Using routine laboratory results to establish valid thyroid reference values

Aas Finn Erik; Oskar Kelp; Magnus Husoy; Thorsby Per Medboe; Alexander Westbye

doi:10.1530/endoabs.99.P577

P577

Prev Next

Section Contents Cite

Endocrine Abstracts (2024) 99 P577 | DOI: 10.1530/endoabs.99.P577

ECE2024 Poster Presentations Thyroid (58 abstracts)

Using routine laboratory results to establish valid thyroid reference values

Finn Erik Aas ¹ , Oskar Kelp ¹ , Magnus Husøy ² , Per Medboe Thorsby ¹ & Alexander Westbye ¹

Views

Author affiliations

¹Hormone Laboratory, Dep. og Medical Biochemistry, Oslo University Hopital, Oslo, Norway; ²Lillehammer Hospital, Lillehammer, Norway

Background: Indirect methods is developed to estimate reference limits from patient data, i.e. data from a mix of sick and healthy people. The indirect methods use a large number of results. A prerequisite for the methods to work satisfactorily is that the proportion of patients is not too large (typically < 20-30%).

Methods: We used direct and indirect methods for estimating reference limits for TSH, FT4, FT3, total T4 and total T3. All analysis were done with Abbot Alinity. We used was refineR and RefLim in addition to the simple non-parametric percentile method. All patients who have had thyroid parameters analyzed at the Hormone Laboratory in Oslo from December 2021 to June 2023 was included (n=21 539).

Results: We analyzed the thyroid parameters by direct method form 140 healthy participants from the population based HUNT4 study in Norway. In the indirect method we first used our entire database (Hlu, n=21 539). Then a selection were we excluded any sample form hospitals, or were TSH<0.1mIU/l and >10, or positive on anti-TPO or TRAb and only the first measurement of a patient were included (HLc, n=3 366). To validate we included patient sample for the county of Innlandet, Norway, only analyzing samples received from GP`s and only one sample from each patient (IH, n=49 433).

Anthropometric data and response to GH therapy in short children with low vs normal IGF1 level.
Analyte	LRL / URL (2.5 / 97.5 percentiles)	Direct method					Indirect methods
		HUNT4		refineR			RefLim			Non-parametric
			IH	HLc	HLu	IH	HLc	HLu	IH	HLc	HLu
TSH mIU/l	LRL	0.44	0.45	0.30	0.33	0.48	0.54	0.40	0.39	0.31	0.01
	URL	3.04	3.0	2.7	4.2	3.3	3.6	4.3	3.3	4.1	5.6
FT4 pmol/l	LRL	9.8	9.7	9.8	9.2	9.5	9.3	9.5	9.7	9.4	9.1
	URL	14	14.7	15.0	15.5	14.3	14.8	16.2	15.1	15.6	21.0
FT3 pmol/l	LRL	3.5	3.4	3.2	3.1	3.4	3.2	3.2	3.3	3.1	2.9
	URL	5.3	5.5	5.3	5.3	5.5	5.1	5.2	5.7	5.3	10
TT4 nmol/l	LRL	66	-	67	64	-	73	65	-	53	53
	URL	117	-	128	146	-	128	145	-	166	172
TT3 nmol/l	LRL	1.1	-	1.1	1.0	-	1.0	0.95	-	0.97	0.81
	URL	2.1	-	1.9	1.9	-	2.2	2.1	-	3.2	3.2

Conclusion: In our Norwegian population it seems like the direct method using healthy individuals form a population based study, is not superior to using indirect methods form our own routine laboratory database.