Treating neurological disease with diet
Precision Xtra and Optium Xceed


Note: new paper

How can we compare the effect of one ketogenic diet to another? One way is by measuring ketone levels. The higher the level, the stronger the diet.

This kind of comparison sounds simple, but until recently it was impossible to do at home. The only available tools for measuring ketones were urine dipsticks and tablets, but they are too inaccurate for this purpose and they measure the wrong thing — urine instead of blood.

The situation changed in 2005 when Abbott Laboratories introduced the first home meter that measures ketones in blood. Abbott calls it the Precision Xtra in North America and Optium Xceed in the rest of the world. As of 2011, no other company has released a similar product.

The Xtra/Xceed makes it possible, for the first time, to measure ketosis accurately at home. The device is extraordinarily accurate and precise, easy to use, and fairly cheap (although not as cheap as dipsticks).

Ease of use

The Xtra/Xceed is easy to use if you don't mind a little pain. Here's the procedure:

1. Wash your hands to avoid infection and to remove skin oils that might contaminate the blood sample.

2. Take a disposable plastic test strip from its foil wrapper and insert it into the slot at the bottom of the machine. The machine turns on automatically.

3. Prick a fingertip with a gadget called a lancet.

4. Allow a drop of blood to ooze onto the bottom of the test strip.

5. Watch the machine count down for ten seconds. Then read the result on the screen. The machine turns off automatically after 60 seconds.

This is the same procedure as the one used by diabetics when they measure blood glucose. In fact, the Xtra/Xceed is a glucose meter for diabetics — it measures both glucose and ketones depending on which type of test strip you insert. However, there is one difference. Ketone test strips require a larger drop of blood than glucose strips, so you may have to set the lancet to prick your finger more deeply than you would for glucose testing.
The Accu-Chek Multiclix is probably the least painful and most convenient lancet on the market. When you push the button, a tiny spring-loaded needle jabs your finger.

Precision and accuracy

The Xtra/Xceed has been tested by numerous independent reviewers. They all report a high degree of precision and accuracy, although some say precision and accuracy decline when sample concentrations rise above a certain point.

One drawback of the reviews is that they mostly look at results in the lower part of the Xtra/Xceed's range.

One reviewer summed up the results of several studies like this:

Three recent research studies… all gave very similar results and all confirm the excellent accuracy of the Precision Xtra blood ketone system for diagnosing ketosis in dairy cattle. These studies involved a total of 622 cows with a 14.1% prevalence of ketosis. The average R2 between hand-held meter and laboratory BHBA results was .94.… results from the meter are only slightly different than results from sending blood serum or plasma samples to a laboratory for BHBA analysis (Oezel and McGuirk 2008B).

I've seen only two studies that report serious problems. Ham et al. say the device is very accurate up to 4.0 mmol/L but above that level, accuracy declines. They tested an earlier version of the product which had a more limited range than the current one so these observations may no longer apply.

Janssen et al. say the device is very accurate with samples below 6 mmol/L, but once levels go above that point, test strips become saturated and the Xtra/Xceed can report erroneous values between 3 and 6.5 mmol/L. This means (for example) that if you see a reading of 4 mmol/L on the machine, you can't trust it because you may be seeing an erroneous result from a sample that is really 6.5 mmol/L. No other study reported this, but most studies didn't examine samples with such high levels. The manufacturer says the device is accurate up to 8 mmol/L.

I am skeptical about Janssen et al.'s results because they report extreme variance for samples that contain more than 6 mmol/L. But I haven't seen this with my three Xtras. On the contrary, several times they have reported values as high as 7.5 mmol/L and I immediately retested with three fresh test strips on all three machines simultaneously. In every case, the three machines generated results within a small range.

On the other hand, the manufacturer's own numbers for accuracy are confined to the range from 0 to 5.6 mmol/L (see Table 1). Given that the manufacturer claims the machine has a working range from 0 to 8 mmol/L, this seems very strange.
According to one study (Yoon et al.) the Xtra/Xceed and this expensive laboratory machine have a correlation coefficient (R) of 98 percent in the range from 0 to 4.6 mmol/L. This group of researchers stopped at 4.6 because they didn't have any samples with higher levels.

Table 1. Manufacturer's claimed accuracy
Number of samples
Ketone range mmol/L 0.0–5.6
Slope 0.958
Intercept 0.067
R (correlation coefficient) 0.959
Source: Abbot Diabetes Care Limited, "Precision Xtra Blood β-Ketone Test Strips" (package insert).

Table 2. Manufacturer's claimed precision
Low Mid
Mean mmol/L 0.34 2.36 6.32
SD mmol/L 0.03 0.09 0.2
CV % 3.8 3.1
Source: Abbot Diabetes Care Limited, "Precision Xtra Blood β-Ketone Test Strips" (package insert).

Table 3. Accuracy and precision reported by various reviewers
Lab reference R
Forrow, et al. 0.97 3.2% at 3.55 mmol/L The authors work for the manufacturer.
Ham, et al. Cobas Fara II 0.94 This study used an early version of the product which was different from the current one. The authors say accuracy decreased above 4.0 mmol/L.
Iwersen, et al. .95
Janssen, et al. LC-MS/MS1 0.97 The authors say readings over 3 mmol/L may be misleading (see text).
Pardon, et al. 0.88 Mean difference between meter and reference was 0 mmol/L.
Rewers, et al. Cobas Mira Plus 0.92
Voulgari, et al. 0.99 Mean difference between meter and reference was 0.49 mmol/L.
Voyvoda, et al. 0.97
Yoon et al. Beckman Coulter
Unicel DxC 800
0.98 1.6% at 3.81 mmol/L This study is unusually throrough.
1. Liquid chromatography tandem mass spectrometry.

Technical stuff

The Xtra/Xceed measures the concentration of beta-hydroxybutyrate (BHB) in blood plasma. The manufacturer calls it a "blood ketone meter" but it's really a plasma meter. Plasma concentrations of beta-hydroxybutyrate are about 5–20 percent higher than blood concentrations. (Janssen et al.)

The Xtra/Xceed's test strips contain an enzyme called beta-hydroxybutyrate dehydrogenase and a coenzyme called nicotinamide adenine dinucleotide (NAD). When blood soaks into the test strip, BHB in the sample reacts with NAD, releasing electrons and creating a current that is proportional to the concentration of BHB. The Xtra/Xceed calculates the level of BHB from the strength of the current (Abbott Diabetes Care Limited).

The Xtra/Xceed's results are calibrated to match the Ranbut venous plasma D-beta-hydroxybutyrate enzymatic assay on the RX Daytona system manufactured by Randox Laboratories. (Abbot Diabetes Care Limited).
The Xtra/Xceed is calibrated to match the Randox RX Daytona clinical chemistry analyzer.


Abbot Diabetes Care Limited.
"Precision Xtra Blood β-Ketone Test Strips."
Package insert. 2006.

Burke CM, Raphael W, Leslie KE, Neuder LM.
"Test comparison of Precision Xtra and Ketostix for ketosis in dairy cows."
Proceedings of the 41st Annual Convention of the American Association of Bovine Practitioners. 2008 Sep 25–27; p. 287.

Forrow NJ, Sanghera GS, Walters SJ, Watkin JL.
Development of a commercial amperometric biosensor electrode for the ketone D-3-hydroxybutyrate.
Biosens Biolectron. 2005 Feb;20(8):1617–25.

Ham MR, Okada P, White P.
Bedside ketone determination in diabetic children with hyperglycemia and ketosis in the acute care setting
Pediatric Diabetes. 2004;5:39–43

Iwersen M, Falkenberg U, Voigtsberger R, Forderung D, Heuwieser W
Evaluation of an electronic cowside test to detect subclinical ketosis in dairy cows
J Dairy Sci. 2009 Jun;92(6):2618–24.

Janssen MJ, Hendrickx BH, Habets-van der Poel CD, van den Bergh JP, Haagen AA, Bakker JA.
Accuracy of the Precision® point-of-care ketone test examined by liquid chromatography tandem-mass spectrometry (LC-MS/MS) in the same fingerstick sample.
Clin Chem Lab Med. 2010 Dec;48(12):1781–4.

Oetzel GR and McGuirk SM.
2008 A.
"Evaluation of a hand-held meter for cowside evaluation of blood beta-hydroxybutyrate and glucose concentrations in dairy cows."
Proceedings of the 41st Annual Convention of the American Association of Bovine Practitioners. 2008; 41:31.

Oetzel G and McGuirk S.
2008 B.
Fact Sheet — Cowside blood BHBA testing with a hand-held "ketometer"
Univ of Wisconsin-Madison School of Vertinary Medicine, 2008 Sep 9, version 3

Pardon B, Hostens M, Ribbers L, De Bleecker K, Opsomer G, Deprez P.
Ear vein sampling procedure with a commercial beta-hydroxybutyrate meter as a cow side test for ketosis.
1st European Buiatrics Forum (EBF 2009).

Rewers A, McFann K, Chase HP.
Bedside monitoring of blood beta-hydroxybutyrate levels in the management of diabetic ketoacidosis in children.
Diabetes Technol Ther. 2006 Dec;8(6):671–6.

Voulgari C and Tentolouris N.
The performance of a glucose-ketone meter in the diagnosis of diabetic ketoacidosis in patients with type 2 diabetes in the emergency room
Diabetes Technology & Therapeutics. 2010 July; 1(7) 529–35.

Voyvoda H and Erdogan H
Use of a hand-held meter for detecting subclinical ketosis in dairy cows
Research in Veterinary Science. 2010 Dec;89(3):344–351

Yoon NG, Crystal YWMC, Jacob E.
Blood ketone testing in the clinical laboratory — technical evaluation of test-strips
SGH Proceedings. 2008;17(2):88–92

This page was last revised on January 20, 2011.