Skip to content

Skou ST, et al. A randomized, controlled trial of total knee replacement. N Engl J Med. 2015;373:1597–1606. PMID 26488691

One-paragraph summary

One hundred people with moderate-to-severe knee OA eligible for unilateral replacement were randomized to total knee replacement plus a 12-week nonsurgical program or the nonsurgical program alone; 95 completed 12 months. KOOS4 improved 32.5 versus 16.0 points, adjusted mean difference 15.8 (95% CI 10.0–21.5). Serious adverse events numbered 24 versus 6 (P=.005).

Key findings

  • Thirteen of 50 participants assigned nonsurgical care crossed to replacement within 12 months; one assigned surgery received nonsurgical care only.
  • The nonsurgical program included exercise, education, dietary advice, insoles and pain medication—not usual-care neglect.
  • Most people allocated to nonsurgical care did not undergo replacement within 12 months.
  • Benefit and harm moved in opposite directions, making preference and baseline severity central.

Limitations

  • Small pragmatic trial; it estimates a strategy in surgery-eligible patients, not all knee OA.
  • Crossover complicates per-protocol interpretation but was handled by intention-to-treat for the primary comparison.
  • Twelve months does not quantify lifetime implant revision risk.

Why it matters

This was rare randomized comparative-effectiveness evidence for arthroplasty and established that an effective operation can add both substantial benefit and substantial adverse-event burden.

Cited by wiki pages

  • overview
  • knee OA
  • surgery and joint replacement