Home/Guides/YourProtocol/Say It First: Being Heard in a Doctor's Appointment
Guides · YourProtocol
YourProtocol

Say It First: Being Heard in a Doctor's Appointment

What audio recordings of real appointments show about how the first minute works, and the one change that follows honestly from it.

Mixed: the problem is well measured across decades, and the patient-side fix has never been tested in a trial
In-house · synthesized from the cited primary sources
At a glance
Time
7 min
Difficulty
Beginner
Strong The mechanism that justifies the advice is stated in the original 1984 paper: clinicians treat the first concern raised as the most important one, and the authors explicitly note there is no empirical evidence that patients order their concerns by importance.
What the evidence says

Understand the finding before you use it

74 audiotaped visits. Opening statements were completed in 23% of visits. When an opening was redirected, it happened on average 18 seconds after the patient began to speak. The load-bearing number, and the one almost nobody quotes: of 51 redirected openings, exactly 1 was ever completed.
Why ↓

The famous 18 seconds is not the finding that matters. The finding that matters is that once the opening is redirected, the conversation essentially never comes back to it. EVIDENCE TIER A for what was recorded, in a small primary care sample from the early 1980s.

Beckman & Frankel, Ann Intern Med, 1984;101(5):692-6 (PMID 6486600)

Read the word "interruption" precisely

What was measured is redirection to a single concern, usually with a closed-ended question. It is not a measure of attention, interest or care, and nothing in this literature measures how much a clinician is listening.
Why ↓

"Doctors stop listening after 18 seconds" is the version that went viral, and it is not what was measured. Getting this right is most of the page's credibility, and it is also why the advice here is about sequencing your own sentence, not about doubting the person opposite you. EVIDENCE TIER A for what the studies did and did not measure.

Beckman & Frankel, Ann Intern Med, 1984, methods

Check whether the finding replicated

264 interviews with 29 physicians: opening statements were completed 28.0% of the time, and redirection came at a mean of 23.1 seconds.
Why ↓

Different decade, different sample, close to the same numbers. EVIDENCE TIER B: convenience sample.

Marvel, Epstein, Flowers & Beckman, JAMA, 1999;281(3):283-7 (PMID 9918487)

Notice what was not asked at all

In 112 recorded encounters, the patient's agenda was elicited in only 40 of them (36%): 30 of 61 in primary care, 10 of 51 in specialty care. Of the 40 who were asked, 27 were redirected, at a median of 11 seconds.
Why ↓

"Doctors interrupt after 11 seconds" describes 27 of 112 encounters, about 24% of that sample. The truer and worse finding is that in most encounters nobody asked what the patient had come in for. That is exactly why saying it first, unprompted, is the move. EVIDENCE TIER B.

Singh Ospina et al., J Gen Intern Med, 2019;34(1):36-40 (PMID 29968051)

Drop the assumption that you are taking up too much time

Across 335 patients allowed to speak without interruption, mean spontaneous talking time was 92 seconds (SD 105), the median was 59 seconds, and 78% finished within two minutes.
Why ↓

The common belief, including among patients, is that a full opening would eat the appointment. It does not: a minute is the typical ask. Caveat worth stating: those doctors had an hour of training in active listening, so this is a best case, not an average clinic. EVIDENCE TIER A for the direction.

Langewitz et al., BMJ, 2002;325(7366):682-3 (PMID 12351359)

Know that the biggest measured lever is not yours

In 224 patients seen by 20 family physicians, asking "is there SOMEthing else you want to address today?" rather than "is there ANYthing else?" cut unmet concerns from 37% to 9% (odds ratio 0.154, p=0.001), with no effect on visit length.
Why ↓

We include this because it is the cleanest experimental result in the whole literature, and because it is honest to say the strongest lever measured is a clinician-side one, not a patient-side one. EVIDENCE TIER B.

Heritage, Robinson, Elliott, Beckett & Wilkes, J Gen Intern Med, 2007;22(10):1429-33 (PMID 17674111)

Know where the gaps have been measured

A prospective cohort of 981 adults arriving at one hospital with acute abdominal pain found women received opioid pain relief less often than men (45% versus 56%) and waited longer for it (median 65 versus 49 minutes) despite similar recorded pain scores. Adjusted, women were 13% to 25% less likely to receive it.
Why ↓

This is linked to, not caused by: it is observational, from a single centre, and it cannot tell you why the difference occurred. It is included because it is the measured version of an experience many people describe, and it is why the first sentence carries more weight for some people than others. EVIDENCE TIER B.

Chen, Shofer, Dean et al., Acad Emerg Med, 2008;15(5):414-8 (PMID 18439195)

Hold the advice loosely

No trial has ever randomised patients to lead with their most important concern and then measured an outcome. The closest evidence: training 48 physicians in upfront agenda setting, across 1,460 patients, meant patients raised fewer concerns late in the visit with no increase in visit length, while the authors state plainly that it "is not sufficient to enhance patient satisfaction, trust or functional status".
Why ↓

We are not going to pretend a communication tip is a tested intervention. What is documented is the asymmetry: the first concern stated is read as the most important, and there is no evidence patients order their concerns that way. EVIDENCE TIER B for the trial, and NO TIER for the patient-side action, because it has never been tested.

Brock, Mauksch et al., J Gen Intern Med, 2011;26(11):1317-23 (PMID 21735348)

Do not write everything down

A meta-analysis of 14 randomized trials in cancer settings (n=2,678) tested written question lists. Questions asked rose by a standardised mean difference of 0.24 (95% CI 0.00 to 0.48, p=0.05), which is borderline, and it was significant only when the clinician endorsed the list (SMD 0.31, 95% CI 0.10 to 0.52, p=0.003). Consultation length was unchanged, satisfaction was unchanged, and anxiety rose slightly (SMD 0.12, 95% CI 0.02 to 0.22, p=0.02).
Why ↓

The instinct is to arrive with a long list. The trials say a list barely moves anything unless the clinician picks it up, and it nudged anxiety upward. Sequencing one sentence is a different act from bringing ten. EVIDENCE TIER A for the meta-analysis, in cancer settings specifically.

Han, Liu, Chen et al., Asia Pac J Oncol Nurs, 2025;12:100765 (PMID 40896753)

Write one sentence and say it first

Before the appointment, write ONE sentence: the single thing that would make this visit worth it. Say that sentence first, before the history and before the apology. One sentence, not a list.
Why ↓

Clinicians treat the first concern stated as the most important one, and the 1984 paper explicitly notes there is no empirical evidence that patients order their concerns by importance. Saying the important thing first closes that gap. It costs nothing, it has never been tested as an intervention, and we would rather say so.

Beckman & Frankel, Ann Intern Med, 1984
The evidence 8
Strong

The mechanism that justifies the advice is stated in the original 1984 paper: clinicians treat the first concern raised as the most important one, and the authors explicitly note there is no empirical evidence that patients order their concerns by importance.

The Effect of Physician Behavior on the Collection of Data (Beckman & Frankel, Ann Intern Med, 1984) Read Ann Intern Med, 1984;101(5):692-6
Soliciting the Patient's Agenda: Have We Improved? (Marvel, Epstein, Flowers & Beckman, JAMA, 1999) Read JAMA, 1999;281(3):283-7
Eliciting the Patient's Agenda: Secondary Analysis of Recorded Clinical Encounters (Singh Ospina et al., J Gen Intern Med, 2019) Read J Gen Intern Med, 2019;34(1):36-40
Spontaneous Talking Time at Start of Consultation in Outpatient Clinic (Langewitz et al., BMJ, 2002) Read BMJ, 2002;325(7366):682-3
Reducing Patients' Unmet Concerns in Primary Care (Heritage, Robinson, Elliott, Beckett & Wilkes, J Gen Intern Med, 2007) Read J Gen Intern Med, 2007;22(10):1429-33
Gender Disparity in Analgesic Treatment of Emergency Department Patients With Acute Abdominal Pain (Chen et al., Acad Emerg Med, 2008) Read Acad Emerg Med, 2008;15(5):414-8
Effectiveness of Intensive Physician Training in Upfront Agenda Setting (Brock, Mauksch et al., J Gen Intern Med, 2011) Read J Gen Intern Med, 2011;26(11):1317-23
Question Prompt Lists in Cancer Care: a Meta-analysis (Han, Liu, Chen et al., Asia Pac J Oncol Nurs, 2025) Read Asia Pac J Oncol Nurs, 2025;12:100765
Related protocols
Guides Explainer

How to Read a Health Headline

YourProtocol8 min
Women's Health Limited

Women's Brain Health & Perimenopause Protocol

Louisa NicolaDaily + weekly
Recovery Moderate

Migraine Lifestyle Triggers: The SEEDS Framework

Amaal StarlingDaily habit

Not medical advice. This page is for education only and is not a substitute for professional medical care. Consult a qualified clinician before changing your health routine.
Editorial disclosure. This protocol is written and fact-checked by the YourProtocol editorial team directly from the primary sources cited below; it is not written or reviewed by any outside expert.

Is this for you
  • Anyone who has left an appointment without saying the thing they came to say
  • People preparing for a short appointment with more than one concern
  • Anyone who has been told they are "a lot" in a consultation
Cautions
  • This is a communication guide, not medical advice, and it does not diagnose anything
  • No trial has tested leading with your most important concern and measured an outcome; the problem is well measured and the fix is not
  • The studies here describe specific settings and time periods, mostly primary care in the United States and Switzerland
  • If something is urgent, say so immediately and do not wait for a structure
Common questions
Do doctors really interrupt after 18 seconds?
In a 1984 study of 74 recorded visits, when an opening statement was redirected it happened on average 18 seconds after the patient began speaking. Redirection is not the same as not listening: what was measured is the conversation being steered to one concern, usually with a closed question. The more striking number is that of 51 redirected openings, only one was ever completed.
Is it true doctors stop listening after 11 seconds?
No. That figure comes from a 2019 analysis in which the patient's agenda was elicited in only 40 of 112 encounters, and of those 40, 27 were redirected at a median of 11 seconds. The bigger finding in that study is that in most encounters the patient was never asked what they had come in for.
Will I take up too much time if I explain everything?
Probably not as much as you think. When 335 patients were allowed to speak without interruption, the median opening ran 59 seconds and 78% finished inside two minutes. Those doctors had an hour of training in active listening, so treat it as a best case rather than a guarantee.
Should I bring a written list of questions?
Order helps more than length. A meta-analysis of 14 randomized trials found written question lists barely moved the number of questions asked unless the clinician actively endorsed the list, did not change satisfaction, and nudged anxiety slightly upward. One clear opening sentence is a different move from a long list.
What should I actually say first?
One sentence naming the single thing that would make the visit worth it, said before the history and before any apology for taking up time. The reason is documented: clinicians treat the first concern raised as the most important, and the original researchers noted there is no evidence patients raise concerns in order of importance.
Does this mean my doctor is not on my side?
No, and nothing here measures that. These studies describe the shape of a short conversation under time pressure, not anyone's intent or competence. The point is that the first minute carries more weight than most people expect, so it is worth deciding in advance what goes in it.
Get the next protocol first
Evidence updates, cited to the source. No spam.