PPIE examples: what good evidence looks like to reviewers

The shape every good example shares

Strong PPIE evidence is not better-written involvement. It is the same activity, reported against a structure that a vague account skips. Four parts, in this order:

1. What was proposed. The design, wording, or plan as it stood before contributors saw it. 2. What contributors said. The specific concern or suggestion, ideally close to their own words. 3. What changed. The concrete difference in the protocol, materials, or plan afterwards. 4. How you know it stuck. Where the change is visible in the application: the protocol version, the materials pack, the outcome list.

Leave out step three and you have a consultation log, not evidence. Leave out step one and a reviewer cannot tell whether anything actually moved. Most weak PPIE sections are missing one of these four, almost always the same one: what changed.

Five places reviewers look for it

These are composite, illustrative examples built from patterns we see across applications, not quotes from a specific study. Use them as a template for your own sentences, not as text to copy in.

1. The research question or protocol.

Weak: "The research question was refined following patient input."

Good: "The protocol proposed a single primary outcome of symptom severity at 12 weeks. Contributors felt this missed what mattered most to them day to day, so a secondary patient-reported outcome measure of impact on daily activities was added, and appears in section 4.2 of the protocol."

2. Recruitment and eligibility criteria.

Weak: "Contributors were consulted on the recruitment strategy."

Good: "The draft criteria excluded people currently receiving informal care, on the assumption they could not commit to weekly visits. Two contributors who are carers pointed out this excluded exactly the group with the most to say about care burden. The exclusion was removed and replaced with a flexible visit schedule, reflected in the eligibility table in the application."

3. Participant-facing materials.

Weak: "The participant information sheet was reviewed by our PPI group."

Good: "The first draft of the information sheet ran to six pages and used the phrase 'randomised controlled trial' without explanation on page one. Contributors said they would not have read past that. The sheet was rewritten to two pages with a plain-English summary first, and randomisation explained using the coin-toss analogy contributors suggested themselves."

4. Outcome measures and what "success" means.

Weak: "Patient priorities were considered when selecting outcomes."

Good: "The team had planned to measure success solely by reduction in clinic attendance. Contributors argued that fewer appointments could equally mean people had given up trying to get help, and asked for a measure of whether people felt more in control of their condition. A validated self-efficacy scale was added alongside the attendance measure."

5. Dissemination and reporting findings back.

Weak: "Findings will be disseminated to patients and the public."

Good: "The original dissemination plan was a journal article and a conference poster. Contributors pointed out that neither would reach anyone actually living with the condition, and helped design a one-page plain-English summary to be sent to everyone who took part, plus a short video for the condition-specific charity's members' newsletter."

When nothing changed, and that is fine

Not every good example ends in a change. Sometimes contributors review something and confirm it is right as it stands, and that is a legitimate outcome, provided it reads as a genuine test rather than a formality.

Weak: "Contributors were happy with the proposed approach."

Good: "Contributors were specifically asked whether the six-month follow-up window felt too long, given similar studies use three months. They considered both options and confirmed six months was preferable, since it allowed time for symptoms to settle before assessment. No change was made, but the decision is now evidenced rather than assumed."

The difference is specificity about the question that was actually put to them. "Were you happy with this?" invites agreement. "Is six months too long, and why not three?" invites a real answer, and a real answer either way is evidence.

Building your own bank of examples as you go

The reason so many applications end up with weak, generic sentences is not a lack of real influence. It is that the specific detail gets lost between the session and the write-up, usually because nobody wrote down "before" and "after" at the time.

Note the four-part structure straight after each session, while the specifics are still sharp: what was on the table, what was said, what changed, and where that change now lives in the document. A rough note taken on the day beats a polished paragraph reconstructed three months later from memory, because the reconstructed version is exactly where "contributors were supportive" creeps back in.

This is also what GRIPP2 reporting is built to capture, and keeping a running log against its headings during the study, rather than compiling one retrospectively before submission, is the difference between an impact section that writes itself and one that has to be invented at the last minute.

How we help

Producing evidence like this is a by-product of how we run PPIE, not an extra step. Every session we deliver as part of our PPIE service is written up against what was proposed, what was said, and what changed, so the material for your application or your funder report already exists when the session ends.

For the section this evidence goes into, see writing the PPIE section of an NIHR application, and for what it costs to generate, budgeting for PPIE in NIHR grant applications. Our guide to what NIHR expects from patient involvement covers the standards this evidence is judged against.