By
Adam Simmons
· Last checked
August 2026
The honest answer is that it depends on what the message says, and that "I didn't look" is a much weaker position than it feels like at eleven at night.
The worry is real and practitioners state it directly. One describes being ethically bound to respond if they read certain texts — a client texting that they were suicidal, read and not responded to, being treated as unethical. A former trainee describes being told they weren't to handle emergencies, but that if they saw something at three in the morning they had to tend to it.
That is why "the safest thing is never to see it" circulates as advice. And it is why it is fragile: it works until the one night you do see it, and it does nothing about the message you were always going to read the next morning.
The durable version of this is not avoidance. It is arranging things so the question rarely arises — an agreed channel that clinical content isn't supposed to travel through, an automatic reply that redirects urgency somewhere staffed, and a decision made in advance about what you will do rather than one made at midnight.
On responding, the published expectations are thinner than the worry suggests. The social work College asks that registrants "respond to client queries, concerns, and/or complaints in a timely and reasonable manner" (OCSWSSW, Interpretation 3.4), and the psychotherapy College asks that you tell clients your own hours and average response times at the outset (CRPO Electronic Practice Guideline). Neither sets a clock running when you open a message.
On documenting it, the position is firmer, and it is the part worth acting on.
The psychotherapy College's list of what belongs in a clinical record includes "a notation of all in-session and out-of-session contacts with a client or their authorized representative," with the published examples being "letters, emails, texts, and telephone calls" (CRPO Professional Practice Standards, Standard 5.1). And where there is no written communications agreement, the terms themselves have to be charted: "Where no written contract exists, communications with the client regarding terms for services provided via electronic communications technology must be documented in the clinical record" (CRPO Electronic Practice Guideline).
The social work College's practice note on communication technology makes the same point through a worked example, in which the practitioner tells the client at the next session "that the email is part of the client's clinical record" (OCSWSSW Practice Notes).
Which turns a midnight question into a morning one. Nobody is asking you to reply at two in the morning. What is asked for is a record of what arrived, what you decided and why — and that is a note you can write at nine.
The honest answer is that it depends on what the message says, and that "I didn't look" is a much weaker position than it feels like at eleven at night.
The worry is real and practitioners state it directly. One describes being ethically bound to respond if they read certain texts — a client texting that they were suicidal, read and not responded to, being treated as unethical. A former trainee describes being told they weren't to handle emergencies, but that if they saw something at three in the morning they had to tend to it.
That is why "the safest thing is never to see it" circulates as advice. And it is why it is fragile: it works until the one night you do see it, and it does nothing about the message you were always going to read the next morning.
The durable version of this is not avoidance. It is arranging things so the question rarely arises — an agreed channel that clinical content isn't supposed to travel through, an automatic reply that redirects urgency somewhere staffed, and a decision made in advance about what you will do rather than one made at midnight.
On responding, the published expectations are thinner than the worry suggests. The social work College asks that registrants "respond to client queries, concerns, and/or complaints in a timely and reasonable manner" (OCSWSSW, Interpretation 3.4), and the psychotherapy College asks that you tell clients your own hours and average response times at the outset (CRPO Electronic Practice Guideline). Neither sets a clock running when you open a message.
On documenting it, the position is firmer, and it is the part worth acting on.
The psychotherapy College's list of what belongs in a clinical record includes "a notation of all in-session and out-of-session contacts with a client or their authorized representative," with the published examples being "letters, emails, texts, and telephone calls" (CRPO Professional Practice Standards, Standard 5.1). And where there is no written communications agreement, the terms themselves have to be charted: "Where no written contract exists, communications with the client regarding terms for services provided via electronic communications technology must be documented in the clinical record" (CRPO Electronic Practice Guideline).
The social work College's practice note on communication technology makes the same point through a worked example, in which the practitioner tells the client at the next session "that the email is part of the client's clinical record" (OCSWSSW Practice Notes).
Which turns a midnight question into a morning one. Nobody is asking you to reply at two in the morning. What is asked for is a record of what arrived, what you decided and why — and that is a note you can write at nine.
Why does reading it change anything?
Because knowledge is what obligations attach to, and this is intuitive to anyone who has thought about it for five minutes at midnight.
If you didn't see it, you couldn't act. If you saw it and chose not to act, you made a decision, and decisions are the thing that gets reviewed later. That's the structure of the fear, and it isn't paranoid.
What practitioners disagree about is the size of it. Some treat any read message as creating a duty. Others hold that a written agreement stating the channel is not monitored outside hours, given to the client in advance, is what determines the expectation. That disagreement isn't resolvable in a guide, and the published standards don't resolve it either — which is a reason to have the agreement, since it is the only part of this you control.
Is "don't look" a real strategy?
As a habit, sometimes. As a plan, no.
Its failure modes are all the same shape: it depends on you never breaking it, and you will break it. You'll check something else and see the preview. A client will call rather than text. The message will arrive at six on a Friday, which is after hours and also plainly still the working week.
It also has a cost that people underestimate. Several practitioners describe the mere possibility of a message being the stressor — knowing a client could reach them means never switching off, on quiet evenings as much as busy ones. Deliberate not-looking doesn't remove that. It adds to it, because now you are also monitoring your own monitoring.
And it does nothing for the client, who is sitting with a message that went nowhere and no information about what to do instead.
What's the alternative?
Three things, and they work together.
Decide in advance what you do. Write down, for yourself, what categories of message you will act on outside hours and what you'll do — call, contact emergency services, wait until morning. The value isn't the document. It's that you're not constructing an ethical position at midnight while frightened.
Make the channel the wrong place for it. A written communications agreement that says text is for scheduling, and that clinical content and crises go elsewhere, changes what turns up. It doesn't change it completely. It changes it a lot.
Put an automatic reply behind the hours. This is the piece most people skip and it does the most work. A message that goes back immediately — the practice is closed, I'll reply on {{day}}, if this can't wait then here's where to go — means the client is never sitting in silence, and it means the redirection happened whether or not you read anything.
That last point is the one that answers the worry. If the redirection is automatic, it happened at 11:04pm regardless of when you opened your phone.
What about documenting it?
Assume anything meaningful that arrives about a client belongs in the record, and find out what your own College publishes rather than working from a rule of thumb. One practitioner's version — any text should have a clinical note attached to it — is a reasonable default and not a substitute for the actual guidance.
The practical implication is about where messages live. A thread inside a consumer messaging app on a personal handset is the hardest possible version of this to satisfy later, because the record is mixed in with your own life and tied to a device. That's an argument for a separate line independent of every other argument for one.
And if I've already read something worrying?
Act on the content, not on the hour. If you'd act on it at two in the afternoon, act on it now. If you wouldn't, note it and take it to the next session.
Then, afterwards and not during, do the boring part: write down what you did and why. The thing that protects you is a documented decision, not an unread message.
The short version
Reading a message can create an obligation, and not reading it is not a plan. Set the channel up so clinical urgency doesn't arrive there, put an automatic redirect behind your hours so the client is never left with silence, decide in advance what you'd do — and read what your own College actually publishes rather than anyone's paraphrase of it.
About Rivet
Rivet is a Canadian practice line built for therapists — a separate number for calls, texts and voicemail, with video sessions and clinical tools in the same place. Your data stays in Canada, and transcription runs on Rivet's own hardware rather than a third-party AI service.
One plan, $65 CAD a month, everything included. Fourteen-day trial, no card.