Quality Tolerance Limit (QTL) Check
Check whether a monitored clinical-trial parameter has breached its pre-defined quality tolerance limit — a trial-level signal of a possible systematic issue. Enter the event count, subjects at risk, and the QTL; with a baseline, see how far past expected it sits. Aligned to ICH E6(R3) and E8(R1).
OUTPUT
TIME
Limitations — read before you rely on this
- A QTL is only as meaningful as its pre-specification and justification; the tool checks the arithmetic but cannot judge whether the limit or baseline was set defensibly.
- Breach flags at small n are dominated by sampling variation; the tool does not model the enrolment trajectory or apply group-sequential corrections, so early breaches need confirmation.
- This calculator is transparent but it is not a validated system, and its z output is a descriptive heuristic, not a statistical test.
- Reproduce the check in your own qualified system (a validated statistical environment or RBQM platform) before a QTL breach enters the monitoring record or triggers a protocol action.
WHAT THIS CALCULATES
Checks whether an observed clinical-trial parameter rate has breached a pre-defined quality tolerance limit that signals a possible systematic issue affecting participant safety or data reliability.
THE METHOD
Observed_rate = x / n Breach = Observed_rate > QTL z = (Observed_rate − p0) / √(p0·(1−p0)/n)- Observed_rate
- The current proportion of the monitored event across the trial (0–1).
- x
- The count of events observed (e.g. important protocol deviations, or subjects lost to follow-up).
- n
- The number of subjects or observations at risk — the denominator over which the rate is computed.
- QTL
- The pre-defined quality tolerance limit: the protocol threshold beyond which a systematic problem is signalled.
- Breach
- True when the observed rate crosses the QTL in the adverse direction.
- z
- An optional descriptive standardised distance of the observed rate from baseline — a "how far past expected" heuristic, not a significance test.
- p0
- The expected/baseline rate assumed when the QTL was set (0–1).
ICH E6(R3) frames QTLs as parameters with pre-specified limits that trigger evaluation at the trial/systematic level — not thresholds for acting on individual sites or subjects. The z term is a descriptive secondary limit only; a QTL breach is a signal to investigate, not a statistical decision.
THE INPUTS, AND WHAT THEY MEAN
- Events observed (x)
- The count of the quality parameter monitored across the trial — e.g. subjects with a major protocol deviation, lost to follow-up, or with missing primary-endpoint data.
- Subjects / observations at risk (n)
- The denominator: total subjects or evaluable observations to date over which the event rate is calculated.
- Quality tolerance limit (QTL)
- The protocol-defined threshold proportion, agreed before monitoring, beyond which a systematic issue is considered signalled.
- Baseline rate (p0, optional)
- The anticipated event rate used when the QTL was justified, needed only for the secondary z-distance readout.
Has a parameter breached its quality tolerance limit?
Enter the monitored event count, the subjects at risk, and the pre-defined quality tolerance limit. SPEQ flags whether the trial-level rate has breached the limit and, with a baseline, how far past expected it sits — a signal to investigate systematically, per ICH E6(R3).
HOW TO READ THE OUTPUT
- The pass/breach output reflects the sponsor’s own pre-defined QTL, not a regulatory threshold; ICH E6(R3) requires the sponsor to set and justify the limit and defines what a breach means procedurally.
- A breach is a trigger for root-cause evaluation and possible protocol/process action recorded in the monitoring plan — it is not, by itself, evidence of misconduct or a reason to act on a single site.
- The z distance is descriptive only: a large z at small n means the rate is unstable, not that the finding is significant — use it to prioritise which near-limit parameters to watch.
- QTLs are meant for a small number of critical-to-quality parameters; a breach early at low n usually reflects sampling noise and should be confirmed as n grows.
WORKED EXAMPLE
A Phase 3 trial pre-set a 5% QTL for subjects lost to follow-up, and the data-management lead runs an interim check.
- Events (x)
- 22
- At risk (n)
- 300
- QTL
- 0.05
- Baseline p0
- 0.03
RESULT
Loss to follow-up has crossed the pre-defined limit and sits well above the planned rate, so the sponsor documents a root-cause assessment and any action in the monitoring record — while resisting reading the breach as any single site’s fault.
REGULATORY BASIS
- ICH E6(R3) Good Clinical Practice
- Frames quality tolerance limits as pre-defined parameters whose breach triggers evaluation at the systematic level.
- ICH E8(R1) General Considerations for Clinical Studies
- Establishes the quality-by-design / critical-to-quality-factors framing that QTLs operationalise.