Two errors, both common. Sitting on something for six weeks because it did not seem big enough to interrupt anybody. And reaching out constantly, which makes it harder for anybody to see what matters.
The way out of both is to decide in advance, once, what your threshold is. Then you are not making the judgement in the moment, which is exactly when judgement is least available.
Decide the threshold at the appointment
The best time to agree this is when a plan is being made, and the question to ask is direct:
"What would make you want to hear from me before the next appointment?"
Most clinicians will give you a specific list. Write it down. It usually includes some combination of: a particular side effect, a lack of improvement by a certain point, a worsening of something specific, and a safety item.
Having it in writing does two things. It tells you what to watch for, and — the underrated half — it tells you what not to worry about. A named list is permission to stop monitoring everything else.
Things that should not wait, regardless
Whatever your personal list says, these are not appointment material:
- Thoughts of harming yourself, or of not being here. Local emergency number, or in the US call or text 988. This is a common feature of treatable conditions and interrupting your day for it is always the right call.
- A severe reaction to a medication — rash, swelling, difficulty breathing, fever with a new drug, or new and worsening confusion.
- Anything that could be acute: sudden severe headache, weakness or numbness on one side, difficulty speaking or understanding, chest pain, fainting.
- A rapid change in thinking or behaviour over hours to days rather than weeks.
- Elevated mood with reduced need for sleep, racing thoughts, unusually rapid speech, or uncharacteristic risk-taking. Specifically worth raising promptly, because it changes what treatment is appropriate and some treatments can worsen it.
- New thoughts or perceptions that frighten you.
A reasonable default threshold
If nobody gave you a list, this is a defensible one for anything being treated or adjusted:
Reach out within a few days if:
- A new symptom appears after a change in treatment
- Something has been clearly worse for three days or more, without an obvious cause
- A side effect is bothering you enough that you are thinking about stopping
- You have missed doses in a pattern rather than occasionally
- Something practical is blocking treatment — the pharmacy, an authorisation, the cost
- Someone close to you has said they are worried
- Whatever the plan was supposed to achieve has not happened by the point you were told to expect it
Reach out immediately if anything on the previous list applies.
Note it for next time if: it is mild, it is not new, and it is not getting worse.
The two reasons people wait, and why neither holds
"It is not serious enough to bother them." A care team exists to be contacted. A five-minute answer in week two is better care and less work than a stopped medication and a wasted three months. The most common thing clinicians say about this is that they would rather hear the small thing.
"I will mention it at my appointment." Six weeks is long enough for the detail to evaporate, for the problem to become normal, and for the treatment to have been silently abandoned. Also: the appointment is when decisions are made, and a decision made on a clear account of week two is a better decision than one made on a vague memory of it.
The one people do not expect
Practical obstacles count. The prescription you cannot fill, the authorisation that was denied, the appointment you could not get, the cost you cannot manage. These stall treatment exactly as effectively as a clinical problem, and people are far more reluctant to raise them because they feel like admin rather than medicine.
They are not admin. A treatment you cannot obtain is not a treatment, and it is a solvable problem that somebody on the team is there to solve.
Say what you want
When you do reach out, name the request. "Is this worth changing something, or should I keep going?" gets a clearer answer than a description with no question attached. So does "I do not need anything, I just want this on the record."
Both are legitimate. The second is genuinely useful — an observation logged in week two is what makes the next decision better, and it costs nobody anything.
What if you are unsure whether it is urgent
Treat the uncertainty itself as the answer and use the faster route. Nobody minds the false alarm; the failure mode everyone minds is the message that sat in a queue over a weekend.
If you are unsure who to ask, ask anyone on the team. Routing is their problem. The only bad outcome is the question that never gets asked.