“No Plan” Isn’t “No Risk”: Suicide & Homicide Risk Assessment Lessons from the Clancy Case

 

RISK ASSESSMENT  ·  MEDICATION SAFETY  ·  PROFESSIONAL PRACTICE

“No Plan” Isn’t “No Risk”: Suicide & Homicide Risk Assessment Lessons from the Clancy Case

A companion piece to my earlier post — this one goes deeper on mixed states, the SSRI black box warning, structured risk assessment, and a practice-management lesson about your own professional bio.

My first post on the Lindsay Clancy trial covered the treatment timeline, care fragmentation, and the DSM-5-TR diagnostic gaps around postpartum psychosis. As testimony has continued, several more specific clinical issues have come up that deserve their own discussion: how a mixed bipolar state actually differs from standard postpartum depression, what the FDA's black box warning on antidepressants really covers, how suicidal and homicidal ideation should be assessed structurally rather than as a single yes/no question, and a practice-management lesson that has nothing to do with prescribing at all.

As before: this is an active criminal trial, and Clancy has also filed a malpractice suit against several of her treating clinicians, while her ex-husband has filed a separate wrongful death suit. The complaints contain allegations that have not been adjudicated, and the defendants have not been found liable for anything. Nothing here is a comment on the outcome of that litigation — it's a look at the general clinical and professional-practice principles the testimony happens to illustrate.

⚠ Before We Go Further This post discusses suicidal and homicidal ideation assessment in clinical detail, for a professional audience of NP students and new graduates. If you are personally struggling with thoughts of suicide or harming yourself or others, please reach out for support — the 988 Suicide & Crisis Lifeline (call or text 988) is available 24/7 in the U.S.

1. What a Mixed Bipolar State Actually Looks Like — and Why It's Easy to Miss

Testimony has described Jollotta's initial working diagnosis as postpartum depression and anxiety. She raised an underlying bipolar disorder or mixed manic/hypomanic state as a differential consideration after Clancy described going roughly 48 hours without sleep following a reaction to Zoloft, and she testified she disclosed this concern to Clancy and her husband directly, in person, on December 6th — Patrick responded, “my wife is not bipolar”; Lindsay said nothing. Worth noting precisely, since it changes the lesson: a structured screen (the Mood Disorder Questionnaire, administered earlier by a colleague) had come back negative — Clancy endorsed only 3 of the 13 listed symptoms, well under the threshold for a positive screen. Jollotta testified she stayed concerned anyway, citing the MDQ's known lower sensitivity for bipolar II and NOS presentations, and continued to act on the differential afterward: holding Prozac and titrating Seroquel on the FDA's mania-dosing schedule rather than the depression-dosing schedule, over the family's stated skepticism.

In uncomplicated postpartum depression, sleep disturbance is congruent with the depressive picture: the patient is exhausted, whether or not she's actually sleeping. A mixed state looks different. DSM-5-TR defines a mixed features specifier as a major depressive (or manic/hypomanic) episode with at least three symptoms from the opposite pole present concurrently. The hallmark most relevant here is a dramatically reduced need for sleep that is not accompanied by fatigue — the patient may describe feeling wired, restless, or flooded with emotion rather than depleted, even after 48 hours awake.

The teaching point this transcript actually supports is more specific than “watch for this symptom”: a negative structured screen doesn't close the question when the tool's own sensitivity is known to be lower for the exact subtype in play. Documenting that reasoning — why you're staying concerned despite a negative screen, and what you did about it — is what turns a clinical hunch into a defensible clinical judgment. It's also worth naming plainly: disclosing a bipolar concern to a patient and her spouse, and having them push back on it, doesn't obligate a clinician to drop the differential, and this transcript shows Jollotta didn't — she kept adjusting the medication plan around it even as the family disagreed.

Other features worth actively screening for alongside sleep changes: racing thoughts, pressured speech, psychomotor agitation, irritability, and increased impulsivity or risk-taking. That combination — depressive hopelessness plus manic-spectrum energy and impulsivity — is considered one of the highest-risk presentations in psychiatry, because it can pair the motivation to act with the activation to act on it. This is precisely why a single atypical reaction to an antidepressant, especially one involving profound sleep loss without fatigue, deserves a structured bipolar screen (the Mood Disorder Questionnaire is a quick, validated option) rather than being folded into the existing PPD diagnosis and treated with reassurance or another sedative.

2. The FDA Black Box Warning: What It Actually Covers

The FDA's black box warning on antidepressants, in place since 2004 and updated since, is often summarized loosely as “SSRIs can cause suicidal thoughts,” which undersells what it's actually about. The warning specifically addresses increased risk of suicidal thinking and behavior in children, adolescents, and young adults up to age 24 who are started on an antidepressant, concentrated in the first one to two months of treatment or shortly after a dose change.

The mechanism the warning is built around is essentially activation syndrome: agitation, akathisia-like restlessness, insomnia, irritability, hostility, and impulsivity emerging early in treatment — sometimes before mood has actually improved. That creates a specific high-risk window where a patient has more energy and activation but is still deeply depressed, which is a materially different risk profile than either symptom alone.

Two points worth being precise about for practice: first, even outside the FDA's specific age cutoff, the underlying principle — that early antidepressant treatment can transiently increase agitation and risk before benefit appears — is considered clinically relevant across ages and should be part of informed consent and early monitoring for any patient starting or changing an SSRI/SNRI, postpartum patients very much included. Second, the warning is a reason for closer follow-up in the first weeks of treatment — checking in within days, not weeks — not a reason to avoid antidepressants altogether in patients who need them.

3. Structured Risk Assessment: Why “No Plan” Isn't the End of the Conversation

This is arguably the most practically important section for day-to-day charting. “Denies plan” is one of the most common phrases in psychiatric documentation — and one of the most commonly over-relied-upon. Absence of a stated plan is one input into a risk assessment, not a stand-alone reassurance that closes the inquiry.

Structured tools like the Columbia-Suicide Severity Rating Scale (C-SSRS) treat ideation, intent, plan, means, and history of attempts as separate dimensions that each need to be asked about and documented individually. A patient can have passive or even active ideation with no articulated plan and still carry meaningfully elevated risk, particularly when other risk factors are stacking up — sleep deprivation, mixed-state features, a recent hospitalization, access to lethal means, and the postpartum period itself, which independently elevates risk.

In the postpartum context specifically, risk assessment has a second, distinct track that has to be asked about directly rather than inferred from a mood screen: thoughts of harm toward the infant. These thoughts occur on a real spectrum. On one end, intrusive, unwanted, ego-dystonic thoughts of harm are extremely common in postpartum anxiety and postpartum OCD — the patient finds them horrifying and would never act on them, and simply having them is not, by itself, a sign of psychosis. On the other end, thoughts that feel ego-syntonic, commanding, or paired with a plan or intent are a psychiatric emergency. Distinguishing between those two ends requires direct, explicit questioning: What is the content of the thought? Is it distressing and unwanted, or does it feel right or necessary? Is there intent? Access to the infant? Who else is supervising or available? A generic “denies harming baby” note doesn't capture where on that spectrum a patient actually sits.

The documentation lesson: a defensible risk note documents ideation, intent, plan, means, protective factors, and the clinician's stated reasoning for the risk level assigned — not just the single data point of “no plan.” When a chart shows only “denies plan” with nothing else, that tends to read, in hindsight, as a superficial screen rather than a clinically reasoned risk assessment, regardless of how much thought actually went into it at the time. Write down the reasoning, not just the conclusion.

Worth noting: Jollotta's own testimony describes documentation more granular than a bare “denied” — she testified to repeatedly asking about plan and intent specifically, documenting “SI, no plan, no intent” at multiple points in December, and asking directly whether Clancy had any thoughts of harming her children, which Clancy denied every time she was asked. That's closer to the defensible end of the spectrum than the thin end. The general principle still stands — plan-only documentation is a common and real failure mode worth guarding against — but on the primary record here, this particular case shows a more structured line of questioning than “denies plan” alone.

✓ Clinical Bottom Line •     Severe insomnia without accompanying fatigue after starting or increasing an antidepressant is a red flag for a mixed or manic state, not just “bad sleep” — screen formally rather than treating it as a side effect to wait out. •     The SSRI black box warning is about early-treatment activation syndrome creating a window of increased energy paired with persistent depression — it's a cue for closer follow-up in the first weeks, not a reason to avoid treatment. •     “Denies plan” is one data point in a multi-dimensional risk assessment (ideation, intent, plan, means, protective factors) — never document it as the whole assessment. •     Postpartum harm-related intrusive thoughts require their own direct line of questioning, separate from a general mood or suicide screen — content, distress level, intent, access, and supervision all need to be asked about explicitly. •     Document your risk-level reasoning, not just your risk-level conclusion — the “why” is what makes a risk assessment defensible later.

4. A Lesson That Has Nothing to Do With Prescribing: Your Own Bio

One exchange from testimony is worth pulling out separately because it's not a clinical issue at all — it's a practice-management one. On cross-examination, testimony indicates Jollotta was confronted with website language describing her as having expertise in postpartum depression and postpartum psychosis; she characterized her actual background as more of a personal clinical interest, and said she was not aware of the specific language on her website. We're not going to speculate here about how that exchange affects the litigation — that's a legal judgment for the attorneys and experts actually involved in the case, not something to analyze from the outside.

What is worth taking from it, as a general practice principle: your public-facing professional bio — on a practice website, a hospital system page, a directory listing, LinkedIn — is a document that can be read back to you, and holding yourself out (even unintentionally, even via a marketing vendor's language you never reviewed) as having expertise in a specific condition can raise the standard you're compared against if something goes wrong with a patient who has that condition.

•     Review your own bio, on every site it appears, at least annually — don't assume a practice manager, marketing vendor, or hospital system got the language right, and don't assume it stays accurate once written.

•     Be precise about the difference between clinical interest, additional continuing education, formal certification, and specialization — these terms aren't interchangeable, and “expert” or “specialist” language implies a level of credentialing that most general FNPs or PMHNPs, even ones who see a lot of a given patient population, don't formally hold without specific fellowship training or certification.

•     If your practice's marketing overstates your scope, get it corrected. This isn't only about litigation exposure — a mismatch between what your bio promises and what you're actually positioned to manage can set patients up with the wrong expectations from the first visit, which is a patient-safety issue in its own right.

◆ A Nuance Worth Sitting With It's tempting to read the mixed-state and black-box-warning content above and conclude every postpartum patient with insomnia after starting an SSRI needs an urgent bipolar workup. That's not the right takeaway, and treating it that way would flood an already-thin perinatal psychiatric referral system with low-yield urgent evaluations. The skill being asked for here is specificity: it's the combination of features — reduced sleep need without fatigue, plus racing thoughts or agitation, plus a symptom trajectory that's getting worse rather than better on treatment — that should raise concern, not any single symptom in isolation. Most patients who report trouble sleeping on a new antidepressant are having ordinary insomnia, not a mixed episode.

Is Writing “Denied” Enough? A Closer Look at Outpatient Documentation

This deserves its own section because it's one of the most common documentation habits in outpatient behavioral health, and one of the thinnest. “Denied SI/HI” by itself tells a reviewer that you asked something and got a negative answer — it doesn't show what you actually asked, whether you assessed passive versus active ideation separately, whether known risk factors were weighed, or why you landed on the disposition you chose. In a chart review, a peer audit, or a hindsight analysis after a bad outcome, a bare “denied” tends to read as a box that was clicked, not a clinical judgment that was made.

A more defensible outpatient note includes several distinct elements, not just a global conclusion:

•     The screening method itself — name the tool used (e.g., C-SSRS, PHQ-9 item 9) or the actual questions asked, rather than only the result.

•     Ideation, intent, plan, means, and history documented as separate dimensions — passive versus active ideation in particular, since they carry different risk weight.

•     Homicidal/infant-harm ideation asked about and documented on its own line, distinct from the general SI screen — and, when thoughts are present, whether they are distressing/unwanted (anxiety-spectrum, lower acute concern) or ego-syntonic/commanding (higher concern), plus intent, access, and available supervision.

•     Protective factors named explicitly — support system, insight, future orientation, treatment engagement — rather than left for the reader to infer.

•     The clinician's actual reasoning and disposition, in a sentence — what risk level was assigned, why, and what happens next (follow-up interval, safety plan, crisis resources provided).

•     Context from that specific visit — recent medication changes, sleep status, symptom trend since the last encounter — so the note reflects this visit rather than reading as identical boilerplate every time.

A quick before-and-after illustrates the gap. Thin: “Denies SI/HI.” Defensible: “Denies active suicidal ideation, plan, or intent; endorses occasional passive thoughts of wanting to sleep and not wake up, present intermittently over the past 2 weeks, no worsening. No history of attempts. Directly asked about thoughts of harming infant — denies any such thoughts, including intrusive/unwanted ones. Protective factors: engaged in care, supportive spouse in home, future-oriented regarding children. Risk assessed as low at this time. Safety plan and crisis resources reviewed; follow-up in 1 week or sooner if symptoms escalate.” The second version takes perhaps 45 extra seconds to write. That's the difference between a note that demonstrates clinical judgment and one that only demonstrates a question was asked.

Quick Reference: Documentation Checklist

•     Risk assessment includes ideation, intent, plan, means, protective factors, and stated clinical reasoning — not just a plan status.

•     Harm-related intrusive thoughts (postpartum) are screened with direct questions about content, distress level, intent, access, and supervision — documented separately from a general SI screen.

•     Any new or worsening agitation, insomnia, or activation within weeks of starting/changing an antidepressant is documented with a follow-up interval, not just a reassurance note.

•     An atypical reaction to an antidepressant (e.g., severe insomnia without fatigue) is documented as a differential-diagnosis trigger, with the screening tool used (or the reasoning for not using one) noted explicitly.

•     Your own professional bio and marketing language are reviewed periodically and kept consistent with your actual training, certification, and scope.

•     SI/HI screening notes name the method used and document ideation, intent, plan, means, and protective factors as separate elements — never “denied” alone.

The Defense's Core Critique: Why “Checklist-Style” Care Is a Structural Question, Not Just an Individual One

Clancy's defense has argued more broadly that her medical team relied too heavily on brief, checklist-style assessments and failed to recognize the underlying severity of what the defense characterizes as postpartum psychosis and adverse medication reactions. It's worth closing on this critique directly, because it isn't really an argument that one symptom got missed — it's an argument that the format of care itself is structurally mismatched to catching a condition that is rare, fast-moving, and doesn't present in a checklist-friendly way. That's a useful lesson independent of how the litigation resolves.

•     A checklist can confirm a diagnosis; it can't rule one out. Tools like the EPDS, PHQ-9, or a structured SI screen are validated for detecting the condition they're built to detect — not a rarer, faster-moving condition hiding underneath a more common one. A reassuring score answers a narrow question (“does this look like standard PPD today”) and should never be read as answering the broader one (“is something more serious developing”).

•     Brevity is the enemy of trajectory. A single well-conducted visit is a snapshot; postpartum psychosis and mixed bipolar states are defined by their trajectory — worsening despite treatment, atypical reactions, symptoms that don't fit the working diagnosis. That trajectory is only visible by deliberately comparing this visit to the last several, not by any one screen administered fresh each time.

•     Checklists measure symptoms; they don't measure treatment-resistance or paradoxical worsening. A pattern of repeated medication changes without improvement — or reactions that make things worse rather than better — is often a stronger signal than any single symptom score, but it lives in longitudinal clinical judgment, not in a tool's output, and has to be actively tracked and documented as its own finding.

•     Standardized tools are a floor, not a ceiling. They ensure a baseline set of questions gets asked reliably, especially in time-pressured primary care or telehealth settings — they are not a substitute for clinical judgment once a presentation stops fitting the pattern the tool was validated on. Atypical insomnia, medication reactions moving the wrong direction, or escalating contact frequency are all cues to lean on judgment and consultation rather than the next scheduled screen.

•     “Checklist-style” care is often a symptom of system design, not individual failure. Short visit windows, high caseloads, and portal-message triage are structural features of how much behavioral health is delivered, particularly in perinatal programs stretched thin on specialist capacity. The more durable fix isn't asking individual clinicians to be more vigilant in a 15-minute visit — it's building structural safeguards: mandatory trajectory review at defined intervals, low-threshold psychiatric consult triggers, and explicit criteria for escalating a patient out of brief-visit format.

The one-line version worth remembering: a validated screening tool tells you what's present today; it can't tell you what's building underneath. Treating a reassuring score as closure, rather than as one data point in an ongoing trajectory, is where checklist-style care stops being a safeguard and starts being a liability.

One caveat worth stating plainly: Jollotta's own testimony describes something more active than a single static checklist applied on repeat. She tracked Clancy's Edinburgh and GAD-7 scores visit to visit, adjusted her differential as the trajectory changed, disclosed a bipolar concern to the family, and titrated medication against a bipolar-specific dosing schedule rather than the depression schedule once she suspected a mixed state. Whatever the jury ultimately concludes about the adequacy of that response, the record doesn't show a clinician mechanically re-running the same screen and calling it done — it shows ongoing clinical judgment operating alongside the screens. The structural critique above is worth taking seriously as a general principle for how perinatal psychiatric care is often delivered; it's a separate question from whether it's a fully accurate description of this specific clinician's actual process, and the primary testimony suggests the latter is more nuanced than the checklist framing implies.

References

MDLinx. “Lindsay Clancy’s malpractice suits put postpartum psychiatric care under a microscope.” July 30, 2026.

U.S. Food and Drug Administration. Antidepressant Use in Children, Adolescents, and Adults — boxed warning information. Updated guidance.

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR). 2022.

Columbia Lighthouse Project. Columbia-Suicide Severity Rating Scale (C-SSRS).

American College of Obstetricians and Gynecologists (ACOG). Screening and Diagnosis of Mental Health Conditions During Pregnancy and Postpartum, Clinical Practice Guideline. 2023.

Commonwealth v. Lindsay Clancy. Plymouth Superior Court trial testimony of Rebecca Jollotta, CNP, and related proceedings, August 2026.

This post is for clinical education purposes only and does not constitute legal commentary, medical advice, or a judgment of any individual involved in ongoing litigation. Case facts are drawn from public court reporting and may evolve as the trial continues.