Mobility Atlas research: knee-scooter-guide
Data and method for the article at mobilitatlas.com/blog/knee-scooter-guide/
License: CC BY 4.0 (credit "Mobility Atlas")
Prepared: 19 October 2026

WHAT WAS READ
Every FDA MAUDE medical device report, under any product code, whose event text contains "knee scooter", "knee scooters", "knee walker", "knee walkers" or "kneewalker" (hyphenated forms included), read through openFDA on 19 October 2026: 252 reports, received from 26 March 2008 to 28 September 2026, all coded. 242 are about a knee scooter (227 of them filed under product code ITJ) and 10 about another device. Four reports whose brand name, but not their text, contains these words were not in the pool (report keys 21873249, 25019438, 8466873 and 14696901).
The FDA recall records that mention a knee walker or knee scooter, downloaded through openFDA on 19 October 2026: 3 records.
Makers: grouped from the manufacturer and brand names in each report (Medline, HL, Kenstone, Lin Kou, other).

HOW IT WAS CODED
1. A codebook (codebook.txt) was written before any report was coded. Its SHA-256 fingerprint, recorded at 2026-10-19T13:42:03Z, is 82e8bcb71de6ea89d9902b4582041440f469bb3369560fdb61b97d3ae3cdd16c. Any later change to the codebook would change this fingerprint.
2. The reports were split into chunks. Two coders (AI language models working separately, each in a private scratch folder, each without access to the other's answers, told to read every report and not to use keyword scripts) coded every report in every chunk against the codebook.
3. Agreement between the two coders: device 100.0% (kappa 1.0); problem 94.0% (kappa 0.929); injury 100.0% (kappa 1.0); fall 99.2% (kappa 0.984); user 99.2% (kappa 0.937); what was found 95.2% (kappa 0.9).
4. Every report where the coders differed was decided by the lead analyst under the written rules below (31 reports marked adjudicated). A consistency pass then re-applied the rules across the whole table and changed codes in 17 more reports (marked consistency).
5. Counts in the article were produced by script from the final table, not by hand. Some sub-counts (parts, places or words named in reports) are keyword counts over the report text; the exact word patterns are listed under KEYWORD COUNTS. Different wording would give somewhat different counts.

ADJUDICATION RULES
K1: DEVICE O and PROBLEM: when another device is the subject (device O), PROBLEM is X unless the report alleges that the other device itself failed or malfunctioned (a plate that broke, a pump that stopped, a metal-on-metal hip revised for metal debris and tissue reaction, a crutch that split, a stimulator lead and strap that would not hold); then it is Q. The letters S, B, W, P, F and T name knee-scooter parts and events and are not used for another device. Damage done to the other device by an outside event (pumps that fell off the scooter and broke their tubing) stays X.
K2: PROBLEM (several failures): code what the report says went wrong first in the chain of events. When one failure is said to cause another (a rear axle shifting so the brake loses its drum), the causing failure comes first. Damage that happens in or after the fall (a caster breaking after the user fell, a grip breaking in the fall) does not count. List order decides only when the report names several failures together with no sequence (handlebar screw and brake; wobbly wheels and brakes).
K3: PROBLEM (no part named): a scooter that "collapsed" or "snapped and collapsed" with no part named is F, because the codebook names collapse under F. It is Q when the reporter only says the scooter was defective, failed, broke, malfunctioned, or was wobbly or unstable, design complaints about stability included, without naming a failed part. It is T when the report describes only a tip or fall and its circumstances and makes no complaint about the device.
K4: PROBLEM (obstacles): when a wheel hits or catches a crack and the handlebar keeps turning, or turns a different way from the wheels, it is S (the handlebar turned or the steering did not steer), even with no part named. It is T when the scooter only stopped, flipped or tipped on the obstacle or clothing caught a wheel. It is W when the reporter says the wheels locked.
K5: PROBLEM (handlebar height and fold vs frame): a failure in the handlebar or steering column's own height adjustment or fold is S, because S and F both apply and S comes first. Examples: the handle telescoping down, a missing height pin for the handle, a broken weld where the push handle folds, a height adjuster letting the steering column collapse. F covers the frame, a frame folding latch, and height adjustments or locks whose part the report does not identify.
K6: PROBLEM (knee pad supports): a bar, brace or beam that the report places under the knee pad, knee rest or seat is the support of the pad, so it is P, not F.
K7: PROBLEM (allegation vs inspection): code the failure the reporter alleges, even when the maker found a different part (a back hinge that snapped is F although the maker found the tiller bolt missing). Use the part an inspection names only when the reporter says no more than that the unit was damaged or broken in some area without saying what failed. A "cable" on a knee scooter is the brake cable (B), the only cable these scooters carry.
K8: FALL: y when the user fell, went to the ground, or was thrown or launched forward, or when the scooter tipped over while she rode it. n when she caught herself before reaching the ground, or when only the operated foot or leg struck the floor while she stayed on or beside the scooter.
K9: USER: y when the report states that the user or a third party caused the event, as the cause or as the maker's conclusion ("most likely cause", "suggests", a cause code traced to the user). Also y when an inspection documents the user's wrong assembly or misuse, or use beyond a stated product limit. n when the user's part is raised only as an unconfirmed possibility (possible over-tightening with no one named; a user unsure whether her own use caused it), when an IFU caution is quoted without saying the user ignored it, or when the user or a relative acted only after the problem.
K10: FINDING: D when the maker, a dealer or a service inspection confirms the reported fault, or finds a broken, bent, missing or defective part, on the device or on photos, video or images of it, or when a part or the unit is replaced as faulty. This holds whatever cause the maker then names (shipping damage, mishandling, user knowledge, unknown). N when an inspection finds no fault or only wrong assembly, or when the maker concludes that the cause lay outside the device without confirming any fault. X when nothing was inspected and no conclusion is drawn, or the fault appears only in the account of the event with no inspection stated (a new walker given to the user is not an inspection). X also when the maker says it could not confirm or determine the fault, cannot tie its photo to the unit, or cannot say whether the damage came before the event or from it.

KEYWORD COUNTS (Python regular expressions, matched against the upper-case report text without its header line)
KW = {'fracture': 'FRACTUR', 'head': '\\bHEAD\\b', 'reinjury': 'RE-?INJUR|SURGICAL SITE|INCISION|DEHISC|HARDWARE', 'curb': 'CURB', 'slope': 'SLOPE|INCLINE|RAMP|HILL|DRIVEWAY', 'stairs': 'STAIR|STEP', 'brake': 'BRAKE', 'assembly': 'ASSEMBL|TIGHTEN'}

QUALITY NOTES (disclosed in full)
- The article went through fact-check passes by checkers who had not written it (a full pass with a reader review, then a control pass over the changes), each recomputing numbers from the files here and reading quotes against their sources; every error found was corrected before publication.

AMOUNTS
Every report text was searched for money figures; none gives one. Prices in the article are prices shown on makers', dealers' and retailers' pages and in the Mobility Atlas catalog, listed in knee-scooter-guide-documents.csv and the article sources.

LIMITS
- Reports are not rates. Nobody publishes how many devices of each model are in use.
- A report is an allegation or a description; the FDA says the submission of a report "itself does not necessarily demonstrate that the device caused or contributed to the adverse outcome or event."
- Rule change disclosed: in this study, a fault confirmed by the maker, a dealer or an inspection counts as a confirmed fault even when the maker then names an outside cause such as shipping damage; earlier Mobility Atlas studies counted those with the outside causes.
- Filing differs by year: some years are dominated by one maker forwarding every complaint, and many 2026 reports are short complaint notes filed by one company.
- Five reports repeat another report's text and are counted as filed.

FILES
codebook.txt        the codebook as fingerprinted
coded_reports.csv   one row per coded report: both coders' answers, the final code, and how a disagreement was resolved (resolution, rule_applied)
report_pool.csv     the reports in the pool with product code, brand, date received and event type
The full report texts are public: search MAUDE or openFDA by report key.
knee-scooter-guide-documents.csv  the documents and pages read, with links