Your Draft LCD Backlog Is Smaller Than the Database Says
The Medicare Coverage Database shows 93 proposed LCDs sitting past their comment close date, a median 345 days each. One column in the download export says 68 of them were released as final policy already, which puts the real queue at 25 drafts and 93 days. Twelve of the finished ones drop off the report between now and October 30.
Ninety-three proposed LCDs in the Medicare Coverage Database are sitting past their comment close date. The median one closed comments 345 days ago, and 29 of them closed more than a year ago.
Every one of those numbers is correct.
This desk counted the same list two weeks ago and read it as a finalization backlog. One column in the download export says that read was wrong.
"How many of these drafts are actually still waiting on a decision?"
Twenty-five.
What One Column Does to the Count
lcd.csv inside current_lcd.zip carries 48 columns. The data dictionary CMS ships in the same zip defines one of them in a single line: draft_released_date is "the date on which the Proposed LCD was released to a Final LCD."
68 of the 93 have one.
They still carry status P and they still appear on the proposed-LCD report, because a draft does not leave the database the day its final goes live. The policy is settled and the draft record outlives it by a year.
Pull those 68 out and the queue changes shape (chart below). Median wait since comment close drops from 345 days to 93, and the count of drafts more than a year past comment close drops from 29 to exactly one.

That one is DL38571, Colon Capsule Endoscopy, National Government Services, the draft this desk flagged on October 5 as the oldest thing on the report.
Posted September 30, 2021, comments closed that November, no release date and no retirement date against it. 1,845 days and counting.
The other 24 have a median of 93 days since their comment windows shut, which is an ordinary turnaround rather than a queue.
Against the Manual's Own Clock
Program Integrity Manual chapter 13 sets the deadline, and sets it twice. Section 13.5.4 and section 13.5.6 both say that "MACs shall finalize or retire all proposed LCDs within a rolling year of publication date of the proposed LCD on the MCD (365 days)."
Which makes those 68 released drafts a scorecard. Posting date in one column, release-to-final in another, subtract.
Median 193.5 days. The fastest two cleared in 48 days, a Molecular Pathology Procedures draft at National Government Services and a brain MRI quantification draft at CGS.
Four of the 68 went past 365 (chart below). All four are titled Botulinum Toxin Injections, all four posted the same day in May 2024, and they released at 511 days (CGS), 512 twice (Noridian), and 567 (Wisconsin Physicians Service).

The manual leaves a door open for that. A MAC facing "an unusual circumstance" can ask for an exception by notifying its COR and the LCD business function leads at least 21 business days before the one-year expiration date, and nothing public says whether these four walked through it.
Here is the catch.
Every one of those four display IDs is back in the current drafts list, reposted in April 2026 with fresh comment windows that closed in May and June. Two more joined them, at Palmetto and National Government Services.
Six open Botulinum drafts, none released, the whole family back on a clock it has already overrun once.
The Twelve That Drop Off This Month
The year a released draft spends in the database is exact. Subtract draft_released_date from the scheduled retirement date across all 68 and you get 365 days in 66 of them and 364 in the other two.
Twelve of those fuses burn out between October 13 and October 30:
- Three Transurethral Waterjet Ablation of the Prostate drafts, gone on the 13th and 14th
- Five Superficial Radiation Therapy drafts across WPS, Noridian, Palmetto, National Government Services and CGS, on the 23rd and 29th and 30th
- Three Botulinum Toxin Injections drafts on the 23rd and 24th
- The joint DME Urological Supplies draft on the 28th
Nothing is being withdrawn. Those policies went final months ago and the drafts are being swept.
If you monitor the proposed-LCD report for what is coming, those twelve disappearances are the kind of thing that generates a false alarm, and the same report has 19 rows with comment periods genuinely still open, closing November 7 and November 14.
Final Thoughts
The field does none of your reading for you. It tells you a final LCD exists, and it says nothing about what that final said, which codes moved, or whether the MAC adopted the draft language or walked it back.
What it does is cut the list you have to actually read by two thirds, and point at the one draft in the whole database that has been open since 2021.
A database that keeps finished work in the pending pile is not unusual. The useful part is that CMS ships the column that separates them, inside the same zip file, for free.
Thanks for reading.