Introduction: The Specification That Doesn’t Exist Yet
The assignment arrives four months before launch. Define the through-focus acceptance criteria for the new EDOF IOL line. The process engineer opens ISO 11979-2 and finds power tolerance, MTF threshold at best focus, and model eye configurations. Everything needed for monofocal QC is there. Everything needed for EDOF-specific quality verification is not.
No international standard specifies what plateau width is acceptable for an EDOF IOL. No published guideline defines the minimum contrast within the extended range. No industry consensus establishes how much plateau asymmetry is tolerable, how performance should be verified at larger pupil sizes, or how much ripple is acceptable in a diffractive EDOF design.
The process engineer must build these EDOF acceptance criteria from scratch. Not by guessing. Not by copying a competitor’s unpublished internal specification. By deriving each threshold from three sources that together define where acceptable performance lives: the design intent from R&D, the manufacturing capability from production data, and the clinical correlation that links measurement values to patient outcomes.
This article provides the systematic framework for that derivation. It identifies the five parameters to specify, explains how to derive each threshold from the three sources, accounts for measurement uncertainty, and provides an 8-week timeline from first measurement to operational specification. It is written for the process engineer who has been handed this assignment and needs a methodology, not just motivation.
The Three Sources: Where Acceptance Criteria Come From
Every EDOF acceptance criterion is derived from the intersection of three inputs. Omitting any one of them produces a specification that is either unachievable, inadequate, or unsupported.
Source 1: Design intent
The optical designer provides the target through-focus profile: the ideal plateau width, the minimum MTF within the designed range, the intended symmetry, and the expected aperture dependence. These targets represent what the lens is designed to achieve-the center of the performance distribution.
The design target is not the acceptance criterion. It is the starting point. If the design target for plateau width is 1.5D, the acceptance criterion will be lower-because manufacturing variation, measurement uncertainty, and a margin of safety all consume part of the tolerance budget. The design target is the nominal. The acceptance criterion is the boundary.
R&D also provides sensitivity analysis: what happens to the through-focus profile when each parameter deviates by 5%, 10%, and 20% from design? This sensitivity data identifies where the cliff edge is-the parameter deviation beyond which clinical performance degrades unacceptably. The acceptance criterion must sit above that cliff edge with adequate margin.
Source 2: Manufacturing capability
The manufacturing process delivers lenses with a distribution of through-focus characteristics, not a single deterministic value. The mean of that distribution should be close to the design target. The standard deviation determines how tightly the process clusters around the target.
Setting acceptance criteria tighter than the manufacturing process can reliably deliver produces excessive reject rates. Setting them looser than the clinical requirement produces field complaints. The manufacturing capability study-measuring actual production lenses and computing the distribution statistics-provides the data needed to find the balance.
The key metric is process capability index (Cpk). A Cpk of 1.33 or higher indicates the process is well-centered within the specification with adequate margin. Below 1.0, the specification exceeds current manufacturing capability and must be either relaxed (with clinical justification) or achieved through process improvement.
Source 3: Clinical correlation
The clinical threshold is the hard boundary. Below this value, patients experience the performance degradation that generates complaints. Above it, performance is clinically adequate even if not optimal.
For a new EDOF launch, direct clinical correlation data may not yet exist. The initial clinical thresholds are estimated from three sources: the design sensitivity analysis (at what deviation does the simulation predict clinically meaningful degradation), published literature on EDOF through-focus requirements, and clinical experience with similar products from the manufacturer’s portfolio.
These initial estimates are refined as production runs accumulate and clinical feedback arrives. The criteria are living thresholds, not permanent fixtures. The first version is the best available estimate. The third version, updated with actual field data, is evidence-based.
The acceptance criterion lives where all three sources converge: above the clinical threshold, within the manufacturing capability, and aligned with the design intent. This intersection is the specification.
The Five Parameters to Specify
EDOF through-focus acceptance criteria require specification of five parameters. Together, these five capture the complete extended-range performance that distinguishes an EDOF IOL from a monofocal. Omitting any parameter leaves a gap that can allow clinically deficient lenses to pass.
Parameter 1: Plateau width
Plateau width is the defocus range over which MTF remains above a defined threshold-typically 0.15 at 50 lp/mm, corresponding to approximately 20/30 visual acuity equivalent. This is the primary EDOF quality metric. It directly represents the range of clear vision the patient will experience.
To derive the acceptance threshold: begin with the design target (e.g., 1.5D). Measure 30 or more pre-production lenses to establish the manufacturing distribution (e.g., mean 1.45D, σ = 0.08D). Identify the clinical boundary from sensitivity analysis and literature (e.g., complaints begin below 1.0D). Set the acceptance threshold at a point that maintains adequate Cpk while staying safely above the clinical boundary.
Example derivation: design target 1.5D, manufacturing mean 1.45D with σ of 0.08D. Accepting at ≥ 1.2D places the threshold 3.1σ below the mean (Cpk ≈ 1.04) and 0.2D above the clinical boundary. This yields a projected reject rate below 1% while maintaining a meaningful safety margin above the complaint threshold.
Parameter 2: Minimum MTF within the designed range
Even a wide plateau is clinically useless if it contains a deep dip where contrast drops below the usable threshold. Minimum MTF captures the worst point anywhere on the plateau-the contrast dead zone that patients reliably notice.
Design target: minimum ≥ 0.15 everywhere within the designed range. Manufacturing distribution: the minimum MTF across lenses follows its own distribution, typically with a mean slightly below the plateau average and higher variability. Clinical boundary: contrast dead zones become noticeable to patients below approximately 0.08–0.10 at 50 lp/mm.
The acceptance threshold should be set above 0.10 (clinical comfort zone) with adequate margin for measurement uncertainty. A practical starting point: accept ≥ 0.12, review 0.08–0.12, reject < 0.08.
Parameter 3: Plateau center position
The center of the plateau should align with the design intent. Some EDOF designs are deliberately asymmetric-biased toward intermediate distance. The specification must reflect the intended center, not assume symmetry around zero defocus.
R&D provides the design reference center position. Manufacturing distributes the actual center position around a mean that should be close to the design reference. The clinical impact of center shift: a hyperopic shift moves useful range away from intermediate toward distance; a myopic shift does the opposite. Shifts exceeding 0.50D from design compromise the intended clinical balance.
Practical threshold: accept within ±0.25D of the design reference center. Review ±0.25–0.50D. Reject shifts exceeding ±0.50D.
Parameter 4: Multi-aperture consistency
Wavefront-shaping EDOF designs are inherently pupil-dependent. Performance at 3mm (photopic) may differ substantially from performance at 4.5mm (mesopic). The specification must address both conditions because patients use their lenses across a range of lighting environments.
The specification approach: define an acceptable ratio between the plateau width at 4.5mm and the plateau width at 3mm. A 4.5mm plateau that is 65% or more of the 3mm plateau indicates the EDOF effect survives at mesopic pupils. Below 50%, the lens loses its extended-range characteristic under low light.
Practical threshold: accept if the 4.5mm plateau ≥ 1.0D when the 3mm plateau ≥ 1.5D. Reject if the 4.5mm plateau drops below 0.5D regardless of the 3mm result.
Parameter 5: Ripple depth (diffractive EDOF designs)
Diffractive EDOF designs produce characteristic oscillations within the plateau. These are design features, not defects-but manufacturing errors can deepen them beyond the designed amplitude, creating narrow contrast gaps at specific working distances.
The specification targets the ripple amplitude (peak-to-valley of the oscillation) and the absolute minimum within any ripple trough. A ripple amplitude below 0.05 MTF units is indistinguishable from the smooth-plateau ideal. Above 0.08, patients may notice a specific distance where contrast dips.
Practical threshold: accept ripple amplitude < 0.05. Review 0.05–0.08. Reject > 0.08 or any ripple trough below 0.08 absolute MTF. For non-diffractive wavefront-shaping EDOF designs, the plateau should be smooth-any oscillation in these designs is a manufacturing artifact, not a design feature, and the tolerance should be correspondingly tighter.
Table 1: EDOF Acceptance Criteria Derivation Matrix
| Parameter | Design Target | Mfg Capability (example) | Clinical Threshold | Accept | Review | Reject |
| Plateau width (MTF ≥ 0.15 at 50 lp/mm) | 1.5D | Mean 1.45D, σ 0.08D | Complaints below 1.0D | ≥ 1.2D | 1.0–1.2D | < 1.0D |
| Minimum MTF within range | ≥ 0.15 | Mean 0.18, σ 0.03 | Dead zones below 0.08–0.10 | ≥ 0.12 | 0.08–0.12 | < 0.08 |
| Plateau center position | Per design ref. | Mean +0.02D, σ 0.06D | Shift > 0.50D compromises balance | ±0.25D of design | ±0.25–0.50D | > ±0.50D |
| Multi-aperture (4.5mm plateau) | ≥ 65% of 3mm | Ratio mean 0.72, σ 0.08 | Mesopic complaints below 50% | ≥ 1.0D | 0.5–1.0D | < 0.5D |
| Ripple depth (diffractive only) | < 0.03 MTF | Mean 0.025, σ 0.012 | Gaps noticeable > 0.08 | < 0.05 | 0.05–0.08 | > 0.08 |
[Note: All values in this table are illustrative examples demonstrating the derivation methodology. Actual design targets, manufacturing distributions, and clinical thresholds will differ for each EDOF product. The framework is universal; the numbers must be derived from your specific design, production process, and clinical data.]
The Manufacturing Capability Study
Setting EDOF acceptance criteria without production data is guessing. The manufacturing capability study provides the empirical foundation: what does the process actually deliver for each through-focus parameter?
Study protocol
Measure a minimum of 30 lenses from at least three production batches. The sample must represent the normal range of manufacturing conditions-different material batches, different machine setups, different operators if applicable. A sample drawn entirely from one ideal batch will underestimate the true process variability.
Each lens is measured once on the IOLA MFD. The single 9-second measurement captures the complete wavefront and automatically computes all five through-focus parameters: plateau width, minimum MTF, plateau center, multi-aperture performance at any specified pupil sizes, and ripple depth. Thirty lenses require 4.5 minutes of total measurement time plus data export and analysis.
For each parameter, compute the mean, standard deviation, and histogram. These statistics reveal the process center (is it close to the design target?), the process spread (how tightly does production cluster?), and the distribution shape (is it normal, skewed, or bimodal?).
Interpreting capability indices
For each parameter, compute the process capability index (Cpk) against the proposed acceptance threshold. Cpk quantifies how well the process fits within the specification relative to both centering and spread.
A Cpk of 1.33 or higher indicates the process is well-centered with adequate margin-fewer than 64 parts per million fall outside the specification. This is the target for each through-focus parameter. A Cpk between 1.0 and 1.33 indicates the specification is achievable but tight. Expected reject rate is 0.3–1.0%, which may be acceptable for a premium product or may warrant process improvement. Below 1.0, the current process cannot reliably meet the specification. Either the specification must be relaxed (if clinical evidence supports a wider tolerance) or the process must be improved (if the clinical threshold demands the tighter criterion).
The capability study also reveals which parameter has the lowest Cpk. This parameter is the constraint-the most likely source of future rejects and the highest-priority target for process improvement. Directing engineering effort at the constraint parameter yields the largest return in reject rate reduction.
Accounting for Measurement Uncertainty
Every through-focus parameter measurement carries uncertainty. If the acceptance criterion is set without accounting for this uncertainty, the QC system will make incorrect decisions-accepting lenses that are actually out of specification and rejecting lenses that are actually conforming.
The gauge R&R study
Before finalizing any acceptance threshold, characterize the measurement system’s repeatability and reproducibility for each through-focus parameter. The standard protocol: 10 lenses measured by 3 operators with 2 repetitions each, producing 60 data points per parameter.
Compute the gauge R&R as a percentage of the tolerance band. The AIAG Measurement Systems Analysis manual provides the industry-standard decision criteria: gauge R&R at or below 10% of the tolerance band is excellent-measurement contributes negligibly to accept/reject errors. Between 10% and 20% is acceptable for most production applications. Above 20%, the measurement system consumes too much of the tolerance budget, and either measurement precision must be improved or the tolerance band must be widened.
Guardbanding
When measurement uncertainty is significant relative to the tolerance band, guardbanding provides a principled approach to accept/reject decisions. The guardband shifts the acceptance threshold inward from the specification limit by an amount proportional to the measurement uncertainty.
Example: if the specification acceptance threshold for plateau width is 1.2D and the measurement repeatability is ±0.05D (at 95% confidence), applying a guardband shifts the definitive acceptance threshold to 1.25D. Lenses measuring between 1.20D and 1.25D enter the review zone and are remeasured for confirmation.
The guardband approach prevents two types of errors: shipping lenses that are actually below specification (consumer risk) and rejecting lenses that are actually above specification (producer risk). The balance between these risks depends on the product’s criticality-for an implantable medical device, minimizing consumer risk takes priority.
Run the gauge R&R study early in the implementation timeline-week 2 at the latest. If measurement uncertainty is larger than anticipated, it affects every acceptance threshold in the specification. Discovering this in week 7 forces rework of the entire criteria derivation.
The 8-Week Implementation Timeline
Weeks 1–2: Baseline measurement and gauge R&R
Configure the IOLA MFD for the specific EDOF product. Establish measurement protocols: aperture sizes, defocus scan range, spatial frequency for MTF evaluation. Measure 30 or more pre-production lenses across multiple batches. Record all five through-focus parameters. Simultaneously execute the gauge R&R study: 10 lenses, 3 operators, 2 repetitions.
Deliverables: manufacturing distribution statistics (mean, σ, histograms) for all five parameters. Gauge R&R values for each parameter. These two data sets are the empirical foundation for everything that follows.
Weeks 3–4: Design sensitivity and clinical boundary mapping
R&D provides the sensitivity analysis: through-focus performance as a function of parameter deviation from design target. The process engineer maps the cliff edge for each parameter-the deviation beyond which simulation predicts clinically meaningful degradation.
Clinical affairs provides whatever historical or literature-based clinical threshold data is available. For a first EDOF launch, this may be limited to published studies on similar designs and the manufacturer’s clinical experience with adjacent products.
Deliverable: for each parameter, a documented mapping from design target through manufacturing distribution to clinical boundary, identifying the range within which the acceptance threshold must fall.
Weeks 5–6: Draft specification
Set accept, review, and reject thresholds for each parameter based on the three-source derivation. Apply guardbands based on gauge R&R data. Calculate the expected reject rate using the manufacturing distribution and the proposed thresholds. Verify that the reject rate is operationally sustainable-typically below 2% for a mature process, potentially higher for an initial launch with process optimization ongoing.
Circulate the draft specification to R&D, QC, regulatory, and clinical affairs for cross-functional review. Each stakeholder validates from their perspective: R&D confirms alignment with design intent, QC confirms operability, regulatory confirms audit defensibility, clinical confirms the safety margin.
Deliverable: draft EDOF through-focus acceptance specification with documented derivation rationale for each threshold.
Weeks 7–8: Validation and finalization
Measure 50 or more additional lenses from new production batches. Apply the draft criteria to every lens. Verify correct classification: lenses that are known to be good (from design reference or R&D-confirmed samples) should pass. Lenses that are known to be marginal or below standard should be caught by the criteria.
Examine the review zone: are lenses that fall between accept and reject thresholds genuinely borderline, or is the review zone capturing measurement noise? If the review zone is populated primarily by measurement variability rather than genuine process variation, the guardband may need adjustment.
Finalize the specification. Compile the validation report including: derivation rationale, capability study data, gauge R&R results, validation measurement results, and the final accept/review/reject thresholds. This documentation package is the audit-ready evidence that the criteria are derived from data, not from assumption.
Table 2: 8-Week Implementation Timeline
| Week | Activity | Deliverable | Key Metric |
| 1–2 | Baseline measurement (30+ lenses, ≥3 batches) + gauge R&R study (10 lenses, 3 operators, 2 reps) | Manufacturing distribution: mean, σ, histogram per parameter. Gauge R&R per parameter. | Gauge R&R ≤ 20% of tolerance band for all parameters |
| 3–4 | Design sensitivity analysis (from R&D) + clinical threshold mapping (from literature/clinical affairs) | Parameter boundary map: design target → cliff edge → clinical threshold per parameter | Clinical thresholds documented with source traceability |
| 5–6 | Draft acceptance criteria: thresholds with guardbands. Reject rate calculation. Cross-functional review. | Draft specification with derivation rationale for each threshold | Cpk ≥ 1.33 for each parameter; reject rate < 2% |
| 7–8 | Validation: 50+ lenses from new batches. Apply criteria. Verify correct classification. Finalize. | Final specification + validation report + audit-ready documentation package | Zero false accepts on known-marginal lenses; false reject rate < 1% |
Living Criteria: Updating as Evidence Accumulates
The initial EDOF acceptance criteria are the best available estimate derived from pre-production data and limited clinical input. They are version 1.0, not the permanent specification. As production runs, measurement data, and clinical feedback accumulate, the criteria should evolve.
Quarterly manufacturing data refresh. As production volume grows, the manufacturing distribution statistics become more precise. A capability study based on 30 lenses carries meaningful statistical uncertainty. After 6 months of production with 500 or more lenses measured, the mean and standard deviation estimates are much more reliable. Update the Cpk calculations quarterly and adjust the specification if the process has improved (tighter σ enables tighter criteria or lower reject rate) or degraded (wider σ requires investigation and correction).
Clinical correlation feedback loop. As surgeon feedback and complaint data accumulate, correlate through-focus measurement parameters from shipped lenses with clinical outcomes. This correlation is the most valuable data in the specification development process. It replaces estimated clinical thresholds with empirically validated ones. If complaints cluster below a specific plateau width that is different from the initial estimate, adjust the clinical threshold accordingly.
Version control. Maintain a revision history for the acceptance specification. Each revision documents what changed, why it changed (new capability data, updated clinical threshold, process improvement), and the supporting evidence. Auditors view specifications that evolve with evidence as signs of a mature quality system. Specifications that were set once and never revisited suggest the opposite.
The criteria are not static documents. They are living instruments that sharpen with use. The first version gets the EDOF line to market with defensible quality controls. The third version, refined with production-scale data and real clinical feedback, represents the mature EDOF acceptance criteria that protect both patients and the manufacturer’s reputation.
Conclusion
No standard provides EDOF through-focus acceptance criteria. The process engineer builds them from three sources: the design intent that defines the target, the manufacturing capability that defines the achievable, and the clinical correlation that defines the necessary. The acceptance criterion lives at the intersection-above the clinical threshold, within the manufacturing capability, and aligned with the design.
Five parameters capture the complete EDOF performance: plateau width, minimum MTF within the range, plateau center position, multi-aperture consistency, and ripple depth. Each parameter requires its own derivation from the three sources, its own gauge R&R assessment, and its own guardband.
The timeline is eight weeks from first measurement to operational specification. The methodology is systematic and repeatable. The documentation is audit-ready from day one. And the criteria are designed to evolve-refined with production data and clinical feedback as the EDOF line matures.
The standard tells you how to measure power. The design tells you what the plateau should look like. The production data tells you what it actually looks like. The acceptance criterion lives where all three converge-and it is your job to find that point.
Disclaimer: This document is intended for educational use only. It does not represent legal, regulatory, or certification advice, and should not be interpreted as a declaration of compliance or approval by Rotlex or any regulatory authority. All numerical examples are illustrative and must be validated against your specific EDOF design, manufacturing process, and clinical data.
