Critical Appraisal of an Article about a Diagnostic Test

A) Basic Points

Q1. What is the Aim of the Study?

Validating a predictive clinical model for the diagnosis of Group A –hemolytic streptococcal pharyngitis in children in different clinical settings and during different seasons..

Q2. What are the Criteria for Selecting Patients?

Cases: Children aged between 1 and 18 years with signs and symptoms of pharyngitis on initial examination attending a children's hospital ED and 2 practicing pediatric OP clinics between April 1, 1999, and March 31, 2000.

Controls: Children aged between 1 and 18 years with noninfectious, non-respiratory tract complaints such as lacerations, and musculoskeletal injuries, or well-child visits attending a children's hospital ED and 2 practicing pediatric OP clinics between April 1, 1999, and March 31, 2000.

Exclusion criteria:All patients and controls who had received antibiotic therapy within 5 days and those who were previously enrolled.

Q3. What is the Diagnostic Test Used?

Clinical predictive model including 4 variables: Moderate to severe tonsillar swelling, Cervical lymph-adenopathy (moderate to severe tenderness and enlargement of cervical lymph nodes), scarletiniform rash, and the absence of coryza.

Q4. What is the Gold (Reference) Standard?

Microbiological testing and culture for tonsillo-pharyngeal specimens.

Q5. What is the Pretest Probability of the Condition (Prevalence) in the Population Studied before the Application of the Diagnostic Test?

Thirty seven percent.

B) Worksheet for Using an Article About Assessing Diagnostic Tests

I) Are the Result of the Study Valid?

1. Was there an independent, blind comparison with a reference Standard?

Yes.

Is reference standard used acceptable?

Yes.

Were both reference standard and test applied to all patients?

Yes.

Were assessors blind to other test results?

Not mentioned clearly.

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

Yes.

Setting:

Community based!

Children attending children's hospital ED and 2 practicing pediatric OP clinics.

Spectrum of patients:

Various disease levels and characteristics.

Yes.

Various patient characteristics (age, ethnic gp …).

Yes.

Timing:

All through the year.

3. Did the result of the test being evaluated influence the decision to perform the reference standard?

"Verification" or "work-up" bias!

No.

4. Were the methods for performing the test described in sufficient details to permit replication?

Yes, but subjective evaluation.

Preparation of patients!

Yes.

Performing the test!

Subjectivity (tonsillar swelling: absent/mild, moderate/severe).

Analysis and interpretation of results!

Yes.

5. Overall, are the results of the study valid?

Yes.

II) What were the results?

1. Are likely hood ratios for the test results presented or data necessary for their calculation provided?

Yes.

How big or small is the likelihood ratio?

Likelihood ratio of a positive test: 5.9 (3.9-11.5).

Likelihood ratio of a negative test: 0.2 (0.1-0.8).

Magnitude of change from pre- to post-test probability.

For a positive test result: from 37% to 79%.

For a negative test result: from 37% to 12%.

III) Will the results help me in caring for my patients?

1. Will the reproducibility of  the test result and its interpretation be satisfactory in my setting?

Yes, but it is subjective evaluation.

Does the test require special skills?

No, simple evaluation carried out by physicians with different and wide range of experience.

Does it require interpretation that causes disagreement (e.g. ECG)?

Yes, (subjective evaluation with possible disagreement).

Will it vary in my setting (e.g. different reagent)?

No, but subjective evaluation.

2. Are the results applicable to my patient?

Yes, but the prevalence of rheumatic heart diseases and other complications could be different.

Similar distribution of disease severity!

Yes.

Similar distribution of competing diseases!

Yes.

Compelling reasons why the results should not be applied?

No.

3. Will the results change my management?

Yes.

The way to manage similar cases.

Develop protocol.

Test and treatment threshold!

Test threshold: score of zero.

Treatment threshold: score of four or more.

High or low Likelihood ratios!

Moderate likelihood ratio (5-10): 5.9 for a positive test.

4. Will patients be better off as a result of the test?

Yes.

Is target disorder dangerous if left undiagnosed?

Yes, (complications).

Is test risk acceptable?

Yes.

Does effective treatment exist?

Yes.

Information from test will lead to change of management beneficial to patient!

Yes.