Critical Analysis of A Report of a Diagnostic or Screening ...



M8120 INFORMATICS FOR EVIDENCE-BASED PRACTICE

Critical Analysis of A Report of a Diagnostic or Screening Test

Answers

Tribble, D.R., Rodier, G.R., Saad, M.D., Binson, G., Marrot, F., Salah, S., Omar, C., Arthur, R.R. (1997) Comparative field evaluation of HIV rapid diagnostic assays using serum, urine, and oral mucosal transudate specimens. Clinical and Diagnostic Virology, 7, 127-132.

Are the results of the study valid?

1. What diagnostic tests were being evaluated?

Serum: Murex suds, Span combaids

Saliva OMT: EIA, Murex suds

Urine (immediate and repeat): EIA, Murex suds, Urine testpack

2. Was there an independent, blind comparison with a reference (gold standard)? If so, what was the gold standard test?

HIV 1 / 2 enzyme immunoassay (EIA) with confirmatory Western blot assay was the gold standard. It is not clear from the description whether the comparison was independent and blind since it was done by personnel from the same unit (NAMRU-3).

3. Who was the sample?

Consenting adults from a tuberculosis hospital (n=114) or STD clinic (n=189) in the Republic of Djibouti.

4. What were the inclusion and exclusion criteria?

No inclusion or exclusion criteria were reported beyond the fact that the subjects were consenting adults from the two settings.

5. Did the patient sample include an appropriate spectrum of patients to whom the diagnostic test will be applied in clinical practice? How or how not?

The sample did not include an appropriate spectrum of patients to whom the diagnostic test would be applied in clinical practice. The sample included groups with high prevalence rates. Persons beyond those with or at risk for TB and/or STDs are at risk for HIV disease. Although data was not provided, given the location of the study the sample was most likely not representative in terms of race.

6. Were the methods for performing the tests described in sufficient detail to permit replication? How or how not?

Yes, the methods of specimen collection, processing, and laboratory test procedures were described in sufficient detail. Specific devices, samples, temperatures, times of storage, etc. were provided.

What are the results?

7. Are likelihood ratios for the tests presented? If not, what data is necessary to calculate the likelihood ratios and is it included?

Likelihood ratios are not presented. Likelihood ratios can be calculated from the TPR and FPR by using the formula TPR/FPR. Sensitivity = TPR and is given in the article. Specificity = 1 – FPR and specificity is given in the article. Thus, likelihood ratios can be calculated from the information in the article.

8. Which test(s) had the lowest false positive rate(s)?

Specificity = 1 – FPR, thus tests with highest specificity have lowest FPRs. Three tests had a specificity of 99.6 (FPR=.04): Span combaids, urine EIA, and Murex suds.

Will the results help me in caring for my patient?

9. What would you tell the patient who received a negative test result using Murex suds in response to the question: Do I have HIV?

The negative predictive values (NPV) of Murex suds tests on all specimens is high (96.9-100). Thus, there is a high level of confidence that the patient does not have HIV. No repeat test would be warranted, particularly if the test was done on serum.

10. What would you tell the patient who received a positive test result using EIA on a preserved urine specimen in response to the question: Do I have HIV?

The positive predictive value (PPV) of EIA in urine is 75.7. The patient should be re-tested with a serum EIA and if positive, a confirmatory Western blot.

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

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download