidelity drowning signal precisely because it most closely mimics the drowning paradigm and requires low-intesnity pre-staging.
If lifeguards are not present when silhouettes are positioned, this might relieve some of the problems but be cautious as any pattern associated with drill preparation has the potential to generate anticipatory signals that vie with the drowning signal search.
I recommend changing things up to prevent an emerging pattern and to execute drills that are less onerous so lifeguards do not come to rely on pretreatment signal capture. It would probably be very helpful if a new species of drowning dolls and silhouettes were marketed that are easy to handle and inconspicuous, to minimize the manifestation of staging signals. A market can only emerge when there are demands for the product.
Thanks for pointing out my clerical error. I will correct it for the next iteration. And thanks again for your feedback.
I agree, some easy to handle dolls/silhouettes would be helpful. I can sometimes get my regular lap swimmers who wear trunks to hide the silhouettes in them and drop while they swim, but most regular lap swimmers don't wear loose, big trunks.
Unfortunately, I do not accept that unembellished substitute victims exist. Even patrons or staff pretending to be a victim do not display the behaviors to the extent necessary to be considered non-cued. For example, can a victim hold his or her breath and remain motionless with muscles relaxed for the time it takes to be recognized and rescued? Does a victim pretending to be an active drowning victim correctly display the instinctive drowning response? If a lifeguard can determine that a patron is faking a drowning in the pool, how does he or she know that a drill is in progress and not just a patron trying to bait the lifeguard? Certainly, inaccurate portrayals of drowning victims can be a cue but they can also be a patron goofing around.
A silhouette is most certainly a "cued" false victim. I have seen silhouettes and manikins used as false victims in such drills, and I have never seen a lifeguard enter the water thinking he or she was rescuing a real victim.
You cannot avoid the potential for so called "signal bifurcation" or the division of attention and focus caused by the implementation of such drills. My consistent message though out this thread has been to point out the problems with integrating intrusive drilling that competes with the lifeguard's attention on his/her primary duties.
In my own experience, as a first year lifeguards (many, many, many years ago), i was well trained before I ever climbed onto a tower using techniques that are still applicable today. When our manager introduced the red shirt and blue shirt drills and let us know that our times would be a large part of our evaluation, I realized even back then that this was a distraction. I made up my mind that I was going to keep my eye on the prize (i.e., focus on actual victims, hazards, etc.) and let the chips fall where they may. I wasn't going to devote energy to a daily focused search for these decoys. Other lifeguards felt the same way I did (because we discussed it in private afterwards).
As I mentioned in an earlier post, I was lucky enough to notice behaviors of the false victim and the manager that tipped me off and allowed me to switch gears and conduct a quick search for the drill. I imagine if I had not seen those signs that my times would have been much longer and this would have affected my performance review even though I was actively scanning the pool and maintaining a focus on my primary task. The quality of my work was not reflected in how fast I recognized the drill except in the most general way, and a delayed time would also not accurately reflect the quality of my work.