Industries / Health / Healthcare / Chiropractic
Assessment
Assessment is how a chiropractor decides whether to treat and where: taking the history, feeling the spine for stiff or tender joints, testing how it moves, and scoring pain and disability with questionnaires.
6
Milestones achieved, from 1924 to 2020.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1924
A heat meter for subluxations
In 1924 B. J. Palmer introduced the neurocalometer, a heat-sensing instrument said to detect "nerve interference", the subluxation that chiropractic theory held to be a spinal joint out of place and pressing on a nerve. He leased the instruments to chiropractors rather than selling them, made the device the centrepiece of a "back to straight chiropractic" movement and pursued makers of rival instruments as infringers; the historian Joseph C. Keating Jr. calls its launch "a model of unethical promotions in health care". The device was invented by Dossa Dixon Evins, by J. Nash's account. A 2013 review of the methods chiropractors use to decide where to manipulate rated the evidence for skin temperature measurement as unfavourable.
Lessonan instrument promoted before it was tested split the profession, and it did not hold up when it was tested.
Credit
B. J. Palmer; Dossa Dixon Evins
1991
A score for neck disability
H. Vernon and S. Mior, of the Canadian Memorial Chiropractic College in Toronto, adapted the Oswestry low back pain questionnaire into the Neck Disability Index (NDI), ten questions on how neck pain limits daily life, and tested it on whiplash patients (a test-retest correlation of 0.89 in 17 patients). An outside review of 37 studies in 2009 called it the most commonly used self-report measure for neck pain and found its reliability acceptable in most studies.
Lessona Rasch analysis, a test of whether questionnaire items measure one quantity on an even scale, of 521 trial patients by Gabrielle van der Velde and colleagues found that the ten items do not, so adding them up and comparing change scores is open to doubt; dropping two items fixed it.
Credit
H. Vernon and S. Mior (Journal of Manipulative and Physiological Therapeutics, 1991); review by Joy C. MacDermid and colleagues (2009)
2004
Palpation put to the test
Motion palpation, feeling how far and how freely each spinal joint moves to find one that is "fixed", grew from decades of work in Belgium by Henri Gillet, a 1928 Palmer graduate who began rough tests of agreement between examiners soon after graduating, with his brother Marcel and a colleague named Liekens; by the 1980s it had become a school of chiropractic of its own, by the historian Simon Senzon's account. A review of 49 studies published from 1966 to 2001 found pain provocation tests the most reliable, motion tests acceptable in 58% of the studies that used the kappa statistic (a measure of agreement beyond chance) but often not reproduced by other examiners, and soft-tissue tests not reliable.
Lessonexaminers agreed with themselves more than with one another, and neither profession, experience nor training just before a study improved agreement.
Credit
Henri Gillet, Marcel Gillet and Liekens; review by Michael Seffinger and colleagues (Spine, 2004)
2010
Subluxation called a historical concept
In May 2010 the UK's General Chiropractic Council (GCC), the body with which chiropractors there must register each year, told registrants that the chiropractic vertebral subluxation complex "is an historical concept, but it remains a theoretical model", and that it "is not supported by any clinical research evidence that would allow claims to be made that it is the cause of disease". It revised the guidance in August 2010, reissued it in February 2017 and again with effect from 1 January 2026, tying it to its rules on evidence-based care and honest advertising.
Credit
General Chiropractic Council
2013
Methods for the spot graded
John Triano and nine colleagues reviewed 201 studies of the methods chiropractors use to choose where on the spine to apply manipulation. The most favourable evidence was for methods that provoke or confirm pain at a level. They rated the evidence as high quality and supportive, with limitations, for static and motion palpation and for measures of leg length difference; mixed for posture; unclear for stiffness measures and spinal X-rays; and unfavourable for manual muscle testing, skin conductance, surface electromyography (muscle electrical activity read through the skin) and skin temperature.
Lessonmethods that examine the sore spot directly held up better than indirect signals such as skin heat.
Credit
John J. Triano and colleagues (Chiropractic & Manual Therapies, 2013)
2020
Stiffest or sorest, same result
At the Spine Centre of Southern Denmark, 132 people with persistent low back pain had the stiffness of each lumbar level measured with VerteTrack, a device that rolls along the back and records stiffness level by level, and their pain thresholds measured with pressure. They were then randomly assigned to four sessions of manipulation at either the stiffest level or the most pain-sensitive one. The level chosen made no difference to reported back pain or to stiffness, though it did change the pressure pain threshold.
Lessonthe effort put into finding the exact level to treat did not show up in the patient's pain in this trial.
Credit
Casper Glissmann Nim, Gregory Neil Kawchuk, Berit Schiøttz-Christensen and Søren O'Neill (Scientific Reports, 2020)
Watch for
Machine learning picks likely responders
Researchers in the chiropractic medicine department of Balgrist University Hospital in Zurich trained machine learning models on the baseline questionnaires of 96 people with spinal pain under chiropractic care to predict who would recover by three months. The area under the curve, a 0.5-to-1 score of how well a model separates two groups, was 0.93 to 0.99 in one validation method but fell to 0.62 to 0.90 in a stricter one. Positive expectations of treatment, confidence in managing one's own pain, younger age, lower body mass index and fewer other illnesses went with recovery. The authors call the results hypothesis-generating.
Achieved whena larger group of patients recruited by another team confirms the same predictors, with accuracy reported on patients the model has not seen.
Claimed by
Michael L. Meier and colleagues, Pain Management, September 2026
Open
01
An outside test of CORESCORE
An independent, pre-registered test of the INSiGHT neuroTECH scanners sold by the Chiropractic Leadership Alliance (CLA), which combine heart rate variability, surface electromyography and paraspinal thermography (skin temperature beside the spine) into a single "CORESCORE" shown to patients. Volunteers with and without neck or back pain are scanned twice on one day and again a week later by two operators who cannot see earlier results, and again after changes that should not matter, such as a warmer room or a short brisk walk. No one is treated.
The briefCredit · unclaimed02
Simulator palpation on real backs
A training study of whether practice on a haptic simulator, a robot-arm device that pushes back on the fingertips to imitate tissue of different stiffness, makes chiropractic students agree better when they palpate real backs. One group of students practises finding a stiff level on the simulator; a matched group practises as usual; both are tested before and after on the same volunteers, palpating only.
The briefCredit · unclaimed
Open 01 · the brief
An outside test of CORESCORE
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An independent, pre-registered test of the INSiGHT neuroTECH scanners sold by the Chiropractic Leadership Alliance (CLA), which combine heart rate variability, surface electromyography and paraspinal thermography (skin temperature beside the spine) into a single "CORESCORE" shown to patients. Volunteers with and without neck or back pain are scanned twice on one day and again a week later by two operators who cannot see earlier results, and again after changes that should not matter, such as a warmer room or a short brisk walk. No one is treated.
Why it is still open
CLA says the scans give "objective examination results" and that its scanners are Class II medical devices registered with the FDA and Health Canada, offered from $399 a month. The one reproducibility study of the scanner's thermal function, on 100 people, found high agreement (intraclass correlations of 0.959 to 0.976) but was co-written by a consultant to CLA and measured temperatures, not the combined score. Triano's 2013 review found the evidence unfavourable for skin conductance, surface electromyography and skin temperature as ways to decide where to treat. Europe PMC and web searches in November 2026 found no test of the CORESCORE by anyone independent of CLA.
insightcla.cominsightcla.compmc.ncbi.nlm.nih.govpmc.ncbi.nlm.nih.gov
Built with what exists
INSiGHT scanners at a chiropractic college clinic that already owns them, run by its licensed staff as the devices' registration requires; a FLIR thermal camera and a room thermometer as independent checks; a protocol registered on the Open Science Framework; R or Python for intraclass correlations and Bland-Altman limits of agreement (the range within which repeat readings fall).
First step
Write and register the protocol, and ask two chiropractic colleges whether their teaching clinics' INSiGHT units can be used for a measurement-only study.
Done looks like
60 volunteers, 30 with pain, scanned at least four times each; published within-day and week-to-week agreement for the CORESCORE and each part, with the smallest change that exceeds measurement noise; how well the score separates the pain and no-pain groups; and how far room temperature and the walk move it. Data and protocol public.
Open 02 · the brief
Simulator palpation on real backs
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A training study of whether practice on a haptic simulator, a robot-arm device that pushes back on the fingertips to imitate tissue of different stiffness, makes chiropractic students agree better when they palpate real backs. One group of students practises finding a stiff level on the simulator; a matched group practises as usual; both are tested before and after on the same volunteers, palpating only.
Why it is still open
Ohio University's Virtual Haptic Back, built for 89 first-year osteopathic medical students, taught them to find a region of abnormal stiffness placed at random among 12 locations: over six practice sessions the smallest stiffness difference they could detect fell from 28% to 14% and their time from 39 to 17 seconds. Its 2008 studies measured performance inside the simulator. On real spines, agreement between examiners stays low: Seffinger's review found motion tests often not reproduced, and a 2015 study of 25 people found poor to low agreement between two raters palpating the thoracic spine. A Europe PMC search in November 2026 for haptic or virtual palpation trainers found only the Ohio studies and none that tested whether the skill transfers to real backs.
Built with what exists
two 3D Systems Touch haptic devices; CHAI3D, an open-source haptics software framework; the continuous measures of agreement used by R. Cooperstein and M. Young, such as the median distance between two examiners' choices of the stiffest level, alongside kappa.
First step
Build a one-level stiffness scene in CHAI3D with one haptic arm and check that five students can tell three stiffness settings apart.
Done looks like
40 students randomly assigned to the two groups; pairs palpate the same 10 volunteers for the stiffest lumbar level before and after six practice sessions; the change in between-examiner agreement published for each group with confidence intervals; the simulator's code released.