The first call: an inbound script that respects the patient

01

The person answering the phone is doing clinical work

Nobody writes that on the job description. Look at what actually happens on the call, though.

A 2024 study in Health Communication by Weatherall and Grattan examined audio recordings of calls to medical reception at a New Zealand practice using conversation analysis. Across that small set of calls, the authors found receptionists assessing the possible urgency of a caller's problem and beginning a triage process while simultaneously operating a booking system, and they described this part of the work as largely unrecognised.

That matches what I see. The person on the phone is deciding, in real time and usually without training, whether this caller waits eleven days or gets a slot tomorrow. If your script treats them as a data entry function, you have buried a clinical decision inside an administrative role and given it no support.

So the script is not there to control that person. It is there to protect the caller from a bad day, a new starter, or a queue of forty.

02

Silence is the cheapest thing in the script

The strongest evidence about openings does not come from contact centres. It comes from consultations, and it transfers cleanly.

A 2019 study in the Journal of General Internal Medicine, titled Eliciting the Patient's Agenda, analysed recorded clinical encounters and found that clinicians asked patients what they had come about in only 36 percent of them. Where they did ask, they interrupted in 67 percent of cases, at a median of eleven seconds. The figure I keep returning to is the other one in that paper. When patients were left alone, they finished their opening statement in a median of six seconds.

Six seconds. That is the entire cost of letting somebody say what is wrong with them.

Format matters as much as patience. A 2006 study in Patient Education and Counseling by Robinson and Heritage, covering visits between 28 primary care physicians and 142 patients, found that opening with a general open inquiry rather than a closed request for confirmation was associated with better patient ratings of the physician's listening and of the relational side of the encounter. The practice implication the authors drew was plain: open with an inquiry, not with a guess at the answer.

On the phone that means the first line is some version of how can I help you today, and then nothing at all. Not are you calling about an appointment. That version invites a yes, and the yes hides everything you needed to hear.

Almost every bad first call I have reviewed was a person being interrupted at second nine to be asked for a postcode.

03

The first call card

Six lines. Print it, put it next to the monitor, use it for calibration, and resist adding a seventh. If it does not fit on one card it will not survive a busy Tuesday.

One. Answer inside three rings with the clinic name, your own first name, and an offer of help. A first name is a commitment. It tells the caller there is a specific human at the other end who can be asked for again.

Two. Ask one open question, then say nothing until they stop. Count to two after you think they have finished. People put the real reason in the second sentence more often than the first.

Three. Say it back. Something like: so this started about three weeks ago and it is worse in the mornings, have I got that right. Eight seconds, and it is the moment the caller decides whether you were listening.

Four. Answer the question they actually asked before you ask yours. If they wanted to know the cost, give the range, or say honestly that nobody can quote until a clinician has assessed them. Then collect what you need.

Five. Give one next step carrying a name and a time. Not somebody will call you. A clinician will call you between four and six today, and here is my name if that does not happen. Ownership plus a window.

Six. Close by asking whether anything is still unclear, and then wait for the answer. Same rule as line two.

04

Write down what was said, not what the form wants

Three fields do more work than the twenty most systems ship with.

The reason for the call, in the caller's own words, copied not paraphrased. What was promised, by whom, and by when. And what you did not know, written as a question for whoever picks this up next.

Everything else can be captured later or inferred. I have seen teams spend forty seconds of a ninety second call on source attribution and preferred contact method while the actual clinical concern went into a free text box nobody opened again. If a field is not read by a named person within a week, delete it. The reception team is paying for it in attention, and attention is the only thing the caller can hear.

05

What the script can carry, and what it cannot

A 2022 cross-case analysis in the Journal of General Internal Medicine by Chuang and colleagues looked at telephone access at six Veterans Administration medical centres with strong appointment availability. Sites where patients were more satisfied had call handling better integrated with primary care workflow, all six sites reported chronic understaffing of the call function, and the authors noted that standard call centre performance metrics do not capture how well the call function is integrated with clinical work at all. That last point is the one worth pinning up. Measure only handle time and abandonment, and you will punish exactly the behaviour this article asks for.

So here is the honest split. You do not control how many people ring at once, whether the clinical diary has space this week, what a competitor quoted yesterday, or whether the caller was already frightened before they dialled. Those are conditions, and arguing with them at the desk is wasted energy.