Missing Data: How to Best Account for What Is Not Known.

Missingdata are common in clinical research, particularly for variables requiring complex, time-sensitive, resource-intensive, or longitudinal data collectionmethods. However, even seemingly readily available information canbemissing. There aremany reasons for “missingness,” includingmissed study visits, patients lost to follow-up, missing information in sourcedocuments, lackof availability (eg, laboratory tests that were not performed), and clinical scenariospreventingcollectionofcertainvariables (eg,missingcoma scaledata in sedatedpatients). It is particularly challenging to interpretstudieswhenprimaryoutcomedataaremissing.However,many methods commonly used for handling missing values during data analysis can yield biased results, decrease study power, or lead to underestimates of uncertainty, all reducing the chance of drawing valid conclusions. In this issue of JAMA, Bakris et al evaluated the effect of finerenoneonurinaryalbumin-creatinine ratio (UACR) inpatientswithdiabetic nephropathy in a randomized, phase 2B, dose-finding clinical trial conducted in 148 sites in 23 countries.1 Because of the logistical complexity of the study, it is not surprising that some of the intended data collection could not be completed, resulting in missing outcomedata. Bakris et al used several analysis and imputation techniques (ie,methods for replacingmissingdatawith specific values) to assess theeffects ofdifferent approaches forhandlingmissing data. Thesemethods included complete case analysis (restricting theanalysis to includeonlypatientswithobserved90-dayUACR values); last observation carried forward (LOCF; typically this involvesusing the last recordeddatapointas the finaloutcome;Bakris et al used the higher of 2 UACR values and, separately, themost recent UACR obtained prior to study discontinuation); baseline observationcarried forward (using thebaselineUACRvalueas theoutcomeUACR value, therefore assuming no treatment effect for that patient);mean value imputation (replacingmissing valueswith the meanofobservedUACRvalues); and randomimputation (using randomly selectedUACRvalues to replacemissingUACRvalues).1Multiple imputation2 tohandlemissingvalueswasalsoperformed.With the exception of multiple imputation, each of the imputation approaches replaces a missing value with a single number (termed “single”or “simple” imputation)andcanthreatenthevalidityof study results.3,4 The authors concluded that finerenone improved the UACR,a result thatwasconsistent regardlessof themethod forhandling missing data.