Inspired by Wang et al 2024.[1]
LGTD – longitudinal generational tracking database
aHR – adjusted hazard ratio
CCI – Charleston comorbidity index (see previous blog)
PD – Parkinson’s disease
GPT – generative pre-trained transformer
CI – confidence interval
PSA – post-stroke aphasia– key to acronyms
This is another large retrospective cohort study from Taiwan, this time looking at the incidence of injuries (mainly falls) following a stroke.
The LGTD 2000 was used to find patients hospitalised with stroke between 2001 and 2012. Patients were followed up for at least 6 years to estimate accidental injury incidence and risk.
108 196 patients were found who were at risk of accidental injury following a stroke. Of these nearly 40% had acupuncture (n=42 038) and just over 60% did not (n=66 113).
During the follow-up period (up to the end of 2018), there were 4 264 (10.1%) accidental injuries in the acupuncture cohort and 8 802 (13.3%) in the no-acupuncture cohort.
Propensity score matching reduced the number in each group to 31 249. The incidence of falls was just under 30% lower in the matched acupuncture cohort (aHR 0.71).
Apart from acupuncture, the only other factor that reduced risk of accidental injury was being in the highest health insurance bracket. The risk was increased with age and comorbidity (CCI), as well as specific comorbidities (in order from highest risk) – epilepsy, dementia, PD, fracture, osteoporosis, anaemia, hyperlipidaemia, coronary heart disease, chronic kidney disease. The risk was also increased in groups taking certain drugs (in order from highest risk) – barbiturates, benzodiazepines, Z-drugs.
I asked ChatGPT about Z-drugs and it gave me zolpidem, zopiclone, and zaleplon, which seems to fit here under the category used in the paper – sedative hypnotic drugs.
The risk reduction associated with having acupuncture after a stroke was not really affected by any other categories or comorbidities apart from where the CIs expanded, presumably due to smaller numbers in these subgroups.
I asked ChatGPT about the cost of a fall in a patient following a stroke. It gave me a variety of different costs including acute healthcare costs, post-fall rehabilitation costs, and social and long-term care costs. It then teased me with a 2022 UK study that gave an estimate for aggregated fall-related healthcare costs for stroke survivors. The cost range was stated as £3k to £5k per fall episode. I thought I better check the original reference, but (shock horror) when I asked for the reference, ChatGTP said that it …could not be identified in full detail. It then suggested I use PubMed, so I did.
I only found one paper with my first search attempt, but it had some useful data.[2] Unsurprisingly, falls are more common following a stroke (49.5% vs 35.1%), and injurious falls are too (16.0% vs 10.3%). In 2015, the cost of informal caregiving resulting from falls in stroke survivors in the US was estimated to be 2.9 billion dollars and 6.5 billion for those who never had a stroke.
Using ChatGPTs aggregated fall-related healthcare costs for stroke survivors (£4k per fall) plus the rate of falls (over the previous 2 years) at 49.5%, and the prevalence of stroke in the UK population (2%), I estimate that acupuncture for all stroke survivors might save 373 million pounds a year in just the cost attributed to falls.
Providing everyone in the UK with 10 sessions of acupuncture following a stroke would cost roughly 30 million pounds per year (£30 per session and 100 000 strokes per year in the UK).
It seems like a good deal, and of course there could be other benefits from using acupuncture following a stroke, such as an improvement in PSA (see Acupuncture for PSA 2024). Obviously, NICE is not going to get its health economic estimates from ChatGPT, and the figures above should be viewed as entirely speculative, not to mention the fact that they are predicated upon the implicit assumption of a causal relationship between acupuncture and a reduced risk of falls in patients following a stroke. This retrospective observational research, whilst having over 30 000 subjects in each group (cohort), can only provide data on the association between having acupuncture after a stroke and a reduced incidence of accidental injuries (mostly falls).
References
1 Wang M-J, Chou H-J, Lin S-K. Efficacy of acupuncture in reducing accidental injury risk in stroke patients: A national-scale cohort study. Heliyon. 2024;10:e40081. doi: 10.1016/j.heliyon.2024.e40081
2 Joo H, Wang G, Yee SL, et al. Economic Burden of Informal Caregiving Associated With History of Stroke and Falls Among Older Adults in the U.S. Am J Prev Med. 2017;53:S197–204. doi: 10.1016/j.amepre.2017.07.020
The BMAS Blog 
You must be logged in to post a comment.