Repair Training & Education

After Repair Training: Documenting Competence, Work Records, and Audit Readiness

1 2

The course certificate goes in a drawer; the audit finding usually comes from what was never written down. Sending a technician to a repair course is the easy half. The half that survives an audit — or a customer’s supplier review — is converting what was learnt into four artefacts: a competence matrix by device family, work records with enough fields to be traceable, objective test data with pre-set acceptance criteria, and a reassessment cycle that catches decay. This guide sets out what the standards actually require, the minimum record fields, and how to prove that the training worked, using published evidence where it exists.

Repair training discussion with endoscope equipment, documents, and optics on the bench

What ISO 13485 and AAMI EQ56 actually ask for

Two reference points frame the requirements, and both apply to repair organisations, not only to manufacturers. ISO 13485:2016 — the quality management system standard for medical devices, currently in its third edition and confirmed again in October 2025 — states in its scope that it applies to organizations involved in the design, production, installation, and servicing of medical devices, which covers independent repair companies and hospital HTM departments alike. Its clause 6.2 requires that the organization document the processes for establishing competence, provide the needed training, and then do three things buyers of training often forget: evaluate the effectiveness of the actions taken, ensure personnel are aware of the relevance of their work, and retain records of education, training, skills, and experience.

On the operations side, ANSI/AAMI EQ56:2024 establishes minimum requirements for a medical equipment management program and applies explicitly to healthcare delivery organizations, device manufacturers, and independent service providers. It requires that repair programs have written procedures defining the service provider’s activities and the necessary documentation, that inventory management include evaluation, inspection, and record-keeping processes, and that equipment safety programs cover training, recall management, inspections, and planned maintenance — with annexes addressing computerized maintenance management systems (CMMS). AAMI’s EQ committee has also published ANSI/AAMI EQ110:2024, a dedicated standard for healthcare technology management educational programs, which is the natural reference when the training itself needs a documented structure.

RequirementSource clauseEvidence a repair shop should hold
Determine required competence per role and device familyISO 13485:2016 cl. 6.2 a)Competence matrix approved by management
Provide training or take other actions to achieve competencecl. 6.2 b)Course outlines, attendance records, theory scores
Evaluate the effectiveness of the actions takencl. 6.2 c)Practical assessments, post-course rework and return data
Ensure awareness of the relevance and importance of the workcl. 6.2 d)Procedure briefings and acknowledgements on file
Retain education, training, skills, and experience recordscl. 6.2 e)Personnel files with dates, evidence, and signatures
Written repair procedures and defined documentationANSI/AAMI EQ56:2024 (repair programs)SOP set per device family; standard work-record template
Inventory with evaluation, inspection, and record-keepingEQ56:2024 (inventory management)Device registry with service history per serial number

Build a competence matrix by device family

“Trained” is not a competence level; it is an event. The matrix replaces the event with a map: for each device family, the specific tasks a person is authorised to perform, the level of authorisation, the evidence that supported the sign-off, and the next reassessment date. Authorisation levels should be written plainly — for example, trainee (works only under direct supervision), supervised (performs tasks with a senior technician reviewing the result), independent (performs and releases within the family), and trainer (trains and signs off others). A cascading structure has published precedent: in the SIRT programme’s second phase, 12 new trainees were taught by three technicians trained in the first phase, and the programme’s outcome measures held. Training your own trainers is what makes competence scale without sending everyone back to the vendor every year.

Device familyTasks covered by the matrixAuthorisation levelsTypical sign-off evidenceReassessment
Flexible endoscopesLeak testing; insertion tube replacement; bending section and angulation wire service; biopsy channel replacement; air/water and suction valve serviceTrainee / supervised / independent / trainerObserved task sign-off plus a leak test record on a scope with known faultsAnnual
Rigid endoscopesOptical assessment; distal window replacement; sealing and leak testingTrainee / supervised / independentMeasurement records against stated optical limitsAnnual
Camera heads and cablesDisassembly and reassembly; sensor handling; optical alignment; waterproofing; electrical testingTrainee / supervised / independentElectrical safety test record and alignment verificationAnnual
Ultrasound probesCable and strain-relief replacement; acoustic lens repair; piezo element diagnostics; sensitivity testingTrainee / supervised / independentSensitivity baseline comparison against a reference probeAnnual

Minimum fields for every repair record

A work record has one job: let a competent stranger reconstruct what was done, on which unit, with which parts, and how the result was verified. That is also what a recall response or a customer complaint investigation will ask for, usually under time pressure. The field list below is the minimum we recommend for any repair shop, and it maps directly onto the sourcing and incoming-inspection discipline described in our endoscope repair parts sourcing guide.

Record fieldWhy it mattersCommon failure in practice
Device identification: brand, model, serial numberTies the record to the physical unit and its service historyModel recorded without the serial number
Reported symptom versus observed faultSeparates the user’s complaint from the bench findingOnly the complaint is logged
Diagnosis steps and intermediate test resultsShows the method and enables peer reviewA conclusion with no steps
Parts replaced: description, part number, batch or lot, supplierTraceability for recalls and quality claims“Insertion tube replaced” with no lot or source data
Acceptance test results against pre-set criteriaProves the unit left the bench within specification“Checked OK” with no numbers
Technician, reviewer, and datesAccountability and training evidence in one placeA single signature with no review step
Release status and next scheduled maintenanceFeeds fleet planning and PM schedulingField left blank

Two supporting disciplines keep the parts field honest. Torque-controlled fastening on scopes and instrument repairs should use tools calibrated on a documented schedule — the calibration requirements for hand torque tools are set out in ISO 6789-2:2017, and test equipment whose results enter the record should trace to a calibration laboratory working to ISO/IEC 17025:2017. A record that cites an uncalibrated instrument is a finding, not evidence.

Objective test data beats opinion

The strongest argument for insisting on measured results is what happens when a facility collects them systematically. A 2024 retrospective study by Ofstead and colleagues in Biomedical Instrumentation & Technology analysed three years of repair data from a large academic medical center performing borescope inspections of every flexible endoscope after manual cleaning — approximately 109,000 inspections across a fleet of more than 500 endoscopes in 25 departments. The proportion of repairs classified as major fell from 12.1% to 3.2%, and the mean cost per repair fell from $4,426 to $2,337 (P<0.001). Mean turnaround time fell from 24.1 to 15.5 days, the mean number of uses between repairs rose from 52.1 to 87.2, and the facility reduced its endoscope fleet from 593 to 508 units.

Two details in that study matter for training documentation. First, the technicians performing the inspections were certified and extensively trained, with the study noting that guidelines recommend using borescopes during training and competency testing — the test instrument and the training are explicitly linked. Second, the improvements emerged over years of recorded inspections, not from a single intervention. The practical reading for a repair shop: define the acceptance criteria before the work (leak test pressure hold, optical limits, electrical safety thresholds), record the raw readings rather than checkboxes, and the record set becomes something an auditor or customer can act on.

Checking whether the training actually worked

ISO 13485 asks for an effectiveness evaluation, which means the question “was the course worth it” needs a written answer. We recommend three layers of evidence, each with a threshold recorded in the training SOP before the course: immediate evidence (theory score against a fixed pass mark and observed task sign-offs), short-term evidence (rework and return rates in the first 30 days on the device family trained), and ongoing evidence (quarterly record sampling and reassessment scores against the baseline). Perception surveys belong in the file too, but as supporting material: in the SIRT evaluation, 83.3% of participants felt more confident and 95.8% reported adequate repair skills, yet the programme’s headline evidence was objective — 1,623 instruments evaluated with a 99.6% successful repair rate. Confidence proves engagement; outcomes prove competence.

Reassessment, nonconformance, and audit day

Three routines close the loop. Schedule reassessment annually per device family, or sooner after a long absence from the bench, a new scope generation, or a process change. Define the nonconformance path in advance: when a repair fails at incoming inspection or in the field, the record is reviewed, the unit is isolated, the root cause is documented, and the event either triggers targeted retraining or a procedure correction — both of which then appear in the training file as evidence the system responds. Finally, prepare for the audit itself by rehearsing the exercise a surveyor will actually run: pull three completed work orders at random and walk each one from the complaint, through diagnosis, parts with lot numbers, test readings, and review signatures, to release. If any of those links is missing on the paper, the competence matrix will not save the finding. Facilities handling reusable instruments should also keep their processing competencies aligned with the training expectations in standards such as ANSI/AAMI ST79:2017, because repair and reprocessing competencies are usually audited together.

Making the record set part of your sourcing story

The finished documentation is not only an audit defence; it is a commercial asset. A repair company or hospital department that can show a competence matrix, sample work orders with full part traceability, and calibrated test evidence answers the supplier-qualification questions before they are asked, which shortens RFQ cycles with independent service buyers. The same applies in reverse: when you select a training provider, treat their own documentation discipline as part of the product. Our repair training programs and course listings are supported by the same parts supply our bench uses — including the flexible endoscope parts catalogue referenced throughout training — and you can request a course outline and quotation by email or WhatsApp with your device categories and dates.

Frequently asked questions

How long should repair training records be kept?

ISO 13485 requires that records of education, training, skills, and experience be maintained, but the retention period is defined by your own documented record-control procedure and by any contractual or regulatory requirements that apply to your customers. The practical anchor most shops use is to keep the training file for as long as the person remains authorised on that device family, plus the retention period the procedure states.

Can one competence matrix cover every device family?

It can carry them, but authorisation should not. Skills do not transfer automatically between a flexible scope bench and an ultrasound probe bench, and an authorisation on one family says nothing about another. Keep the matrix in one document for overview, but sign off tasks and levels per family, with evidence recorded per family.

Who is qualified to sign off competence?

Someone whose own competence on that device family is documented at trainer level — typically a senior technician or the service manager, supported where needed by the training provider’s instructor. The signature means two things, and the record should carry both: the evidence was reviewed, and the reviewer takes responsibility for the authorisation decision.