Efficacy of Prolotherapy for Osteoarthritis: A Systematic Review.

All 14 trials were judged to carry an overall high risk of bias (one low risk on its own, four unclear, nine high), mostly because participants and the people assessing outcomes were not blinded. The trials used different dextrose concentrations, injection sites, numbers of injections and intervals between them, so their results cannot be meaningfully combined. The authors give this as their own explanation for the inconsistent findings. Nothing was pooled. The review counts how many trials pointed in each direction rather than producing an estimate of how much prolotherapy helps. Eleven of the 14 trials were in knee osteoarthritis, with only two in hand and one in hip, so anything said about other joints rests on very little. In some trials the prolotherapy group also did exercise or rehabilitation, so the injections' own contribution cannot be separated from the rest of the treatment. Data were missing from several included trials, and in two trials the outcomes reported in the protocol and the publication did not match. The advice the authors give about dextrose concentration and injection intervals comes from comparing one trial with another, not from a trial that tested one dose against another.

The text available does not include a funding statement or a declaration of competing interests, so it is not possible to say who paid for this review or whether the authors had ties to any maker of the products compared.

Do not take this review as proof that prolotherapy works for osteoarthritis. Nine of the 14 trials were rated high risk of bias and the overall judgement across trials was high risk, mostly because participants and outcome assessors were not blinded, the injection protocols varied widely, and the results were never pooled into a single estimate. The authors themselves concluded that this makes it difficult to consolidate the treatment's benefit.