I-821D Medians Drop 18 Days, But a 13,000-Case Backlog Rise Clouds the Picture

The I-821D final-stage median fell 18 days this week, from 148 to 130 days, the largest single-week improvement in the 12-week window. That number is real, but it sits inside a trend where the 8-week regression's R² is 0.09, meaning the slope explains almost none of the week-to-week movement. Over the same 12 weeks, medians have ranged from 21 days to 169 days, a 148-day spread. Any single reading, including this one, carries very little predictive weight.

The 12-week pattern rewards closer inspection. Weeks 23 and 24 posted medians of 33 and 21 days, respectively, but those readings coincided with sample counts of 1,481 and 5,019 cases, far above the prior weeks' 213 to 498. That volume pattern is consistent with a burst of older cases clearing in a short window rather than a structural acceleration. This week's sample of 7,585 approved cases is the largest in the 12-week window, which gives the 130-day reading more weight than most prior weeks, though the overall volatility remains.

Looking at the stage-level data from the last 30 days (Stage durations, raw percentiles), 51% of all pending I-821D cases sit at "Case Is Still Being Processed By USCIS," with a median raw wait of 164 days at that stage across 81,462 cases and an interquartile range (IQR, the spread between the 25th and 75th percentiles) of 76 to 195 days. Another 25% are at biometrics (the fingerprint and photo appointment USCIS uses before a case can reach the adjudication (the case decision) step), where the median is 20 days and the IQR is 18 to 23 days across 40,402 cases. The "still being processed" label is the normal adjudicator queue; it does not mean a case is stalled.

A smaller but notable share, 13.9% of cases, sits at "Case Was Received and A Receipt Notice Was Sent," with a median raw wait of 288 days and an IQR of 113 to 309 days across 22,291 cases (Stage durations, 30-day window). This group likely includes cases that received an initial receipt notice but have not yet had biometrics scheduled, a possible intake bottleneck. The data does not tell us whether this reflects routing delays or a separate renewal-cycle wave.

The rfe_fast anomaly flag fired this week: cases with an RFE (Request for Evidence, USCIS's standard follow-up letter asking for missing documents) posted a median final-stage wait of just 11 days, against 97 days for cases that went through plain processing without an RFE. That 8.77x inversion is counterintuitive. Two candidate explanations exist, though the data brief does not distinguish between them. First, a batch of older RFE-response cases whose responses arrived many months earlier may be clearing together this week, pulling the RFE median down. Second, adjudicators may be routing resolved RFE responses ahead of the general queue this month. Either way, the full filing-to-approval arc for RFE cases remains longer; this week's reading reflects timing, not a policy shift.

Receiving an RFE is not a denial signal. Most I-821D RFEs involve residence continuity or education documentation gaps, and the majority of cases that respond are ultimately approved. The faster clearing this week appears to follow a batch-timing effect rather than any change in adjudication standards. Applicants who have responded to an RFE should not read this week's 11-day figure as a guarantee of quick resolution.