Troubleshooting / Common Issues
Study fails — Upper Confidence Limit exceeds the Error Limit
Possible causes
- –Instrument sample probe not washing adequately between specimens
- –Residue in sample tubing, cuvettes, or flow path from the high-concentration pool
- –High-concentration pool was too high relative to the low pool — the signal contrast amplifies even small absolute carryover
- –Instrument requires preventive maintenance — worn seals, degraded tubing, or contaminated probe
Recommended action
First, verify the raw data for entry errors. If the data are correct, run the instrument's internal probe wash or decontamination cycle if available, then repeat the carryover study. If the study fails again, escalate to instrument maintenance — inspect the sample probe, tubing, and cuvette wash system. Review the instrument's service log for recent maintenance history. Document all findings and corrective actions before returning the instrument to routine use.
Low (Control) SD is unusually high
Possible causes
- –The low-concentration pool is near the instrument's detection limit, where measurement noise is proportionally larger
- –Pipetting or sample handling inconsistency during the baseline measurements
- –Sample material (matrix) instability — the pool may have degraded between measurements
Recommended action
High control SD directly inflates the Auto error limit (3 × SD) and widens the UCL. Review the three baseline values individually — if one is an outlier relative to the other two, investigate whether a sample handling error occurred for that measurement. If the pool is near the detection limit, consider using a higher low-concentration pool that is well within the instrument's quantitative range.
Carryover Mean is near zero but UCL still fails
Possible causes
- –High variability in the Low (Test) group produces a wide confidence interval that extends above the error limit even when the point estimate is small
- –Very tight Auto error limit (3 × a small control SD) that does not accommodate normal measurement variability in the test group
Recommended action
Review the Low (Test) results individually. If two or three test specimens are substantially elevated while others are close to baseline, the variability is real and carryover may be occurring inconsistently — which itself warrants investigation. If all test values are close to the control mean, consider whether the Auto error limit is appropriate or whether an absolute/percent-based limit better reflects the clinical acceptance criterion for this analyte.
Study sequence was disrupted or a specimen result is missing
Possible causes
- –Instrument error voided one result mid-sequence
- –Operator ran specimens out of order
- –A specimen was inadvertently re-run, displacing the sequence
Recommended action
The 21-specimen sequence must be complete and in order. Partial or reordered sequences cannot be analyzed correctly because the statistical model depends on the specific positions of Low specimens relative to preceding High specimens. If the sequence was interrupted, the study must be repeated from the beginning with freshly prepared specimens in the correct order.