This Australian RCT randomized 76 adults with chronic non-specific low back pain to either an individualized 8-week swimming program plus education (delivered via four telehealth physiotherapy sessions) or education alone (up to two sessions). At 8 weeks, the swimming group had significantly less disability on the Roland Morris Disability Questionnaire than the education-only group (mean difference -2.5 points, 95% CI -4.5 to -0.5), but this advantage shrank over time and was no longer statistically significant by 52 weeks; pain intensity showed only a small, possibly non-meaningful benefit, and the swimming group had more adverse events (56% vs 30%), almost all non-serious.
AI summary · from the full text
Why it mattersSwimming is widely recommended by clinicians for chronic low back pain despite almost no trial evidence, and this is the first RCT to test it directly as the primary intervention, giving learners a concrete example of how weak evidence can underlie common practice recommendations.
Effect exceeds the MCID on Roland Morris Disability Questionnaire (RMDQ) and patient-important outcome.
Presenting this at rounds? Start here
?The swimming group received four telehealth sessions versus up to two in the education-only arm, so the intervention was not attention-matched; how much of the short-term benefit might reflect extra clinician contact and health-coaching rather than swimming itself?
?Given that the benefit was present mainly at 8 weeks and largely dissipated by 26-52 weeks alongside declining swimming adherence, how would you counsel a patient about whether to start a swimming program and how to sustain it long term?
This retrospective study from three Japanese spine centers compared complication rates between 407 patients who had corpectomy-containing anterior cervical surgery (ACCF or hybrid ACDF/ACCF) and 617 patients who had non-corpectomy anterior surgery (mostly ACDF) for degenerative cervical myelopathy. Corpectomy patients had far more perioperative local complications (38.1% vs 16.5%), dural injury (17.4% vs 1.5%), reoperation (9.8% vs 2.4%), and upper-extremity weakness (13.0% vs 4.2%), but after adjusting for how many levels were operated and whether the upper cervical spine was involved, the extra risk from corpectomy itself largely disappeared (adjusted OR fell from 1.82 to 1.12, no longer significant). The authors conclude that much of the added morbidity reflects the underlying OPLL pathology and greater surgical extent rather than corpectomy as a technique per se.
AI summary · from the full text
Why it mattersHelps residents and spine surgeons frame corpectomy's higher complication burden as a marker of surgical complexity and OPLL severity rather than an independent reason to avoid the procedure, which matters for informed consent and case selection.
Patient-important outcome and fragility index 96, offset by no power calculation.
Presenting this at rounds? Start here
?Because the choice between corpectomy and non-corpectomy surgery was driven by pathology severity and surgeon judgment rather than randomization, how confident can we be that adjusting for operated levels and upper cervical involvement fully removes confounding by indication, especially since important variables like canal-occupying ratio, K-line status, and dural ossification were not modeled?
?When counseling a patient with massive, dura-adherent OPLL who may need corpectomy for adequate decompression, how should this data on higher dural injury, reoperation, and airway complication rates be weighed against the potential neurological benefit of more complete anterior decompression?
Journal of the American Academy of Orthopaedic Surgeons · IF 2.9 · Jul 21, 2026 · LoE II
This study combined an umbrella review of 16 prior systematic reviews/meta-analyses with a newly reconstructed meta-analysis of 38 primary studies to ask whether NSAIDs impair bone healing, and if so, in whom. Pooling roughly 82,000 NSAID-exposed and 756,000 unexposed patients, NSAID use was associated with higher odds of nonunion (OR 1.56, 95% CI 1.18-2.11), with the risk concentrated in adults, traumatic (not elective) fractures, long bones (not spine), and higher-dose regimens, while short courses of 14 days or less and pediatric patients showed no significant increase in risk.
AI summary · from the full text
Why it mattersThis gives clerkship students and residents a risk-stratified framework for the common question of whether to give NSAIDs after a fracture, suggesting short low-dose courses are likely safe while caution is warranted with high-dose or prolonged use in adult long-bone fractures.
Patient-important outcome and fragility index 12304, offset by no power calculation and high heterogeneity.
Presenting this at rounds? Start here
?The umbrella review found most source reviews were rated low or critically low quality by AMSTAR-2, and the reconstructed meta-analysis pooled RCTs, cohorts, and case-control studies with I2 as high as 83-93% for several key comparisons; how much should we trust pooled odds ratios built on such heterogeneous, largely observational data with likely confounding by indication (sicker or more severely injured patients may both need more analgesia and heal worse)?
?Given the nonsignificant finding for delayed union, no signal in pediatric patients, and no significant harm with short-term (<=14 day) use even in adults, how would you counsel a patient after an isolated long-bone fracture regarding a brief postoperative NSAID course for pain control?
This retrospective registry-based cohort study followed 4,412 adults who had multilevel posterior cervical spine surgery (2009-2023) at a large US integrated health system, comparing laminoplasty (PCL, 29.3%) to laminectomy with fusion (PLF, 70.7%). After adjusting for patient and surgical factors, PCL was associated with a lower risk of reoperation for adjacent segment disease (hazard ratio 0.42) and lower all-cause reoperation (hazard ratio 0.51) over a median 4.7 years of follow-up, with no difference in mortality, 90-day ED visits, or readmissions.
AI summary · from the full text
Why it mattersThis large, long-term comparison adds weight to the argument that laminoplasty should be considered more often than laminectomy/fusion for multilevel cervical spondylotic myelopathy, a debate residents will encounter on spine rotations and boards.
Patient-important outcome and 56.4-month follow-up, offset by no power calculation and no correction for multiple comparisons.
Presenting this at rounds? Start here
?Given the significant baseline imbalances in myelopathy diagnosis, spondylolisthesis, number of levels, and operative time between groups, how confident can we be that multivariable adjustment fully accounts for confounding by indication in choosing PCL versus PLF?
?If you were counseling a patient with multilevel cervical myelopathy and preserved sagittal alignment, would this study's reoperation data alone be enough to favor laminoplasty over laminectomy/fusion, or would you want patient-reported outcome and alignment data before changing your usual approach?
This multicenter study followed 154 skeletally mature adolescents with bilateral pars defects or low-grade spondylolisthesis for a median of 10.5 years to see how the condition behaves after growth stops. Vertebral slip itself rarely worsened (7.1%), but MRI-detected disc degeneration progressed in 30.5%, low back pain worsened in 44.2%, and new neurologic symptoms appeared in 4.5%, so that 56.5% of patients showed progression in at least one domain; L5-level disease, paraspinal sarcopenia, and sedentary behavior independently predicted this overall progression.
AI summary · from the full text
Why it mattersIt challenges the long-standing teaching that spondylolysis and low-grade spondylolisthesis become mechanically 'burned out' after skeletal maturity, suggesting these patients still need longitudinal monitoring for disc degeneration and pain rather than reassurance based on stable slip alone.
Contradicts current practice and 126-month follow-up, offset by no power calculation and no comparator.
Presenting this at rounds? Start here
?The study had no healthy control group and relied on retrospective extraction of symptom data from questionnaires and records; how might recall bias and the absence of a comparator affect confidence in the 44.2% LBP progression rate and the identified risk factors?
?If vertebral slip itself rarely progresses but disc degeneration and pain commonly do, should adolescent spondylolysis follow-up protocols shift emphasis away from serial radiographs toward symptom tracking and MRI surveillance, especially in patients with L5-level disease, sarcopenia, or sedentary lifestyles?
JAMA Network Open · IF 10.5 · Jul 24, 2026 · LoE II
This retrospective database study used TriNetX electronic health records to compare fragility fracture risk in about 133,600 propensity-matched adults with type 2 diabetes who started a GLP-1 receptor agonist versus a DPP-4 inhibitor. Over 3 years, fragility fractures occurred in 2,030 GLP-1 RA users versus 2,484 DPP-4i users, corresponding to a 21% lower relative risk (HR 0.79) that appeared independent of weight loss or glucose control, though the effect weakened by year 3 and was not seen in patients without diabetes.
AI summary · from the full text
Why it mattersOrthopaedic surgeons are increasingly seeing patients on GLP-1 RAs and need to know whether these drugs raise or lower fragility fracture risk, especially since weight loss itself is a known risk factor.
Contradicts current practice and patient-important outcome, offset by only 36 months of follow-up and more patients lost than the result can absorb.
Presenting this at rounds? Start here
?Given this is a retrospective, non-randomized comparison using administrative EHR data with unmeasured confounders like physical activity, fall risk, and baseline bone mineral density, how much can propensity-score matching really substitute for randomization here?
?If you were counseling a patient with T2D and osteopenia who is starting a GLP-1 RA, would this study change how you approach bone health monitoring, and why or why not?
This retrospective single-center study followed 243 patients aged 75 or older undergoing lumbar fusion, comparing 57 with preoperative cognitive impairment (CI, by education-adjusted MMSE) to 186 without. Major postoperative complications were more common in the CI group (42.1% vs 33.9%), and CI amplified the harmful effects of frailty, diabetes, and intraoperative transfusion volume on complication risk (significant interaction terms), while also independently predicting worse functional recovery (lower MCID achievement on ODI and VAS scores).
AI summary · from the full text
Why it mattersThis suggests that preoperative cognitive screening could help identify elderly spine surgery patients in whom frailty, diabetes, and blood management deserve extra perioperative attention, though the retrospective, single-center, exploratory design limits how much should change in practice today.
Effect exceeds the MCID on ODI, VAS back, VAS leg and adequately powered, offset by only 6 months of follow-up and no correction for multiple comparisons.
Presenting this at rounds? Start here
?Given only 57 patients in the CI subgroup and a stated limitation of reduced model stability, how confident can we be in the three specific interaction effects (diabetes, frailty, transfusion) versus finding them by chance among many variables tested without correction for multiplicity?
?If a patient screens positive for cognitive impairment preoperatively, how would you translate this study's findings into a concrete perioperative plan around glycemic control, frailty prehabilitation, and blood management, and what would you tell the patient about expected functional recovery?
