Mobility Atlas research: standing-wheelchair-guide
Data and method for the article at mobilitatlas.com/blog/standing-wheelchair-guide/
License: CC BY 4.0 (credit "Mobility Atlas")
Prepared: 22 October 2026

WHAT WAS READ
Every FDA MAUDE medical device report with text filed under product code IPL (Wheelchair, Standup; 21 CFR 890.3900), plus every report under any product code whose text contains "stander", "standers", "standing frame", "EasyStand", "standing wheelchair", "stand-up wheelchair" or "standup wheelchair", read through openFDA on 22 October 2026. Reports matched only by the word bystander (also spelled "by stander" or "by-stander") were dropped, and 5 IPL reports with no text were left out. Pool: 113 reports, all coded: 54 about standers or standing frames, 41 about standing wheelchairs, 18 about other devices (coded O and left out of device claims). The stratum of each report (ipl or text) is in strata.csv.
FDA recall records for product code IPL or for standers, from the local copy of openFDA recalls (2 records: Z-0243-2023 EasyStand Bantam, filed under INO; Z-0993-2014 Superstand).
Makers: grouped from the manufacturer and brand names in each report (Altimate/EasyStand, Permobil, Rifton, The Standing Company, other).

HOW IT WAS CODED
1. A codebook (codebook.txt) was written before any report was coded. Its SHA-256 fingerprint, recorded at 2026-10-22T05:17:01Z, is 3baac388ccc2be4a432a29216e1f5269365bb314291dec72244b8e6bbbab58c0. 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 98.2% (kappa 0.971); problem 95.6% (kappa 0.95); injury 97.3% (kappa 0.928); fall 99.1% (kappa 0.98); standing 98.2% (kappa 0.964); user 96.5% (kappa 0.928); what was found 83.2% (kappa 0.728).
4. Every report where the coders differed was decided by the lead analyst under the written rules below (33 reports marked adjudicated). A consistency pass then re-applied the rules across the whole table and changed codes in 13 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
S1: DEVICE (W vs F vs O): decide from the text first, then the brand in the header; the product code alone (IPL) never makes a report W. W is a chair that carries the user seated and raises them to standing (Corpus VS, LifeStand, Superstand, Redman, Tek RMD). F is a frame in which the user is held upright, including a mobile or self-propelled stander and one with a power-drive add-on (Standing Dani WheelStand). A chair model that has no stand function and whose text never mentions standing (Permobil F5 Corpus without VS, Quickie) is an ordinary wheelchair, O. Exoskeletons (REX, Ekso), implanted pumps and catheters, infusion pumps, stair lifts, bed rails, standing slings, implants and surgical instruments are O.
S2: DEVICE O and PROBLEM (follows knee rule K1): when another device is the subject (device O), PROBLEM is X unless the report alleges or suspects that the other device itself failed or malfunctioned (a pump that offered boluses by itself or failed calibration, a catheter clogged, kinked or suspected of malfunction, a stair lift that jerks and leaks, a sling closure that would not stay shut, a chair armrest link rod that broke, a surgical shaft that broke); then it is Q. The letters P, M, K, S, E, D, T and U name standing-device parts and events and are not used for another device. A body event or injury with no failure of the other device (drug side effects, pain or fracture in an exoskeleton, entrapment in a bed rail, a finger caught during a repair) stays X.
S3: PROBLEM (pinned after a collapse or tip): P is a body part caught, pinched or crushed by the mechanism, a strap or the frame as the event itself (a pinch point, a gap, a squeeze against a wall). A person pinned under or struck by the device after it folded, collapsed, tipped or drove over them is a consequence; code the cause that came first (S for a frame that folded or collapsed, T for a tip with no failed part, D for a chair that veered). A stander that folded up while pushed and pinned the child is S.
S4: PROBLEM (straps and supports, K vs T): K when the report says a strap, belt, chest bar or support slipped, loosened, came loose, opened or released, even if the maker later traces it to wrong fastening (the allegation is coded), and when the reporter complains that it opened too easily. T when the report says only that the user or another person deliberately undid it or never fastened it, the user then fell, and no complaint is made about the part (a child who reached around and undid the strap).
S5: PROBLEM (M vs T, chair rose by itself): a seat or frame that rose, went into stand or moved without the user commanding it is M ("moved by itself" is named under M, which comes before T), even when the cause is that the safety lock was not engaged or the chair was not in its locked-down position. The user flag carries the unlocked chair; the problem letter records what the device did.
S6: PROBLEM (T vs U): a fall or tip with no failed part is T even when a medical event (cardiac arrest, seizure, fainting) caused it. U is only for a body event with no fall (the user stays in the device, as when found folded over in the frame).
S7: PROBLEM (allegation silent, investigation names a part): when the reporter names no failure and only describes a fall, and the maker's own investigation ties the event to a failed part (a loose set screw in the arm swivel housing, with the arm assembly replaced and corrective action opened), code that part's letter instead of T, because T requires that no part is said to have failed anywhere in the report.
S8: INJURY: y when harm, pain or medical care is stated, including "not seriously injured" (it implies a minor injury), a trip to the ER, and a body event that happened during the incident (seizure, fainting, blood-pressure drop). A blow or drop by itself (hit the head, dropped onto the floor, pinned) with no harm, pain or care stated is n, as is "hit the head but was not injured".
S9: FALL: y when the user reached the ground or floor uncontrolled, was thrown or fell out, or the device tipped over with the user, including a fall the report gives as the cause of the event ("this last time was a hard fall"). n when someone caught the user, lowered them to the floor in control, or they ended up near the floor (an arm's length from it) without landing; n when the user fell back into the wheelchair behind the frame.
S10: STANDING: y when the user was standing, being raised or lowered, or the stand function was in use, and also when the report implies the user was up in the frame (the wheelchair positioned behind them caught them when the strap gave). A seat elevator, seat tilt or leg rest moving on its own while the user sits, with the backrest left upright and no stand function named, is n.
S11: USER: y when the report states, or the maker concludes, that the user or a third party caused the event, including a failure that the maker attributes to inspection or maintenance not done as the manual instructs (visible cracks that should have been seen, bolts allowed to loosen, lack of maintenance). n when the maker only notes the device is past its expected life or quotes the inspection instruction while saying maintenance was done; when the user's part is one unconfirmed possibility among others with no determination ("use error or the rainy conditions"; operator training merely questioned); and when the user or a friend acted only after the event (swapping in an old motor after the drive motor locked).
S12: FINDING D: D when the maker, a dealer, a service technician or a clinical test (dye study, images) confirms the reported fault or finds or sees (on the device, a returned part or photos) a broken, cracked, missing or loose part, whatever cause the maker then names (age, wear, lack of maintenance, misuse, unknown, "could not determine root cause"). D also when a part is replaced as the failed part in response to the failure (a hydraulic actuator replaced after the stander would not lower; replacement lift parts or straps shipped for the part that failed) even if no root cause is found. A part sent as a preventive measure, as goodwill after an inspection found nothing, or a whole unit swapped before anyone examined it, is not D.
S13: FINDING N vs X: N when an inspection finds no fault or only wrong setup or adjustment, even if the maker adds that it cannot determine the root cause; N also when, with no inspection, the report itself establishes that nothing malfunctioned (the user says the device worked normally, no malfunction is alleged, and the user's own act or the surroundings caused the event). X when nothing was inspected and the account says nothing either way, when the whole unit was swapped or the parts were not returned, or when the maker says it could not confirm the fault and no inspection result is given.

KEYWORD COUNTS (Python regular expressions, matched against the upper-case report text without its header line)
KW = {'fracture': 'FRACTUR|BROKEN (LEG|ANKLE|FEMUR|TIBIA|HIP|ARM|BONE)', 'tibia': 'TIBIA|TIBIAL|FEMUR', 'child': '\\bCHILD|\\bSON\\b|\\bDAUGHTER|STUDENT|\\bKID|TODDLER|PEDIATRIC', 'chest_belt': 'CHEST (BAR|SUPPORT|PAD|STRAP|BELT)|SEAT ?BELT|BELT|STRAP', 'knee': 'KNEE (BLOCK|PAD|SUPPORT|BRACE)', 'school': 'SCHOOL|CLASSROOM|THERAPIST|THERAPY', 'weight_limit': 'WEIGHT (LIMIT|CAPACITY)|OVER ?WEIGHT', 'tip': 'TIP(PED|S)? OVER|TIPPED|FELL OVER|TIPPING', 'head': '\\bHEAD\\b', 'skin': 'SKIN|PRESSURE (SORE|INJURY|ULCER)|BLISTER', 'blood_pressure': 'BLOOD PRESSURE|FAINT|PASSED OUT|DIZZ|SYNCOPE|HYPOTENS'}

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 standing-wheelchair-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."
- One maker filed most reports in each group (Altimate for standers, Permobil for standing wheelchairs), which reflects sales and filing, not safety.
- The text match caught reports about other devices (a bed rail and a security pole sold under the name Stander, pumps, catheters); they are coded O and not counted as standing-device reports.
- Many stander reports come from schools and therapy settings, where events are watched and reported more closely than at home.

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.
standing-wheelchair-guide-documents.csv  the documents and pages read, with links