←back to thread

239 points giuliomagnifico | 2 comments | | HN request time: 0.476s | source
Show context
barbegal ◴[] No.36213140[source]
The headline is misleading. The actual study proved that the recorded date of admission to hospital in Ireland with ST-segment elevation myocardial infarction was increased on a Sunday and Monday. Increased admissions on a Monday is not that unusual given that people often seek medical attention after the weekend but maybe more surprising is the increase on a Sunday. https://heart.bmj.com/content/109/Suppl_3/A78
replies(8): >>36213241 #>>36213290 #>>36213562 #>>36213909 #>>36214664 #>>36216341 #>>36217855 #>>36219532 #
haldujai ◴[] No.36213241[source]
From the methods section of the abstract: “We excluded post-fibrinolysis patients, patients with old stents, and those who presented more than 24 hours after the onset of pain.”[edit: I misread the PDF version which included multiple abstracts, the methods I’m referring to was from a separate study with the title cutoff, this specific abstract didn’t specify. But from below and table 1 in: https://jamanetwork.com/journals/jama/article-abstract/20140... which looked at 68,000 STEMIs, 3.1% presented > 12 hours and 8.4% had an unknown time of symptom onset. Wouldn’t explain the magnitude of effect seen in this study. Circadian effects on STEMI and increased incidence on Monday are not new observations.]

Don’t think late presentation STEMIs are that common to begin with for your argument to have logical sense, this is the worst form of a “heart attack”.

From this single center study presentations > 12 hours only comprised 10%.

https://www.ingentaconnect.com/content/wk/jcarm/2017/0000001...

replies(2): >>36214640 #>>36215022 #
1. barbegal ◴[] No.36214640[source]
Thanks for links to those extra studies. The 3.1% and 10% presented > 12 hours are averaged across all days of the week not just the Sunday to Monday gap which is likely to be greater given the reduction in public services in Ireland on a Sunday. Do you know of previous studies which report Monday as being particularly risky? I can only find references to time of day (circadian cycle) which obviously make sense given how many bodily processes are linked to a circadian cycle but I'm skeptical about a weekly cycle (which I find confusing to be referred to as circadian also)
replies(1): >>36216507 #
2. haldujai ◴[] No.36216507[source]
They include an off-hours subgroup which comprises of weekday after hours as well as weekends and represents 2/3rds of cases. They don't provide a further breakdown but in this after-hours group delayed presentation was even lower (2.9%) and they report statistical significance (although it looks like a multivariate P value at a quick glance). Unknown (10%) is hard to interpret what that actually means.

Just knowing the pathology of STEMI it's hard to buy that an effect of this size (in the Ireland study) is largely due patient's not seeking care on weekends unless you're somewhere extremely rural as this isn't your average heart attack.

If this was about ACS (acute coronary syndrome) in general I'd be more suspicious that patient delays are a relevant confounder, but we have other literature to support the trend (granted with some conflicting studies).

from [1]: > Many studies have shown an excess of cardiovascular events on Mondays (1,3,10,16,18,19). A relative trough has been seen on Saturdays and Sundays I compared with the expected number of cases. A similar pattern was seen in most subgroups irrespective of age, gender, cardiac medication, and in-fart characteristics (first or recurrent, Q or non-Q, site). The frequency of morning infarction is greater during the working week than on weekends, suggesting a superimposition of work-related stress on endogenous circadian rhythms.

>Circadian variation is found on all days of the week including weekends' when the morning peak is less obvious.

I haven't looked at the methodology of the cited studies but they include 6 references for your perusal.

[0]https://onlinelibrary.wiley.com/doi/abs/10.1002/clc.22 [1]https://onlinelibrary.wiley.com/doi/epdf/10.1002/clc.4960261... [2]https://europepmc.org/article/med/12061302