Table 1

Summary of proposed approaches for reporting MPD

AuthorMethods used to prevent missing dataNumber of participants with MPDDifferences in rates of MPD between trial armsReasons for MPDFlow of participantsDifferences between participants with and without MPDPattern of missingness (eg, whether at random)Methods for handling MPD in analysisResults of any sensitivity analysesImplication of MPD on interpreting the results
Staquet et al13✓*
Bernhard et al12
Troxel et al20✓†
Liu et al21
Amico et al22✓*
Sterne et al23
Polit and Gillespie24
Al-Shurafa et al25
Gewandter et al26
CONSORT‡3✓*
Number of studies recommending it11037423422
  • *These approaches further recommended reporting missing data by study arm.

  • †Troxel et al recommended examining patient-related and institution-related factors affecting missing data rates descriptively or by using logistic regression models. ‡Elements recommended by CONSORT extensions, not already included in the main statement: (1) extension for harms, none; (2) extension for cluster trials: reporting for each group, the missing participant data for both clusters and individual cluster members; (3) extension for patient reported outcomes, reporting outcome data at baseline and at subsequent time points, interpreting any supportive (eg, sensitivity) analyses.

  • MPD, missing participant data.