Stalking and fixation: the response plan you write before it happens

First, treat it as safety — not a clinical failure

If a patient is stalking, following, or fixating on you, the first move is a reframe: this is a safety situation, not a clinical problem you are obligated to solve inside the therapy. Clinicians often freeze here because the instinct to understand and help collides with the need to protect themselves. Both can be true, but safety comes first. If at any point you are in immediate danger — someone is at your door, following your car, or threatening you — call 911 before anything else.

You do not owe a stalker continued access in the name of the therapeutic relationship. Fixation that has crossed into surveillance, unwanted contact, or intimidation has already broken the frame; your job now is to stop feeding it, document it, and follow a plan you ideally wrote before it ever happened. Trying to therapize the behavior in real time usually reinforces it.

Restate the boundary once — then stop responding

Set the boundary a single time, in writing where you can, and then stop responding to contact. A common and defensible approach is one clear, unemotional message stating that the professional relationship has ended or must stay strictly professional, that further contact is unwelcome, and that continued contact will be reported to the authorities. Keep a copy. After that one message, do not negotiate, explain, or reply again — each response is a reward that teaches the behavior to continue.

Expect it to get temporarily worse before it improves. When reinforcement stops, fixated contact often spikes — more messages, more attempts — before it fades; this burst is common and is not a sign your approach is failing. Warn your front desk or answering service, if you have one, so no one inadvertently passes along your schedule or confirms your whereabouts. The single boundary plus consistent non-response is the pattern; mixed signals are what prolong these situations.

Document every intrusion the day it happens

Keep a running log of every intrusion — the date, the time, what happened, any witnesses, and screenshots or copies of messages. This log is what turns a vague sense of being harassed into the evidence a court or the police can act on, and memories of individual incidents blur quickly once they pile up. Save voicemails and texts, photograph any gifts left for you, and note sightings near your home or office.

Documenting the patient's conduct is lawful and does not conflict with their privacy, but your own obligations do not disappear. You may not retaliate by disclosing the patient's protected health information beyond what a permitted purpose allows, and the ordinary Privacy Rule limits still govern what you share and with whom 1Ref 1HHS Office for Civil Rights (2026).Summary of the HIPAA Privacy Rule.The Privacy Rule's use-and-disclosure limits and minimum-necessary standard, governing what a clinician may disclose to law enforcement and may not retaliate by releasing.. When you report to law enforcement, disclose the minimum necessary to convey the threat — the conduct, the risk, and what you need — not the person's clinical history. Keep the safety log distinct from the clinical record so the two are not confused.

The written response plan (build it before you need it)

Write your response plan while you are calm, not during a crisis — the federal small-business continuity framework maps neatly onto a practice of one: assess the risk, decide the response in advance, and set your emergency communications so you are not improvising under fear 2Ref 2U.S. Department of Homeland Security (2026).Ready.gov Business.The federal business-continuity planning structure — risk assessment, response plan, emergency communications — applied to a solo practice's safety plan.. A one-page plan you can execute mechanically is worth more than a perfect plan you have to invent mid-incident.

Build the plan around a few concrete decisions and controls:

• Who you call, in order — a trusted colleague, your own attorney, the local police non-emergency line, and 911 for immediate danger.
• Front-desk protocol — train whoever answers your phone or greets clients never to confirm your schedule or location, and mind incidental disclosures in the waiting room so one patient cannot learn another's timing.
• Physical controls — lighting, locks, where you park, and cameras: a lobby camera can be reasonable, while a clinical space never is, given confidentiality.
• Telehealth or location shifts — moving a fixated patient's care off-site, or lowering your own visibility, when in-person contact is the vector.

If your practice has staff, fold this into a broader workplace violence-prevention plan rather than keeping it in your head. A plan on paper is also what lets someone else act if you are the one being targeted.

When to involve law enforcement and seek a protective order

Involve law enforcement when the conduct crosses from unwanted into threatening or persistent: explicit threats, showing up at your home, or contact that continues after a clear warning. File a police report even if you are unsure it will lead anywhere — it creates the record that later steps depend on, and patterns matter more than any single incident. For immediate danger, 911; otherwise use your local non-emergency line to document.

There is a point when you need the order — a civil protection, restraining, or anti-harassment order that a court issues against the threatening patient. The standards and names vary by state, so this is where a brief consult with your own attorney earns its cost. Note one interaction with your records: if the court process generates a subpoena or a court order for the patient's records, a court order authorizes only what it specifies, and a bare subpoena requires the usual assurances before you produce anything — the litigation does not waive the patient's privacy 3Ref 3HHS Office for Civil Rights (2026).Court Orders and Subpoenas.The distinction between a court order (produce only what it specifies) and a subpoena without a court order (assurances required first) if litigation seeks the patient's records.. Protecting yourself and protecting their record are not in conflict.

Terminating care without abandoning the patient

Deciding to end treatment with a patient who is stalking you is legitimate — you cannot be an effective clinician to someone who has turned the relationship into pursuit, and you are not required to keep seeing someone who endangers you. The ethical requirement is to end care carefully rather than vanish. A common standard is written notice, an offer of referrals, making records available for continuity, and, where feasible, coverage for an interim period.

Safety can change how you deliver that termination — you might send it in writing rather than in a room, route it through your attorney, or coordinate timing with a protective-order filing so the notice does not trigger an escalation. Document the clinical reasoning and the safety reasoning together. Ending treatment for cause, done with notice and referrals, is defensible; abandoning a patient without any handoff is the thing that turns a safety problem into a licensing problem too.