Biological monitoring protocol

You should select a laboratory that follows the nonmandatory protocol and participates in recognized proficiency-testing programs for cadmium analyses. OSHA’s Appendix F recommends using a “participating laboratory” that (1) takes part in an ongoing proficiency-testing program (for CDB, CDU, and B2MU), (2) provides QA/QC documentation on request, and (3) sends proficiency results to the physician responsible for medical monitoring (1910.1027AppF).

• Ask the lab whether it participates in the Centre de Toxicologie du Quebec (CTQ) interlaboratory program (preferred for CDB, CDU, B2MU) or an equivalent program.
• For creatinine in urine (CRTU), request CAP proficiency or CAP accreditation as recommended in Appendix F.
• Require the laboratory to provide proficiency reports and internal QA/QC summaries to the physician responsible for monitoring.

OSHA’s Appendix F recommends that laboratories analyze CDB, CDU, and B2MU through the Centre de Toxicologie du Quebec (CTQ) interlaboratory program or an equivalent proficiency-testing program. The protocol explains that CTQ currently sends 18 samples per analyte annually and compares returned results to consensus means to evaluate laboratory performance (1910.1027AppF).

• If a U.S. lab does not participate in CTQ, verify that it participates in an equivalent program covering the same analytes and sample frequency.
• Confirm the lab returns results within the CTQ-required time window (about four to five weeks) if it participates in CTQ.

Yes — CRTU must be analyzed with CDU and B2MU so results can be normalized, and Appendix F recommends participating laboratories obtain proficiency testing or accreditation for CRTU. In the U.S., the protocol notes that the College of American Pathologists (CAP) conducts CRTU proficiency testing and recommends CAP accreditation for CRTU analysis (1910.1027AppF).

• Request the lab’s CAP accreditation or proficiency reports for creatinine.
• If CRTU is subcontracted, Appendix F says the contract lab should provide QA/QC documentation to the physician.

A participating laboratory should provide internal QA/QC documentation for each analytic procedure (CDB, CDU, B2MU) and CRTU when requested by the physician. Appendix F specifies that QA/QC data and proficiency results should be provided in standard formats so the physician can compare laboratory performance and interpret worker monitoring results (1910.1027AppF).

• QA/QC records should include control charts, standard values used for calibration, recovery (F/T) data, limits of detection, and run-to-run precision.
• If CRTU is performed by a contract lab, ask that contract lab to send the CRTU QA/QC directly to the physician.

Under Appendix F, the CTQ interlaboratory program sends 18 samples annually for each analyte (CDB, CDU, and/or B2MU), and participating laboratories must return results within about four to five weeks of receiving the samples (1910.1027AppF).

• Confirm with the lab that it complies with that schedule or an equivalent program timeline.
• Ensure the physician receives the proficiency evaluations sent by the proficiency-testing body.

Control samples are quality-control specimens whose values are unknown to the analyst but known to QA/QC personnel; Appendix F requires QC samples to be run in sets containing one sample from each predefined pool for the analyte (1910.1027AppF).

• Labs should set up pools with at least four theoretical values across the analyte range.
• Run a QC set that includes one QC sample from each pool during analytic runs and chart results on control charts with predefined control limits.

Target values are specific concentrations that trigger medical actions in the cadmium surveillance program; Appendix F lists the target values and notes they may change over time. The target values are CDB: 5, 10, and 15 µg/L; CDU: 3, 7, and 15 µg/g CRTU; and B2MU: 300, 750, and 1500 µg/g CRTU (1910.1027AppF).

• Use these target values as triggers for the physician to consider job-placement, exposure control, or medical removal actions, per the regulatory medical surveillance requirements in 1910.1027(l).
• Document any change in laboratory methods or QA/QC that might affect an employee’s result interpretation.

If a lab’s proficiency reports show poor performance, the physician responsible for monitoring should evaluate whether the lab’s data are reliable for medical decisions and consider switching to a better-performing laboratory. Appendix F states that physicians may collate proficiency and QA/QC information to compare laboratories and to help decide whether results can be used to support continued work or medical removal (1910.1027AppF).

• Request corrective-action documentation from the lab and re-review subsequent proficiency results before relying on its data.
• If doubts remain, obtain repeat samples analyzed by a different accredited/proficient lab to confirm results.

No — Appendix F is a nonmandatory protocol that provides guidance to improve analysis quality, but it does not create mandatory requirements by itself. OSHA calls it a nonmandatory protocol intended to supplement the cadmium rule and help characterize and maintain data quality (1910.1027AppF).

• Even though Appendix F is nonmandatory, compliance and medical decisions must still meet the regulatory requirements in 1910.1027.
• Employers and physicians should treat Appendix F as best-practice guidance when choosing labs and evaluating results.

Appendix F includes specifications and recommendations for analytical methods to ensure accurate CDB, CDU, and B2MU measurements, including use of appropriate standards, calibration across the expected result range, and limits of detection near the lowest standard (1910.1027AppF).

• Use standards of at least three different values constructed from at least two sources; the lowest standard should be near the limit of detection and the highest standard above expected maximum compliance or QC sample values.
• Document method detection limits, calibration procedures, and any matrix effects that could bias results.

Appendix F specifies that CDU and B2MU must be normalized to creatinine concentration and reported as µg analyte per g creatinine (µg/g CRTU) so results account for urine dilution. Therefore the laboratory should analyze CRTU alongside CDU and B2MU and report normalized values (1910.1027AppF).

• Ask the lab to provide both the raw analyte concentration (e.g., µg/L) and the creatinine-normalized value (µg/g CRTU).
• Request CRTU QA/QC and proficiency documentation, especially if CRTU is subcontracted.

Proficiency is the laboratory’s ability to meet a specified level of analyte performance and is assessed by comparing the lab’s results for proficiency-test samples against consensus or theoretical values derived from the interlaboratory program. Appendix F explains that proficiency is a function of agreement between the lab’s reported results and the consensus means for distributed samples (1910.1027AppF).

• Look for statistical measures in proficiency reports (e.g., bias, coefficient of variation, z-scores) that indicate performance across the range of target values.
• A physician should use those measures plus the lab’s internal QA/QC to judge reliability for medical decisions.

The physician responsible for biological monitoring must evaluate laboratory performance by reviewing proficiency-test reports and QA/QC documentation to determine if the lab’s results are reliable for making medical decisions, including continued work or removal actions (1910.1027AppF).

• Physicians should request proficiency results be sent directly to them and should collate those with the lab’s QA/QC records.
• If a physician finds a lab’s results unreliable, they should request reanalysis or change to another participating lab before making medical removal decisions per 1910.1027(l).

Laboratories should provide biological specimen measurements together with QA/QC information and proficiency results in standard formats so physicians can collate and interpret results. Appendix F emphasizes that both the analytical values and QA/QC/proficiency documentation be provided to the responsible physician (1910.1027AppF).

• Standard reports should include: specimen ID, analyte values (raw and creatinine-normalized), detection limits, calibration standards used, control-chart summaries, and proficiency-test summaries.
• Ensure the lab uses consistent units (e.g., µg/L, µg/g CRTU) and clearly labels target values that trigger action.

If CRTU analysis is performed by a contract laboratory, Appendix F requires that the contract lab provide QA/QC documentation for CRTU directly to the requesting physician, and the participating laboratory should make this information available on request (1910.1027AppF).

• Verify chain-of-custody and that both labs’ QA/QC and proficiency documents are provided to the physician.
• The primary participating laboratory should transparently identify subcontractors and their accreditation/proficiency status.

Appendix F recommends establishing pools of QC samples for each target value with at least four theoretical values per pool and running QC samples in sets that include one QC sample from each pool during analytic runs (1910.1027AppF).

• Pools should cover the analytical range, including low near LOD and high values above expected maximum compliance samples.
• Track QC results on control charts and define control limits to detect bias or loss of precision.

Appendix F advises using at least three different calibration standards constructed from at least two different sources, with the lowest standard near the limit of detection and the highest exceeding the expected maximum compliance or QC sample values (1910.1027AppF).

• Maintain documents showing concentrations and traceability of each standard.
• Recalibrate instruments and verify performance with known standards during runs, especially when measuring near target action levels.

If a lab reports results below the limit of detection (LOD), employers or physicians should ask the lab to provide the laboratory’s defined LOD, the method used to calculate it, and how nondetects are reported or handled, because Appendix F treats the limit of detection as an important component of sensitivity (1910.1027AppF).

• Request the raw measurement value, the LOD, and whether the lab reports a numerical estimate (e.g., <LOD) or substitutes a value for statistical purposes.
• If results are near target values, consider reanalyzing the sample at a lab with lower LOD or confirmatory methods.

A physician should combine proficiency-test results, internal QA/QC (control charts, bias, precision), and the lab’s LOD to judge if a borderline or near-target result is reliable for action. Appendix F states physicians may collate proficiency and QA/QC information to evaluate worker monitoring data and support decisions about continued work or removal (1910.1027AppF).

• If QA/QC shows recent bias or high variability, treat borderline results with caution and consider repeat testing.
• Document the physician’s rationale for any medical removal or continued assignment when relying on borderline lab data.

Appendix F describes procedures for ensuring analytic quality but refers back to the cadmium rule for the timing of sample collection; the medical monitoring program requires blood and urine samples be collected at defined intervals by the physician responsible for monitoring (1910.1027AppF).

• For the specific sample-collection intervals and who is eligible, follow the medical surveillance schedule in the regulatory text of 1910.1027(l).
• Appendix F’s role is to ensure that once samples are collected, the laboratory measurements are reliable for interpreting those periodic checks.

If a lab changes analytic methods, the physician and employer should require the lab to provide method validation data, comparative results to the prior method, and updated QA/QC/proficiency documentation so the physician can assess continuity of results; Appendix F emphasizes that changes affecting data quality should be documented and supplied to the responsible physician (1910.1027AppF).

• Request a method-comparison study (split-sample testing) showing equivalence or documented bias and how it will be corrected.
• Until equivalency is demonstrated, treat new-method results carefully when making medical decisions tied to established target values.

Appendix F notes that collection and handling affect data quality and recommends clear documentation of collection time, handling, transport conditions, and chain-of-custody so physicians can interpret lab results reliably (1910.1027AppF).

• Keep records showing who collected samples, date/time, storage temperature, shipment conditions, and any deviations from protocol.
• Provide those records with analytical results and QA/QC data to the physician reviewing the monitoring program.

The recommended DQOs for cadmium in blood are a detection limit of 0.5 µg/l and accuracy of ±1 µg/l or ±15% of the mean, with precision requirements that depend on concentration. Specifically, Table 1 in 1910.1027AppF lists a limit of detection of 0.5 µg/l; for results ≤2 µg/l a precision (CV) of 40% is acceptable, and for results >2 µg/l a precision (CV) of 20% is expected.

The recommended DQOs for cadmium in urine are a detection limit of 0.5 µg/g creatinine and accuracy of ±1 µg/l or ±15% of the mean, with precision that varies by concentration. Table 1 in 1910.1027AppF specifies the detection limit of 0.5 µg/g creatinine; for urine results ≤2 µg/l creatinine a precision (CV) of 40% is acceptable, and for results >2 µg/l creatinine a precision (CV) of 20% is expected.

The recommended DQO for urinary β2-microglobulin (B2MU) is a detection limit of 100 µg/g creatinine, precision of about 5% CV, and accuracy of ±15% of the mean. Table 1 in 1910.1027AppF lists these values for B2MU (100 µg/g creatinine detection limit, 5% precision, ±15% accuracy).

OSHA recommends specific analytic methods: the Stoeppler and Brandt (1980) method for cadmium in blood (CDB), the Pruszkowska et al. (1983) method for cadmium in urine (CDU), and the Pharmacia Delphia test kit for urinary β2-microglobulin (B2MU). These recommended methods and their detection limits are listed in 1910.1027AppF.

The protocol requires an independent determination of urine creatinine (CRTU) so CDU and B2MU can be reported normalized to creatinine (µg/g CRTU). 1910.1027AppF specifies either the OSHA Salt Lake City Technical Center (OSLTC) method or the Jaffe method may be used for CRTU, and states that CDU and B2MU reporting (except for CTQ proficiency samples) should be accompanied by CRTU results and combined to give µg/g CRTU.

Yes — laboratories may use alternate methods, but they must submit data and a QA/QC plan to the responsible physician showing the alternate method meets the program's DQOs and demonstrates equivalent QA/QC performance. The protocol in 1910.1027AppF requires sufficient method performance data and an internal QA/QC plan for acceptance.

To show they have performed regular analyses for an analyte, a laboratory must provide at least 8 consecutive quarters of documentation such as laboratory reports with results, signed contracts for regular analyses, or invoices to clients requesting payment for the analyses; each document must identify the specific analytic procedures used. These documentation options are described in 1910.1027AppF.

A participating laboratory must submit its internal QA/QC plan that details standard operating procedures for each analyte and must provide the results of its internal QA/QC monitoring program. 1910.1027AppF specifies submission of QA/QC plans and ongoing QA/QC records to the responsible physician.

Laboratories should enroll in the CTQ interlaboratory proficiency program for each analyte (CDB, CDU and/or B2MU) and, if performing CRTU for CDU or B2MU, should be accredited by the College of American Pathologists (CAP) and participate in the CAP CRTU survey. These proficiency and accreditation expectations are stated in 1910.1027AppF.

The responsible physician should review a laboratory's status six months after initial selection and then repeat reviews every six months or whenever new proficiency or QA/QC documentation arrives. 1910.1027AppF explains this six-month initial review followed by periodic reassessments tied to proficiency testing rounds and QA/QC submissions.

During the first six months, a laboratory must satisfy performance requirements for at least 2 of the 3 proficiency samples in each of the three rounds (i.e., demonstrate acceptable results for at least 2 out of 3 samples in each round) to maintain competency for each analyte it analyzes. This initial-proficiency expectation is described in 1910.1027AppF.

After one year, the protocol adds a rolling performance criterion: a laboratory should not have failed performance requirements for more than 4 samples out of the 6 most recent consecutive rounds (18 discrete proficiency samples) for each analyte. This longer-term proficiency rule is set out in 1910.1027AppF.

For samples with a consensus mean less than 1 µg/l (low-level CDB or CDU), the acceptable performance interval is defined as a result between the method detection limit and a maximum of 2 µg/l; this allows laboratories to report measured values below 1 µg/l without being penalized. 1910.1027AppF explains this special interval for low consensus means.

The numeric performance requirements are: for CDB and CDU an analytical result within ±1 µg/l or ±15% of the consensus mean (whichever is greater), and for B2MU results within ±15% of the consensus mean. These performance criteria are specified in 1910.1027AppF.

Laboratories analyzing CDU or B2MU must demonstrate proficiency in creatinine (CRTU) analysis as defined by CAP; they should be CAP-accredited for CRTU and participate in the CAP survey, or they must contract with a laboratory that meets those CAP requirements. This requirement is stated in 1910.1027AppF.

The responsible physician must evaluate laboratories' proficiency and QA/QC information, review submitted proficiency and QA/QC data, and determine which laboratories meet the DQO and QA/QC recommendations before using them for biological monitoring. 1910.1027AppF directs physicians to review proficiency and internal QA/QC program results when choosing laboratories and to reassess labs periodically.

Laboratories enrolling in the CTQ program should submit an enrollment application with an initial fee (approximately $100 per analyte, subject to change) and indicate agreement to have proficiency results sent directly to designated physicians; CTQ will assign a code number and confirm analyte participation. This enrollment process and fee guidance are described in 1910.1027AppF.

Although specific detection limits are not given for the CRTU methods, the protocol notes that expected creatinine urine concentrations (approximately 0.9–1.7 g/l) are well above likely method detection limits for either the OSHA Salt Lake City Technical Center method or the Jaffe method. 1910.1027AppF discusses the CRTU methods and expected ranges.

When reporting CDU or B2MU results (other than CTQ proficiency samples), laboratories must also report CRTU results so the analyte can be expressed as µg/g CRTU; the two results should be combined to provide the normalized measure. This reporting requirement is stated in 1910.1027AppF.

A newly-formed laboratory must submit key documentation so the responsible physician can determine whether the lab meets the protocol guidelines. Section 3.2.3 of 1910.1027AppF explains the requirements.

• The lab must provide a written internal QA/QC plan describing the standard operating procedures it will use to meet QA/QC guidelines (see QA/QC content requirements in 1910.1027AppF).
• It must submit the results of its initial characterization analyses for each analyte it will measure.
• If the lab has no prior commercial analyses, it should supply results from a minimum of two mock standardization trials per analyte per month for a 6-month initial internal QA/QC period, and then submit those internal QA/QC results from that initial 6‑month period to the physician.
• The lab must enroll in the appropriate CTQ proficiency program and arrange for the CTQ program to send initial participation confirmation and proficiency test results directly to the designated physician; the physician should receive results from three completed CTQ rounds before approving the laboratory.

Reference: see 1910.1027AppF for the full selection criteria.

A laboratory must demonstrate acceptable proficiency across CTQ rounds before it can be approved to analyze biological monitoring samples. 1910.1027AppF Section 3.2.3–3.2.4 sets out these expectations.

• The lab should perform satisfactorily on at least 2 of the 3 proficiency samples in each of the three CTQ rounds over a 6‑month period for each analyte (CDB, CDU, or B2MU).
• The designated physician should receive results from three completed CTQ rounds before approving the lab to participate.
• Proficiency must be maintained for each analyte to continue participation; failing to meet the guidance may lead the physician to disapprove further analyses until the lab corrects deficiencies.

See 1910.1027AppF for details on rounds and reevaluation.

The numeric acceptance intervals are specified so that CTQ proficiency results can be judged consistently. 1910.1027AppF gives the performance requirements for each analyte.

• For CDB (cadmium in blood): the analytical result must be within ±1 µg/l or ±15% of the consensus mean, whichever is greater. For samples with consensus mean less than 1 µg/l, the acceptable result is between the method detection limit and a maximum of 2 µg/l.
• For CDU (cadmium in urine): the analytical result must be within ±1 µg/l urine or ±15% of the consensus mean, whichever is greater. For consensus means <1 µg/l urine, the acceptable result is between the detection limit and a maximum of 2 µg/l urine.
• For B2MU (beta-2-microglobulin): the analytical result must be within ±15% of the consensus mean.

These numeric criteria are described in 1910.1027AppF.

Yes — laboratories performing CDU or B2MU must also meet requirements for CRTU (creatinine in urine) analysis and demonstrate appropriate accreditation. 1910.1027AppF explains these requirements.

• The lab must provide documentation of accreditation by the College of American Pathologists (CAP) for CRTU analyses performed with CDU and/or B2MU determinations. If CRTU is run by a contract lab, that contract lab must provide proof of CAP accreditation.
• When reporting CDU or B2MU results (except CTQ proficiency testing), the laboratory must include the companion CRTU result and combine the two to express CDU or B2MU normalized to creatinine units (µg/g CRTU).

See 1910.1027AppF for the accreditation and reporting expectations.

The QA/QC plan must document procedures that keep analytic work under control and show how the lab will detect and correct problems. 1910.1027AppF Section 3.3.1 lists the required elements.

At a minimum the written plan should include:

• Sample acceptance and handling procedures (chain-of-custody).
• Sample preparation procedures and detailed instrument parameters.
• Calibration procedures and how calibrations will be verified.
• Calculations and how results are derived and recorded.
• Internal QC procedures (frequency, types of QC materials, blinding of QC levels to analysts).
• Control limits, how they were set (statistical or proficiency-based), and how charts or plots will be maintained.
• Defined criteria that will trigger suspension of analysis of compliance samples.
• Corrective-action procedures, including investigation, corrective steps, and reanalysis of affected compliance samples.

These elements are described in 1910.1027AppF to ensure consistent, auditable QA/QC.

Instruments must be calibrated daily on days when QC or compliance samples are run, and standards must be checked regularly during a run. 1910.1027AppF specifies calibration and standard-check procedures.

• Calibration should be performed at the start of each day that QC or compliance samples are analyzed.
• During a run, approximately every fifth sample should be a standard to ensure the calibration holds.
• Calibration is considered maintained if the standard value is within ±15% of its theoretical value; if a standard exceeds ±15% the run has exceeded control limits due to calibration error and the entire set of samples should be reanalyzed after recalibration or results recalculated from a valid standard curve.
• The highest standard analyzed should exceed the highest sample concentration; laboratories should refresh standards as they age and compare old standards to new ones.

See 1910.1027AppF for full guidance on calibration and standards.

Internal QC samples must be run frequently and control limits must be meaningful so the lab can detect bias or loss of control. 1910.1027AppF describes internal QC frequency and control-limit approaches.

• Internal QC samples should be interspersed with compliance samples at a minimum rate of 5% of compliance samples or at least one set of QC samples per analysis of compliance samples, whichever is greater.
• If only two QC samples are run in a batch, they should contain different levels of cadmium.
• QC materials may be commercial reference materials or well-characterized internally prepared materials; internally prepared QC must be traceable or compared to a consensus reference.
• Analysts should not know QC target values before reporting results; QC results should be plotted or charted to show recovery and control limits.
• Control limits may be statistical (for example ±2 standard deviations from the laboratory mean) or based on proficiency testing limits (e.g., ±1 µg or ±15% of the mean). Laboratory statistical limits that exceed ±40% should be reevaluated to find sources of error.

See 1910.1027AppF for additional detail on QC design and limits.

If control limits are exceeded, the laboratory must stop analytic work, find and correct the cause, and reanalyze affected compliance samples. 1910.1027AppF sets out corrective-action expectations.

• Analytic work should be suspended immediately until the source of the error is identified and corrected.
• Compliance samples affected by the error must be reanalyzed after corrective measures are implemented (for example, recalibration, instrument maintenance, or replacement of degraded standards).
• The QA/QC plan should describe how unusual trends (e.g., consecutive results biased high or low) are investigated and corrected.
• The lab should document corrective actions and the disposition of affected results; the responsible physician may disapprove further analyses by the lab until compliance is demonstrated.

Refer to 1910.1027AppF for the recommended corrective-action protocols and reporting expectations.

A laboratory run exceeds control limits when any control sample shows a measurement error greater than 15% or when the average coefficient of variation (CV) for duplicate samples is greater than 5%. This triggers the need for documented corrective actions and further checks to assure accurate data are reported to the responsible physician.

• What the limits mean: the 15% figure applies to the measurement error for one or more control samples (i.e., measured value versus target/reference value), and the 5% limit applies to the average CV calculated from duplicate analyses. See 1910.1027AppF.

• Required documentation and follow-up: the guidance says the QA/QC plan "should document in detail specific actions taken if control limits are exceeded or unusual trends develop," and corrective actions "should be noted on an appropriate form, accompanied by supporting documentation." Labs should also include any additional steps necessary to ensure accurate data are provided to the responsible physicians. See 1910.1027AppF and the overall cadmium standard at 1910.1027.

• Practical examples of steps to take (based on the protocol recommendations):

  • Review recent instrument calibrations and QC sample performance (keep records per the App F recordkeeping recommendations).
  • Repeat the affected run or re-analyze suspect samples if appropriate.
  • Compare accuracy and precision to available kit performance (the protocol recommends comparing to the Pharmacia Delphia kit benchmarks).
  • Review proficiency testing results (CTQ) and notify the responsible physician of findings and any accreditation-status changes promptly.

• Recordkeeping and reporting reminders: keep QC and calibration records as described in the guidance (e.g., tabulations of in/out-of-control determinations over the past 2 years and copies of data reports and proficiency reports provided to physicians). See the reporting and recordkeeping sections in 1910.1027AppF.