Acronyms
RCT
Randomized Controlled Trial
Patients from which clinical trial were included in this study?
ARCTIC Trial
Complete case analysis
What proportion of real-world (i.e., non-trial) patients did not have complete follow-up data for the main outcome?
61.0%
Acronym: RA
Rheumatoid arthritis
CC
Complete case
Trial and real-world patients in this study came from which country?
Norway
Which of the 5 missing data methods used censoring (set as drop-out) to account for intermittent missing visits?
(1) String censoring + IPCW
(2) MI + censoring + IPCW
Which of the 5 missing data methods gave the most conservative estimate for achievement of remission at 6, 12, and 24 months?
MI for all missing data
MICE
Multiple imputation by chained equations
What was the follow-up period in the ARCTIC trial?
2010-2013
In the strict censoring + IPCW approach, the IPCW methods assigns weights to individuals with complete follow-up data corresponding to the inverse of what?
Inverse of their estimated probability of having complete data.
IPCW
Inverse Probability of Censoring Weighting
What type of study design was used to create the real-world patient cohort (i.e., non-trial patients)? Be specific!
Prospective observational study
When do intermittent missing visits arise in the data?
When patients do not meet for one or more visits, but reenter the study at a later time point.
Name 2 advantages of multiple imputation described in this article.
1. MI can provide unbiased estimates under the missing at random assumption.
2. MI efficiently uses information from individuals with partially missing data.
3. All available and relevant data can be included in the imputation model, including both variables related to the outcome analyses and variables associated with missingness.
DMARD
Disease-modifying Antirheumatic
Drug (therapy)
How many total patients were included between the trial and real-world studies?
230 (trial) + 429 (real-world)=659 patients!
State the three assumptions of the MI + IPCW model. (Hint: Look at table 1)
1. Both the MI model and the IPCW model are correctly specified.
2. MI models missing outcome variables at visits and intermittent missing visits given information in the dataset.
3. IPCW models the missing mechanism given the observed past for missing data due to naturally occurring drop-out