Abstract
Background
An evaluation of currently available in‐clinic diagnostic tests for Giardia duodenalis infection of dogs and cats has not been performed. In addition, there is discordance among published diagnostic comparisons. The absence of a true gold standard for detecting Giardia duodenalis also complicates diagnostic evaluations.
Objectives
To evaluate diagnostic tests commercially available in the United States for detecting Giardia duodenalis in dogs and cats, in comparison to a widely used reference test, the direct immunofluorescent assay (IFA), and also to compare the results of 2 methods of analysis: comparison of diagnostic tests to a reference test (IFA) and Bayesian analysis.
Animals
Fecal samples from a convenience sample of 388 cats and dogs located in Colorado, Oklahoma, and Virginia.
Methods
Fecal samples were tested for Giardia duodenalis by zinc sulfate centrifugal fecal flotation and 4 different commercial diagnostic immunoassays. Results were analyzed via Bayesian analysis and by comparison to the IFA as the reference test.
Results
Sensitivity and specificity by comparison to IFA was ≥82% and ≥90%, respectively, for all diagnostic tests in dogs and cats. When analyzed via Bayesian analysis, sensitivity and specificity were ≥83% and ≥95%, respectively. When ZnSO4 centrifugal fecal flotation results were combined with immunoassay results, there was no longer a significant difference between the sensitivities of the commercial in‐clinic immunoassays.
Conclusion and Clinical Relevance
The Bayesian analysis validates using IFA as the reference test. Differences in commercial in‐clinic immunoassay sensitivities can be mitigated when the results are combined with ZnSO4 centrifugal fecal flotation results.
Keywords: canine, direct immunofluorescent, fecal flotation, feline, immunoassay, protozoa
Abbreviations
- CAPC
Companion Animal Parasite Council
- IFA
MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay
- SNAP
IDEXX Giardia Antigen Test (IDEXX Laboratories, Westbrook, ME)
- VetChek
TECHLAB VETCHEK ELISA (TechLab Inc., Blacksburg, VA)
- VetScan
Abaxis VetScan Giardia Antigen test (Abaxis, Inc., Union City, CA)
- ZnSO4
Zinc sulfate centrifugal fecal flotation
1. INTRODUCTION
Giardia duodenalis is a common enteric protozoan parasite of dogs and cats. Most infections are subclinical, but acute or chronic diarrhea can also occur.1, 2, 3 The prevalence of G. duodenalis infection varies depending on the age, clinical status, housing, and geographic region of the animals surveyed and is influenced by the detection method used.1, 4 Historically, diagnosis of G. duodenalis in dogs and cats has been via microscopic examination of feces for trophozoites or cysts.5 However, accuracy of microscopic diagnosis of G. duodenalis is limited by the infrequent presence of trophozoites in diarrheic feces, intermittent passage of cysts, and the requirement that an experienced technician perform the examination.6, 7, 8 The direct immunofluorescent assay (IFA) is more sensitive and specific for diagnosing G. duodenalis than conventional flotation tests9, 10 and is the reference test for evaluating G. duodenalis tests in companion animals.4, 11, 12 Immunoassays are also available that detect a soluble cyst antigen of G. duodenalis. The Companion Animal Parasite Council (CAPC), which is widely cited as a source of guidelines for parasite control in the United States, recommends that centrifugal fecal flotation be used in conjunction with an immunoassay for diagnosing G. duodenalis infections in veterinary practices (www.capcvet.org). This is supported by the evidence that the sensitivity of fecal flotation is improved when used with a commercial immunoassay.13
Diagnostic test evaluation is performed by comparison to a gold standard reference test. However, there is often not a true gold standard for test comparison, and this is the case for G. duodenalis. The IFA is widely accepted as the most sensitive and specific test for G. duodenalis and is often used as the reference test.9, 10 But, it is not a true gold standard (ie, “…absolutely accurate”),14 and using an imperfect reference test as the gold standard for diagnostic comparison can lead to miscalculation of test performance for the test(s) being evaluated. To overcome this problem, a Bayesian analysis can be performed. This statistical method allows for diagnostic test evaluation in the absence of a gold standard and has been used for G. duodenalis diagnostic test evaluations.10, 14, 15, 16, 17
Various G. duodenalis diagnostic tests have been evaluated for detection of infection in small animals, but not all comparisons have included a reference test.4, 10, 11, 13, 15 The purpose of this study was to evaluate diagnostic tests commercially available in the United States that are optimized for detecting G. duodenalis in dogs and cats, by comparison to a reference test (IFA) and also to compare all the diagnostic tests, including the IFA, using a Bayesian analysis.
2. MATERIALS AND METHODS
2.1. Fecal specimen collection and screening
Fecal samples (n = 388) were collected from dogs and cats at 3 distinct study sites from 2012 to 2014. These sites were parasitology laboratories at veterinary teaching hospitals in Fort Collins, Colorado; Stillwater, Oklahoma; and Blacksburg, Virginia. Samples at all 3 sites consisted of hospital submissions (wellness examinations and clinical cases), plus collection surveys from animal shelters, and rescue organizations.
At each study site, samples were screened for the presence of G. duodenalis cysts using zinc sulfate centrifugal fecal flotation (ZnSO4) as outlined below. Giardia duodenalis positive samples of sufficient quantity to perform all diagnostic tests were included in the study. Fecal consistency and the presence of other parasites were recorded for each sample. For each G. duodenalis positive sample, a matching G. duodenalis negative sample (by ZnSO4 flotation) of the same fecal consistency from the same geographic location was included in the study. This selection procedure was used to ensure an adequate number of G. duodenalis positive samples for meaningful analysis, because prevalence in the general dog and cat population is relatively low. Near the end of the study period, all fecal samples from shelter collections in Virginia, regardless of positive or negative status, were included in the sample pool to increase the overall sample size. Samples collected from Oklahoma and Colorado were refrigerated and shipped weekly with ice packs to the main study site. Sample selection and fecal flotations were performed at each study site, and all immunoassays were performed weekly at the main study site (Blacksburg, Virginia) upon sample arrival. Zinc sulfate fecal flotations were not repeated at the main study site.
2.2. Fecal flotation
Zinc sulfate centrifugal fecal flotation was performed as described by Zajac and Conboy18 using 2 to 4 g of feces. The feces and zinc sulfate solution (specific gravity 1.18) were centrifuged with a coverslip in place for 5 minutes at 200 × g. After centrifugation, the coverslip was removed to a glass slide and scanned for G. duodenalis cysts with a compound microscope at 100×, and if no cysts were detected then also at 200×. Giardia duodenalis and any other parasite species present were recorded as present, but no quantitation was performed.
2.3. Immunoassays
The VETCHEK ELISA (TECHLAB Inc, Blacksburg, Virginia) (VetChek) is an enzyme immunoassay developed for the qualitative detection of G. duodenalis cyst antigen in canine and feline fecal specimens. The test was performed following the manufacturer's instructions. This is the first well‐plate ELISA that is commercially available and optimized for use in dogs and cats.
The SNAP Giardia Antigen Test (IDEXX Laboratories, Westbrook, Maine) (SNAP) is a rapid in‐clinic enzyme immunoassay for the detection of G. duodenalis antigen in canine and feline feces. Tests were performed following the manufacturer's instructions.
The Abaxis VetScan Giardia Antigen test (VetScan), is a rapid in‐clinic enzyme immunoassay for the detection of G. duodenalis antigen in canine feces only. Tests were performed following the manufacturer's instructions. Although this test is not intended for use in cats, it was included in this feline comparison because practitioners might consider using the VetScan for cats if the test is available in their clinic. The VetScan cat results were excluded from statistical analysis because the manufacturer does not intend for the test to be used in cats.
The MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay (Meridian Bioscience Inc, Cincinnati, Ohio) (IFA) was used as the reference test and performed following the manufacturer's instructions to identify G. duodenalis cysts in feces. Samples were run in batches, with positive and negative controls each time the test was run. Slides were examined at 100×, and if no cysts were detected, then also at 200× magnification using a fluorescence microscope. A sample was considered positive if any G. duodenalis cysts were detected. The presence of Cryptosporidium oocysts was also recorded if any were observed, but results are not reported here.
2.4. Statistical analysis
The data was analyzed by multiple methods. Sensitivity and specificity of the fecal flotation and immunoassays were calculated by comparison to IFA.4, 9, 10 The sensitivities and specificities of each diagnostic test were then compared for differences using McNemar's test for significance of changes.19 These analyses were conducted on data from dogs and cats separately. Additionally, the in‐clinic immunoassay results (SNAP and VetScan) were analyzed in combination with the ZnSO4 fecal flotation results in dogs (ie, the sample is considered positive if at least 1 test is positive) to mimic their use in clinics as recommended by CAPC. The combined results were then compared for differences in sensitivities and specificities using McNemar's test for significance of changes. Lastly, a Bayesian analysis, in which no test is considered a reference standard, was performed to estimate the sensitivities and specificities of each diagnostic test for dogs and cats separately.10, 20, 21
The Bayesian analysis framework expanded the 3‐test script BayesDiagnosticTests\src\tt3.txt by Lawrence Joseph and colleagues (www.medicine.mcgill.ca/epidemiology/Joseph/Bayesian-Software-Diagnostic-Testing.html) to 5 tests for data from canine samples and to 4 tests for data from feline samples. Theoretical foundations of the actual software (BayesDiagnosticTests version 3.10.2, January 2016) have been reported.20, 21 The new scripts were implemented and executed using WinBugs version 1.4.3. Parameters to be estimated included overall prevalence for either canines or felines, and sensitivity and specificity for each of the diagnostic tests. Pairwise comparisons between diagnostic tests for sensitivity and specificity were included as differences within the scripts. Prior information for the model was obtained from published studies and manufacturers' information.4, 10, 11, 13 The prior information was collated and summarized as mean (±SD) that was subsequently converted into alpha and beta parameters (Table 1) of the beta prior density. Research evidence to update the prior information is presented in the form of cross‐classification for canine and feline samples (Table 2). After initialization, all models were run for 25 000 iterations (5000 for burn‐in and 20 000 for parameter estimation). Trace plots were inspected to verify convergence, and the Monte Carlo error for each parameter was also verified to be less than 5% of the sample SD. Finally, for each parameter of interest, a median together with the 2.5 and 97.5 percentiles were obtained from the posterior distribution.
Table 1.
Prior information for each of the 11 parameters to be estimated by Bayesian analysis
| Test | Parameter | Mean (%) | SD (%) | Alpha | Beta |
|---|---|---|---|---|---|
| Prevalence | 7.69 | 6.42 | 1.25 | 14.98 | |
| ZnSO4 | Sensitivity | 67.15 | 25.67 | 1.58 | 0.77 |
| Specificity | 96.85 | 4.04 | 17.13 | 0.56 | |
| Vet Chek | Sensitivity | 93 | N/A | 9.80 | 0.74 |
| Specificity | 99.7 | N/A | 0.44 | 0.001 | |
| SNAP | Sensitivity | 84.77 | 7.51 | 18.56 | 3.33 |
| Specificity | 97.27 | 4.56 | 11.45 | 0.32 | |
| VetScan | Sensitivity | 98.1 | N/A | 2.26 | 0.04 |
| Specificity | 99.3 | N/A | 2.33* | 0.016* | |
| IFA | Sensitivity | 100† | 0† | 997.00 | 1.00 |
| Specificity | 100† | 0† | 997.00 | 1.00 |
N/A, not available. N/A values were replaced with 7.51 and 4.56 for sensitivity and specificity, respectively. These values were approximately mid‐way for all SDs.
The model would not estimate this parameter. As such noninformative values of (1) were used instead.
To compute alpha and beta, a mean of 99.9% and an SD of 0.1% were used.
Table 2.
Results from 5 diagnostic tests for the detection of G. duodenalis
| ZnSO4 | VetChek | SNAP | VetScan | IFA | Number of samples | ||
|---|---|---|---|---|---|---|---|
| Canine | Feline | Total | |||||
| − | − | − | − | − | 128 | 82 | 210 |
| + | + | + | + | + | 84 | 24 | 108 |
| + | + | + | − | + | 10 | 11 | 21 |
| − | + | + | + | + | 9 | 0 | 9 |
| + | − | − | − | − | 4 | 1 | 5 |
| − | + | − | − | − | 2 | 3 | 5 |
| + | + | − | + | + | 4 | 0 | 4 |
| + | + | + | + | − | 4 | 0 | 4 |
| + | − | − | − | + | 3 | 1 | 4 |
| + | + | − | − | + | 2 | 1 | 3 |
| − | − | + | − | + | 2 | 1 | 3 |
| − | + | + | − | − | 2 | 1 | 3 |
| − | + | − | − | + | 2 | 0 | 2 |
| + | + | − | − | − | 2 | 0 | 2 |
| − | − | − | − | + | 1 | 1 | 2 |
| − | + | + | − | + | 0 | 1 | 1 |
| + | − | + | − | + | 1 | 0 | 1 |
| + | + | + | − | − | 1 | 0 | 1 |
| Total | 261 | 127 | 388 | ||||
Abbreviations: +, positive result; −, negative result; IFA, MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay; SNAP, IDEXX Giardia Antigen Test; VetChek, TECHLAB VETCHEK ELISA; VetScan, Abaxis Giardia Antigen test; ZnSO4, zinc sulfate centrifugal fecal flotation.
3. RESULTS
Overall 388 samples (127 feline and 261 canine) were included in the study and evaluated for the presence of G. duodenalis by each diagnostic test. The cross‐classified results from each diagnostic test are outlined in Table 2. Of the 388 samples tested, 108 (24 feline and 84 canine) were positive for G. duodenalis on all 5 diagnostic tests, whereas 210 (82 feline and 128 canine) were negative on all diagnostic tests. In total, 318 of 388 samples (82%) had concordant results across all 5 diagnostic tests. With regards to the geographic origin of each sample, there was no difference in sensitivity or specificity of the initial ZnSO4 fecal flotation when compared to IFA among the different study sites.
3.1. Comparison to IFA
The sensitivity and specificity of each test compared to the IFA reference test for dogs and cats are presented in Tables 3 and 4. Analysis of canine data alone shows that the test with the highest sensitivity when compared to IFA was VetChek (94.1%) followed by SNAP (89.8%). The results of the McNemar's test in dogs showed that the ZnSO4 flotation, VetChek, and SNAP all had significantly higher sensitivities than the VetScan. When the in‐clinic immunoassay results in dogs were combined with the ZnSO4 fecal flotation result with a dog being considered positive if either immunoassay or flotation was positive, there was no longer a significant difference in sensitivity between the SNAP and VetScan immunoassays in dogs. There was no statistically significant difference in specificities between the in‐clinic immunoassays (SNAP and VetScan) in dogs. VetScan (97.2%) and SNAP (95.1%) both had significantly higher specificities than the VetChek and ZnSO4 fecal flotation (both 92.3%).
Table 3.
Diagnostic test performance in dogs compared to IFA
| Diagnostic test | 95% confidence interval | |
|---|---|---|
| Sensitivity | Specificity | |
| SNAP | 89.8b (83.1‐94.1) | 95.1bc (90.2‐97.6) |
| VetChek | 94.1b (88.3‐97.1) | 92.31a (86.8‐95.7) |
| VetScan | 82.2a (74.3‐88.1) | 97.2b (93.0‐98.9) |
| SNAP w/ZnSO4 | 97.5ab (92.8‐99.1) | 90.9abc (85.1‐94.6) |
| VetScan w/ZnSO4 | 95.8ab (90.5‐98.2) | 92.31abc (86.8‐95.7) |
| ZnSO4 | 88.1ab (81.1‐92.8) | 92.3ac (86.8‐95.7) |
The sensitivity and specificity of each test with 95% confidence intervals (CI) when compared to the IFA reference test in dogs. Within columns, different letters are significantly different (McNemar's P < .05). Cells without superscripts had no significant differences.
Abbreviations: IFA, MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay; SNAP, IDEXX Giardia Antigen Test; Vet Chek, TECHLAB VETCHEK ELISA; VetScan, Abaxis Giardia Antigen test; ZnSO4, zinc sulfate centrifugal fecal flotation.
Table 4.
Diagnostic test performance in cats compared to IFA
| Diagnostic test | 95% confidence interval | |
|---|---|---|
| Sensitivity | Specificity | |
| SNAP | 92.5 (87.6‐95.9) | 98.9 (95.9‐99.9) |
| Vet Chek | 92.5 (92.9‐99.1) | 95.4 (90.2‐97.8) |
| ZnSO4 | 92.5 (85.7‐95.1) | 98.9 (90.7‐97.8) |
| SNAP w/ZnSO4 | 97.5 (87.1‐99.6) | 97.7 (92.0‐99.4) |
The sensitivity and specificity of each test with 95% confidence intervals (CI) when compared to the IFA reference test in cats. Within columns, different letters are significantly different (McNemar's P < .05). Cells without superscripts had no significant differences.
Abbreviations: IFA, MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay; SNAP, IDEXX Giardia Antigen Test; Vet Chek, TECHLAB VETCHEK ELISA; VetScan, Abaxis Giardia Antigen test; ZnSO4, zinc sulfate centrifugal fecal flotation.
In cats, all 3 tests had the same sensitivity (92.5%) when compared to the IFA (Table 4), and specificity was ≥95% for each diagnostic test. There were no statistically significant differences between the sensitivities and specificities of any of the tests in cats. When the immunoassay (SNAP) result for cats was combined with the ZnSO4 fecal flotation result, sensitivity did improve from 92.5% to 97.5%, but this difference was not statistically significant.
3.2. Bayesian analysis
The estimated sensitivity, specificity, and 95% confidence interval of each diagnostic test for dogs using the Bayesian analysis are shown in Table 5. The most sensitive diagnostic test was the IFA (99.4%), which was also the most specific (99.7%). Although the VetScan had the lowest sensitivity (83.3%), it had the highest specificity (99.3%) after the IFA. Specificity was ≥95% for all other diagnostic tests.
Table 5.
Bayesian analysis estimates of diagnostic test parameters in dogs
| Diagnostic test | 95% confidence interval | |
|---|---|---|
| Sensitivity | Specificity | |
| IFA | 99.4 (98.86‐99.7) | 99.8b (99.3‐99.9) |
| SNAP | 90.5ab (85.0‐94.9) | 98.7ab (95.9‐100) |
| VetChek | 94.5a (90.9‐98.7) | 96.0a (91.9‐98.8) |
| VetScan | 83.3b (75.9‐89.6) | 99.3b (97.4‐99.9) |
| ZnSO4 | 88.7b (82.3‐93.7) | 95.5a (91.6‐98.3) |
The sensitivity and specificity of each test in dogs with 95% confidence intervals (CI) estimated by Bayesian analysis. Within columns, different letters are significantly different (McNemar's P < .05). Cells without superscripts had no significant differences.
Abbreviations: IFA, MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay; SNAP, IDEXX Giardia Antigen Test; Vet Chek, TECHLAB VETCHEK ELISA; VetScan, Abaxis Giardia Antigen test; ZnSO4, zinc sulfate centrifugal fecal flotation.
The analysis of the feline data using the Bayesian analysis is presented in Table 6. The most sensitive and specific diagnostic test was the IFA (99.9% and 99.8%, respectively). In cats, the sensitivity of the ZnSO4 flotation and SNAP were similar (92.9% and 91.1%, respectively) as were their specificities (98.5% and 98.8%, respectively). The sensitivity of the VetChek was 94.4%, and specificity was 95.7%. Similar to its performance in dogs, the SNAP (98.8%) followed the IFA in specificity.
Table 6.
Bayesian analysis estimates of diagnostic test parameters in cats
| Diagnostic test | 95% confidence interval | |
|---|---|---|
| Sensitivity | Specificity | |
| IFA | 99.9 (99.6‐100) | 99.8b (99.8‐99.9) |
| SNAP | 91.1a (82.7‐96.9) | 98.8ab (95.7‐100) |
| VetChek | 94.4a (86.4‐99.1) | 95.7a (90.5‐98.9) |
| ZnSO4 | 92.9a (83.3‐98.6) | 98.5ab (95.5‐99.9) |
The sensitivity and specificity of each test in cats with 95% confidence intervals (CI) estimated by Bayesian analysis. Within columns different letters are significantly different (McNemar's P < .05). Cells without superscripts had no significant differences.
Abbreviations: IFA, MERIFLUOR Cryptosporidium/Giardia direct immunofluorescent assay; SNAP, IDEXX Giardia Antigen Test; Vet Chek, TECHLAB VETCHEK ELISA; VetScan, Abaxis Giardia Antigen test; ZnSO4, zinc sulfate centrifugal fecal flotation.
4. DISCUSSION
4.1. Importance of results
This study evaluated diagnostic tests for G. duodenalis detection in dogs and cats using both comparison to a reference test and the Bayesian methodology for data analysis. The Bayesian analysis in this study correlated well with the direct comparison to the IFA, which provides support for previous studies comparing results to the IFA. Overall, when the diagnostic tests were compared to the IFA as the reference test in both cats and dogs, no single test stood out as an obvious best diagnostic, as all of the tests had relatively high sensitivities and specificities. Although there were significant differences in sensitivity between the in‐clinic immunoassays (SNAP and VetScan) when compared to the IFA in dogs, this difference was no longer significant when the antigen test results were combined with the ZnSO4 flotation results, and it was not statistically significant on the Bayesian analysis either. This data strongly supports the recommendation to use centrifugal fecal flotation in conjunction with an immunoassay for diagnosing G. duodenalis infections in veterinary practices. Compared to other diagnostic evaluations in dogs, our study had a much higher ZnSO4 flotation sensitivity.4, 10, 11 However, because we screened samples using ZnSO4 and included positives with a matching negative, this could have increased the ZnSO4 sensitivity in this study compared to other studies. Additionally, positive and negative predictive values are not displayed for this study as these variables are reliant on the prevalence of the disease being tested for and can vary greatly among populations. This study had an artificially high prevalence of G. duodenalis as we selected positive samples for inclusion in the study to allow for a more robust comparison.
This study also confirms the value of the performing more than 1 test when G. duodenalis infection is suspected. There were several samples with discordant results (Table 2). In particular, there were 5 samples (4 canine and 1 feline) positive by ZnSO4 fecal flotation but negative on the IFA and all other immunoassays, and an additional 2 samples (both canine) that were positive on the ZnSO4 fecal flotation and the VetChek but negative on the IFA, SNAP, and VetScan. There were also 4 samples (all canine) that were negative on the IFA but positive on all other diagnostic tests. It seems likely that for at least some of these samples, the IFA results are false negatives. However, it is also possible that the refrigeration, storage, and shipping of samples between the time of fecal flotation and the performance of the immunoassays contributed to some of the discrepancies between the positive ZnSO4 fecal flotation and negative immunoassay results, particularly in samples with low cyst numbers. These discordant results demonstrate the difficulty in determining which diagnostic test is most accurate when comparing diagnostic tests without a true gold standard for detecting G. duodenalis.
4.2. Comparison to IFA
VetChek, the most recently licensed immunoassay included in the study, was the most sensitive test in dogs when compared to IFA. When compared to IFA, the sensitivity of the SNAP test was determined to be 92.5% in cats and 89.8% in dogs, which is lower than reported on the package insert (95%) but is similar to other reports.4, 10, 11
4.3. Bayesian analysis
Sensitivities and specificities for the IFA, ZnSO4, and SNAP were higher in the Bayesian analysis than have been reported by others.4, 10, 11 Sensitivity for the SNAP test was 92.2%, which is much higher than the reported sensitivity (52% and 67%) in a previous analysis.10 The estimated sensitivity of the ZnSO4 from the Bayesian analysis was also much higher in this study (90.95%) compared to others (34% and 65%).10 A possible explanation for these differences is that more noninformative priors were used in previous studies,10 which has minimal impact on the variables in question when building a model. We were able to utilize prior information from the previous studies,4, 10, 11 which could have resulted in a more rigorous analysis in the present study.
Other studies have also found the IFA to be the most sensitive test when comparing tests using a Bayesian analysis.10, 17 This underscores the high performance of the IFA and provides more evidence to support the use of IFA as the de facto reference test even though it is not a true gold standard with potential for false negatives/positives.
4.4. Practical application of results
When evaluating diagnostic tests, it is important to consider the test purpose. The SNAP and VetScan are both rapid in‐clinic tests that require no additional equipment. Although the ZnSO4 centrifugal fecal flotation test has a short turnaround time, it does require a centrifuge. Although not designed as an in‐clinic rapid test, VetChek performed as well as other currently available diagnostics by both direct comparison to the IFA and Bayesian analysis and can be considered a sensitive and specific test for G. duodenalis detection. The IFA requires the most specialized equipment and training and is not available in the veterinary practice.
In conclusion, there are now several highly sensitive and specific antigen tests that are optimized for detecting G. duodenalis in companion animals. However, only centrifugal fecal flotation has the ability to detect other parasites that could be present, and a combination of both immunoassay and microscopic techniques provides the most sensitive procedure for detection of infection with G. duodenalis and other internal parasites.
CONFLICT OF INTEREST DECLARATION
Anne M. Zajac has received honoraria from IDEXX and research support from TechLab, Inc. Meriam N. Saleh received travel funds and partial stipend support from TechLab, Inc. David S. Lindsay has received research support from IDEXX. Joel F. Herbein is currently an employee of TechLab, Inc, and at the time the research was conducted, Jack R. Heptinstall was employed by TechLab, Inc.
OFF‐LABEL ANTIMICROBIAL DECLARATION
Authors declare no off‐label use of antimicrobials.
INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Authors declare no IACUC or other approval was needed. Fecal samples were collected by animal owners or environments without animal handling.
HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.
ACKNOWLEDGMENTS
This study was supported by funds from TechLab, Inc, Blacksburg, VA. Portions of this work were presented by Meriam Saleh at the 60th Annual Meeting of the American Association of Veterinary Parasitologists in Boston, MA, July 10‐14, 2014, and the 64th Annual Meeting of the American Association of Veterinary Parasitologists in Denver CO July 14‐17, 2018.
Saleh MN, Heptinstall JR, Johnson EM, et al. Comparison of diagnostic techniques for detection of Giardia duodenalis in dogs and cats. J Vet Intern Med. 2019;33:1272–1277. 10.1111/jvim.15491
Present address Meriam N. Saleh, Oklahoma State University, Stillwater, Oklahoma.
REFERENCES
- 1. Ballweber LR, Xiao LH, Bowman DD, Kahn G, Cama VA. Giardiasis in dogs and cats: update on epidemiology and public health significance. Trends Parasitol. 2010;26:180‐189. [DOI] [PubMed] [Google Scholar]
- 2. Leib MS, Zajac AM. Giardiasis in dogs and cats. Vet Med. 1999;94:793. [Google Scholar]
- 3. Thompson RCA, Palmer CS, O'Handley R. The public health and clinical significance of Giardia and Cryptosporidium in domestic animals. Vet J. 2008;177:18‐25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Rishniw M, Liotta J, Bellosa M, Bowman D, Simpson KW. Comparison of 4 Giardia diagnostic tests in diagnosis of naturally acquired canine chronic subclinical giardiasis. J Vet Intern Med. 2010;24:293‐297. [DOI] [PubMed] [Google Scholar]
- 5. Bowman DD. Georgis' Parasitology for Veterinarians. 10th ed. St. Louis, MO: Elsevier; 2014. [Google Scholar]
- 6. Zimmer JF, Burrington DB. Comparison of four protocols for the treatment of canine giardiasis. J Am Anim Hosp Assoc. 1986;22:168‐172. [Google Scholar]
- 7. Kirkpatrick CE, Farrell JP. Feline giardiasis: observations on natural and induced infections. Am J Vet Res. 1984;45:2182‐2188. [PubMed] [Google Scholar]
- 8. Barr SC, Bowman DD. Giardiasis in dogs and cats. Comp Cont Educ Pract Vet. 1994;16:603. [Google Scholar]
- 9. Alles AJ, Waldron MA, Sierra LS, Mattia AR. Prospective comparison of direct immunofluorescence and conventional staining methods for detection of Giardia and Cryptosporidium spp. in human fecal specimens. J Clin Microbiol. 1995;33:1632‐1634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Geurden T, Berkvens D, Casaert S, Vercruysse J, Claerebout E. A Bayesian evaluation of three diagnostic assays for the detection of Giardia duodenalis in symptomatic and asymptomatic dogs. Vet Parasitol. 2008;157:14‐20. [DOI] [PubMed] [Google Scholar]
- 11. Mekaru SR, Marks SL, Felley AJ, Chouicha N, Kass PH. Comparison of direct immunofluorescence, immunoassays, and fecal flotation for detection of Cryptosporidium spp. and Giardia spp. in naturally exposed cats in 4 northern California animal shelters. J Vet Intern Med. 2007;21:959‐965. [DOI] [PubMed] [Google Scholar]
- 12. Rimhanen‐Finne R, Enemark HL, Kolehmainen J, Toropainen P, Hänninen ML. Evaluation of immunofluorescence microscopy and enzyme‐linked immunosorbent assay in detection of Cryptosporidium and Giardia infections in asymptomatic dogs. Vet Parasitol. 2007;145:345‐348. [DOI] [PubMed] [Google Scholar]
- 13. Dryden MW, Payne PA, Smith V. Accurate diagnosis of Giardia spp and proper fecal examination procedures. Vet Ther. 2006;7:4‐14. [PubMed] [Google Scholar]
- 14. Dohoo IR, Martin SW, Stryhn H. Methods in Epidemiologic Research. Charlottetown, Canada: VER Inc; 2012. [Google Scholar]
- 15. Papini R, Carreras G, Marangi M, Mancianti F, Giangaspero A. Use of a commercial enzyme‐linked immunosorbent assay for rapid detection of Giardia duodenalis in dog stools in the environment: a Bayesian evaluation. J Vet Diagn Invest. 2013;25:418‐422. [DOI] [PubMed] [Google Scholar]
- 16. Tangtrongsup S, Scorza V. Update on the diagnosis and management of Giardia spp infections in dogs and cats. Top Companion Anim Med. 2010;25:155‐162. [DOI] [PubMed] [Google Scholar]
- 17. Geurden T, Levecke B, Pohle H, de Wilde N, Vercruysse J, Claerebout E. A Bayesian evaluation of two dip‐stick assays for the on‐site diagnosis of infection in calves suspected of clinical giardiasis. Vet Parasitol. 2010;172:337‐340. [DOI] [PubMed] [Google Scholar]
- 18. Zajac AM, Conboy GA. Veterinary Clinical Parasitology. 8th ed. Chichester, West Sussex, England: John Wiley & Sons Inc; 2012:4‐8. [Google Scholar]
- 19. Adedokun OA, Burgess WD. Analysis of paired dichotomous data: a gentle introduction to the McNemar test in SPSS. J Multidiscip Eval. 2012;8(17):2011. [Google Scholar]
- 20. Joseph L, Gyorkos TW. Inferences for likelihood ratios in the absence of a “gold standard”. Med Decis Making. 1996;16:412‐417. [DOI] [PubMed] [Google Scholar]
- 21. Joseph L, Gyorkos TW, Coupal L. Bayesian estimation of disease prevalence and the parameters of diagnostic tests in the absence of a gold standard. Am J Epidemiol. 1995;141:263‐272. [DOI] [PubMed] [Google Scholar]