The patient agenda: listen before interrupting

A 45-year-old woman walks into the office and begins: "Doctor, I came because of this knee pain, but I also wanted to talk about...". Before she finishes the sentence, the doctor asks how long the knee has been hurting. The visit proceeds around the knee, with an exam, guidance and a prescription. When he stands up to wrap things up, she says, already with her hand on her bag, that she has been sleeping badly and crying for no reason since her husband lost his job. There are two minutes left before the next patient.

Scenes like this are common in any health service. The person arrives with several concerns, but does not always start with the most important one, and the doctor, pressed for time, tends to close the conversation on the first topic that comes up. This text discusses what research shows about interruptions at the start of the visit, how much time people really need to tell their story, how to bring up the full agenda, how to ask about ideas, concerns and expectations, and how to negotiate what fits into a visit.

Key points

• Doctors rarely ask about all of the person's concerns and, when they do, they interrupt a few seconds later.
• Left to speak without interruption, most people finish their opening account in under two minutes.
• Changing one word in the final question, from "anything else" to "something else", greatly reduced unmet concerns.
• Unspoken concerns are linked to misunderstandings, unwanted prescriptions and treatments not followed.
• Asking about ideas, concerns and expectations reveals what the person thinks and changes clinical decisions.
• When there are many problems, the way out is to list everything at the start and negotiate what will be addressed now and what is left for later.

Summary

• What the patient agenda is
• Interrupted within seconds
• How much time the person needs to speak
• Bringing up all the concerns
• Ideas, concerns and expectations
• Many problems, little time
• Negotiating the agenda
• What gets in the way of listening
• Frequently asked questions
• Final checklist

What the patient agenda is

The patient agenda is the set of reasons, concerns and expectations the person brings to the visit. It includes the symptom that prompted the visit, but also what the person thinks is happening, what they fear, what they expect the doctor to do and how the problem affects their life.

A qualitative study with 35 patients of 20 family doctors showed how complex this agenda is: people arrive with varied needs and many are not spoken (Barry, 2000). Only 4 of the 35 patients set out their whole agenda during the visit (Barry, 2000). The items most often mentioned were symptoms and requests for diagnosis and prescriptions; the ones kept quietest were concerns about the diagnosis and the future, the person's own ideas about what was wrong, side effects, the wish not to receive a prescription and information about social context (Barry, 2000).

What is not said cannot be addressed. In the same study, there were 14 visits with problems in the outcome, such as serious misunderstandings, unwanted prescriptions, unused prescriptions and treatments not followed, and in all of them at least one problem was linked to an agenda item that was not spoken (Barry, 2000).

Bringing up the agenda, therefore, is not an optional kindness. It is part of clinical reasoning. The text on why empathy belongs in reasoning discusses how listening changes the quality of the hypotheses.

Interrupted within seconds

An analysis of 112 recorded visits showed that doctors asked about the person's agenda in only 40 of them, about one in three (Singh Ospina, 2019). In primary care this happened in 30 of 61 visits, about half, and in specialty care in 10 of 51, about one in five (Singh Ospina, 2019).

Even when the agenda was asked about, the listening was short. In 27 of the 40 visits in which the doctor asked, he interrupted the person after a median of 11 seconds (Singh Ospina, 2019). When they were not interrupted, people took a median of 6 seconds to set out their concern (Singh Ospina, 2019). The interruption, in other words, often came before the person needed more time.

A study of 94 consultations with 27 hospital physicians detailed the type of interruption. There was a median of 9 interruptions per minute during the opening statement, and the first came at a median of 6.5 seconds (Mulder-Vos, 2023). About two thirds of these interruptions were encouraging signals, such as "mm-hmm" or "go on", which prompt the person to continue (Mulder-Vos, 2023).

Not every interruption, therefore, is bad. The problem is the ones that change the course of the conversation. In 52 of the 94 consultations, more than half, the person could not finish their opening statement because the physician took the floor, with a median of 31.4 seconds (Mulder-Vos, 2023).

The encouraging signals deserve attention because they show that interrupting is not the same as taking part. A nod, an attentive look or a short word tell the person that they are being heard and can continue. A closed question about a detail of the first symptom, asked before the person finishes, moves the conversation onto the physician's ground and usually ends the list before it appears in full.

How long a person needs to speak

The fear that the person will talk without stopping is one of the reasons for interrupting early. A study set out to measure this time. In a general internal medicine outpatient clinic of a university hospital, 14 physicians trained in active listening were instructed not to ask questions at the start and to time, without the patient knowing, how long it took them to indicate they had finished (Langewitz, 2002).

Among the 335 patients analyzed, the mean spontaneous speaking time was 92 seconds, with a median of 59 seconds, and about eight in ten finished within two minutes (Langewitz, 2002). Only 7 patients spoke for more than five minutes, and in all these cases the physicians felt the information was important and should not have been interrupted (Langewitz, 2002).

The authors concluded that the physician does not risk being overwhelmed by complaints if they listen until the person indicates they have finished, and that two minutes of listening should be possible even in a pressured service (Langewitz, 2002). It is worth remembering that the study was done in a referral center, with patients with complex histories; in less selected populations, the time may be even shorter (Langewitz, 2002).

In practice, two minutes of attentive silence seem long to someone in a hurry. A few simple techniques help sustain listening without losing focus:

• look at the person and not at the screen while they speak;
• use short encouraging signals, such as nodding or saying "mm-hmm";
• repeat a key word to encourage them to continue;
• wait a few seconds after a pause before asking;
• note the topics that come up, without interrupting to go deeper into each one.

The opening listening also helps clinical reasoning. When the person tells their story in their own way, they choose the order, the words and the details, and these choices carry information. The way someone describes a pain, what they mention in passing and what they repeat more than once can suggest hypotheses that a directed questionnaire would not reach. Interrupting early trades this information for a sequence of short answers to the questions the physician already had in mind.

Bringing up all the concerns

Listening to the opening statement is the first step. The second is checking whether there are other concerns before starting to investigate the first. A study with 20 family physicians and 224 patients with acute complaints tested two questions for this, asked after the person presented their main complaint (Heritage, 2007).

The difference was a single word. One version asked whether there was "something else" the person wanted to address; the other asked whether there was "anything else" (Heritage, 2007). The second question eliminated about eight in ten unmet concerns compared with consultations without the intervention, while the first did not differ from the control (Heritage, 2007). Neither increased the length of the consultation (Heritage, 2007).

The study was done in another language, and the exact translation of the questions does not convey the full nuance. The central idea, however, applies: a question that assumes there is something more invites the person to speak, while a question that sounds like a closing invites them to say no. "What else did you want to see today?" works better than "Is that all?".

Even with good questions, some concerns only appear in the middle or at the end of the consultation. A study with 2,243 patients in 182 primary care practices found an unexpected psychosocial agenda in about one in six consultations and an unexpected biomedical agenda in about one in seven (Peltenburg, 2004). These unexpected agendas were not associated with consultation time, but with the person's expectations and the content of the conversation (Peltenburg, 2004).

The text on how to map the agenda before investigating shows how to organize this list in practice.

In practice, these questions work better when the doctor explains why they are asking. Saying that they want to make sure nothing important is left out reassures the person and shows that there is room for other topics. It also helps to write the list where it can be seen, so the person sees that each item was recorded, and to return to it at the end, checking what was addressed and what was left for later.

Ideas, concerns and expectations

Knowing what the problems are is not enough. You need to know what the person thinks about them. A widely recommended technique is to ask about the person's ideas about what is happening, their concerns, their expectations and the effect of the problem on their life. A study with 92 recorded consultations by family doctors analyzed how often these components appear (Edwards, 2023).

About nine in ten consultations had at least one component (Edwards, 2023). The person's ideas appeared in about eight in ten consultations, concerns in a little more than half, expectations in about half and the effect on life in about four in ten (Edwards, 2023). But most of the time it was the person who raised the topic, and the doctor asked directly about expectations in only 3 consultations (Edwards, 2023).

The same study found fewer of these components in problems addressed later in the consultation, in people aged 75 or older and in the poorest (Edwards, 2023). Precisely those who may have the most difficulty speaking up receive fewer questions. The text on what to do with your own bias discusses how this kind of difference appears without the doctor noticing.

These questions change decisions. A study with 613 consultations in 36 family medicine practices showed that, in 104 of them, about one in six, none of these components was expressed, and the doctor decided without knowing what the person thought (Matthys, 2009). When concerns and expectations were expressed, there was a greater chance the doctor would not prescribe a new medication (Matthys, 2009). Sometimes the person just wants to know whether it is serious, and an explanation solves what a prescription would not.

Question	What it reveals	Example wording
Ideas	What the person thinks is happening	"What do you think it could be?"
Concerns	What they fear	"Is there anything that worries you in particular?"
Expectations	What they expect from the consultation	"What did you expect us to do today?"
Effect on life	How the problem changes daily life	"How has this been getting in the way of your routine?"
Other concerns	What has not yet been said	"What else did you want to see today?"

These questions do not need to sound like a script. They can come up naturally over the conversation, at the moment the person describes the symptom or when the doctor proposes a test. What matters is that, before deciding, the doctor knows what the person thinks, what they fear and what they expect, so that the explanation and the plan answer these questions.

Many problems, little time

Raising the whole agenda can be frightening: what if the list is too long? A study with 29 family physicians and 572 patients showed that physicians handled, on average, about three problems per visit (Beasley, 2004). A little more than a third of visits dealt with more than three problems, and among people over 65 the average reached 3.88 (Beasley, 2004).

Addressing several problems at once is therefore the rule of primary care, not the exception. The question is not whether the list will be long, but what to do with it.

Interviews with 40 patients with multiple illnesses and 17 physicians showed how both sides deal with this (Kowalski, 2018). Many patients prepared with written lists, messages to the physician or conversations with the family, so as not to forget anything or feel intimidated at the moment (Kowalski, 2018). One patient described the white coat effect: "You don't think about it when you're in there, you kind of freeze" (Kowalski, 2018).

The physicians said they had learned to like lists. One of them said that early in his career he hated lists and that today he loves them, because they prevent the "by the way" at the door, when the person reveals something serious just as they leave (Kowalski, 2018). The study title refers precisely to that moment: the hand on the doorknob (Kowalski, 2018).

The same study recorded the other side. One patient said that, in a 15-minute visit, they did not ask about a limitation in their arm; another admitted being very depressed and not having told the physician (Kowalski, 2018). Hurry pushes out precisely what is hardest to say.

Negotiating the agenda

With the list in hand, the visit needs an agreement. Not everything fits in one encounter, and the choice of what to address first should be made together with the person. The text on how to prioritize the serious, the urgent, the important and the postponable offers a clinical roadmap for this choice.

A practical way to conduct the negotiation follows a few steps:

• summarize the list of concerns out loud and check whether anything is missing;
• say frankly how much time there is and what seems most urgent from the clinical point of view;
• ask what is most important to the person;
• agree on what will be addressed today and what is left for a scheduled follow-up;
• record the postponed items so they are not lost;
• at the end, check whether the person left with what they most needed.

The physicians interviewed in the study on lists said they had learned over time to help the person and themselves separate what needs to be said that day (Kowalski, 2018). This negotiation is a form of shared decision making applied to the visit itself.

The shape of the conversation also matters. A study with 865 patients from three practices showed that the perception of communication and partnership, with a physician interested in the concerns and expectations and who discusses and agrees on the problem and the treatment, was linked to satisfaction (Little, 2001). A positive approach, clear about the problem and about when it should improve, was associated with fewer symptoms one month later, and the feeling of having a personal relationship with the physician was associated with fewer referrals (Little, 2001).

When an important item is left for later, the person needs to know it has not been forgotten. The text on grouping, separating and re-anchoring the problem shows how to keep the list alive between visits.

The negotiation does not always end in full agreement. Sometimes the person considers something urgent that the physician sees as postponable, or the opposite. Explaining the reasoning, acknowledging the concern and agreeing on a concrete date for the postponed item is usually enough. When it is not, it is worth adjusting the plan, because an ignored concern tends to come back, whether in the same visit, at the door, or as a new visit to the emergency department.

What gets in the way of listening

Several factors push the physician to interrupt. Short time is the most obvious, but not the only one. The computer screen, the habit of following a fixed question script, tiredness at the end of the day and the feeling of already knowing what the person will say also weigh in. The text on what gets in the way of listening details these factors.

Research suggests that listening at the start costs less than it seems. Spontaneous speech lasts, on average, less than two minutes (Langewitz, 2002), and a well-formulated question about other concerns did not increase the length of the consultation (Heritage, 2007). What costs a lot is the concern that shows up at the door, or that does not show up and comes back as a misunderstanding (Barry, 2000).

Some situations call for extra care. In consultations with an interpreter, speaking time is split and the risk of interrupting increases, as discussed in the text about the errors that happen in interpretation. In remote consultations, the lack of non-verbal cues makes it even more important to ask directly what the person wants to address, as the text about history, alarm signs and plan in telemedicine reminds us.

Finally, there are topics a person only brings up when they trust you. Sadness, family problems, alcohol use and fear of a serious illness are rarely the first sentence. The cues the patient offers are usually discreet, and an unhurried start to the consultation is what allows you to notice them.

Frequently asked questions

How long should I let the patient speak at the start?

Until they indicate they have finished. In one study, spontaneous speech lasted on average 92 seconds, and about eight in ten people finished within two minutes (Langewitz, 2002).

Is every interruption bad?

No. About two thirds of the interruptions observed in one study were signs of encouragement (Mulder-Vos, 2023). The problem is the ones that change direction before the person has finished.

How should I ask about other concerns?

With a question that assumes there may be more. The "anything else" version greatly reduced unmet concerns without increasing the length of the consultation (Heritage, 2007).

Why ask what the patient expects?

Because unspoken expectations lead to misunderstandings and unwanted prescriptions (Barry, 2000), and because knowing concerns and expectations was associated with fewer prescriptions of new medicines (Matthys, 2009).

What if the problem list is too long?

List everything at the start, negotiate what will be addressed today and schedule a return visit for the rest. Physicians manage, on average, about three problems per consultation (Beasley, 2004).

Is it worth asking the patient to bring a list?

Yes. Patients and physicians interviewed said the list helps them not forget anything and prevents a last-minute revelation at the door (Kowalski, 2018).

Final checklist

• Open the consultation with an open question and let the person speak until they indicate they have finished.
• Use short signs of encouragement instead of questions that change direction.
• Before investigating, ask what else the person wants to address today.
• Ask about ideas, concerns, expectations and the effect of the problem on their life.
• Summarize the list, negotiate priorities and schedule a return visit for what is left out.
• Encourage the person to bring a written list to the next consultation.

References

1. Barry CA, Bradley CP, Britten N, Stevenson FA, Barber N. Patients' unvoiced agendas in general practice consultations: qualitative study. BMJ. 2000. PMID 10797036. DOI 10.1136/bmj.320.7244.1246.
2. Beasley JW, Hankey TH, Erickson R, et al. How many problems do family physicians manage at each encounter? A WReN study. Ann Fam Med. 2004. PMID 15506571. DOI 10.1370/afm.94.
3. Edwards PJ, Sellers GM, Leach I, et al. Ideas, concerns, expectations, and effects on life (ICEE) in GP consultations: an observational study using video-recorded UK consultations. BJGP Open. 2023. PMID 37277172. DOI 10.3399/BJGPO.2023.0008.
4. Heritage J, Robinson JD, Elliott MN, Beckett M, Wilkes M. Reducing patients' unmet concerns in primary care: the difference one word can make. J Gen Intern Med. 2007. PMID 17674111. DOI 10.1007/s11606-007-0279-0.
5. Kowalski CP, McQuillan DB, Chawla N, et al. 'The Hand on the Doorknob': Visit Agenda Setting by Complex Patients and Their Primary Care Physicians. J Am Board Fam Med. 2018. PMID 29330237. DOI 10.3122/jabfm.2018.01.170167.
6. Langewitz W, Denz M, Keller A, Kiss A, Rüttimann S, Wössmer B. Spontaneous talking time at start of consultation in outpatient clinic: cohort study. BMJ. 2002. PMID 12351359. DOI 10.1136/bmj.325.7366.682.
7. Little P, Everitt H, Williamson I, et al. Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. BMJ. 2001. PMID 11668137. DOI 10.1136/bmj.323.7318.908.
8. Matthys J, Elwyn G, Van Nuland M, et al. Patients' ideas, concerns, and expectations (ICE) in general practice: impact on prescribing. Br J Gen Pract. 2009. PMID 19105913. DOI 10.3399/bjgp09X394833.
9. Mulder-Vos I, Driever EM, Brand PLP. Observational study on the timing and method of interruption by hospital consultants during the opening statement in outpatient consultations. BMJ Open. 2023. PMID 37770276. DOI 10.1136/bmjopen-2022-066678.
10. Peltenburg M, Fischer JE, Bahrs O, van Dulmen S, van den Brink-Muinen A. The unexpected in primary care: a multicenter study on the emergence of unvoiced patient agenda. Ann Fam Med. 2004. PMID 15576537. DOI 10.1370/afm.241.
11. Singh Ospina N, Phillips KA, Rodriguez-Gutierrez R, et al. Eliciting the Patient's Agenda- Secondary Analysis of Recorded Clinical Encounters. J Gen Intern Med. 2019. PMID 29968051. DOI 10.1007/s11606-018-4540-5.