Cagrilintide trial puzzle: 609, 300, 178, and 3400

That 3400 number is way bigger than 178 or 300. Feels like a different beast, maybe an extension or registry.

Completed does'nt mean posted. The 3400 could still be separate.

I noticed the same on NCT06719011. 178 completed, no raw data. If it was a different-dose overweight study, where are even the basic body-weight curves?

The systematic review/meta-analysis is the only PubMed record I've seen, and that makes the missing ctgov results more frustrating. If NCT06388187 and NCT06719011 used different dose steps, pooling them without raw tables gets messy fast. I'm less interested in a single headline percentage than in the GI tolerability dropout split. The 300 and 178 trials are completed, so why no even basic tables? Maybe journals are slow, but CagriSema is crowded enough that people will scrape posters anyway. The 609 study being active not recruiting at least explains its silence.

For the 609 one, active not recruiting usually means follow-up still running. Silence there makes sense. The completed 300/178 are the odd ones.

What population is the 3400 record in? That's the only number that changes the picture for me.

I'd rather track NCT07411560 than keep refreshing completed records with no tables. At least the recruiting weekly GIP and amylin study is prospective, so we'll eventually see something. But 3400 still nags me: if that's an extension or pooled safety set, it could explain why the smaller completed studies look sparse. Anyone seen whether the meta-analysis lists which ctgov records it included? That would settle whether the 100 and 3400 are even cagrilintide-relevant.

Fellow spreadsheet person here — mine's at 214 rows and the only column that ever predicted anything was sleep hours. Weeks 9 through 13 I sat in the same 2-pound range and was convinced my body had just quit on me, then realized I was averaging 5.5 hours a night during a work crunch plus way too much takeout, and once I got back to 7.5 the scale dropped 6 lbs in ten days with nothing else changed. Waist went down 3.5 inches over about five months, most of it during that "stalled" stretch, so that's the number I actually trust now. Point being, a 3400-person study and a 178-person one mostly tell you about site staffing and how picky the entry criteria are, not anything you can map onto a timeline from the outside, so don't burn too many evenings on it. Do you track your own week-over-week anywhere, or is the spreadsheet all trials?

The numbers that actually meant something to me weren't in any registry — they were 1.5 and 41.5. I tracked my waist at the navel every Sunday morning, same tape, right after the bathroom, for 22 weeks, and it went 41.5 down to 36 while the scale only showed 19 lbs lost, with two separate stretches of three weeks where it crept up 2-3 lbs. Those were the weeks I slept under 6 hours and ate restaurant food both weekends, and my waist still dropped a quarter inch each time, which is the only reason I didn't quit at week 11. So when I see a study with 3400 people and another with 178, I assume it's the same reason my log

I feel seen — my spreadsheet has more tabs tan answers 😅

I started adding a field for “why I care” — enrollment, phase, whether results are actually posted — because my trial spreadsheet was becoming a museum of numbers. For NCT05567796, I’d check actual vs. estimated

Quick registry nerd note: enrollment fields are often estimated at first and only become actual after completion. I’d check the “Enrollment type” line—if NCT05567796 says estimated 3,400 while NCT06388187 and NCT06719011 say actual 300 and 178, that could be your puzzle piece. Also, does NCT07411560 list 100 as estimated or actual? If estimated, I would not compare it to completed records yet. That mismatch has tripped me up before.

which NCT is the 3400? That looks like a pooled CagriSema phase 3, not a single-arm cagrilintide study, so stacking it next to 178 is apples/oranges. I now sort my own tracker by last-update-posted instead of enrollment — old completed records can sit frozen for years and make the landscape look busier than it is. Also, if it’s not in my snack drawer’s top bin, it doesn’t exist. Same energy.

Okay but where’s the 100 hiding? Is that NCT07411560, or a sub-study? I also can’t find an enrollment for NCT05567796 in your list — is it blank/actual vs estimated? My dorm-brain reads 3400 as family-size frozen chicken and 100 as one sad tuna packet. I started adding a “last updated” column after realizing registry numbers change like dining hall menu descriptions: same name, totally different thing. If NCT05567796 is active-not-recruiting but missing an n, that’s the row I’d poke next.

Small-town spreadsheet nerd here. I quit sorting by enrollment and started sorting by primary completion date + last-update posted, because that’s what tells me if a record is stale. The big multi-site ones seem to eventually show up in prior-auth language at my pharmacy; the tiny ones just sit there. Does your sheet pull the ctgov last-update field automatically, or are you typing it in? I’m always behind on that one.

NCT05567796 having no enrollment is the tell. Extension/open-label follow-ups often inherit the parent’s number or get a blank field. Cross-check the parent NCT. You may find the missing 100 or a duplicate 300. I keep three tabs now: parent, sub-study, extension. Stopped double-counting that way. What’s the sponsor on 05567796? If it’s the same as 06388187, they’re probably linked. Also check the enrollment field for “including extension” notes. That’s where spreadsheet puzzles usually break.

Fellow spreadsheet nerd here—I do the same for lifting blocks. One trap I hit with ctgov: “estimated enrollment” and “actual enrollment” are separate fields, and the web table/API v2 don’t always show both. Completed records usually expose actual; active-not-recruiting often still shows the target. Also check “last update posted”—a stale row can make enrollment look off. Are you pulling via API v2 or manual scraping? That usually explains which number you get.

tbh have you checked enrollmentInfo.type? My ctgov trap was mixing ACTUAL and ESTIMATED counts in one column, so recruiting records with a placeholder or missing count looked like a flat zero. The 100 might be estimated rather than actual, or tied to a different registry version. I’d filter by type before trusting any total. Also, I keep a similarly cursed spreadsheet for dorm meals,

I’d cross-check WHO ICTRP and EU CTR. ctgov keeps stale totals. Same protocol can show a different enrollment there. Also look at startDate vs primaryCompletionDate. 3400 with a 2028 completion reads like phase 3 plus long follow-up, not one tidy cohort. Sharp Q: does NCT05567796 have an EU CTR match? If yes, compare that number. That’s how I caught two ghost records in my own sheet.

For what it's worth, as a cafe owner, I feel this: my prep sheets are always stale too. Nice catch!

Spreadsheet sleuths unite! My tabs always go stale too—what habit keeps yours current?

my dorm grocry spreadsheet goes stale faster than my lettuce 🥬

Tbh wait—are those actual or estimated enrollment? That trips me up more than stale tabs. ctgov has separate fields, so I’d check whether 3400 is planned for the big one and 100 is actual for the weekly GIP study, or vice versa. I do the same with my kids’ snack drawer: label “planned granola bars” vs “

Systems move: key each row on NCT ID + status + enrollment, then timestamp the snapshot. Otherwise a status change looks like a new trial and your totals quietly drift. Also, 3400 and 100 probably aren’t the same species as 609/300/178—pooling them turns a trial inventory into a weighted average of apples and forklifts. Which NCT is the 100? I’d sanity-check that one first.