Thirty-four recommendations on talking about side effects, and the grade next to each one
- The PANACEA Placebo Project is an Erasmus+ cooperation partnership, grant number IT02-KA220-HED-000088065, whose outputs are published by the University of Bologna. Its centrepiece is a consensus statement, "Managing Placebo and Nocebo Effects during Patient-provider interactions: A Consensus Statement and Clinical Recommendations for Healthcare Professionals", posted to the project results page on 16 January 2026.
- That document runs to 95 pages. It sets out 34 individual recommendations across five domains, each carrying a class and a level of evidence, plus one overall rating per domain. Beside it sit a 5-page syllabus, posted 23 October 2025, and a 187-page compendium of the eight teaching modules, last updated on the server on 4 February 2026.
- All three files download straight from the results page as PDFs, with no account and no fee. The companion learning platform at eduonplacebo.unibo.it does require registration, asking for name, email and password; no charge for it is stated anywhere on the project site. No reuse licence, Creative Commons or otherwise, is stated on the results page or inside the PDFs; the university's Legal notes page, linked from every project page, requires prior consent from the rights holders for reproduction and requires the source to be cited. No CPD, CME or ECTS credit is offered or claimed anywhere on the site.
- The consensus statement says on its own first page that it is a pre-peer-reviewed version that will be submitted to a journal. Only 2 of the 34 recommendations reach Class I; 23 are Class IIa and 9 are Class IIb, and 13 rest on limited data or expert opinion alone.
Most of what a clinician does about nocebo happens inside one sentence: the one in which a side effect is named. Advice on how to phrase that sentence has circulated for years in consensus papers, usually without any indication of how well each piece of it is evidenced. The PANACEA consensus statement supplies both halves, it is free, and it is honest about which half is thin.
What is actually in it
The recommendations are organised around the arc of a consultation. Optimising the clinician-patient relationship carries 7 recommendations, asking about expectations 3, discussing the treatment rationale 8, discussing risks and side effects 7, and explaining placebo and nocebo effects to patients 9. Each row of Table 1, which begins on page 5, gives the recommendation, its class and its level of evidence, using a grading scheme adapted from the American College of Cardiology and American Heart Association system. A note under the table points to an extended version in the supplementary materials with references and do's and don'ts for each item.
The evidence behind the grades is documented rather than asserted. The consortium ran a systematic review registered on PROSPERO as CRD420250642824, searching PubMed, Embase via Ovid, Web of Science, the Cochrane Library and PsycINFO from 1 January 2000 to 15 May 2024. That yielded 19,114 records, 12,186 unique after de-duplication, and 197 included articles, with 5 more added after discussion with an expert panel. The relationship domain alone drew on 96 studies with a combined n of 133,608. The methods name two evidence streams and put them in this order: the first is prior consensus work and an existing eLearning module whose literature stopped in 2021, and the systematic review is the second.
Three tables are the part a working clinician will actually use. Table 2 lays out the NURSE framework for empathic responses with example sentences. Table 3 gives example questions for eliciting a patient's expectations of a treatment. Table 4 gives example phrasing for explaining placebo and nocebo effects to a patient without either overselling them or implying that the symptom is imaginary.
How to use it, and where to be careful
Start with the fourth domain, discussing risks and side effects, because that is where nocebo is generated and where the guidance is most specific. It contains items such as balancing side effect information with positive information, framing outcomes positively rather than negatively, discussing risk in statistical terms, and stating symptoms briefly rather than dwelling on each one. Read the grade column while you read them. Discussing risks in statistical terms is Class IIa with randomised evidence behind it; framing and brief symptom listing are graded on limited data, and several items in the final domain rest on expert opinion alone. That distribution is the useful information, not an embarrassment: it tells you which sentence you can change on Monday and which one is still a hypothesis.
Two cautions belong on the same page as the enthusiasm. The document is not peer reviewed and says so, which means the grades themselves have not been through external review. And no reuse licence is stated, which is more restrictive than it sounds: the university's Legal notes require prior consent from the rights holders to reproduce the content, with the source cited. Reading and citing are safe. Reproducing the tables in your own handout, or redistributing the modules inside a training course, is a question for the project's contact address rather than an assumption.
The companion survey explains why the material exists. Published in PLOS ONE on 14 January 2026 under a Creative Commons Attribution 4.0 licence, DOI 10.1371/journal.pone.0338905, it drew 807 participants across European countries. Of those, 71.7% reported taking advantage of placebo effects in practice and 55.8% reported observing nocebo effects, while 53.3% felt somewhat confident in harnessing placebo effects and 47.5% felt confident in preventing nocebo ones. Formal training was largely absent: 63.1% had received none on placebo effects and 67.7% none on nocebo. The authors flag their own sampling problem plainly, with only 212 respondents completing the survey in full, 55.19% of participants Italian and 31.6% working in pain.
If you are starting, tapering or switching a medication for anyone this week, open the 95-page PDF at Table 1 on page 5 and read the seven rows under discussing risks and side effects together with their grades, then Table 4 for the wording. That is fifteen minutes, it costs nothing, and it gives you specific wording for the next side-effect conversation rather than a principle to apply on your own.
Only 2 of the 34 recommendations carry the strongest recommendation class, and even those rest on level B evidence, which the document prints in the same table as the advice.
The consensus statement is explicitly a pre-peer-reviewed version awaiting journal submission, and 13 of its 34 recommendations rest on limited data or expert opinion. No reuse licence is stated for any of the materials, and the learning platform requires a free account. Nothing on the site offers CPD, CME or ECTS credit.