Authors: De Cassai A. et al.
Anesthesiology, September 25, 2025. DOI: 10.1097/ALN.0000000000005776
This meta-research study examined the persistence of the so-called “Table 1 fallacy”—the practice of performing statistical significance testing on baseline characteristics in randomized controlled trials (RCTs). Because randomization itself ensures that baseline characteristics are balanced on average, such tests are unnecessary and may foster misinterpretation of trial validity. Despite long-standing CONSORT and methodological guidance advising against this practice, its frequency in anesthesiology and pain medicine literature has not been well characterized.
Researchers reviewed 2,453 parallel-group RCTs published between 1996 and 2025 across 101 journals indexed in Scopus under the anesthesiology and pain medicine category. Data were extracted on whether baseline testing was performed, how many variables were tested, and whether significant differences were reported or discussed. Statistical models assessed factors associated with baseline testing and the expected false-positive rate.
Nearly half of all RCTs (48.3%) still performed statistical testing of baseline characteristics, and 19.2% of those reported at least one significant difference. Only a quarter of these studies (25.4%) acknowledged such differences as potential limitations. Out of more than 11,500 baseline variables analyzed, 3.7% were reported significant—less than the 5% expected by chance—suggesting that most “differences” arose randomly. Larger author teams were slightly less likely to conduct baseline testing (odds ratio 0.95 per additional author), while the number of tested variables was a strong predictor of reporting at least one spurious difference.
The findings confirm that baseline significance testing remains pervasive in anesthesiology RCTs despite clear methodological consensus against it. This ongoing problem likely stems from misunderstanding the purpose of randomization and an overemphasis on statistical testing in reporting standards. The authors call for improved researcher education and stronger journal editorial policies to eliminate this outdated practice and enhance trial transparency and methodological rigor.
What You Should Know
-
Nearly half of anesthesiology and pain medicine RCTs still perform statistical testing of baseline characteristics.
-
Randomization renders such testing unnecessary and potentially misleading.
-
Larger author teams were less likely to conduct baseline testing, while testing more variables increased false positives.
-
Education and editorial oversight are needed to align RCT reporting with CONSORT guidelines.
Thank you to Anesthesiology for publishing this important meta-research analysis addressing persistent statistical misconceptions in anesthesiology and pain trial reporting.