Changing Reference Values in the Clinical Laboratory ... Any experience?

[Torsed]

---

Members don't see this ad.

I have a clinical issue I would like to hear other's honest opinion. I have been dealing with our endocrinologist that wants us to raise our free T4 level upper limit of normal as he feels he is getting too many "abnormal" values referred to him by primary care and he doesn't seem (in my opinion) to want these one stop referrals.

Our current instrument has a free t4 of 0.6 to 1.12. (BC Unicel DxC 600 I). The "other" inpatient labs in town have changed their reference ranges (one even did a 1 month study with abnormal and normal free T4 to get their ranges) their range went from ours of 1.12 to 1.6 based upon this (I don't think normal reference ranges should include abnormal values in the validation studies but who listens to me). He brought me example labs reports to show me their reference ranges.

I've reviewed many articles and cannot find anything solid addressing endocrinologists wishing to change reference ranges. One unifying thought in the literature is that reference ranges may be all over the place for different instruments and clinical judgment needs to come into play (I'll leave my own thoughts of this clinician's judgment out). Our lab has done routine validation studies on this instrument to make sure it jives with the manufacturer's ranges. It is up to par on all proficiency testing that we do. The CLSI has recommendations for validating reference ranges but I feel these are too time and cost consuming to do again given we have already done validation of the instrument and PT. I cannot fathom changing these reference ranges at the whim of a single clinician, but I'm trying to give this clinician as scientific explanation of the why's and wherefores as I can.

Would like to hear your thoughts and wondering if any of you have ever bumped into a related issue.

 

[mlw03]

I'm sorry, we only discuss the pathology job market on here. Discussions related to the actual practice of pathology are not welcome here.

Just kidding... thanks for starting a real thread. I have nothing to contribute, as an AP/FP guy, but hopefully you have a real discussion about medicine.

 

[BamaAlum]

I know that you have validated your assay, but have you conducted a normal range study? Your reference range should reflect your patient population. You can't rely solely on the manufacturer's reference range and I surely wouldnt change based on the request of one clinician. We change reference ranges periodically after any significant change to methodology. Pretty simple to do.

 

[BU Pathology]

We did studies when our new lab opened to establish our own reference ranges.

 

[icpshootyz]

We just went through new reference range studies to validate our new lab machines. Basically asked anyone in the hospital to "donate" some tubes of blood to figure out the appropriate ranges, had to change a couple from the manufacturer's. Maybe do a normal range study using your hospital staff and see what if anything changes.

 

[Torsed]

Last time we tried a normal range study here we (the lab) could only get 6 people in the entire hospital (700 employees) to participate. Unfortunate.

 

[flindophile]

This is not an unusual scenario. It is surprising but calibration problems are often first identified by clinicians. In our laboratory, we started getting calls from clinicians that they suspected our measurements for a specific analyte might be off because they were seeing more positive cases than usual. We contacted the supplier of the calibrators and, indeed, other labs had made similar inquiries and it turned out that the supplier had changed lots of calibrator and there was an issue with the new calibrator.

You might try plotting the percentage of abnormal results as a function of time. That may signal a change. Also, there is a difference between a reference limit and a decision limit.

Sorry for this interruption. We now return to our regularly scheduled programming on the job market.

I have a clinical issue I would like to hear other's honest opinion. I have been dealing with our endocrinologist that wants us to raise our free T4 level upper limit of normal as he feels he is getting too many "abnormal" values referred to him by primary care and he doesn't seem (in my opinion) to want these one stop referrals.

Our current instrument has a free t4 of 0.6 to 1.12. (BC Unicel DxC 600 I). The "other" inpatient labs in town have changed their reference ranges (one even did a 1 month study with abnormal and normal free T4 to get their ranges) their range went from ours of 1.12 to 1.6 based upon this (I don't think normal reference ranges should include abnormal values in the validation studies but who listens to me). He brought me example labs reports to show me their reference ranges.

I've reviewed many articles and cannot find anything solid addressing endocrinologists wishing to change reference ranges. One unifying thought in the literature is that reference ranges may be all over the place for different instruments and clinical judgment needs to come into play (I'll leave my own thoughts of this clinician's judgment out). Our lab has done routine validation studies on this instrument to make sure it jives with the manufacturer's ranges. It is up to par on all proficiency testing that we do. The CLSI has recommendations for validating reference ranges but I feel these are too time and cost consuming to do again given we have already done validation of the instrument and PT. I cannot fathom changing these reference ranges at the whim of a single clinician, but I'm trying to give this clinician as scientific explanation of the why's and wherefores as I can.

Would like to hear your thoughts and wondering if any of you have ever bumped into a related issue.

 

[icpshootyz]

Last time we tried a normal range study here we (the lab) could only get 6 people in the entire hospital (700 employees) to participate. Unfortunate.

Did everyone in the lab at least contribute? We started with all the lab employees and then spread out to the rest of the hospital. Usually easier to browbeat the lab staff first since they're right there.