From Near Miss to Never Again: Two Decades of Risk management and Error Prevention in an IVF Laboratory

Abstract

Objective: To evaluate 20-year outcomes of a structured quality improvement (QI) and risk-management program in a single IVF laboratory, with emphasis on never events, near misses, and longitudinal performance on predefined quality indicators (QIIs).
Design: Longitudinal, single-center quality improvement study (2004–2024).
Setting: Large healthcare network -affiliated IVF laboratory operating under CAP accreditation.
Patients/Cycles: 15,956 ART cycles (8,320 fresh IVF; 7,636 frozen embryo transfer).
Interventions: Implementation and continuous refinement of a laboratory QI framework comprising high-risk process mapping; QIIs with thresholds; standardized reporting (verbal escalation → SBAR); structured investigations (RCA) with corrective/preventive actions (CAPA); competency-based staff training; electronic/dual witnessing; cryoinventory reconciliation; equipment maintenance and alarm testing; and a non-punitive reporting culture.
Main Outcome Measures: Incidence of never events and intercepted near misses; protocol non-compliance; report errors; cryoinventory accuracy (QIR07); gamete/embryo traceability (QIR10); equipment/handling issues affecting care (QIR16).
Results: Across 20 years, one true “never event” occurred (erroneous discard of an embryo intended for cryopreservation with freezing of a lower-quality embryo instead; ≈0.006% of cycles). The event was disclosed, investigated via RCA, corrected per SOPs, and remediated with a no-cost IVF cycle. One intercepted near miss (thaw of an undesired-gender embryo detected pre-transfer) was identified, disclosed, and resolved (refreeze and correct embryo transfer) without clinical impact. Protocol non-compliance declined from 8 cases (2004) to 0 by 2008 and remained at or near zero thereafter. Report errors decreased to 0% in recent years. Cryoinventory performance remained near 0% error with one easily resolved misplacement. Gamete/embryo traceability (QIR10) stayed well below thresholds with no significant missing/untraceable specimens. QIR16 recorded one handling incident (faulty pipette), causing loss of several oocytes, prompting protocol revision, equipment checks, and retraining via RCA/CAPA.
Conclusions: A structured, data-driven QI program–embedding SBAR, RCA/CAPA, traceability safeguards, and a just culture–was associated with sustained near-zero serious events and progressive reliability gains over two decades. This reproducible model can inform benchmarking and multi-center learning aimed at further reducing latent risk in IVF laboratories.