ESGUERRA, STEPHANIE DL.

YL 3; FCM III

CLINICAL SCENARIO A 19-year old single female from Quezon City came in for intermittent fever of six days duration accompanied by headache, muscle and joint pains and body malaise. No bleeding, abdominal pain, cough, colds or vomiting was noted. The patient self-medicated with paracetamol, which gave partial relief to symptoms. Patient denies any wading in flooded waters. She came in with a CBC: Hb 140, Hct 0.44, WBC 4.1, platelet count 205. She is seeking second opinion to make sure she is not having dengue and requested for a better test. On physical exam, patient looks acutely ill. Vital signs: BP 110/70, HR 101, RR 21, temp 38.8C. There was no tonsillopharyngeal congestion, no lymphadenopathies, no nasoaural discharge. No significant findings on chest and heart. Abdomen was soft, no tenderness, no ascites, and no organomegaly. There was no edema, no rashes in the extremities. Pulses are full and equal. Tourniquest test was not done. You are considering NS1 test which is available in your hospital but is quite reluctant. Questions: What is your therapeutic threshold for Dengue Fever? 70% What is your diagnostic threshold for Dengue Fever? 20% What is your pre-test probability for Dengue in this particular case? 50% Searching the internet, you found this particular journal. Appraise the article.

Journal Title:

Multi-Country Evaluation of the Sensitivity and Specificity of Two Commercially-Available NS1 ELISA Assays for Dengue Diagnosis Journal Name: PLoS Neglected Tropical Diseases Date Published: August 31, 2010 Pre-Test Probability = 50% Therapeutic Threshold = 70% Diagnostic Threshold = 20% I. Are the results of the study valid? A. Primary Validity Guides 1. Was there an independent, blind Yes comparison with a reference No standard? Cannot tell

2. Did the patient sample include an appropriate spectrum of patients to whom the diagnostic test will be applied in clinical practice?

Yes No Cannot tell

There was comparison of the intervention to be tested (NS1 Antigen) with a reference standard: virus isolation, RT/PCR, IgM, IgG, and total antibody seroconversion. Statistical analysis of the data obtained was done by two independent data entry facilities: in Guatamala and in Thailand; which was then merged by a group from Germany. It was not clear though whether blinding was done to any of the data collectors and analyzers, but the use of a reference test and independent analyzers is sufficient for me to answer yes. The study included children (from 6 months) to adult age (which presents a good spectrum of differences in age) whom were suspected clinically as having dengue, and fever for less than 7 days. Patients were then classified using a former WHO criteria for DF, DHF, and DSS (which represents a good spectrum of dengue disease presentation). All the patients, from the initial inclusion criteria (clinical suspects of having dengue) were all tested using the reference tests regardless if they really had the disease or not. It was only after being confirmed with the disease were they divided to two groups for the interventions. The interventions and comparators (reference standards) were all stated, and the instruction of their use as well as geographic location on where they are found and utilized.

B. Secondary Validity Guides 1. Did the results of the test being evaluated influence the decision to perform the reference standard?

Yes No Cannot tell

2. Were the methods for Yes performing the test described in No sufficient detail to permit Cannot tell replication? II. What were the results? Are the likelihood ratios for the test results presented or data necessary for their calculation provided?

Sensitivity and specificity values comparing the interventions (Pan E and Patella) to the reference tests were given. Pan-E Dengue Early ELISA Sensitivity = 52% Specificity = 100% (+) LR = .52 / 1-1 = 0.52 Platelia Dengue NS1 Ag Assay Sensitivity = 66% Specificity = 100% (+) LR = .66 / 1-1 = 0.67 *Note: 100% specificity poses a problem in rejection of the true null hypothesis

III. Will the results help me in caring for my patients? 1. Will the reproducibility of the test Yes The diagnostic test seems fairly easy to use and result and its interpretation be interpretation can easily be done (with a guide from the No satisfactory in my setting? Cannot tell manufacturers protocol). Information on where to acquire the diagnostic tests was given in the study (from Australia and France). 2. Are the results applicable to my Yes My patient is within the inclusion criteria of the study. patient? The population size was also obtained from South East No Cannot tell Asia. Although it seemed that there is low sensitivity for the test from the population sample from The Philippines, I preferred to use the overall result to remove any possible bias on choosing a specific population for the use of the NS1 tests. 3. Will the results change my Based on the nomogram (plotted at the end of the Yes management? page), both NS1 tests yielded ~40% post-test No Cannot tell probability values. It seems that I would still need other tests to truly diagnose my patient with having or not having the disease. 4. Will patients be better off as a Since the diagnostic test would not yield a strong Yes result of the test? diagnosis for me to either treat my patient with dengue No Cannot tell or to give me a decision to change my differential, I would not use NS1 (my patient would not benefit from the test if I use it, I would need a test with higher sensitivity or one with less possibility of having a type 1 error [since the study indicated a 100% specificity for the diagnostic tests])

Pan-E Dengue Early ELISA

Platelia Dengue NS1 Ag Assay