FOCUSED REPORTS

Evaluation of a New Generation Automated

Assay for 25-Hydroxy Vitamin D Based on
Competitive Protein Binding
Maryam Asif,1 Sarah E. Groboske,2 Edward K.Y. Leung,1,2† Kiang-Teck J. Yeo,1,2
and Xander M.R. van Wijk1,2*

Background: The interest for vitamin D has exponentially increased testing demand for 25-hydroxy vitamin
D [25(OH)D]. Consequently, many laboratories are switching from LC-MS/MS methods to automated, high-
throughput immunoassays. One of the major potential issues with these assays has been the lack of cross-
reactivity with 25(OH)D2.
Methods: We have evaluated the Roche Elecsys vitamin D total II assay for accuracy by comparing 79 patient
samples with LC-MS/MS. The cross-reactivity for 25(OH)D2 was evaluated by analyzing samples with high
25(OH)D2 separately and estimating 25(OH)D2 recovery, as well as by spiking of 25(OH)D2. The assay was further
evaluated for precision, linearity, sample type, and common interferences.
Results: There was mostly good agreement between the Elecsys and LC-MS/MS assays (Deming regression: y =
0.95x + 0.70), with an overall bias of 2.3% (−0.84 ng/mL). However, there were 6 out of 79 (7.6%) discordant
samples. The Deming regression for samples with high 25(OH)D2 compared to LC-MS/MS showed similar slope
and intercept (y = 0.97x − 1.1). The average recovery of 25(OH)D2 for these samples was 90%. The initial precision
studies were in general agreement with the package insert, but long-term clinical use showed higher-than-claimed
imprecision (11.7%–14.4% at 12 ng/mL and 6.9%–7.6% at 27 ng/mL; claimed: 7.2% and 5.0%, respectively). We
observed 1 falsely high result in plasma, an issue previously addressed by Roche in a medical device correction.
Conclusions: The analytical performance of the Roche Vitamin D assay was acceptable, and the assay had a good
cross-reactivity for 25(OH)D2.

IMPACT STATEMENT
Increased testing requests for vitamin D forces clinical laboratories to consider automated
immunoassay or competitive binding assay methods to meet the demands. This study will provide
laboratories with valuable information on the performance characteristics of a widely available
new-generation automated assay for total 25-hydroxy vitamin D by Roche Diagnostics. It addresses
several issues laboratories need to be aware of, including accuracy, with emphasis on cross-
reactivity with 25-hydroxy vitamin D2, assay precision, and false-high results.

1
Department of Pathology, Pritzker School of Medicine, The University of Chicago, Chicago, IL; 2Section of Clinical Chemistry, The University of
Chicago Medicine, Chicago, IL.
*Address correspondence to this author at: The University of Chicago Medicine & Biological Sciences, 5841 S. Maryland Ave., Rm. TW 010-B,
MC 0004, Chicago, IL 60637. Fax 773-702-6268; e-mail xvanwijk@bsd.uchicago.edu.

..............................................................................................................

September 2019 | 04:02 | 247–253 | JALM 247

FOCUSED REPORTS
Vitamin D has always been associated with its
role in bone and calcium homeostasis (1). Its pos-
sible involvement in nonskeletal diseases, espe-
cially cardiovascular or autoimmune diseases and
cancer (2), has led to new interest in this hormone
during the last decade or so. This interest has pro-
portionally increased the testing demand for vita-
min D to a point that it has now been ordered as
part of routine medical care and general screening
of healthy individuals. Although this practice is not
evidence based and is being negated by studies
including those by the US Preventive Services Task
Force (3–5), it persists. LC-MS/MS testing for 25-
hydroxy vitamin D [25(OH)D]3 is considered the
gold standard method; however, many laborato-
ries worldwide are switching to automated immu-
noassay or competitive binding assay methods to
meet the increased demand of vitamin D testing
(6–8). The advantage of these methods is their au-
tomated and time-effective format (9); however,
their main issue is that they generally do not rec-
ognize the D2 form of 25(OH)D with high recovery
(9–13). Although a recent letter reported the char-
acteristics of the Roche Elecsys® vitamin D total II
competitive binding assay (a new generation as-
say), the authors did not evaluate the cross-
reactivity with 25(OH)D2 [only one sample had de-
tectable 25(OH)D2] (14). We evaluated this new
assay and specifically its cross-reactivity with
25(OH)D2.
METHODS
Assay principle
The Elecsys vitamin D total II assay (Roche
Diagnostics) is based on a competition principle
and uses electrochemiluminescent detection.
†
Present affiliation: Department of Pathology and Laboratory Medicine, Children's Hospital Los Angeles, Keck School of Medicine, University of
Southern California, Los Angeles, CA.
DOI: 10.1373/jalm.2018.028415
© 2019 American Association for Clinical Chemistry
3
Nonstandard abbreviations: 25(OH)D, 25-hydroxy vitamin D; IQR, interquartile range.
...............................................................................................................
248 JALM | 247–253 | 04:02 | September 2019
Roche Elecsys Vitamin D Total II Assay Evaluation
After release of 25(OH)D from vitamin D– binding
protein by a pretreatment reagent, vitamin D–
binding protein labeled with a ruthenium complex
is used as capture protein to bind 25(OH)D. Cross-
reactivity to 24, 25-dihydroxyvitamin D is blocked
by a specific unlabeled antibody. Addition of
streptavidin-coated microparticles and 25(OH)D
labeled with biotin allows a complex to form be-
tween unoccupied ruthenylated vitamin D– bind-
ing protein and biotinylated 25(OH)D, which is
attached to the solid phase via the biotin–strepta-
vidin interaction. The microparticles are magneti-
cally captured onto the surface of the electrode
and application of a voltage to the electrode in-
duces chemiluminescent emission, which is mea-
sured. According to the manufacturer, the assay
has 91.4% and 112.8% normalized cross-reactivity
to 3-epi-25(OH) D2 and D3, respectively. We eval-
uated the assay on a Roche COBAS-8000 e602
module.
Accuracy and spike-recovery studies
The accuracy of the assay was evaluated by com-
paring 79 patient serum samples analyzed on
the Elecsys assay with an in-house laboratory– de-
veloped, clinically validated LC-MS/MS assay de-
scribed in the Data Supplement that accompanies
the online version of this article at http://www.jalm.
org/content/vol4/issue2. Recovery studies were
performed for both 25(OH)D2 and 25(OH)D3 with
certified reference materials from Cerilliant. Solu-
tions of 100 ng/mL 25(OH)D2 and 25(OH)D3 were
prepared in vitamin D-free serum (VD-DDC Mass
Spect Gold; Golden West Biologicals) and then
mixed with vitamin D-free serum to generate tar-
get concentrations of 10, 20, 40, 60, and 80 ng/mL.
The assay was run in duplicate and the percentage
Roche Elecsys Vitamin D Total II Assay Evaluation
cross-reactivity was calculated as the mean recov-
ery from all 5 concentrations.
Precision studies
Roche QC PreciControl vitamin D total II levels 1
and 2 were analyzed 20 times on the same day,
and twice daily for 5 days, to assess within-run and
between-day precision, respectively. After imple-
mentation, the between-day precision was also
evaluated in clinical operation over 72 days (6/19/
2018 – 8/29/2018), run at least twice daily with
PreciControl QCs. The 25(OH)D concentration at
which the CV is 10% was determined with 4 sets of
pooled patient serum analyzed once daily over 20
days. The samples used in experiments described
in this section were not tested by LC-MS/MS; there-
fore, the constitution of 25(OH)D2 and 25(OH)D3 is
not known.
Analytical measuring range (AMR), dilution,
and interference studies
The AMR was determined by intermixing 2 pa-
tient samples (9.9 and 89.2 ng/mL) at 5 different
ratios (100%:0%, 75%:25%, 50%:50%, 25%:75%,
and 0%:100%). The AMR claimed by the manufac-
turer is 5–100 ng/mL. Dilution studies were per-
formed in duplicate using 4 patient samples by
making 2× dilutions with vitamin D-free serum. In-
terference studies were performed as described
previously (15). The unspiked control pooled pa-
tient serum for the hemolysis, lipemia, and icterus
experiments had 25(OH)D concentrations of 28.7,
27.5, and 31.9 ng/mL, respectively. No significant
interference was defined as recovery between
90% and 110%. The samples used in experiments
described in this section were not tested by LC-
MS/MS; therefore, the constitution of 25(OH)D2
and 25(OH)D3 is not known.
Data analyses
Data analyses were performed with Prism 7
(GraphPad). Deming regression was used for
..............................................................................................................
September 2019 | 04:02 | 247–253 | JALM
FOCUSED REPORTS
analysis of method comparison and linear regres-
sion for spike-recovery and AMR studies. The
25(OH)D concentration at which the CV is 10%
was interpolated with a nonlinear one-phase de-
cay fit.
RESULTS AND DISCUSSION
Method comparison with LC-MS/MS showed
that the overall accuracy of the Elecsys assay was
acceptable, as indicated by the slope (0.95) and the
intercept (0.70) of the Deming regression line and
by the low mean bias in the Bland–Altman plots
(−0.84 ng/mL or 2.3%) (Fig. 1). For up to 10 ng/mL
of 25(OH)D, as determined by LC-MS/MS, only
positive differences were observed (n = 9, mean
difference 2.6 ng/mL), which could indicate overes-
timation by the Elecsys assay at these low concen-
trations. This finding may be especially important
when the Institute of Medicine guidelines for eval-
uation of vitamin D status are used, which suggest
that “persons are at risk of deficiency relative to
bone health at serum 25(OH)D levels of below 12
ng/mL” (16). Furthermore, 6 out of 79 (7.6%) sam-
ples showed discordance, as determined with the
same criteria as used by a recently published study
(12), that is, greater than ±5 ng/mL [25(OH)D < 20
ng/mL] or ±20% [25(OH)D > 20 ng/mL]. Tolan et al.
previously found that 38% of samples were
discordant with LC-MS/MS for the first-generation
Elecsys assay owing to underrecovery of 25(OH)D2
(12). Although seemingly improved over the first-
generation assay, a 7.6% discordance rate may
warrant further testing with LC-MS/MS in certain
cases. However, in only 2 out of 6 discordances
would interpretation of the patient's vitamin D sta-
tus have changed according to the Institute of
Medicine (16) or Endocrine Society (17) guidelines,
that is, 17 vs 24 ng/mL and 27 vs 18 ng/mL, by
LC-MS/MS and Elecsys, respectively. Of note, pa-
tient samples in this study were selected only on
the basis of 25(OH)D2 and 25(OH)D3 concentra-
tions. The accuracy of the assay was not specifically
249
FOCUSED REPORTS Roche Elecsys Vitamin D Total II Assay Evaluation

Fig. 1. Method comparison of the Elecsys vitamin D assay vs. LC-MS/MS.

Top: Scatter plot with Deming regression comparing the results of 79 patient serum samples by use of the Elecsys assay with
a laboratory-developed, clinically validated LC-MS/MS assay. Bottom: Bland–Altman plots comparing the absolute (left) and
percentage (right) difference of the 2 assays with LC-MS/MS. LoA, limit of agreement.

investigated for patient populations that have pre-
viously been shown to be possibly problematic for
automated assays, that is, pregnant women, dialy-
sis patients, and intensive care patients (18).
To assess the cross-reactivity of the assay for
25(OH)D2 and 25(OH)D3, we separately compared
the patient samples with high 25(OH)D2 [25(OH)D2 >
25(OH)D3] and samples with only 25(OH)D3 with
LC-MS/MS. The slopes and intercepts of the 2 Dem-
ing regression lines were similar, suggesting the
assay did not preferentially recognize 1 form of
25(OH)D over the other (Fig. 2). This finding contrasts
a significant underrecovery for 25(OH)D2 shown by
prior studies on the first-generation Elecsys vitamin
D assay (11–13). The cross-reactivity for 25(OH)D2
was also estimated by subtracting the 25(OH)D3
concentration obtained by LC-MS/MS from the
Elecsys result [assuming 100% reactivity for 25(OH)D3]
and dividing by the 25(OH)D2 concentration ob-
tained by LC-MS/MS, according to Tolan et al. (12).
The median and average calculated recovery for
the samples with high 25(OH)D2 was 90% (inter-
quartile range, 84%–102%), with 1 sample showing
a recovery of 46%. This result is significantly better

...............................................................................................................

250 JALM | 247–253 | 04:02 | September 2019

Roche Elecsys Vitamin D Total II Assay Evaluation FOCUSED REPORTS

Fig. 2. Method comparison of the Elecsys vitamin D assay vs. LC-MS/MS divided into samples with high
25(OH)D2 [25(OH)D2 > 25(OH)D3] (left) and samples with only 25(OH)D3 (right).
Scatter plot with Deming regression showed similar results for both sample sets.

than the previously reported median of 52% (inter-
quartile range, 49%– 60%) (12). Furthermore, spik-
ing experiments with 25(OH)D2 and 25(OH)D3
showed that the new generation assay recognizes
25(OH)D2 and 25(OH)D3 equally (see Fig. 1 in the
online Data Supplement). This experiment further
showed a tendency toward underrecovery at
the lower end of the assay and overrecovery at the
higher end, with close to 100% recovery in the
range of 20 – 40 ng/mL. This finding was unex-
pected given that this was not observed in the pa-
tient comparison with LC-MS/MS. The average
cross-reactivity was 95.2% for 25(OH)D2 and
102.8% for 25(OH)D3. The cross-reactivity for
25(OH)D2 normalized to 25(OH)D3 was 92.6%,
which is close to the package insert specification
(93.7%).
The within-run precision for both levels of Roche
QC showed a CV of 5.6% and 3.9% with a mean of
14.3 and 30.2 ng/mL, respectively. Similarly, the
between-day precision run on both low and high
concentrations showed a CV of 9.0% and 3.0%,
with a mean of 14.4 and 29.8 ng/mL, respectively.
These CVs were in general agreement with the
package insert claims. Over 72 days in clinical use,
the CVs of the level 1 QC were 11.7% (mean, 12.1
ng/mL) and 14.4% (mean, 11.8 ng/mL), for measur-
ing cell 1 and 2 on the e602 analyzer, respectively.
For the level 2 QC, this was 6.9% (mean, 27.1 ng/mL)
and 7.6% (mean, 27.0 ng/mL). This result was
higher than the package insert claimed CVs (7.2%
and 5.0%, for level 1 and 2, respectively) but very
similar to observations by Batista et al. (14), who
reported a 12.7% CV at 12.8 ng/mL and a 6.7% CV
at 27.6 ng/mL for Roche QCs run 6 times daily for 2
months. To evaluate assay precision at the lower
end, we analyzed 4 patient pools over 20 days. At a
25(OH)D concentration of 10.3 ng/mL, the CV was
10% (see Fig. 2 in the online Data Supplement),
which was slightly higher than the package insert
claim of 8.9% at 10.5 ng/mL.
Verification of the AMR showed an acceptable
linearity with a maximum deviation of 2.6% from
the expected concentration (see Fig. 3 in the online
Data Supplement). Dilution studies of 4 specimens
ranging from 34 to 68 ng/mL with vitamin D-free
serum showed an average overrecovery of 9.6%,
which was better than the 30%– 43% overrecovery
reported by Batista et al. (14), which possibly
resulted from their use of Roche Universal diluent.

..............................................................................................................

September 2019 | 04:02 | 247–253 | JALM 251

FOCUSED REPORTS Roche Elecsys Vitamin D Total II Assay Evaluation

The interference studies showed that there was no
significant interference up to a hemolysis index of
500, a lipemia index of 500, and an icterus index of
60; this result was consistent with the package in-
sert claims (hemolysis ≤600 mg/dL, intralipid ≤300
mg/dL, bilirubin ≤66 mg/dL) (see Fig. 4 in the online
Data Supplement).
We also assessed for plasma vs serum samples.
Twenty serum and plasma samples were com-
pared that were drawn from the same patient
at the same time. Nineteen out of 20 samples
showed an average bias of −3%; however, 1
plasma sample gave a result of >100 ng/mL com-
pared to serum sample at 55 ng/mL. A recent
Roche urgent medical device correction (Roche Di-
agnostics Corporation) has raised this issue; that
is, it has been observed that certain samples, in
particular plasma samples, can give falsely in-
creased results. Roche suggests a thorough in-
spection of preanalytical handling with emphasis
on centrifugation conditions and elimination of
foam and bubbles from the surface of the
sample. They also suggest that if the problem per-
sists, to use serum samples instead. As for now, we
are only using serum for 25(OH)D testing on the
Elecsys assay.
CONCLUSION
The Elecsys vitamin D total II assay was evaluated
for accuracy, precision, linearity, and common in-
terferences, and the overall analytical perfor-
mance of the method was considered acceptable.
In particular, the assay had good reactivity toward
25(OH)D2, which was a major issue with the previ-
ous generation assay. However, imprecision eval-
uated over several months in clinical use showed
“real life” CVs that are greater than package insert
claims, exceeding 10% for the low QC at 12 ng/mL.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following
4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b)
drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for
all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately
investigated and resolved.
M. Asif, administrative support, provision of study material or patients; E.K.Y. Leung, administrative support, provision of study
material or patients; K.-T.J. Yeo, statistical analysis.

Authors’ Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.

Role of Sponsor: No sponsor was declared.

Acknowledgments: The authors would like to thank the clinical laboratory staff and Drs. Maximo Marin and Zhen Mei for
technical assistance.

REFERENCES
1. Deluca HF. History of the discovery of vitamin D and its
active metabolites. Bonekey Rep 2014;3:479.
2. Souberbielle JC, Body JJ, Lappe JM, Plebani M, Shoenfeld Y,
Wang TJ, et al. Vitamin D and musculoskeletal health,
cardiovascular disease, autoimmunity and cancer:
recommendations for clinical practice. Autoimmun Rev
2010;9:709 –15.
3. Canadian Agency for Drugs and Technologies in Health.
Vitamin D testing in the general population: a review of the
clinical and cost-effectiveness and guidelines. Ottawa (ON):
Canadian Agency for Drugs and Technologies in Health; 2015.
4. LeFevre ML, Force USPST. Screening for vitamin D
deficiency in adults: U.S. Preventive Services Task Force
recommendation statement. Ann Intern Med 2015;162:
133– 40.
5. Lindblad AJ, Garrison S, McCormack J. Testing vitamin D
levels. Can Fam Physician 2014;60:351.
6. Bilinski K, Boyages S. The rise and rise of vitamin D
testing. BMJ 2012;345:e4743.
7. Caillet P, Goyer-Joos A, Viprey M, Schott AM. Increase of
vitamin D assays prescriptions and associated factors: a
population-based cohort study. Sci Rep 2017;7:10361.
8. Galior K, Ketha H, Grebe S, Singh RJ. 10 years of 25-
hydroxyvitamin-D testing by LC-MS/MS-trends in

...............................................................................................................

252 JALM | 247–253 | 04:02 | September 2019

Roche Elecsys Vitamin D Total II Assay Evaluation
vitamin-D deficiency and sufficiency. Bone Rep 2018;8:
268 –73.
9. French D. The (sun)light and dark of 25-
hydroxyvitamin D testing. J Appl Lab Med 2018;
3:460 –73.
10. Freeman J, Wilson K, Spears R, Shalhoub V, Sibley P.
Performance evaluation of four 25-hydroxyvitamin D
assays to measure 25-hydroxyvitamin d2. Clin Biochem
2015;48:1097–104.
11. Knudsen CS, Nexo E, Hojskov CS, Heickendorff L.
Analytical validation of the Roche 25-oh vitamin D total
assay. Clin Chem Lab Med 2012;50:1965– 8.
12. Tolan NV, Yoon EJ, Brady AR, Horowitz GL. Price of
high-throughput 25-hydroxyvitamin D immunoassays:
frequency of inaccurate results. J Appl Lab Med 2018;
2:868 –79.
13. Wyness SP, Straseski JA. Performance characteristics of
six automated 25-hydroxyvitamin d assays: mind your 3s
and 2s. Clin Biochem 2015;48:1089 –96.
..............................................................................................................
September 2019 | 04:02 | 247–253 | JALM
FOCUSED REPORTS
14. Batista MC, Menegat FD, Ferreira CES, Faulhaber ACL,
Campos D, Mangueira CLP. Analytical and clinical
validation of the new Roche Elecsys vitamin D total II
assay. Clin Chem Lab Med 2018;56:e298 –301.
15. Leung EK, Yi X, Gloria C, Yeo KT. Clinical evaluation of the
QMS (r) tacrolimus immunoassay. Clin Chim Acta 2014;
431:270 –5.
16. Ross AC, Taylor CL, Yaktine AL, Del Valle HB, editors.
Dietary reference intakes for calcium and vitamin D.
Washington (DC): National Academies Press; 2011.
17. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM,
Hanley DA, Heaney RP, et al. Evaluation, treatment, and
prevention of vitamin D deficiency: an endocrine society
clinical practice guideline. J Clin Endocrinol Metab 2011;
96:1911–30.
18. Heijboer AC, Blankenstein MA, Kema IP, Buijs MM.
Accuracy of 6 routine 25-hydroxyvitamin D assays:
influence of vitamin D binding protein concentration. Clin
Chem 2012;58:543– 8.
253