[pain] further notes on chronic pain statistics

[kaberett]

Summary: nobody seems to have done the data analysis I actually want, because data collection is hard and then actually making it internationally comparable ditto, but the proportion of chronic pain cases that are primarily attributable to back pain Of Some Kind seems to be very roughly in the region of 20%-50%, depending.

Andersson, G. B., 'Epidemiological features of chronic low-back pain', The Lancet, 354(9178), 1999, pp.581--5 [The Painful Truth]: extremely focussed on lost days of work, and as chronic back pain as a proportion of all back pain cases; doesn't consider how back pain compares with other locations/causes of chronic pain.

National Health Interview Survey (US) [The Way Out]: I am struggling to wrap my head around getting useful-to-me data out of the public website but -- AHA NO I HAVE FOUND THE SUMMARY STATISTIC PDFs, GO ME. ohKAY. Let's see. Gordon's using age-adjusted rather than crude numbers, which is fine, but then they stopped providing the relevant summaries in 2019, lolsob. But at least I've ruled that out, even if only as a source of NAm data.

Some further notes: * NHIS-LC does not collect data on underlying causes of chronic pain (NHIS itself might, but it's not publically available AFAICT) * GP survey outcomes from Australia circa 2013: prevalence of chronic pain 19.4%; most commonly reported causal conditions were osteoarthritis (48.1%) and back problems (29.4%); tragically the remaining breakdown appears to consist in its entirety of "other arthritis", "cancer", and "other conditions". * telephone survey in Canada in 2011: prevalence 18.9%, see table 6 for breakdown by body area, with lower back as the primary site of pain for ~22% of responders, and upper back for a further ~10% of responders. * Brevik et al. 2010 used a similar telephone survey in Europe: chronic pain in ~19% of responders, back-unspecified for ~24% of respondents + back lower 18% + back upper 5%, then knees then headache (similar to Australia)... with conditions to which the pain is ascribed in the adjacent figure * Rometsch 2025 on prevalence in Europe frustratingly vague: it sort of talks about various locations but doesn't provide anything remotely resembling a useful list (though entertainingly back pain does not appear to be at the top of it, if it were to exist...) * more NIHS, US, 2022: ~20% prevalence of chronic pain, with knee/hip/foot pain (44%) followed by back (41%) as most common location.

Note that NIHS SHS gives low back pain prevalence as 28% (for the USA, in 2017). Notwithstanding the high uncertainties, chronic pain is usually pegged at around 20% of the population; Andersson up above cheerfully asserts that something like 90% of back pain cases resolve without significant intervention. Ergo this combination of data is not helpful for working out what proportion of chronic pain is associated with low back pain, and I also don't think the GBD stats ranking low back pain as the top cause of global YLDs (p. 1884) is particularly useful either, not least because of the weighting????.

Comments

[lilysea]

I wonder how they count people who at different times have a rotation of

- migraine
- neck/shoulder pain
- lower back pain
- hip pain
- abdominal pain
- stomach pain
- foot pain
- hand/wrist pain
- skin pain

so that in any given week something hurts, but not necessarily the same thing that hurt last week

[kaberett]

In a variety of ways! Some studies include a category for "whole-body pain", some want you to list everything in the last 3/6/12 months, some of those want you to designate the "primary" location only. There are lots of approaches to handling this issue, especially as widespread pain is a hallmark of chronic pain -- but this is part of what makes comparison between datasets hard. :)

[lilysea]

Man, if I listed everything that had been painful over the past 3 months, it would be a very long list,
and include

ears
eyes
shins
knees

as well as the ones I mentioned above...

Sometimes totally random body parts hurt intensely out of the blue for 30 minutes, an hour, a day and it's like

"Why? Who knows why!"

or "I banged my shin badly against a metal trailer in 2011 and ever since then it aches randomly when the weather changes"

[kaberett]

They tend to have structured lists of options, and only sometimes ask what the suspected cause is!

[aldabra]

Haha. I thought surely to God we have actual data, I remember seeing actual health data back in the day, let me go and look, and I got to https://digital.nhs.uk/data and searched for "pain", and you'll never guess what the #1 hit was. "Getting started with cloud adoption for NHS trusts". So, OK, that is not it but I can see where they're coming from.

NICE has some leads, at https://cks.nice.org.uk/topics/chronic-pain/

[kaberett]

... okay, I think the actual data the numbers NICE quotes are from the HSE2017, which is restricted access at the UK Data Service, so I feel a bit better about it not showing up in my searches and also I have now requested a UK Data Service account...

Thank you!

(And also: lolsob.)