Search tips
Search criteria 


Logo of canvetjReference to the Publisher site.Journal Web siteJournal Web siteHow to Submit
Can Vet J. 2010 May; 51(5): 525–526.
PMCID: PMC2857435

Language: English | French

The effect of storage temperature on the accuracy of a cow-side test for ketosis


The objective of this study was to assess the effect of storage conditions on the accuracy of a milk test strip for ketosis. Storage at 21°C for up to 18 wk had little effect on accuracy for diagnosis and classification of subclinical ketosis.


L’effet de la température d’entreposage sur l’exactitude d’un test diagnostic pour l’acétonémie. L’objectif de cette étude était de vérifier l’effet de différentes conditions d’entreposage sur l’efficacité d’une bandelette servant au diagnostic de l’acétonémie. L’entreposage de ce test diagnostic à 21 °C pour une durée allant jusqu’à 18 semaines n’a pas eu d’effet sur son efficacité à diagnostiquer l’acétonémie subclinique.

(Traduit par les auteurs)

Subclinical ketosis (SCK) is characterized by an abnormally high concentration of circulating ketone bodies in the absence of clinical signs of ketosis (1). The prevalence of SCK can vary from 6.9% to 34% in dairy cows (2,3). Subclinical ketosis is an important disease in the dairy industry that has been associated with a higher risk of peripartum diseases such as clinical ketosis and displaced abomasum, decreased milk production, and decreased probability of pregnancy at first service (4,5). Given the disease risks associated with SCK, monitoring of SCK in dairy herds is potentially of value to producers. The gold standard diagnostic test for SCK is plasma or serum concentration of β-hydroxybutyric acid (BHBA) (4). Using this test, a threshold of 1400 μmol/L has been found to be the most accurate for detecting cows with SCK (6). Unfortunately, this gold standard test is not practical as a cow-side test for immediate intervention by producers or veterinarians. Alternatively, Keto-Test milk strips (Sanwa Kagaku Kenkyusho Co., Nagoya, Japan) have been shown to be a very good semi-quantitative test for evaluating the SCK status of early lactation dairy cows (7). Using pooled data from 5 studies, Oetzel (7) found that Keto-Test milk strips have a sensitivity and specificity of 83% and 82%, respectively, when using a cut-off value of 100 μmol/L. The Keto-Test milk strips can be used as a cow-side test for detection of SCK in dairy cows and also for monitoring the prevalence of SCK over time at herd level (8).

The commercial label for Keto-Test milk strips recommends that the test strips be stored at 2°C to 8°C. However, these storage requirements are not always easily maintained, and many Canadian dairy producers and veterinarians store them at room temperature. Unfortunately there is no literature regarding the effect of storage conditions on the accuracy of Keto-Test milk strips. Such knowledge would be of considerable practical use to both dairy producers and practitioners. Therefore, the objective of this study was to determine the accuracy of Keto-Test milk strips for detection of subclinical ketosis in early lactation cows, after being stored at 21°C for 0, 6, 12, or 18 wk.

A 500-cow commercial dairy herd in southwestern Ontario was used for data collection in this study. Cows were housed in a free stall barn, fed a total mixed ration, and milked twice daily. Data were collected from April to August 2008. The farm was visited by a technician twice weekly for the duration of the 18-week study period in order to obtain 40 milk and blood samples per week. Milk and blood samples were collected simultaneously from lactating cows of all parities between 2 and 25 d in milk. The Keto-Test milk strips (4 × 720 strips) were initially stored, as prescribed, at 4°C. Following a pre-determined schedule, each of the 4 groups of milk strips was removed from the refrigerator and stored at room temperature (21°C) for 0, 6, 12, or 18 wk. Group 0 was used as a control reference and these strips were taken out of the refrigerator shortly before use. Group identification was blinded to the technician reading the test strips, and to the statistician.

Each milk sample (30 mL) was collected from one quarter and transported on ice, within 2 h following collection, to the Ruminant Field Service laboratory at the Ontario Veterinary College (Guelph, Ontario). One Keto-Test milk strip from each of the test groups was dipped in each milk sample, for a total of 4 milk strips per sample. The test results were read after 1 min using the color chart provided on the Keto-Test bottle label that corresponds to 0, 50, 100, 200, 500, or 1000 μmol/L of BHBA. Immediately after milk sampling blood samples were collected from the coccygeal vessels into vacuum tubes without anticoagulant (BD Vacutainers, Franklin Lakes, New Jersey, USA). Tubes were centrifuged within 4 h of collection. Sera were separated and frozen at −20°C, and submitted to the Animal Health Laboratory at the University of Guelph for determination of BHBA concentration (Ranbut; Randox Laboratories, Antrion, United Kingdom) using an automated analyzer (Hitachi 911; Roche Diagnostics, Indianapolis, Indiana, USA). Statistical analyses were conducted using SAS (version 9.1; SAS Institute, Cary, North Carolina, USA). The MEANS and UNIVARIATE procedures were used for descriptive statistics. Serum BHBA results were dichotomized using a cut-off value of 1400 μmol/L and were used as the gold standard reference test for the detection of SCK (6). Serum BHBA values ≥ 1400 μmol/L were considered to be positive for SCK. Keto-Test results ≥ 100 μmol/L were classified as positive for SCK (7). The agreement between the 3 test groups (6, 12, and 18) and the control group (0) was calculated with the FREQ procedure using the Kappa statistic for agreement beyond chance (9).

A total of 719 samples from 230 cows were collected during the 18-week trial period. Cows that were sampled more than once during the trial were sampled on a different day in milk each time. One Keto-Test milk strip was lost and could not be used for the study. The average days in milk of cows at sampling was 13.2 [standard deviation (s) = 5.7] and the average serum BHBA was 834 μmol/L (s = 705). The true prevalence of SCK as defined by the gold standard test was 9.8%. Prevalence of SCK using the week 0 control test was 20.0%. The sensitivity and specificity of each of the 4 test groups were calculated using the gold standard serum BHBA concentration and are reported in Table 1. The week 0 control test had sensitivity and specificity of 100% and 88.8%, respectively. These values are both greater than the pooled sensitivity (83%) and specificity (82%) values summarized by Oetzel (7) when using the same cut-point. The other 3 test groups had similar sensitivities and specificities to those of the control group indicating little change in test strip accuracy over the storage times studied. The kappa coefficients for agreement between the control Keto-Test and the 6-, 12-, and 18-week test groups were 0.973 [95% confidence interval (CI) = 0.958–0.989], 0.978 (95% CI = 0.964–0.993), and 0.979 (95% CI = 0.967–0.991), respectively. These kappa values indicate that there was almost perfect agreement between all test groups and the control (9). These findings suggest that Keto-Test milk strips may be stored for up to 18 wk at 21°C with very little effect on the accuracy of the test. Producers and practitioners, therefore, may safely store the test strips for 4.5 mo at controlled room temperature without negative effects on the accuracy of this diagnostic tool.

Table 1
Sensitivity and specificity of Keto-Test milk strips (cut-point ≥ 100 μmol/L) for detection of subclinical ketosis after 0, 6, 12, or 18 weeks of storage at 21°C when compared to serum β-hydroxybutyric acid concentration ...


The authors thank the producer who participated in this study. Additional thanks to the Summer Leadership and Research Program at the Ontario Veterinary College, and Elanco Animal Health for financial support. CVJ


This study was funded by the Summer Leadership and Research Program (Ontario Veterinary College), and Elanco Animal Health (Guelph, Ontario).

Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office ( gro.vmca-amvc@nothguorbh) for additional copies or permission to use this material elsewhere.


1. Andersson L. Subclinical ketosis in dairy cows. Vet Clin North Am Food Anim Pract. 1988;4:233–251. [PubMed]
2. Andersson L, Emanuelson U. An epidemiological study of hyperketonaemia in Swedish dairy cows; determinants and the relation to fertility. Prev Vet Med. 1985;3:449–462.
3. Kauppinen K. Prevalence of bovine ketosis in relation to number and stage of lactation. Acta Vet Scand. 1983;24:349–361. [PubMed]
4. Duffield TF, Sandals D, Leslie KE, et al. Efficacy of monensin for the prevention of subclinical ketosis in lactating dairy cows. J Dairy Sci. 1998;81:2866–2873. [PubMed]
5. Walsh RB, Walton JS, Kelton DF, LeBlanc SJ, Leslie KE, Duffield TF. The effect of subclinical ketosis in early lactation on reproductive performance of postpartum dairy cows. J Dairy Sci. 2007;90:2788–2796. [PubMed]
6. Duffield TF, Lissemore KD, McBride BW, Leslie KE. Impact of hyperketonemia in early lactation dairy cows on health and reproduction. J Dairy Sci. 2009;92:571–580. [PubMed]
7. Oetzel GR. Monitoring and testing dairy herds for metabolic disease. Vet Clin Food Anim. 2004;20:651–674. [PubMed]
8. DesCôteaux L, Duffield T, Bélanger AM, et al. Monitoring subclinical ketosis using milk strip test and control chart in dairy herds. Proc Annu Meet Am Assoc Bov Pract. 2003;172
9. Dohoo I, Martin W, Stryhn H. Veterinary Epidemiologic Research. 1st ed. Charlottetown: AVC Inc; 2003. pp. 85–120.

Articles from The Canadian Veterinary Journal are provided here courtesy of Canadian Veterinary Medical Association