The controls a laboratory runs before it reports a number

The number on your report is the end of a system

A laboratory result looks like a measurement. It is closer to the output of a monitored process, and the rules governing that process are unusually concrete.

A laboratory doing anything beyond the simplest testing has to meet a run of analytic systems requirements, and monitor and evaluate the overall quality of those systems, correcting problems it finds.

The pieces are a written procedure manual, controlled materials and equipment, documented maintenance, calibration and calibration verification, daily control testing, corrective action, and records.

Each has a separate rule. Read together they answer a question a patient reasonably has: what did anyone do to know this instrument was working today.

The procedure manual, and the fourteen things it covers

A written procedure manual for every test has to be available to and followed by staff.

One sentence forecloses a common substitute: "Textbooks may supplement but not replace the laboratory's written procedures" for testing or examining specimens.

The required contents run from patient preparation, specimen handling and acceptability criteria through step-by-step performance including calculations and interpretation.

They include preparation of slides, solutions, calibrators, controls and reagents; calibration procedures; the reportable range; control procedures; corrective action when calibration or controls fail; limitations of the method including interfering substances; and reference intervals.

Two entries are the ones a patient would care about most. "Imminently life-threatening test results, or panic or alert values." And the system for entering results in the patient record and reporting them, including the protocol for reporting those values.

The last is a contingency plan: "Description of the course of action to take if a test system becomes inoperable."

Manufacturer instructions may be used to satisfy most of the list, and anything the manufacturer does not supply the laboratory has to supply itself.

Approval is personal and dated. "Procedures and changes in procedures must be approved, signed, and dated by the current laboratory director before use." A copy of each procedure is kept with its dates of first use and discontinuance.

Conditions, labels and expiry

The laboratory selects its test systems, follows the manufacturer's instructions, and has to get results within its own stated performance specifications.

It also has to define the conditions essential to proper storage, reliable operation and result reporting, consistent with manufacturer instructions, and then monitor and document them.

Four are named where applicable: water quality, temperature, humidity, and "Protection of equipment and instruments from fluctuations and interruptions in electrical current" that would adversely affect results and reports.

Reagents, solutions, media, control and calibration materials carry labels showing identity and, where significant, titer, strength or concentration, storage requirements, preparation and expiration dates, and other information needed for proper use.

The prohibition is flat. Those materials "must not be used when they have exceeded their expiration date, have deteriorated, or are of substandard quality."

And one more, which quietly rules out a common shortcut: "Components of reagent kits of different lot numbers must not be interchanged unless otherwise specified by the manufacturer."

Maintenance, and the gate before patient testing

For unmodified manufacturer equipment, the laboratory performs and documents maintenance as the manufacturer defines it, at least as often as the manufacturer specifies.

Function checks follow the same rule, with a hard gate attached: "Function checks must be within the manufacturer's established limits before patient testing is conducted."

Where equipment was developed in-house, modified by the laboratory, or came without maintenance and function check protocols, the laboratory has to write its own.

That means establishing a maintenance protocol that ensures the performance needed for accurate and reliable results, performing it, and documenting it.

And defining a function check protocol, performing the checks including background or baseline checks, and documenting them. The same gate applies: they have to be within the laboratory's established limits before patient testing.

Calibration verification, and the six-month floor

Calibration and calibration verification exist to substantiate the continued accuracy of a test system across the whole range of results it reports.

Calibration follows the manufacturer's instructions, using the specified materials, at least as often as recommended, and again whenever calibration verification fails the laboratory's limits.

Where the laboratory sets its own criteria, calibration materials should be appropriate to the system and, if possible, traceable to a reference method or material of known value.

Verification has a shape as well as a schedule. It must include at least a minimal or zero value, a mid-point value, and a maximum value near the upper limit of the range.

The schedule is a floor plus triggers. "At least once every 6 months", and also whenever a complete change of reagents is introduced, unless the laboratory can show lot changes do not affect the reporting range or control values.

Also on major preventive maintenance or replacement of critical parts that may influence performance. Also when control materials show an unusual trend or shift, or fall outside acceptable limits, and other means of correcting them fail. Also where the laboratory's own schedule for verifying the reportable range demands it sooner.

Known samples, every day, run like yours

Control procedures are the part a patient never sees and would most want to know about.

The laboratory is responsible for having control procedures that monitor the accuracy and precision of the complete analytic process, and sets the number, type and frequency of control materials.

What those procedures have to accomplish is stated in two parts. They must "Detect immediate errors that occur due to test system failure, adverse environmental conditions, and operator performance."

And they must monitor accuracy and precision over time, as test system performance, environmental conditions and operator performance vary.

Then the frequency rule, which is the sentence worth carrying away: "At least once each day patient specimens are assayed or examined", the laboratory runs controls.

The composition depends on the test. Two control materials of different concentrations for a quantitative procedure. A negative and a positive for a qualitative one. A negative plus a graded or titered control for graded results.

A system with an extraction phase runs two controls, one able to detect extraction errors. A molecular amplification procedure runs two, plus one able to detect inhibition where inhibition is a significant source of false negatives.

The rules that stop controls being theater

Several provisions exist to prevent controls from being run in a way that could not fail.

Controls are repeated before patient testing resumes after a complete reagent change, major preventive maintenance, or replacement of a critical part.

Operators are not allowed to be a hidden variable. "Over time, rotate control material testing among all operators who perform the test."

Handling is not allowed to differ either. "Test control materials in the same manner as patient specimens."

Where a calibration material is used as a control, it has to come from a different lot than the one used to set a cut-off or calibrate the system.

Acceptability criteria have to be established or verified, with statistical parameters defined and available for each batch and lot of quantitative control material. A commercially assayed value may be used only if it is for the laboratory's own method and instrument, and the laboratory verifies it.

Reagents, media and supplies get their own checks. Each batch, lot and shipment is checked on preparation or opening. Staining materials are checked each day of use, and fluorescent and immunohistochemical stains each time of use. Media are checked for sterility and for supporting growth before or with first use.

The gate on release, and the fallback when no control exists

Two sentences close the control section, and both are the kind a reader can use.

The first is a release gate: "Results of control materials must meet the laboratory's and, as applicable, the manufacturer's test system criteria for acceptability before reporting patient test results."

So a passing control is a precondition of your number being sent out, not a parallel record.

All control procedures performed have to be documented.

The second covers the case people assume is a gap. "If control materials are not available, the laboratory must have an alternative mechanism to detect immediate errors and monitor test system performance over time." That alternative's performance is documented too.

There is no exemption for a test with no commercial control material. There is a requirement to invent an equivalent.

What a failure obliges, looking backwards

Corrective action policies must be available and followed as necessary to keep the laboratory operating so that results are accurate and reliable.

All corrective actions are documented, and three situations are named.

The first is a test system not meeting performance specifications. That covers equipment or methods operating outside established parameters, and patient values outside the reportable range. It also covers the case "When the laboratory determines that the reference intervals (normal values) for a test procedure are inappropriate for the laboratory's patient population."

The second is the one with the widest consequences. It applies where controls or calibration materials fail the laboratory's acceptability criteria. Then "All patient test results obtained in the unacceptable test run and since the last acceptable test run must be evaluated to determine if patient test results have been adversely affected."

That is a look-back duty. A failed control does not only stop today's testing, it reopens everything since the last good one.

The third is a failure to meet the storage criteria for reagents and specimens.

The records, and what they let someone reconstruct

The test records rule requires an information or record system with positive identification of the specimen, the date and time it was received, and one item people rarely think to ask about.

"The condition and disposition of specimens that do not meet the laboratory's criteria for specimen acceptability."

A rejected specimen leaves a record. That is how a missing result can later be explained rather than merely noticed.

Above all of it sits a monitoring duty. The laboratory monitors and evaluates the overall quality of its analytic systems, and corrects identified problems, for each specialty and subspecialty it performs.

What a reader can do with this

Three questions become askable of a laboratory once these rules are known.

Whether controls were within acceptable limits on the day and run your specimen was tested. The regulation makes that a precondition of releasing your result and requires it to be documented.

Whether any corrective action covering that period was documented, which is the look-back that a failed control triggers.

And whether the reference interval on your report is one the laboratory established or verified for its own patient population, since an inappropriate interval is itself a listed trigger for corrective action.

One limit on this article. Only these sections of the analytic systems standards were read in full. The specialty-specific standards that sit between them, the performance specification section, and the pre-analytic and post-analytic requirements were not read for this piece.

These are federal minimums. A given laboratory may do more, and an accrediting organization may require more, and neither is described here.

Sources

Each document below is named as it names itself, with the date printed on that document rather than the day it was read.

1. Title 42 Code of Federal Regulations section 493.1250, Condition: Analytic systems, read in full for the scope of the standards and the monitoring duty — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
2. Title 42 Code of Federal Regulations section 493.1251, Standard: Procedure manual, read in full for the fourteen required contents, the textbook limitation and the director signature requirement — Electronic Code of Federal Regulations, Office of the Federal Register, August 2003
3. Title 42 Code of Federal Regulations section 493.1252, Standard: Test systems, equipment, instruments, reagents, materials, and supplies, read in full for the environmental conditions, labeling, expiry prohibition and the reagent kit lot rule — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
4. Title 42 Code of Federal Regulations section 493.1254, Standard: Maintenance and function checks, read in full for both the manufacturer and in-house paths and the pre-testing limits gate — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
5. Title 42 Code of Federal Regulations section 493.1255, Standard: Calibration and calibration verification procedures, read in full for the six month floor, the four triggers and the required value points — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
6. Title 42 Code of Federal Regulations section 493.1256, Standard: Control procedures, read in full for the daily frequency, the composition by test type, operator rotation, identical handling, the release gate and the no-control-available fallback — Electronic Code of Federal Regulations, Office of the Federal Register, August 2003
7. Title 42 Code of Federal Regulations section 493.1282, Standard: Corrective actions, read in full for the three named situations and the look-back to the last acceptable run — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
8. Title 42 Code of Federal Regulations section 493.1283, Standard: Test records, read for the record system contents including rejected specimen disposition — Electronic Code of Federal Regulations, Office of the Federal Register, January 2003
9. Title 42 Code of Federal Regulations section 493.1289, Standard: Analytic systems quality assessment, read for the requirement to monitor and evaluate overall analytic system quality — Electronic Code of Federal Regulations, Office of the Federal Register, August 2003