> ## Content Index
> Fetch the complete content index at: https://pulsecorp.co/blog/llms.txt
> Use this file to discover other available public pages before exploring further.

# How to run patient interviews for digital health research
- URL: https://pulsecorp.co/blog/patient-interviews-for-digital-health-research/
- Published: 2026-11-07T12:00:00.000Z
- Updated: 2026-11-07T12:00:00.000Z
- Description: Patient interviews for digital health research: how to word consent, skip health details you don't need, and ask about missed doses without blame.
- Author: Khyati Ojha

Say you're on the research team at a company that makes a medication reminder app for people with high blood pressure. Many users stop after a few weeks. You need to ask them why, but the topic is their health. So what do you tell them first, and what do you ask?

These are patient interviews for digital health research, and they need extra care.

First, tell patients plainly what you'll ask, how their answers are kept and that they can stop at any time. Then ask about their routine and the app, and skip health details the study doesn't need.

Three choices shape the study: the consent wording, the health details you leave out, and how you ask about missed doses.

## The short version

- Patients may share health details, so consent comes before the first question.
- Write consent in plain words, and repeat it when the recording starts.
- Ask only for the health details your decision needs.
- Ask about missed doses as a normal part of life, never as a test.
- Interview patients, caregivers and clinicians separately.
- Share quotes with reviewers only after personal details are removed.

## Why patient interviews need more care

A patient interview looks like any customer interview, but the person may mention a diagnosis, a medicine or a hard week. That changes what the company must protect and how the questions should sound.

The answers can include health data. Some patients are tired, worried or unwell, so interviews should be short and easy to pause. And family members often help with the app, so the person tapping the reminder may not be the patient.

It helps to test your questions before patients hear them. In a [2024 interview study in Frontiers in Digital Health](https://www.frontiersin.org/journals/digital-health/articles/10.3389/fdgth.2024.1275569/full#:~:text=The%20authors%20developed%20the,patient%20with%20major%20depression.), researchers spoke with 17 patients with mild or moderate depression about digital treatments. They tested their questions with two volunteers first, and the recorded interviews lasted between 24 and 42 minutes. A trial run like that shows you which questions confuse or upset someone before a patient meets them.

Which rules apply depends on where the company works and what data it collects, and your privacy, legal and clinical teams decide that.

Bring those teams in before you write the first question.

## What to tell patients before they agree

Consent for an interview is a short, plain explanation the patient agrees to before any question. It says what the study is for, what topics will come up, how long it takes, how recordings are stored and who sees them. It also says that taking part won't change their care, and that they can skip a question or stop.

Say the note for the reminder app reads: "We're talking to people who use our reminder app to learn how it fits into their day. The interview takes about 20 minutes. We'll ask about your routine and the app, not your medical history. We remove your name and personal details before the product team reads anything. Taking part won't change your care or your account. You can skip any question or stop at any time." Your privacy team approves the final wording.

Give people time to read it. When [30 parents of young children tested a digital consent app](https://bmcmedethics.biomedcentral.com/articles/10.1186/s12910-023-00974-1#:~:text=Participants%20also%20received%20a,for%20the%20test%20children.) for a public health study, they spent between 4 and 15 minutes, 8 on average, reading about the study and giving consent (BMC Medical Ethics, 2023). That was for a whole study, but consent still takes time to plan for.

People may also prefer shorter wording. In a [2025 survey of 79 people eligible for a digital health study](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11803319/), published in the Journal of Medical Internet Research, people were less likely to prefer the original consent wording when it was longer.

Repeat the key points when the recording starts, because some patients read the invitation days earlier.

![Checklist of what a plain-language consent note for a patient interview covers: what the study is for, the topics, how long it takes, how recordings are kept and who sees them, no change to care, and the right to skip or stop at any time](https://pulsecorp.co/blog/content/images/2026/11/patient-interviews-for-digital-health-research-meah6x.webp)

## Ask only for the health details you need

Before you write a question, ask whether the study needs the answer. The product team wants to know why people stop using reminders. That's about daily routine, not test results, so most questions about diagnosis, medicines and doses can go.

Collecting only what the decision needs is called data minimization. Each pair swaps a question that collects too much for one that answers the product team's question.

| Question that collects too much                 | Better question                                      | Why                                               |
| ----------------------------------------------- | ---------------------------------------------------- | ------------------------------------------------- |
| Which medicines do you take, and at what doses? | How many reminders do you get from the app each day? | The study needs the routine, not the prescription |
| What did your doctor say at your last check-up? | When did you last change how you use the app?        | It asks about the app, not the visit              |
| What is your blood pressure now?                | What made you open the app, or ignore it, this week? | A reading doesn't explain what people do          |
| How long have you had high blood pressure?      | How long have you used the reminder app?             | Time with the app is what the team can change     |

Tell patients at the start that they don't need to name a condition or a medicine. Some will anyway, so plan for it: names and health details are removed from transcripts before findings are shared. Leave video out unless the question needs it, because a face or a screen can show health information.

## How to ask about missed doses without blame

If you ask "Why did you stop taking your medicine on time?", many patients give the answer they think you want. The question sounds like a test they failed. So make missing a reminder sound normal before you ask about it, and ask about one real day instead of what they "usually" do.

A better opening is: "Lots of people miss a reminder now and then. When did that last happen for you?" Then ask about the day around it, such as work, travel or a change in medicine.

Taking a medicine as planned is called adherence. Questions about adherence are easier to answer when they ask about events instead of habits. A few that fit the reminder app:

- Walk me through yesterday morning. When did the reminder come?
- What were you doing when it came?
- Tell me about the last time you turned a reminder off.
- What changed in the week you stopped opening the app?
- What would make a reminder easier to act on?

These fit into any [interview guide you write](https://pulsecorp.co/blog/how-to-write-a-user-research-interview-guide-step-by-step-examples/).

Say a patient tells you the reminders were annoying. A second question asks when they came. If they say the reminders arrived in the middle of the school run, the problem is the timing, and the fix is to let people pick their own reminder time.

Some patients will get upset, because health is personal. If that happens, pause, offer to skip the question or stop, and point them to their own doctor or care team for health questions. The interviewer never gives medical advice.

![Before and after diagram: the question "Why did you stop taking your medicine on time?" sounds like a test, while the better approach makes missing a reminder sound normal, asks about the last time it happened, and then asks about the day around it](https://pulsecorp.co/blog/content/images/2026/11/patient-interviews-for-digital-health-research-b8qusq.webp)

## Patients, caregivers and clinicians answer different questions

Interview each group on its own, because each knows a different part of the story. Patients know their routine. Caregivers often set up and run the app for a parent or partner. Clinicians see who stops and hear what patients say about the app in the clinic.

Each row shows what one group can add, and what to keep out of their interview.

| Group      | What they can tell you               | What to ask about                         | What to leave out                    |
| ---------- | ------------------------------------ | ----------------------------------------- | ------------------------------------ |
| Patients   | How the reminder fits their day      | Routine, reminders, the day they stopped  | Diagnosis, doses, test results       |
| Caregivers | How the app works when someone helps | Setting up the app, who acts on reminders | The patient's private health details |
| Clinicians | Patterns across many patients        | What patients say about the app in visits | Any one patient's records            |

You can invite the company's own app users, with their permission, or recruit people who match the profile you need. Each group needs its own screener, a short set of questions that checks someone fits before the interview. Keep the same core questions across groups, so you can compare what each one says.

If a question won't change what the product team builds, leave it out of a patient interview.

## Where Pulse fits

Patients are busy, tired or unwell, and booking calls around their day is slow. [Pulse runs AI-moderated voice interviews](https://pulsecorp.co/product/ai-moderation/) that patients take by opening a link on a phone or laptop, at a time that suits them. Every patient gets the same well-formed questions from one interview guide. The moderator asks a second question when an answer is vague, and its follow-ups build on what the patient said earlier.

Personal data is redacted from transcripts, Pulse is GDPR compliant, and customer data is never used to train AI models. Data residency options are available on the enterprise plan.

Pulse groups answers into themes with the quotes behind each one, and every quote links to its transcript, so clinical and product reviewers can check a theme against what patients said. If you're planning patient interviews for a health app, [book a demo](https://pulsecorp.co/demo/).

## Questions readers ask

### 1\. How long should a patient interview be?

Keep it as short as your questions allow, and make it easy to pause or split into two parts. Patients who are tired or unwell may need a break, so offer one halfway through. Test the guide with one or two volunteers first, and cut any question whose answer won't change a decision.

### 2\. Should you pay patients for taking part?

Many studies offer a thank-you for the patient's time, such as a voucher. Keep it modest and the same for everyone, so it doesn't push people to take part against their own judgment. Your legal or ethics reviewers may have rules about payments to patients, so agree the amount with them first.

### 3\. Can a caregiver answer for the patient?

A caregiver can tell you how the app works at home, but they can't speak for the patient's own experience. If the patient can take part, interview them directly and the caregiver separately. If the caregiver does most of the work with the app, they may be the right person to interview anyway.

### 4\. How do you share patient interview findings with reviewers?

Share themes, with the quotes behind each one, after names and personal details are removed. Clinical reviewers can check that nothing reads as medical advice, legal reviewers can check how the data was handled, and the product team can trace each theme back to the transcripts it came from.