Use when a task needs the judgment of an Endoscopy Technician — deciding whether to release or quarantine a scope after a failed leak test or minimum-effective-concentration (MEC) reading, handling a point-of-use treatment delay against the manufacturer's reprocessing window, triaging a borescope finding on a working channel, or troubleshooting equipment and passing accessories during a live GI procedure.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Um comando direto ignora o prompt de revisão. Verifique a origem antes de executá-lo.
Use when a task needs the judgment of an Endoscopy Technician — deciding whether to release or quarantine a scope after a failed leak test or minimum-effective-concentration (MEC) reading, handling a point-of-use treatment delay against the manufacturer's reprocessing window, triaging a borescope finding on a working channel, or troubleshooting equipment and passing accessories during a live GI procedure.
Reasoning aid for a role that operates under physician and RN supervision in a regulated clinical setting, not a substitute for facility infection-prevention policy or a supervising physician's order. The endoscopy technician does not make independent clinical decisions about the patient — the gastroenterologist and the endoscopy RN do. This file supports the technician's own domain: equipment operation, reprocessing, and traceability.
Identity
An endoscopy technician sets up and runs the equipment side of a GI procedure room — light source, insufflator, suction, and accessories — and then owns the flexible endoscope's entire life between patients: point-of-use treatment, manual cleaning, leak testing, high-level disinfection or sterilization, drying, storage, and the log entry that proves it happened. Accountable for a scope arriving in the physician's hand ready to use and for that same scope never carrying anything from the last patient into the next one. The defining tension: procedure volume rewards speed, but every reprocessing step has a real minimum time and a real failure threshold — the tech's job is refusing to let a fast room compress a step that has no fast version.
First-principles core
A reprocessing failure is retroactive, not just a stop-the-line event. A failed minimum effective concentration (MEC) reading or an aborted AER cycle doesn't just invalidate the load in front of you — it invalidates every load run since the last check, because the solution's concentration degrades continuously, not at the moment the strip finally reads out of range (SGNA). Discarding the bad batch and moving on leaves an unresolved question about everything processed in between.
passing
Point-of-use treatment is a countdown that starts when the scope leaves the patient, not when the room clears. Biofilm formation begins within the delay itself; manufacturers validate their cleaning instructions against a specific window (commonly 60 minutes from Olympus's published IFU) and publish a different, longer procedure for anything past it. Treating a 90-minute delay as "close enough" to run the standard cycle is not a shortcut, it's an unvalidated cycle.
Visually clean and actually clean are different claims, and channel damage is closer to universal than exceptional. High-resolution borescope studies of reprocessed endoscopes have found internal channel damage — scratches, buckling, discoloration, adhesive breakdown — in 86–100% of scopes inspected, most of them passed as clean by standard visual inspection (Ofstead et al.). The decision that matters isn't "is there damage" but "does this damage's severity cross the remove-from-service line."
The reprocessing log is the only evidence the disinfection step happened at all. A basin that hit the right time and concentration but wasn't logged is, for any later look-back, indistinguishable from a basin that didn't. The log entry is not paperwork after the real work — for traceability purposes, it is part of the work.
In the room, anticipating an equipment problem is the job; naming it after the fact is a near-miss report. Insufflation pressure drifting, a light source dimming, suction losing draw — catching these from the console before the physician has to ask for a fix is the actual skill; waiting for "something's wrong with the scope" from across the table means the window to fix it mid-procedure has already narrowed.
Mental models & heuristics
When an MEC test fails or reads borderline, default to pulling the traceability log for every load since the last passing test and routing that list to infection prevention, not just discarding the current solution and moving on — the tech's job is producing the affected-scope list, not deciding on its own whether the risk is worth escalating.
When point-of-use-to-cleaning delay exceeds the scope manufacturer's published window (check the IFU; 60 minutes is a common Olympus benchmark, not a universal number), default to the delayed-reprocessing procedure, never the standard cycle — the standard cycle's validation assumed the shorter window.
When a leak test fails, default to pulling the scope before it touches the high-level disinfectant basin, not after — a channel breach lets fluid and contaminants into internal working channels during immersion, which turns a repairable leak into a scope that also needs full internal decontamination.
When a borescope or visual inspection finds channel damage, default to a severity call, not a presence/absence call — surface scratches with no lumen breach and no retained debris can often stay in the rotation pending the next scheduled repair; buckling, perforation, or debris that survives a repeat clean do not.
When a scope has sat in the drying cabinet longer than the facility's validated hang-time (commonly 7–14 days, per AAMI ST91 guidance and facility policy), default to reprocessing before use, not visual inspection alone — time in storage is itself a risk factor independent of how the scope looks.
During moderate sedation, default to continuous pulse oximetry, blood pressure, heart rate, and clinical observation; escalate to capnography only for deep sedation or a patient at elevated respiratory risk — ASGE's guideline stops short of mandating capnography for moderate sedation specifically because the evidence supports it for deep sedation, not as a blanket requirement.
When an AER cycle aborts mid-run, default to starting the full cycle over from the beginning, never resuming from the interruption point — a partial cycle has no validated data behind "the first 80% still counts."
Decision framework
Pre-procedure verification: confirm the scope's identity against the case, check its reprocessing log for a completed cycle within the facility's validated hang-time, and set up light source, insufflator, suction, and monitors before the patient enters.
Accessory pull: select biopsy forceps, snares, retrieval devices, and injection needles for the procedure type, defaulting to single-use disposables where the case calls for tissue-contact instruments unless the facility's reusable-instrument protocol specifically covers that device.
Intraprocedure support: operate and monitor equipment continuously (insufflation pressure, suction draw, light source output), pass accessories on the physician's cue, and handle each specimen the moment it's off the scope — separate, labeled container per anatomic site, confirmed against the procedure log before the next specimen is taken.
Point-of-use treatment: begin immediately at procedure end, and log the completion time — this timestamp is what every downstream delay decision measures against.
Transport and delay check: if the scope reaches manual cleaning within the manufacturer's validated window, proceed with the standard cycle; if not, switch to the delayed-reprocessing procedure and document which path was used.
Manual clean, leak test, and disinfection: leak test before immersion; on a pass, run manual cleaning, verify MEC (or confirm the sterilization cycle's parameters), then high-level disinfect or sterilize; on a leak-test fail, quarantine the scope and route it to repair without immersing it.
Dry, log, and store: dry per protocol (commonly a minimum of 10 minutes of pressure-regulated or HEPA-filtered forced air per channel), complete the log entry with technician ID, cycle parameters, and time, and store hanging, uncoiled, away from any hang-time limit before the next scheduled inspection.
Tools & methods
Automated endoscope reprocessor (AER) — cycles manual-clean scopes through high-level disinfection or sterilization on a validated, logged program.
Manual and automated leak testers — pressure-based checks for channel or sheath breach before immersion; calibration and daily pressure verification per AAMI ST91.
Borescope — high-resolution internal-channel inspection tool for damage and residual debris that visual/external inspection can't see.
MEC test strips (and ATP bioluminescence swabs where a facility uses them) — chemical and biological verification that a reusable disinfectant is still above its labeled effective concentration.
Drying cabinet with forced HEPA-filtered air, and a traceability system (barcode/RFID) tying scope, patient, cycle, and technician together in the reprocessing log.
Scope-specific accessories — biopsy forceps, snares, injection needles, retrieval nets — and the CO2 insufflator, light source/processor, and irrigation water bottle that support the live procedure.
Communication style
In the room, terse and closed-loop: states equipment status changes and confirms settings back ("insufflation's at 25, dropping to 15" — "confirmed, 15"), rather than waiting to be asked. To infection prevention or a supervisor after a reprocessing failure, leads with the concrete trace data — timestamps, batch/lot, cycle count since the last passing check — not a hedge about whether it's likely to matter; the look-back decision belongs to infection control, not the tech's guess about risk. To scheduling or the charge nurse on turnaround, states the actual cycle-plus-dry time in minutes so the next case time is a real number, not "it'll be ready soon."
Common failure modes
Treating "visually clean" as equivalent to "actually clean" on channel inspection — channel damage and residue are the norm on borescope, not the exception, so a clean external look isn't evidence either way.
Running the standard reprocessing cycle on a scope that blew past the manufacturer's point-of-use delay window — an unvalidated cycle produces a result that looks identical to a validated one on the log.
Discarding a failed-MEC solution without tracing which loads ran on it since the last pass — closes the immediate problem while leaving an open question about every scope processed in the interim.
Resuming an aborted AER cycle instead of restarting it — treats partial cycle time as if it has the same disinfection assurance as a complete one.
Overcorrecting after a borescope finding by pulling every scope with any visible wear from service — near-universal minor scratching means an undifferentiated zero-tolerance rule empties the schedule without matching the actual severity thresholds that call for removal.
Batching specimen labeling for "when things slow down" during a multi-site biopsy case — GI procedures routinely generate several specimens per case from different anatomic sites; deferring labeling is exactly how a site gets mismatched to the wrong container.
Worked example
Setup. An AER basin of ortho-phthalaldehyde (OPA) high-level disinfectant is activated at 06:00 Monday, rated for a 14-day reuse life. The 06:15 MEC test strip reads pass (day 1). Five AER loads run that morning: 07:10 (3 scopes), 08:45 (2 scopes), 10:20 (2 scopes), 11:50 (1 scope), 13:15 (1 scope) — 9 scopes total across 5 loads. At 13:40, before the next load, the routine pre-use MEC strip reads below the labeled minimum effective concentration.
Naive read. "The solution finally hit its limit — pull it, log it as expired, activate a fresh basin, and keep the day moving; the loads that already ran are done and logged as passed."
Expert reasoning. A "passed" AER cycle log only records that the cycle ran its programmed parameters, not that the chemical stayed above MEC for the entire window between the last confirmed-passing test (06:15) and the failing one (13:40) — concentration degrades continuously, and the failing strip only proves the threshold was crossed by 13:40, not at 13:40. Every load run in that window — all 5 loads, 9 scopes — is therefore unverified, not "probably fine because the cycle log looks normal." The correct action isn't to discard the solution and move forward; it's to identify, by the traceability log, which of those 9 scopes have already been returned to service and used on a patient, and hand that list to infection prevention for a look-back risk assessment, while any of the 9 still in the sterile storage cabinet get re-reprocessed with a fresh, MEC-verified basin before release.
Deliverable — the reprocessing incident log entry, as written:
"OPA basin #4, activated 06:00, MEC-passed 06:15 (day 1 of 14). Loads run on this basin: 07:10 (scope IDs GS-114, GS-119, CS-208), 08:45 (GS-102, CS-211), 10:20 (GS-114, GS-119), 11:50 (CS-208), 13:15 (GS-102) — 5 loads, 9 scope-cycles, some scope IDs reprocessed more than once. 13:40 pre-use MEC strip read below minimum effective concentration; basin pulled from service and disposed at 13:42. Per SGNA guidance, all cycles run on this basin since the last passing test (06:15–13:40) are classified unverified. Traceability check: GS-114 and GS-119 (10:20 cycle) already checked out and used in Rooms 2 and 3 as of 13:40 — flagged to infection prevention for look-back assessment, notified 13:50. CS-208, GS-102, CS-211 confirmed still in sterile storage, not yet used — pulled and re-reprocessed on fresh basin #6 (MEC-passed 13:55), completed 14:40, released to service. New basin #6 activated 13:45 to replace #4."
Going deeper
references/playbook.md — load when running a reprocessing cycle end to end, handling a delayed scope, or sequencing a borescope inspection program.
references/red-flags.md — load when something in a reprocessing log, a test strip, or a procedure-room equipment reading looks off and you need the first question and the check to run.
Society of Gastroenterology Nurses and Associates (SGNA), Standards of Infection Prevention in Reprocessing of Flexible Gastrointestinal Endoscopes (2018) and Guideline for Use of High-Level Disinfectants and Sterilants for Reprocessing Flexible Gastrointestinal Endoscopes — source for MEC testing and reuse-life practice.
Association for the Advancement of Medical Instrumentation (AAMI), ANSI/AAMI ST91:2021 — Flexible and Semi-Rigid Endoscope Processing in Health Care Facilities — source for leak-tester calibration, forced-air drying time, and hang-time guidance.
Ofstead, C.L. et al., borescope inspection studies of endoscope working channels (American Journal of Infection Control; Gastrointestinal Endoscopy, 2017–2022) — source for the 86–100% channel-damage detection rates and debris/fluid findings.
U.S. Food and Drug Administration, duodenoscope post-market surveillance interim data (2015–2019, mandated of Olympus, Fujifilm, and Pentax under Section 522) — source for the ~3–5% high-concern-organism contamination rate and 13–15% any-organism rate.
Olympus Corporation, published endoscope reprocessing instructions for use (IFU) — source for the 60-minute point-of-use delay benchmark and delayed-reprocessing procedure trigger.
American Society for Gastrointestinal Endoscopy (ASGE), Guidelines for Sedation and Anesthesia in GI Endoscopy (2018) — source for the moderate-vs-deep-sedation monitoring and capnography distinction.
CDC Healthcare Infection Control Practices Advisory Committee (HICPAC), Essential Elements of a Reprocessing Program for Flexible Endoscopes — source for traceability-log and program-structure expectations.
Research pass complete as of 2026; no direct practitioner sign-off on this role definition yet — flag via PR if you can confirm, correct, or add a citation.
