Why the Limit Must Travel With the Result
A concentration result is interpretable only in relation to the analytical method that produced it. The lower limit of quantification (LLOQ) is the lowest concentration at which an assay has been validated to report a quantitative result under specified performance criteria. If the assay, laboratory, specimen matrix, or method period changes, the applicable LLOQ may change too. A single study-wide value can therefore label records incorrectly even when every recorded concentration is unchanged.
The practical task is to connect each result to the LLOQ that applied to that result, using documented assay information. Then check whether the recorded result and its quantification status agree with that limit. This is a data-construction and quality-control operation: it does not determine how below-quantification observations should be handled in a later analysis.
The exact matching dimensions depend on how the laboratory defines and documents its methods. At minimum, identify the analyte and the assay or method. Include laboratory and matrix when they can distinguish validated limits. Include a period or method-version identifier when the applicable method changed over time. Do not add dimensions mechanically: each one should correspond to a real distinction in the source records or validated method documentation.
Representing and Assigning Limits
A robust limit reference table has one row for each distinct applicability rule. It should record the relevant identifiers, the LLOQ and its units, the interval or version to which it applies, and the source used to establish the value. In the result data, retain the original result and status and add fields for the attached LLOQ, its units, and a reference to the matching rule. Keeping the attachment traceable makes it possible to explain and reproduce the assignment.
Where applicability is defined by dates, specify which timestamp selects the period. For example, a documented method change may apply according to the sample collection date, or according to the date of analysis if the laboratory's records define it that way. Use the documented rule consistently; do not choose whichever date produces a more convenient match. A common way to represent adjacent periods is as half-open intervals: a rule applies from its start date, inclusive, up to but not including its end date. This prevents a sample on a changeover date from matching both periods.
The assignment can be described as a lookup function. Let \(k\) denote the set of applicability fields for a result, and let \(t\) be the date or time used under the documented period rule. The attached limit is \(L(k,t)\). For this function to be well-defined, every included result must match one and only one row in the reference table. In particular, identical keys and overlapping periods must not produce competing limits.
This condition is more than a data-management preference. If there are zero matches, the value of \(L(k,t)\) has not been established. If there are two matches with different limits, the result has no unique applicable value. In either case, downstream comparisons between result and limit are not reliable until the reference information or the matching rule is resolved.
Worked Example: Assigning a Limit Across a Method Change
Suppose a hypothetical study uses the same analyte, laboratory, and plasma matrix throughout. The assay documentation specifies method M1 through June 30, 2025, with an LLOQ of 0.050 mg/L, and method M2 starting July 1, 2025, with an LLOQ of 0.080 mg/L. The documented rule assigns the method by sample collection date. The intervals are therefore \([{\rm start}, {\rm end})\), with M1 ending when M2 begins.
| Sample | Collection date | Method | Result | Attached LLOQ |
|---|---|---|---|---|
| A | June 30, 2025 | M1 | 0.060 mg/L | 0.050 mg/L |
| B | July 1, 2025 | M2 | 0.075 mg/L | 0.080 mg/L |
For A, June 30 falls within M1's period, so \(L=0.050\) mg/L. The numerical comparison is \(0.060 \geq 0.050\), consistent with a quantitative result under the basic limit check. For B, July 1 falls within M2's period, so \(L=0.080\) mg/L. Here \(0.075 < 0.080\), which is inconsistent with an unqualified quantitative result under that same check and should be flagged for review. It would be incorrect to use M1's lower limit for B merely because it is the earlier or smaller value.
Checking Result and Status Consistency
Once a limit is attached, compare it with the recorded result and quantification status. The basic check applies to a numerical value recorded as a quantitative concentration: it should be at or above the applicable LLOQ. A value below the limit does not by itself establish what the correct status should be; the record and laboratory reporting convention must be considered. In particular, do not turn a missing value into a below-quantification result simply because a limit is known.
Use the conventions documented for the data. A below-quantification record may have a missing numerical concentration, or it may retain a numerical instrument or laboratory value while its status indicates that it is not reportable as a quantitative result. Those encodings are not interchangeable. The check should distinguish the reported quantitative result from any retained non-quantitative value and should flag contradictions rather than silently rewriting a record.
The comparison is meaningful only when units agree. For example, \(80\) ng/mL and \(0.080\) mg/L are equivalent because \(1\) mg/L equals \(1000\) ng/L and \(1\) mg/L equals \(1000\) ng/mL? No: the conversion must be checked carefully. Since \(1\) mg is \(10^6\) ng and \(1\) L is \(10^3\) mL, \(1\) mg/L equals \(1000\) ng/mL. Thus \(80\) ng/mL equals \(0.080\) mg/L. A unit mismatch can create a factor-of-1000 error and reverse the apparent result-limit relationship.
Worked Example: Comparing Status, Result, and Units
Consider three hypothetical records with an attached LLOQ of \(0.080\) mg/L. The laboratory's documented convention is that a quantitative result is stored numerically, while a below-quantification record has the status “below quantification”; its numerical field may be missing.
Record 1: Status is quantitative and the result is \(0.080\) mg/L. The comparison is \(0.080 \geq 0.080\). The value is at the limit and passes the basic quantitative-result check.
Record 2: Status is quantitative and the result is \(0.072\) mg/L. The comparison is \(0.072 < 0.080\). The numerical value and status conflict with the expected reporting boundary. Preserve the original record and investigate whether the limit assignment, units, rounding, or status was recorded incorrectly.
Record 3: Status is below quantification and the numerical field is missing. This is consistent with the stated convention. The missing value alone would not justify assigning that status, but here the status is explicitly recorded.
Now suppose a fourth result is \(75\) ng/mL and its status is quantitative. Convert before comparing: \(75\ {\rm ng/mL} \div 1000 = 0.075\ {\rm mg/L}\). Since \(0.075 < 0.080\) mg/L, this record also warrants review. Comparing \(75\) directly with \(0.080\) without converting would be meaningless.
Diagnostics and Failure Modes
Several checks help establish whether the attachment process is working. First, count matches for every result: zero matches indicate a missing rule, while multiple matches indicate a key or interval ambiguity. Second, inspect the distribution of attached LLOQs by assay, laboratory, matrix, and period. An unexpected concentration of missing values, abrupt jumps, or identical values across supposedly different methods may indicate a failed join or an incorrect reference table. Third, review result-status disagreements using the documented reporting convention and consistent units.
- Unmatched result: A required key is absent from the reference table, an identifier was coded differently, or the period rule has a gap. Keep the LLOQ unresolved until the source information is clarified.
- Multiple matches: Reference rows overlap, the key omits a necessary distinction, or period boundaries are handled inconsistently. Do not choose the first row returned by a database join.
- Unexpected boundary cases: A result equal to the LLOQ ordinarily passes the quantitative comparison. A result just below it may reflect rounding, but rounding should not be assumed to excuse a discrepancy without checking the stored precision and reporting rules.
- Implausible attached values: Check units, analyte, matrix, method version, and the source of the limit. A small numerical value is not necessarily correct merely because it resembles a known LLOQ.
- Status conflicts: A quantitative value below its attached limit, or a below-quantification status paired with a value at or above it, is a review flag. It is not automatic permission to change the concentration or status.
A short derivation explains why a unique match is essential. Suppose the same result key and date match two reference rows with limits \(L_1\) and \(L_2\), where \(L_1 \neq L_2\). A comparison such as \(C \geq L\) then has no single defined \(L\): it may pass for one row and fail for the other. Therefore, a result-limit validation is interpretable only after the match is unique or the applicability rule is resolved.
Choosing an Approach and Knowing Its Limits
A study-wide constant is simple but is appropriate only when documentation establishes that the same limit applies to every relevant record. A row-by-row value transcribed from source records can accommodate variation, but it is vulnerable to transcription errors and may be hard to reproduce. A curated reference table with explicit applicability fields is often a practical compromise: it makes the rules visible, supports automated matching, and allows the number of matches to be checked.
Whichever approach is used, the reference table should not hide uncertainty. Retain the original reported information, record the attached limit and source rule separately, and document unresolved cases. A limit assignment is not a substitute for confirming that the method, specimen, and period are applicable. Nor does a successful comparison prove that the measured concentration is otherwise accurate; it only checks consistency with the assigned quantification boundary.
The LLOQ also does not determine how a below-quantification observation should be represented in a pharmacokinetic model. That is a separate modeling decision. Here, the objective is narrower: ensure each result carries the correct documented limit and that the recorded numerical value and status do not contradict it without explanation.
Worked Example: Finding a Period Gap and an Overlap
A hypothetical reference table lists one assay for serum at Laboratory Q. Rule A assigns an LLOQ of \(0.10\) mg/L from March 1 through March 31, 2025, and Rule B assigns \(0.12\) mg/L from April 1 through April 30, 2025. The documented convention is that start and end dates are inclusive. A result collected on April 1 matches Rule B, so its attached LLOQ is \(0.12\) mg/L. A result collected on March 31 matches Rule A and receives \(0.10\) mg/L.
Suppose a result collected on May 1 has no matching row. The assignment count is zero, so neither \(0.10\) nor \(0.12\) mg/L can be justified from this table. Extending Rule B beyond April 30 without documentation would invent an applicability rule.
Now suppose a corrected reference table adds Rule C for April 15 through May 15, also for the same assay, laboratory, and matrix. A sample collected April 20 matches both Rule B and Rule C. Its assignment count is two. Even if both rules happen to state \(0.12\) mg/L, the overlap should be resolved in the reference table because two applicable rows violate the unique-match rule and may diverge for other fields or future records. The correct action is to clarify the rule boundaries or method applicability before validating results against the limit.
Check Your Understanding
Answer each question, then compare your reasoning with the solutions.
- A sample was collected on the first day a new method became effective. What must the period convention establish before an LLOQ can be assigned?
- A quantitative result is \(0.045\) mg/L and its attached LLOQ is \(0.050\) mg/L. What does the comparison show, and what should happen next?
- A result has a below-quantification status and a missing numerical concentration. Can the missing value alone establish that status?
- A result is \(60\) ng/mL and its LLOQ is \(0.050\) mg/L. Convert the result and assess the basic quantitative-result check.
- A result matches two limit-reference rows, both with the same LLOQ. Is the assignment fully resolved? Explain.
Solutions. (1) The documented rule must say which period includes the changeover date and which timestamp determines applicability; half-open intervals are one way to make the boundary unambiguous. (2) Since \(0.045 < 0.050\) mg/L, the quantitative value is below its attached limit. Flag it and review the assignment, units, rounding, and reported status rather than changing it automatically. (3) No. Missingness and below-quantification status are distinct; the status must be supported by the recorded reporting information. (4) \(60\ {\rm ng/mL} \div 1000 = 0.060\ {\rm mg/L}\), and \(0.060 \geq 0.050\) mg/L, so it passes the basic check. (5) No. There are two matching rows, so the match is not unique. Resolve the reference-table overlap even if the current numerical limits agree.