Is Bonesetting Safe? What Studies Actually Show

By Eswar Kumar · 9 August 2026

There is no honest one-word answer, and anybody who gives you one — in either direction — has gone past the evidence. What the research supports is narrower: traditional bonesetting is associated with a documented pattern of severe, sometimes limb-destroying complications, and no study has ever established how often those complications occur across everyone treated. Both halves of that sentence are load-bearing.

This article sets out both bodies of evidence — the community studies reporting satisfied patients, and the hospital studies reporting gangrene and amputation — then explains why neither can be turned into a safety rate. If you have an injury right now, read the “when to see a doctor” section below first.

The honest summary

Established: serious harm is real and documented. Hospital series from India, Nigeria, Ghana, Tanzania and Ethiopia record malunion, nonunion, chronic bone infection, compartment syndrome, gangrene and amputation after treatment by traditional bonesetters; two systematic reviews have pooled these complications; the more recent calls them “limb and life-threatening,” including deaths. Most of the harm traces to three things: no imaging, splints applied too tightly, and injuries taken on that should have been referred.

Not established: any complication rate. Not 2%, not 40%, not any number — the data to calculate one has never been collected, and this article will not manufacture it. Nor has any controlled trial shown that bonesetting corrects an injury. Studied is not the same as proven.

So the case for caution does not rest on a statistic. It rests on the fact that a fracture cannot be graded by hand, and that the failure modes — a dead limb, a permanently bent bone — cannot be undone.

What the Puttur study found — and what it can’t tell you

The most-cited favourable data comes from a 2011 paper in the Journal of Ayurveda and Integrative Medicine by Ashok Kumar Panda and Suvendu Rout, a prospective observational study of the Puttur kattu tradition in Andhra Pradesh — a lineage whose history is told separately.

Researchers interviewed 146 patients between 2005 and 2008 and followed 52 to the end stage of treatment. Of those 52, 37 (71%) were satisfied. Loss of joint movement was recorded in one case (2%), followed by malunion, nonunion and delayed union. No case complained of gangrene or Volkmann’s ischaemic contracture.

That is a genuine finding, and it carries far less weight than people quoting it assume. It is an observational description with no comparison group and no randomisation — which is what its authors say it is. Three things follow.

  • Satisfaction is not union. What was measured is whether patients were content, not whether bones healed in alignment. Someone can be entirely satisfied and walking on a malunited tibia.
  • There were no radiographs. The paper records that the clinics have “no X-ray unit in their campus” and that “patients bring their X-rays, but X-rays are given less importance.” A study of fracture outcomes run where fracture imaging is not used cannot confirm what it reports about bones.
  • The 71% describes 37 people out of the 146 who entered the study. The paper does not explain what became of the other 94.

What hospital studies find — and what they can’t tell you

Point an orthopaedic department at the same practice and the picture inverts.

A 2020 retrospective study by Varman and colleagues in the Indian Journal of Orthopaedics Surgery reviewed 362 patients who reached an Indian tertiary hospital between 2016 and 2019 after treatment by native bonesetters. It recorded malunion in 36.74% (133 patients), nonunion in 19.06% (69), joint stiffness in 16.02% (58), compartment syndrome in 5.80% (21), gangrene in 2.48% (9), chronic osteomyelitis in 2.20% (8) and Volkmann’s ischaemic contracture in 1.10% (4). Only 5.24% — 19 patients — had no complication at all.

The pattern repeats beyond India. An eight-year review from Warri, southern Nigeria (Pan African Medical Journal, 2018) found that of 49 complications recorded in 43 patients, nonunion accounted for 40.8% and malunion for 24.5%, alongside gangrene and compartment syndrome. Two systematic reviews have pooled the literature: Onyemaechi and colleagues (Tropical Medicine & International Health, 2021) covering 1,389 patients with 1,470 complications, and Garikapati and colleagues (Tropical Doctor, 2023) covering 833 patients and reporting “limb and life-threatening complications including mortality.”

One sentence governs how all of it should be read. Onyemaechi states of the included studies: “All were hospital-based, observational studies.” Every one of those patients is, by definition, someone whose treatment went badly enough to send them to an orthopaedic surgeon.

Why the two pictures differ (selection bias, explained plainly)

To state a complication rate you need two numbers: how many people were harmed, and how many were treated. Neither body of evidence has both — and this is not a gap that more of the same research would close.

The hospital studies have a numerator with no denominator. An orthopaedic ward counts complications precisely, and has no idea how many people the bonesetters down the road treated that month. It never meets the ones who walked away fine; they had no reason to come. Reading “36.74% of bonesetter patients get malunion” out of the 2020 study is like calculating the failure rate of parachutes by surveying an emergency department — everyone in the sample is there because something went wrong. Those percentages are shares of the harmed, not shares of the treated.

The community studies have a denominator that quietly empties itself. Follow patients inside a tradition and you are following the people who stayed. Anyone whose limb went cold, whose pain escalated, whose bone failed to knit — the exact people whose outcomes you most need — leaves the sample precisely when things deteriorate. They exit the satisfaction survey and reappear in somebody else’s hospital series. The two datasets are not describing different realities so much as catching one population at opposite ends of a pipeline, each blind to the other end.

So the honest position is not “the truth lies somewhere in the middle.” It is that the question of overall risk has not been answered, because nobody has followed a defined group of injured people from the moment they chose a bonesetter through to a radiograph months later, with dropouts accounted for. Any figure presented as the complication rate — favourable or damning — has been produced by ignoring one of these two problems.

That uncertainty is not reassuring. When the potential loss is a limb, an unmeasured risk is a reason for caution, not a licence to assume the number is small. Nor does it mean patients are irrational to go: the reasons people choose bonesetters are substantial ones, including cost, distance, trust and fear of surgery.

The specific dangers: missed fractures, compartment syndrome, infection

Aggregate percentages are abstract. The mechanisms are not, and they recur in every country in this literature.

Missed and misjudged fractures. The foundational problem is diagnostic. Hands find tenderness, swelling and gross deformity; they cannot see whether a fracture line crosses a joint surface, whether fragments are rotated, or whether an apparent sprain is an avulsion. A 2020 study of bonesetters in northern Tanzania (Annals of Global Health) found closed fractures were located by palpation, with only one of six practitioners obtaining radiographs. Malunion and nonunion — the commonest findings in every series above — are largely downstream of this.

Compartment syndrome and gangrene. This is the mechanism that costs limbs, and it is grimly simple. A limb swells after injury. A tight splint does not yield. Pressure inside the muscle compartment rises past the pressure feeding it with blood, and tissue begins to die. The NHS notes that symptoms “can start suddenly, such as after an injury or if a bandage or plaster cast is too tight,” and that surgery is needed “as soon as possible.” Eshete’s Ethiopian study attributed bonesetter gangrene specifically to tight splintage.

The trap deserves naming: a splinted limb is hidden, and pain after a fracture is expected — so escalating pain under a dressing gets read as healing rather than as an emergency. By the time the bandage comes off, the window for a fasciotomy may have closed and the option left is amputation. A 2025 case report in the International Journal of Surgery Case Reports describes a 30-year-old man in Tanzania who developed compartment syndrome and gangrene a week after bonesetter treatment of a forearm fracture, and lost the arm above the elbow.

Infection. Some traditions incise or puncture the skin, or apply preparations directly to wounds; a 2025 survey of 168 patients attending Mbarara Regional Referral Hospital in Uganda after seeing bonesetters recorded cutting or puncturing of the skin among treatments given. Over an open fracture, where the site is already contaminated, this risks chronic osteomyelitis — the persistent bone infection found in 2.20% of the 2020 Indian series, which can mean repeated operations over years.

What would make the practice safer

The literature is markedly less hostile to bonesetters than its complication tables suggest, and unusually specific about remedies. Both systematic reviews land in the same place: bonesetters, in Onyemaechi’s words, “have the potential to contribute positively to primary fracture care when they are trained.”

  • Imaging before treatment, and again afterwards — the single change addressing the largest share of documented harm, and one the traditions ask for themselves. Practitioners in the Tanzanian study specifically requested X-ray access.
  • Splint discipline and explicit warnings. Room for swelling, review within days rather than weeks, and telling every patient plainly which symptoms mean “remove this and go to hospital tonight.” Nearly all the gangrene in this literature had hours or days in which it was still reversible.
  • Referral for the injuries nobody should be splinting — open fractures, vascular compromise, spinal injury, children’s growth-plate injuries. In northern Ghana, 82.1% of bonesetters already recognised difficulty treating open fractures, but only 21.4% regularly referred complications onwards. The knowledge is present; the pathway is missing.
  • Training, which has actually been tested. A two-year survey in southern Ethiopia found 49 amputations, 25 of them for gangrene. After ten one-day courses attended by 112 traditional healers, a two-year prospective study found amputations down from 49 to 25, with only 7 rather than 25 needed for gangrene (Eshete, 2005). It is an uncontrolled before-and-after comparison — but the clearest signal that the harm tracks specific, teachable errors rather than the tradition itself.
  • Collaboration rather than prohibition. In northern Ghana 96.4% of bonesetters wanted engagement with the formal health sector; in northern Tanzania all six endorsed training. A 2025 cluster-randomised pilot in rural Tanzania (BMJ Global Health) paired bonesetters with hospital imaging and analgesia and reported fewer complications — though at 52 patients it was explicitly underpowered, indicating a direction rather than a conclusion.

For anyone studying this tradition rather than seeking treatment in it, the implication is uncomfortable: the most valuable thing to learn is which injuries to refuse.

When to see a doctor

Go to a hospital now — not to a bonesetter, and not tomorrow — if any of the following apply. Several are surgical emergencies measured in hours.

  • Bone visible through the skin, or any wound over a suspected fracture. This is an open fracture: it needs surgical cleaning and antibiotics urgently, and nothing should be packed or applied to it.
  • A limb that is bent, twisted, shortened or obviously the wrong shape. Deformity means displacement, and displacement needs imaging before anything is done to it.
  • A limb that is cold, pale, blue or mottled, or no pulse below the injury. The blood supply may be cut off.
  • Numbness, pins and needles, weakness, or inability to move fingers or toes below an injury. Nerve or circulation involvement.
  • Pain escalating under a splint or bandage — especially pain out of proportion to the injury, or pain on stretching the fingers or toes. Do not assume this is healing; it is the classic presentation of compartment syndrome. The NHS advises calling 999 or going to A&E for sudden severe pain, and not driving yourself.
  • A dressing that has become tight, or a limb swelling inside one. Loosen or remove it and be seen the same day.
  • Fever, spreading redness, foul smell or discharge from an injury site — infection, potentially reaching bone.
  • Any high-energy trauma — road accident, fall from height, crush injury — or any blow to the head, neck or spine, however well you feel afterwards.
  • Loss of bladder or bowel control, numbness around the groin, or severe progressive back pain. A spinal emergency.
  • An injury in a child, particularly near a joint. Growth plates are involved, and damage to them can affect how a limb grows.
  • You are not sure. Uncertainty is itself the reason to be examined. Imaging is cheap relative to a limb.

Fractures and dislocations need imaging and a clinician. A hands-on assessment, however experienced the hands, cannot tell you what a radiograph can. Whatever else you decide, do not leave a splint on a limb that is getting more painful.

Frequently asked questions

Is bonesetting safe?

No study has established a complication rate for traditional bonesetting, so nobody can honestly tell you how risky it is overall. What is well documented is the pattern of harm when it goes wrong: malunion, nonunion, chronic bone infection, compartment syndrome, gangrene and amputation, recorded in hospital series across India, Nigeria, Ghana, Tanzania and Ethiopia. The severity of those outcomes is the reason for caution, not any particular percentage.

What is the complication rate for bonesetting?

There isn’t a known one, and the figures quoted as such are misread. Hospital studies — such as the 2020 Indian series reporting malunion in 36.74% of 362 patients — describe only people who reached hospital because treatment failed, so their percentages are shares of the harmed, not of everyone treated. Community studies have the opposite problem: patients who deteriorate leave the tradition, and the sample, before they can be counted.

Doesn’t the Puttur study show it is safe?

It shows that 37 of 52 patients followed to the end of treatment (71%) were satisfied, with loss of joint movement in one case and no reported gangrene. It does not show safety in the clinical sense: satisfaction is a patient’s opinion, not confirmation that a bone united in alignment, and the paper notes the clinics have no X-ray unit. It is an observational description, not a trial.

What is bonesetter’s gangrene?

It is tissue death in a limb following traditional bonesetting, most often after a splint or bandage is applied tightly over a limb that then swells, cutting off the blood supply and causing compartment syndrome. It is the most catastrophic complication in this literature and frequently ends in amputation. A study in southern Ethiopia found that training bonesetters in safe splinting reduced gangrene amputations in the study region from 25 to 7 over comparable two-year periods.

Should I see a bonesetter for a broken bone?

A suspected fracture should be assessed by a clinician with imaging first, because the decisions that matter — whether fragments are displaced, whether a joint surface or growth plate is involved, whether surgery is needed — cannot be made by hand. An undiagnosed fracture managed under a splint is the situation from which most of the documented serious harm arises.

Sources

This article is educational and is not medical advice. If you have an injury, see a qualified clinician. Bonesetting techniques should only be practised after supervised, in-person training and in line with the laws where you live.