Reiki Tested in an Intensive Care Unit: An Honest Read of a Sham-Controlled Trial on Ventilated Patients

Key Takeaways

• Sixty patients on invasive mechanical ventilation were randomised: thirty received a thirty-minute Reiki session, thirty received a sham session performed by an uncertified person with no intent to transmit energy.
• Reported results favoured the Reiki group on pain and anxiety, and on diastolic blood pressure and heart rate. Systolic pressure did not reach the usual threshold.
• The design is stronger than most Reiki studies because of the sham arm — and sixty patients at one site still cannot settle the question. Both things are true.

Key Terms Explained

• 靈氣 (Reiki) / Universal Energy — Connecting with the energy of the universe and letting it settle life in a better direction.
• 霊授 (Reiju) / Attunement — The transmission given in person by a teacher, traditionally the true beginning of a practitioner's path.
• 師範 (Shihan) / Master Teacher — The level qualified to teach and to give attunement in the traditional lineage.
• Sham control — A convincing imitation of the treatment, given so that patients cannot tell which group they are in. Here, an uncertified person going through the motions without intent.
• Single-blind — The patients did not know which group they were in. The people delivering the sessions necessarily did.

The Setting Nobody Chooses for a Gentle Study

An intensive care unit is an unusual place to test anything gentle.

Patients on invasive mechanical ventilation cannot speak. They are frequently sedated. Pain and anxiety are measured through scales built for exactly that situation, because the ordinary method — asking — is unavailable.

I mention this because it changes what the study is doing. Most Reiki research asks people how they felt afterwards, and self-report is where the softest evidence lives. Here, some of the outcomes were blood pressure and heart rate, which do not care what anyone believes about the session.

That is the part I found genuinely interesting when I read it, and it is also where I want to be careful. An engineer's habit dies hard: when a result looks favourable, the first question is what else could produce that number.

Related: Reiki in the ICU: An Honest Reading of a 2026 Sham-Controlled Trial on Ventilated Patients explains this in detail.

What the Study Actually Did

The sham arm is the design choice worth pausing on.

Related: The Same Practice, Tested Two Ways: What a Dialysis Trial Without a Sham Arm Can and Cannot Tell Us explains this in detail.

Why a Sham Arm Matters Here

If you compare Reiki against nothing, any result can be explained by attention. Someone came, stayed thirty minutes, and was quiet and present. In an intensive care unit — an environment of noise, interruption and procedure — thirty uninterrupted minutes is itself an intervention.

A sham arm holds that constant. Both groups got the person, the thirty minutes and the quiet. What differed was whether the practitioner was trained and whether there was intent.

Whether that difference is the right thing to isolate is a fair philosophical question, and I do not think it has a clean answer. But as a study design it is considerably more demanding than the usual comparison, and studies that pass a harder test deserve to be read more carefully than studies that pass an easy one.

Related: 272 Patients, No Difference: Reading the Largest Reiki Trial That Found Nothing explains this in detail.

The Reported Results

Two rows deserve comment, and they are the two that are easy to skip past.

Systolic pressure at 0.052. This is the number that did not clear the conventional line. In practice it is barely distinguishable from the ones that did, and I mention it because leaving it out would give a tidier picture than the study supports. A result at 0.052 is not a result; it is an invitation to run the study again with more people.

Respiratory rate. A group effect without a time-by-group interaction is a weaker finding than it first appears. Roughly: the groups differed, but the pattern of change over the session did not clearly differ between them. That is worth noting rather than presenting as a straightforward win.

Heart rate at 0.001 is the strongest single number in the set.

What Sixty Patients Cannot Settle

Here is where I want to be as honest as I am about my own practice.

Sixty patients is small. Sixty split into two arms of thirty is smaller still. Small trials produce more variable results than large ones, and a favourable small trial is a reason to run a bigger one rather than a reason to conclude.

One site. Everything about a single intensive care unit — its staffing, its noise, its sedation practice — travels with the result. Whether the same thing happens elsewhere is unknown until someone checks.

One session. This measured a single thirty-minute session. It says nothing about repeated sessions, and nothing at all about the daily practice that this site is mostly concerned with.

Blinding has a limit. Patients did not know their group. The people delivering the sessions did, and in a study where outcomes include an observer-rated pain scale, that is a real constraint. It is inherent to hands-on interventions rather than a fault of these researchers.

Short horizon. The measurements are around the session. Whether anything persists is outside what this design can see.

None of this makes the study uninteresting. It makes it what it is: one reasonably designed piece of evidence, in one setting, pointing in one direction.

What I Take From It, and What I Do Not

I have practised daily for more than twenty years, and I still make the same limited claim: Reiki is not medical care, it does not diagnose or treat anything, and I make no claims about illness. That does not change because a trial reported favourable numbers.

What I will say is narrower. In a setting where the patients could not know which group they were in, and where two of the outcomes were measured by a machine rather than reported by a person, the trained sessions and the imitation sessions did not produce identical results. That is worth knowing. It is not proof of a mechanism, and the study does not claim to have found one.

I am also aware of how this kind of result gets used. Within a week, a number like p = 0.001 will appear in marketing copy with the sample size removed and the word "proven" attached. That is the thing I would ask readers not to do. A study is only as strong as the part of it people are willing to repeat out loud, and the sample size is part of the study.

There is one more thing worth saying about the setting itself. Patients on ventilators cannot consent in the ordinary way, cannot ask what is being done to them, and cannot decline mid-session. I would want to know that consent arrangements were handled carefully before I got enthusiastic about anyone rolling this out. Consent is not a footnote in this practice — it is the first thing. I do not send distant Reiki to anyone who has not asked for it, and the same instinct applies with more force, not less, when the person cannot speak.

FAQ

Q: Does this study prove Reiki works? A: No, and it does not claim to. Sixty patients at one hospital, one session each, with the practitioners aware of group assignment. It reports a favourable result under a reasonably demanding design, which is a reason for a larger trial rather than a conclusion.

Q: What does a "sham" Reiki session involve? A: In this study, an uncertified person went through the motions of a session without any intent to transmit energy. The patient experienced someone present and quiet for the same length of time. The purpose is to hold attention and presence constant so they cannot explain a difference between the groups.

Q: Why does the systolic blood pressure result matter if it was not significant? A: Because reporting only the results that cleared the line would misrepresent the study. At 0.052 it sits just outside the conventional threshold, which in a study this size means the honest answer is "unresolved" rather than "no effect."

Q: Should hospitals start offering this? A: That is a decision for clinicians and administrators working from the whole body of evidence, not from one trial. My own view is that consent arrangements deserve at least as much attention as the results, particularly for patients who cannot speak.

Q: Does this tell me anything about my own daily practice? A: Very little, honestly. This measured one thirty-minute session on critically ill patients. It has almost nothing in common with five quiet minutes at home each morning, and I would not lean on it to justify anything about ordinary practice.

Key Insights to Remember

• The sham arm is what makes this study worth reading. Comparing Reiki to nothing tells you about attention and quiet; comparing it to a convincing imitation isolates something narrower. Studies that accept the harder comparison deserve more of our attention than the ones that avoid it, whichever way the results fall.
• Every honest reading includes the numbers that did not work out. Systolic pressure at 0.052 and a respiratory-rate finding without a time-by-group interaction belong in any summary of this trial. A version that reports only the strong results is not a summary; it is advertising with citations attached.
• Sixty patients at one site is a beginning, not an answer. The correct response to a favourable small trial is to want a larger one, and the correct response to seeing it quoted as proof is to say so. Being unwilling to overstate the evidence is not a weakness in this practice — it is the only position that survives contact with a sceptical reader.

Sources

• Effect of Reiki on Pain, Anxiety, and Hemodynamic Parameters in Mechanically Ventilated Patients: A Randomized, Single-Blind, and Placebo-Controlled Trial — Journal of Integrative and Complementary Medicine (2026)
• The Effect of Reiki on Pain, Anxiety and Haemodynamic Parameters in Patients Receiving Mechanical Ventilator Support — ClinicalTrials.gov (NCT06526949)
• Reiki: What You Need To Know — National Center for Complementary and Integrative Health (NCCIH)