CBD or Codeine After Eye Surgery? One Trial Compared Both

A trial compared oral CBD with codeine after laser eye surgery

CBD and opioids are rarely tested against each other in the same people, at the same time, for the same pain.

This trial made that comparison. Each patient received both treatments, so every patient served as their own comparison.

What the trial did

Thirty-five adults at the University of Florida had photorefractive keratectomy (PRK), a laser procedure that reshapes the cornea. The recovery can be painful during the first days while the surface of the eye heals.

Each patient had both eyes treated in sequence. For one eye’s recovery, they took oral CBD, 50 mg twice a day. For the other eye, they took codeine with acetaminophen, 30 mg and 300 mg every four hours. The treatment order was randomized, and the CBD product contained a low amount of THC.

Pain was recorded in a diary, along with symptoms and quality-of-life scores, and vision was measured out to three months.

What they found

• Pain peaked on the second day after surgery in both cases, which matches what surgeons expect from this operation
• There was no significant difference in pain between the CBD eye and the opioid eye, whether measured by average pain or worst pain (P = .538)
• Vision improved significantly by three months (P < .001), with no difference between the treatments (P = .928)
• Other symptoms and quality-of-life scores were equivalent
• One difference did show up: patients reported a change in how they felt mentally, described in the paper as sensorium, when taking CBD for the first eye but not the second (P = .035), which the authors read as a first-exposure effect

The authors concluded that low-THC oral CBD provided “noninferior pain control” compared with codeine-acetaminophen after this surgery and may offer an opioid-sparing option in this setting.

Why this design is unusual

Most pain trials compare one group of people with another. The groups may differ in age, pain tolerance, healing, or other ways that affect the result.

In this trial, each patient received both treatments. The surgeon and operation were the same, and the treatment order was randomized. This design reduced the influence of individual differences in pain and healing, although it did not eliminate every possible source of bias.

The trial did not include a placebo group, so it cannot show how either treatment compared with no active pain medicine. It tested CBD against an active treatment instead.

What “non-inferior” does and does not mean

“Noninferior” has a precise meaning, and it is narrower than saying the two treatments are equal.

A finding of no significant difference is not proof that two treatments are equal. With 35 patients, a meaningful difference could remain undetected. The trial looked for a difference between the treatments and did not find one.

It also does not say CBD is a painkiller in general. It says that for the specific pain of corneal healing, in the days after one specific operation, at 50 mg twice daily, the diaries did not separate the two treatments.

The study provides no evidence about back pain, knee pain, or headaches, which involve different types and causes of pain.

What a supporting study adds

High-dose CBD for nerve pain after spinal cord injury (2026, EClinicalMedicine): A University of Sydney team used a randomized, double-blind, placebo-controlled crossover design. Each participant took CBD and placebo in separate treatment periods, with the order randomized and neither participants nor researchers told which one they were taking.

Forty adults with neuropathic pain lasting at least three months were randomized, and 38 were included in the analysis. CBD was titrated up to 800 mg per day during one six-week period. Placebo was used during the other, with a four-week break between periods.

• Pain was significantly lower on CBD than on placebo during the active phase, a mean difference of 0.54 points on a 0 to 10 scale (95% CI 0.21 to 0.88, p < 0.001)
• The two treatments did not differ before dosing started in each phase
• Most adverse events were minor, and the event rate per participant was similar with CBD and placebo (2.58 compared with 2.55)
• The authors described the effect as modest and said the result supports further research

The trials answer different questions. The eye trial compared low-THC oral CBD with an active pain medicine after surgery. The spinal cord trial compared high-dose CBD with placebo for chronic nerve pain and found a small reduction. The daily doses differed eightfold, but the conditions, comparisons, and treatment periods also differed.

Why the results differ

The results differ, but they do not contradict each other.

The eye trial found no significant pain difference between CBD and an opioid during a few days of acute pain. The spinal cord trial found a small benefit over placebo during six weeks of chronic pain.

The trials studied different types of pain, doses, comparisons, and treatment periods. Neither result can be used to predict the other.

What this means for you

The safety picture

The eye study’s authors reported minimal side effects, with mild sensorium changes during some patients’ first CBD exposure. In the spinal cord trial, most adverse events were minor, and event rates per participant were similar with CBD and placebo. CBD can interact with some prescription medicines, so anyone considering it should discuss possible interactions with a pharmacist or doctor.

What the trials did not test

Neither trial tested how CBD produced a change or compared different doses within the same condition. These results therefore cannot identify a pain-relief mechanism or establish which dose should be used.

The everyday connection

When reading CBD research, compare the condition, product, dose, and treatment period with your situation. Results from post-surgical eye pain or chronic nerve pain cannot tell us what to expect from a retail capsule used for an everyday ache.

Worth keeping in mind

• The eye trial had no placebo group. It compared CBD against an active drug, which answers a different question.
• Thirty-five patients and thirty-eight patients. Both are small.
• Post-surgical eye pain and spinal cord nerve pain are specific conditions, neither of which resembles everyday aches.
• The trials used different doses in different conditions. Neither trial compared doses within the same condition.
• The benefit in the spinal cord trial was modest, about 0.5 points on a ten-point scale, and the authors say so.
• Both trials were short, at six weeks or less per treatment period.

The eye study supports a narrow conclusion about pain after one operation; it does not establish that CBD works like an opioid for pain in general. The spinal cord trial concerns a different condition and dose, so it cannot broaden that claim.

---

About the original studies

Title: Randomized, controlled crossover comparison of oral cannabidiol with oral opioid for postoperative PRK pain control

Published: July 2026, Journal of Cataract and Refractive Surgery

Authors:

• Yujia Zhou, Department of Ophthalmology, University of Florida, Gainesville, Florida
• Kayla Swiatek, Department of Ophthalmology, University of Florida
• Andrew Nguyen, University of Florida College of Medicine
• Sonal Tuli, Department of Ophthalmology, University of Florida
• Walter A. Steigleman, Department of Ophthalmology, University of Florida

Link: https://doi.org/10.1097/j.jcrs.0000000000002030

Title: High-dose cannabidiol for chronic neuropathic pain associated with spinal cord injury: a randomized clinical trial

Published: June 2026, EClinicalMedicine

Authors:

• Rebecca V. Robertson, School of Medical Sciences, University of Sydney
• Anastasia Suraev, Brain and Mind Centre, University of Sydney
• Danielle McCartney, Lambert Initiative for Cannabinoid Therapeutics, University of Sydney
• Elizabeth A. Cairns, School of Pharmacy, University of Sydney
• Iain S. McGregor, Lambert Initiative for Cannabinoid Therapeutics, University of Sydney
• Luke A. Henderson, School of Medical Sciences, University of Sydney

Link: https://doi.org/10.1016/j.eclinm.2026.103986

---