A week of antibiotics matched four in bone infection when a local dose was implanted

Orthopaedic infections — of bones, joints and the metalwork used to repair them — have long been treated with surgery followed by weeks of systemic antibiotics. Those weeks come at a cost: side effects, disrupted lives, the burden of intravenous lines, and steady pressure on antibiotic resistance. The convention rests more on tradition and caution than on trial evidence, and much of the field has quietly wondered whether the long courses are necessary. SOLARIO, published in the New England Journal of Medicine on 16 September, asked whether that long tail of antibiotics can be cut short when a local dose is placed directly at the infection site during surgery [s1]. Its answer favours doing less.

What the trial did

SOLARIO was a multicentre, open-label, noninferiority trial [s1]. It enrolled adults who had undergone surgery for an orthopaedic infection and had a local-antibiotic carrier — a product that releases antibiotic at the site — implanted during that operation [s1]. Participants were randomly assigned 1:1 to receive postoperative systemic antibiotics for a long duration of four weeks or more, or a short duration of seven days or less [s1].

The primary outcome was definite treatment failure by 12 months, judged against protocol-defined criteria by an end-point committee whose members did not know which group a patient was in [s1]. The trial was designed as a noninferiority study with a margin of 10 percentage points — meaning the short course would be declared acceptable if failures rose by no more than 10 points [s1].

What it found

Of 500 patients randomised, 475 were assessed in the primary analysis [s1]. Definite treatment failure occurred in 34 of 241 patients (14.1%) in the long-duration group and in 26 of 234 (11.1%) in the short-duration group [s1]. The risk difference was −3.0 percentage points, with a 95% confidence interval from −9.0 to 3.0 — comfortably within the pre-specified margin, so the short course met the bar for noninferiority [s1]. If anything, the point estimate slightly favoured the shorter regimen, though the interval spans zero and does not establish superiority [s1].

The result held up under scrutiny. Sensitivity analyses in the wider included population of 497 patients and the per-protocol population of 457 gave risk differences of −2.4 and −2.7 percentage points, both consistent with the main finding [s1].

Where the two arms clearly diverged was harm. By six weeks after surgery, symptoms potentially related to treatment had occurred in 45.2% of the long-duration group and 17.2% of the short-duration group — a risk difference of −28.0 percentage points (95% CI −36.4 to −19.6) [s1]. Fewer weeks of antibiotics meant far fewer side effects.

How to read it

The clean reading is that, in patients who received a local-antibiotic carrier at surgery, a week or less of systemic antibiotics was as effective as four weeks or more at preventing treatment failure, and caused markedly fewer side effects [s1]. That is a strong result for antibiotic stewardship: it points to a way of removing weeks of unnecessary drug exposure from a common surgical scenario without paying for it in relapses.

Two caveats set the boundary of the claim. First, the finding is conditional on the local-antibiotic carrier — every patient in the trial received one, so SOLARIO does not tell clinicians what happens if the short course is used without that local dose [s1]. Extrapolating to patients who did not get a carrier would go beyond the evidence. Second, the trial was open-label, so patients and treating clinicians knew the assigned duration [s1]. The primary outcome was adjudicated by a blinded committee against fixed criteria, which guards the main result, but open-label designs can still colour softer judgements along the way [s1].

The trial was funded by the European Bone and Joint Infection Society together with the National Institute for Health Research and the British Medical Association, and was sponsored by an NHS trust rather than a manufacturer, per its registry record [s2]. That non-commercial backing matters for a trial whose message — use less of a product — runs against the commercial grain.

What to watch

SOLARIO is registered on ClinicalTrials.gov as NCT03806166, where its protocol and definitions can be checked [s2]. The obvious next question is the one the trial deliberately did not answer: whether short-course systemic antibiotics are safe without a local-antibiotic carrier. Until that is tested, the shortening applies to the specific setting SOLARIO studied — surgery plus a local dose — and not to orthopaedic infection care in general.