Define troubleshooting
Troubleshooting is a systematic approach to solving problems quickly and efficiently
What are the 7 steps for troubleshooting?
- Recognize the error (QC values are out of control?)
- Determine the type of error (systematic or random error)
- Relate the type of error to potential causes
- Relate potential causes to recent changes
- Resolve the issue
- Evaluate the resolution to make sure it is working
- Document all steps in the investigation, resolution and evaluation
For quantitative methods, what rules/ criteria will be normally used? How do we know if a quantitative method failed?
Rules/ criteria normally used for quantitative method is multi-rule QC / Westgard rules. If a quantitative method fails, a control value will fall outside of the establish control range
For qualitative method, what rules/ criteria is used? What happens if a test fails in a qualitative method?
For qualitative method, product inserts/ QC organisms are used and expected results assume to occur (a certain colour will appear, presence of agglutination, or expected reactions of organisms). If a qualitative test fails, the expected results (like a colour) will fail to show.
What type of error occurs more random or systematic error?
Systematic error occurs more
True or false, random and systematic error occur for the same reasons?
False, RE and SE occurs for different reasons
What kind of problems are systematic error normally related to?
Systematic error is normally related to reagent or calibration problems
Why is random error hard to identify and resolve?
Because random error cannot be predicted or quantified like systematic errors
If a random error occurs, what are the 3 steps to fix the RE?
- Watch the instrument in operation
- If no problem then re run QC
- If QC passes then run patient sample 10 times to see imprecision problems
For multi-analyte instruments, if several tests are impacted, what do we have to consider to resolve the problem?
We have to consider factors common to all of the tests
For multi analyte test, if only one test is affected, then what do we have to consider?
We have to consider factors that are unique to that method
When troubleshooting, why is it not a good idea to just recalibrate and rerun controls?
Because it will mask the real problem and can cause inaccurate shift in control ranges
When troubleshooting, what is important to remember?
To only change one thing at a time and to keep track/ a document of all the changes made
What are the 4 things to document when troubleshooting?
- The problem/ what happened
- The error type / recent changes
- The variables that were changed during troubleshooting and whether if it worked or not
- The outcome/ results - even if it didn’t work
