Have I lulled myself into a false sense of security?

[yanks26dmb]

So I do a lot of ER work; in person and tele. I understand the risks of bad outcomes, lawsuits, etc are higher in this environment. I believe I received good training during residency from a well regarded forensic attending who has testified as expert witness in several nationally prominent cases, as well as for and against psychiatrists in malpractice cases. He gave us several lectures on writing suicide risk assessments and limiting medicolegal liability by documenting appropriately. One thing he always stressed was "you can be wrong, but you can't be negligent."

As such, I write pretty detailed suicide risk assessments on patients. I see colleagues and their notes are much more brief than mine. They document something like patient is low risk of self harm, okay to discharge. That's it, no reasoning, no weighing of protective factors or risks.

Whereas I have a lengthier boilerplate template and update with relevant facts. In the case of a patient with clear pattern of malingering, I will write something like this...

Patient is deemed to be low imminent risk of self harm at this time. This determination is made after weighing both their protective factors against their risk factors for imminent self harm. Protectively they are future oriented in they discuss upcoming doctor's appointments and plans, they show no objective signs of depression, anxiety, mania or psychosis on exam, they are close to and know how to access MH resources in the community, they have some family support and connection to social services, they are medication compliant, they have no access to guns or lethal weapons. Further they endorse contingent suicidality based upon housing. Contingent suicidality is not well treated by inpatient hospitalization and studies on the matter do not demonstrate increased rates of suicide at 6 months. Further, hospitalization in such cases can do harm to the patient given the fact it can positively reward maladaptive coping strategies. Their risks include their race, gender, drug use, perception of depressed mood, impulsivity, and low SES. In weighing all these factors I have determined them to be low imminent risk of self harm. Further, based on California LPS statute, they do not meet criteria for involuntary hold and therefore must be released.


So this is quite wordy but I feel like it is both thorough, thoughtful, and protective should a lawyer look through notes to see if a case exists. Curious if I have being naive here or if others would concur. As an attending coming up on the end of my first year out of residency, I appreciate the feedback from my more experienced colleagues.

---

Members don't see this ad.

 

[Bartelby]

I completely agree with the advice your forensic attending gave. It is very much worth the time to write a detailed risk assessment (like the one you mentioned above) in high-risk situations like a discharge of a potentially suicidal patient from the emergency room. One-liner risk assessments like “low risk for discharge, appropriate for outpatient“ without any demonstration that you have utilized reasonable medical judgment can end up causing big problems for you after a bad outcome.

For something like stable low risk outpatient follow ups you of course do not need to document extensive risk assessments every time, but in the emergency room your main job is to assess and appropriately manage acute risk.

 

[Osminog]

For ass-covering purposes, writing a long risk assessment is worthwhile. Something that's worth noting, however, is that it's strictly for documentation purposes and should have little to no impact on the real-life patient care decisions. Patients should be treated on the basis of their specific situations and circumstances. The "high-risk" and "low-risk" lingo used in notes strictly serves as a rationalization, a façade of objectivity that appeals to lawyers and judges.

Individualized risk categorization is not an evidence-based, clinically useful practice, and there is virtually nothing "medical" about it. There isn't a shred of evidence that psychiatrists are able to accurately predict individuals' future behavior. But until the legal system (and public at large) recognize that psychiatrists aren't clairvoyant and aren't responsible for what people choose to do outside hospital walls, careful documentation is a must.

 

[yanks26dmb]

For ass-covering purposes, writing a long risk assessment is worthwhile. Something that's worth noting, however, is that it's strictly for documentation purposes and should have little to no impact on the real-life patient care decisions. Patients should be treated on the basis of their specific situations and circumstance. The "high-risk" and "low-risk" lingo used in notes strictly serves as a rationalization, a façade of objectivity that appeals to lawyers and judges.

Individualized risk categorization is not an evidence-based, clinically useful practice, and there is virtually nothing "medical" about it. There isn't a shred of evidence that psychiatrists are able to accurately predict individuals' future behavior. But until the legal system (and public at large) recognize that psychiatrists aren't clairvoyant and aren't responsible for what people choose to do outside hospital walls, careful documentation is a must.

Oh yes, well aware of this. Primarily asking for legal purposes, is my long winded spiel enough to protect me if a bad outcome happens? I realize all that mumbo jumbo I am writing down are just hoops to jump through and have no bearing on my understanding if the patient is going to go home and off themselves or not.

 

[splik]

Good documentation is important but it is not enough. More important I would argue is the actual clinical care you provide to patients. People come up with such nonsense like "if it's not documented it didn't happen". But we do things we don't document all the time, and clinician frequently (though unethically) document things they never did. Before an attorney will file a malpractice complaint, they will have a psychiatrist review the records to see if there is a case. Good documentation can help prevent a lawsuit being filed. It can also make what you say sound more credible (as reflected in contemporaneous records) in the event of a lawsuit. At the same time, the form of documentation also matters.

Write a lot, and it is less forgivable if you miss something. If you don't write much anyway, an omission is more forgivable. Which is why one of my attendings used to say "the more you write, the more rope you give for the lawyers to hang you with."

From my perspective, I wouldn't give much credence to the example you gave above which is quite weaselly. I would like to see things like were they given any medication, some time to decompress in the ER ("therapeutic rest"), offers of crisis diversion or homeless shelters, referrals to appointments in the community, reduction in clinical risk factors based on the therapeutic aspects of this encounter, how did they respond when you discussed your plan with them etc.

When things are too wordy it can back fire. I do recommend detailed documentation if discharging pts presenting with SI, but it should be very clear without any potential to "read between the lines." For example, I don't like "they show no objective signs of depression, anxiety, mania or psychosis on exam" because it implies they had subjective symptoms. I also don't like "Further, hospitalization in such cases can do harm to the patient given the fact it can positively reward maladaptive coping strategies" because that doesn't really sound like a significant harm when weighed against the potential risk for suicide. And you have to be careful about saying a pt expressing SI does not meet criteria for involuntary treatment, when the bar is very low from a legal perspective in most states to detain someone. You also want to document positive protective factors (pertinent negatives as described are important but they are not protective factors).

In short, the best thing you can do to reduce your risk is provide excellent clinical care to your patients and having your records reflect that. If you have a pt who presents with contingent SI, they have other risk factors and a history of mental disorder, significant psychosocial stressors and you have no documentation showing they have not benefited from hospitalization in the past, and they commit suicide upon your discharge, you will have a tough time defending yourself. On the other hand, if it is clear that the patient repeatedly seeks help in the ER, has had multiple hospitalizations in which they have been documented to not benefit, disrupt the milieu for other patients, or using the hospital for shelter rather than stabilization, you have a compelling case to discharge the patient.

People get too hung up on suicide risk assessment. No one ever killed themselves because you didn't document a risk assessment*. Of course, you are expected to do so. However, when I review a case, what I'm interested in is were there departures from the standard of care that led to this outcome? So I'm more interested in what you actually did for this patient than in your assessment of justifications for discharge.

*I actually reviewed a wrongful death case last week where there was no suicide risk assessment. I didn't ding the doctor for it, and for various other reasons told the attorney there was no case against the psychiatrist.

 

[Candidate2017]

100% yes, you're creating a false sense of security. Go to your local court website and look up "bad" and "good" psychiatrists you know, and you will find there is no correlation between number of lawsuits and "bad" or "good" psychiatrists.

Broadly speaking, lawsuit frequency is determined by your choice of patient pop and work setting (of which the ED setting is higher frequency). But do you know who are the experts on the full extent of risk? Malpractice insurance carriers. They literally practice evidence based science. They crunch extensive empirical data to stratify and reduce their risk. On a malpractice insurance application, they have 50-100 questions for you. Answer yes to any of them and your premium goes up. But I've never seen any application ask whether I write really long risk assessments or none at all.

Beyond controlling for such risk factors, following standard of care, and being likable, there's nothing you can do. Accept the risk and realize that bad outcomes =/= malpractice. We're physicians, we deal with serious diseases that not infrequently have bad outcomes.

Ironically, if you think about it, having a lot of "bad" psychiatrists is the best way to reduce risk. If hundreds of psychiatrists in your community can testify all they do is read someone's chakra in the ED, then that's the standard.

 

[splik]

Ironically, if you think about it, having a lot of "bad" psychiatrists is the best way to reduce risk. If hundreds of psychiatrists in your community can testify all they do is read someone's chakra in the ED, then that's the standard.

I know this is mostly tongue in cheek, but it wouldn't help as most places don't have a community standard it's really a national standard. Where available, you look towards national guidelines, say from APA or other professional associations to set standards. There was a famous case years ago where a psychiatrist argued that since he was the only psychiatrist in the area, his care was the standard of care. He was not successful.

 

[yanks26dmb]

Good documentation is important but it is not enough. More important I would argue is the actual clinical care you provide to patients. People come up with such nonsense like "if it's not documented it didn't happen". But we do things we don't document all the time, and clinician frequently (though unethically) document things they never did. Before an attorney will file a malpractice complaint, they will have a psychiatrist review the records to see if there is a case. Good documentation can help prevent a lawsuit being filed. It can also make what you say sound more credible (as reflected in contemporaneous records) in the event of a lawsuit. At the same time, the form of documentation also matters.

Write a lot, and it is less forgivable if you miss something. If you don't write much anyway, an omission is more forgivable. Which is why one of my attendings used to say "the more you write, the more rope you give for the lawyers to hang you with."

From my perspective, I wouldn't give much credence to the example you gave above which is quite weaselly. I would like to see things like were they given any medication, some time to decompress in the ER ("therapeutic rest"), offers of crisis diversion or homeless shelters, referrals to appointments in the community, reduction in clinical risk factors based on the therapeutic aspects of this encounter, how did they respond when you discussed your plan with them etc.

When things are too wordy it can back fire. I do recommend detailed documentation if discharging pts presenting with SI, but it should be very clear without any potential to "read between the lines." For example, I don't like "they show no objective signs of depression, anxiety, mania or psychosis on exam" because it implies they had subjective symptoms. I also don't like "Further, hospitalization in such cases can do harm to the patient given the fact it can positively reward maladaptive coping strategies" because that doesn't really sound like a significant harm when weighed against the potential risk for suicide. And you have to be careful about saying a pt expressing SI does not meet criteria for involuntary treatment, when the bar is very low from a legal perspective in most states to detain someone. You also want to document positive protective factors (pertinent negatives as described are important but they are not protective factors).

In short, the best thing you can do to reduce your risk is provide excellent clinical care to your patients and having your records reflect that. If you have a pt who presents with contingent SI, they have other risk factors and a history of mental disorder, significant psychosocial stressors and you have no documentation showing they have not benefited from hospitalization in the past, and they commit suicide upon your discharge, you will have a tough time defending yourself. On the other hand, if it is clear that the patient repeatedly seeks help in the ER, has had multiple hospitalizations in which they have been documented to not benefit, disrupt the milieu for other patients, or using the hospital for shelter rather than stabilization, you have a compelling case to discharge the patient.

People get too hung up on suicide risk assessment. No one ever killed themselves because you didn't document a risk assessment*. Of course, you are expected to do so. However, when I review a case, what I'm interested in is were there departures from the standard of care that led to this outcome? So I'm more interested in what you actually did for this patient than in your assessment of justifications for discharge.

*I actually reviewed a wrongful death case last week where there was no suicide risk assessment. I didn't ding the doctor for it, and for various other reasons told the attorney there was no case against the psychiatrist.

I appreciate your input. Just a few questions that now come to mind.

1. What about that risk assessment is weasely? I listed facts pertaining to the case and my basis for no invol hold based on California statute. Not arguing with you, just generally curious.
2. I do agree inclusion of more specific interventions would be more beneficial, I will begin to include these things.
3. Risk of reinforcing maladaptive behavioral patterns obviously pales in comparison to risk of suicide, however don't most things? Why not give every depressed patient an SSRI? The risks of 5-10 mg of lexapro pale when compared to potential suicide too
4. Also agree about discharging patients suspected of malingering, contingent SI without clear history of such. I will be much less likely to d/c in this scenario...the above mentioned case is for a patient with clear history of malingering behavior and >10 hospitalizations without clear benefit.

Thanks for your valuable input.

 

[TheRightMedication]

Ive had a happy career of 10 years with one lawsuit. I document a brief risk assessment listing chronic and acute risk factors, choosing low/med/high risk, and yes/no on imminent threat.

Then a line about interventions - med change? Referalls vs continue to follow with current care? Family will remove guns or help manage meds, or a phone call to pharmacy to give meds in smaller increments?

Like splik says, its important to generate a full list of what you can do for the patient, and address each item if possible.

 

[calvnandhobbs68]

Further they endorse contingent suicidality based upon housing. Contingent suicidality is not well treated by inpatient hospitalization and studies on the matter do not demonstrate increased rates of suicide at 6 months. Further, hospitalization in such cases can do harm to the patient given the fact it can positively reward maladaptive coping strategies. Their risks include their race, gender, drug use, perception of depressed mood, impulsivity, and low SES. In weighing all these factors I have determined them to be low imminent risk of self harm. Further, based on California LPS statute, they do not meet criteria for involuntary hold and therefore must be released.

I would say you're also arguing two different things here. One of the first breakpoints you have to do in your assessment is, is this an assessment for a voluntary admission or an involuntary admission? Your first sentence suggests this is voluntary....sounds like a homeless patient stating they're suicidal in order to obtain housing. If you offered them admission, it sounds like they would admit voluntarily. Your primary concern in this situation would be to argue why you feel an inpatient hospitalization is not the best setting for them to be treated in at this moment, DESPITE their stated suicidality and willingness to voluntarily admit to the hospital.

You then give some reason why they wouldn't meet criteria for involuntary admission, which is essentially irrelevant in this case and would probably only hurt you in a lawsuit as everyone would be quite confused as to why you included this.

I agree that you include too many things (of course ignoring the fact I think suicide risk assessments are ridiculous anyway but I digress....). Your assessment needs to be about this patient right now, not vague references to studies about outcomes 6 months from now or some theoretical stuff about patients' maladaptive coping strategies being reinforced.

 

[comp1]

I agree that it's bizarre to attempt to compare the risks of anything to a completed suicide. That kind of post hoc analysis by expert witnesses I think is what leads judges and juries to think we are supposed to have psychic abilities. Ultimately, you didn't believe the patient was going to commit suicide, so you're comparing the risks and benefits of what you are recommending the patient does, not something that you don't think the patient will do. Most, the vast majority, of patients will not kill themselves. However, a large chunk, possibly a majority, of that same patient population will be harmed by psychiatric admission either by the trauma of it inherently or from keeping them developing functional coping skills. And sure you can argue it's better they experience that harm than commit suicide, but that connection is not particularly relevant. It's better they experience that harm than have a heart attack or get run over by car too. You aren't anticipating any of those things happening.
I do agree that documenting, at length, the lack of benefit from prior hospitalizations is important. Very few specialties get as much pressure to keep trying the same thing over and over again expecting a different outcome than psychiatry. I don't at all related to the idea of word being used as the rope to hang you. I've never thought I wrote too much after a bad outcome. Quite the opposite, I've always thought I should have written more, even though I am extraordinarily verbose as it is.

 

[Psychferlyfe3000]

Are there any examples out there of legally very sound assessments? It is hard to fully grasp all these recommendations without specific examples!

 

[jbomba]

I'm guessing the legally sound documentation examples isn't in our awareness, for the most part, because lawyers aren't taking these cases to begin with.

 

[clausewitz2]

Are there any examples out there of legally very sound assessments? It is hard to fully grasp all these recommendations without specific examples!

Your malpractice insurer probably has online modules and/or resources about this. Worth investigating what they can offer you info wise. These are the people who have the best handle on what and what does not get people in trouble in court.

 

[FlowRate]

Are there any examples out there of legally very sound assessments? It is hard to fully grasp all these recommendations without specific examples!

I always liked this TLP The Last Psychiatrist: How to Write A Suicide Note

You should probably also meet standard of care by using some sort of "evidence based" assessment, primarily CSSRS.

 

[Stagg737]

I always liked this TLP The Last Psychiatrist: How to Write A Suicide Note

You should probably also meet standard of care by using some sort of "evidence based" assessment, primarily CSSRS.

CSSRS is not actually an assessment tool, just a screener. It's also trash. One of my QI projects in residency was on the Columbia at our VA and significant amount of data found it was less consistent than a coin flip. There's a reason that VAs require the comprehensive suicide risk assessment (CSRE) for people with a positive Columbia.

 

[FlowRate]

CSSRS is not actually an assessment tool, just a screener. It's also trash. One of my QI projects in residency was on the Columbia at our VA and significant amount of data found it was less consistent than a coin flip. There's a reason that VAs require the comprehensive suicide risk assessment (CSRE) for people with a positive Columbia.

Fair semantic objection to use of the word assessment. Like you said, it's a risk level screener. And any screener is going to be "trash" as an assessment.

I would assert that there is no tool or person who is particularly great at predicting suicide, so there's a lot of "flipping coins" when it comes to suicide assessments, statistically. I looked at the CSRE and it seems to formalize the components that we all should have learned go into, well, a comprehensive suicide risk assessment. My recollection is that most useful evidence in the space has more to do with lethality mitigation than anything else but would be very open to learn if there's been some sort of advancement more recently.

 

[splik]

CSSRS is not actually an assessment tool, just a screener. It's also trash. One of my QI projects in residency was on the Columbia at our VA and significant amount of data found it was less consistent than a coin flip. There's a reason that VAs require the comprehensive suicide risk assessment (CSRE) for people with a positive Columbia.

Yes I believe at least one study has shown that the CSSRS is worse than useless because it misses the people who actually kill themselves. But TJC really pushed it and now everyone uses it. It does provide some useful info but you are right that it is not a risk assessment tool alone

 

[clausewitz2]

Yes I believe at least one study has shown that the CSSRS is worse than useless because it misses the people who actually kill themselves. But TJC really pushed it and now everyone uses it. It does provide some useful info but you are right that it is not a risk assessment tool alone

The agency I work for part time recently decided that every single outpatient note was going to require completing the questions of the CSSRS at every single visit, and if any of them have positive responses, a box pops up requiring you to explicitly justify why you did not decide to try and hospitalize the patient or push them to a higher level of care.

You cannot make this stuff up.

 

[vanfanal]

CSSRS is not actually an assessment tool, just a screener. It's also trash. One of my QI projects in residency was on the Columbia at our VA and significant amount of data found it was less consistent than a coin flip. There's a reason that VAs require the comprehensive suicide risk assessment (CSRE) for people with a positive Columbia.

They're done all the time at the VA, even though their own 2024 guidelines don't recommend for or against.

 

[FlowRate]

Yes I believe at least one study has shown that the CSSRS is worse than useless because it misses the people who actually kill themselves. But TJC really pushed it and now everyone uses it. It does provide some useful info but you are right that it is not a risk assessment tool alone

I'm doing some searching and not finding studies as damning as you're implying. Which articles are you referring to? Pretty much all suicide prediction tools "miss the people who actually kill themselves," but usually have at least some small predictive value and are roughly in line with similar tools. That's the inherent problem with trying to predict a statistically rare outcome. (Not trying to defend the CSSRS, see "evidence based" being intentionally placed in quotes in my comment above, referring to TJC's "rationale" for forcing everyone to use it, but I'm not aware of it being "worse than useless.")

 

[DOswag]

I had just read through this thread 2 days ago and had been thinking about documenting risk assessment (I do this of course, just thinking about how I do it, our template for it in our emr etc.).

I work outpatient adults, I have patients come in all the time with various levels of risk (as I’m sure we all do) but yesterday I had my first one in a long time (thankfully) that I truly felt was at imminent risk. Hx of suicide attempts, vulnerable developmental profile, extremely toxic relationship with spouse that she is going through a nasty divorce with. Cops called frequently to their house for domestic issues, has nowhere else to live and no support, financial stress, housing stress. She attempted suicide via overdose and asphyxiation at about 9-10 in the evening and was in our clinic at like 6 in the morning for a tms treatment sharing this with our tms technician, which then of course escalated to me. I didn’t have a scheduled appointment with her that morning but it turned into a saga. Adamant she was not going to the hospital, did not want police involved, still endorsing suicidal thoughts with plan, intent, continued access to her meds and ligatures at home, no mitigation of her risk factors and wanted to leave immediately to return right back to her home where her spouse was. No insight into why this is so concerning. Emergency services contacted but patient left our clinic and got in her car, drove off, ems diverted to her house which she was not there, she eventually came back to the clinic, eventually spoke with police in our parking lot and they wouldn’t escort her to the ER 🤦‍♂️. She called the clinic later and said she was at home, suicidal, about to overdose, pretty much repeat of what happened this morning but she was at home at this point. EMS called again and last report I got was she was being emergently detained to the ER. Hopefully she got a 72 hr hold which here (TX) is an order of protective custody or OPC. Spent forever last night documenting a novel about all this in detail and justification for why she was at imminent risk of harm to self in lower level of care.

I feel like I did my best to document this. But she’s had a history of issues with another doctor here in town, complained to the medical board which supposedly got some kind of investigation on the doc. I’m not worried about it in the sense of I’ve done anything wrong just the pain in the ass it would be going through that.

 

[DOswag]

This was first thing in the morning. It was a little tough to turn right around and see my 20 patients for the day without being a little distracted as we didn’t get the word that they’d come to her house until about 5pm.

 

[Stagg737]

They're done all the time at the VA, even though their own 2024 guidelines don't recommend for or against.

Yes, that was my point. PHQ-2 was mandatory for ALL patients and not just psych. If answering yes to either question, CSSRS was triggered. If someone screened positive on CSSRS then it triggers a CSRE as stated above, which was insane levels of bureaucratic inefficiency when first rolled out to the point that my PD actually threatened to pull all residents from the VA. They adjusted their protocol pretty significantly after that and to my knowledge has become much more reasonable since then, but it was initially a disaster.

 

[Stagg737]

I had just read through this thread 2 days ago and had been thinking about documenting risk assessment (I do this of course, just thinking about how I do it, our template for it in our emr etc.).

I work outpatient adults, I have patients come in all the time with various levels of risk (as I’m sure we all do) but yesterday I had my first one in a long time (thankfully) that I truly felt was at imminent risk. Hx of suicide attempts, vulnerable developmental profile, extremely toxic relationship with spouse that she is going through a nasty divorce with. Cops called frequently to their house for domestic issues, has nowhere else to live and no support, financial stress, housing stress. She attempted suicide via overdose and asphyxiation at about 9-10 in the evening and was in our clinic at like 6 in the morning for a tms treatment sharing this with our tms technician, which then of course escalated to me. I didn’t have a scheduled appointment with her that morning but it turned into a saga. Adamant she was not going to the hospital, did not want police involved, still endorsing suicidal thoughts with plan, intent, continued access to her meds and ligatures at home, no mitigation of her risk factors and wanted to leave immediately to return right back to her home where her spouse was. No insight into why this is so concerning. Emergency services contacted but patient left our clinic and got in her car, drove off, ems diverted to her house which she was not there, she eventually came back to the clinic, eventually spoke with police in our parking lot and they wouldn’t escort her to the ER 🤦‍♂️. She called the clinic later and said she was at home, suicidal, about to overdose, pretty much repeat of what happened this morning but she was at home at this point. EMS called again and last report I got was she was being emergently detained to the ER. Hopefully she got a 72 hr hold which here (TX) is an order of protective custody or OPC. Spent forever last night documenting a novel about all this in detail and justification for why she was at imminent risk of harm to self in lower level of care.

I feel like I did my best to document this. But she’s had a history of issues with another doctor here in town, complained to the medical board which supposedly got some kind of investigation on the doc. I’m not worried about it in the sense of I’ve done anything wrong just the pain in the ass it would be going through that.

I'm sorry you had to deal with this, but it is an unfortunate part of many of our jobs. That said, this at least seems as clear cut as it could possibly be in terms of suicide risk assessment short of her trying to kill herself at your office. Glad she was finally brought in and hopefully gets more help, but also get how hard it is when there are so many social factors stacked against her.

 

[Bartelby]

I'm sorry you had to deal with this, but it is an unfortunate part of many of our jobs. That said, this at least seems as clear cut as it could possibly be in terms of suicide risk assessment short of her trying to kill herself at your office. Glad she was finally brought in and hopefully gets more help, but also get how hard it is when there are so many social factors stacked against her.

I agree, this is one of those rare risk situations with pretty much no ambiguity. I can't imagine a board complaint about that hold could ever go anywhere.