Syllabi - Child Health - Diagnosis Scenario

A five week old baby boy has been admitted with projectile vomiting. You are unable to palpate a pyloric tumour. You decide to admit the child and observe him at least for the next 24 hours. However, the parents are keen to take their child home now.

You ask for an ultrasound of the pylorus and the radiologist reports the result as negative. Do you send the child home? You pose the question, in young infants with projectile vomiting and no palpable pyloric tumour, what is the probability of pyloric stenosis with a negative or a positive ultrasound of the pylorus?

You search Medline using the terms "pyloric stenosis" and "ultrasound" and find the following paper:

Neilson D, Hollman AS.
The ultrasonic diagnosis of infantile hypertrophic pyloric stenosis: technique and accuracy.

Clinical Radiology 1994;49:246-7.

Read the article and decide:

• Are the results of this diagnostic article valid?
• Are the valid results of this diagnostic study important?
• Can you apply this valid, important evidence about a diagnostic test in caring for your patient?

Completed Diagnosis Worksheet for Child Health

Clinical Question

In infants with projectile vomiting in whom there is no palpable tumour, does ultrasound aid in diagnosis (rule in or out) of pyloric stenosis?

Are the results of this diagnostic study valid?

Was there an independent, blind comparison with a reference ("gold") standard of diagnosis?

Yes. All followed until hospital discharge. Length for follow up not given.

Was the diagnostic test evaluated in an appropriate spectrum of patients (like those in whom it would be used in practice)?

Yes. In infants with an equivocal diagnosis.

Was the reference standard applied regardless of the diagnostic test result?

Yes. All followed up until discharge. We assume that pyloric stenosis will not resolve spontaneously, this may not be true.

Are the valid results of this diagnostic study important?

Your calculations:

		Target Disorder		Totals
		Present	Absent
Diagnostic test result (exam)	Positive	66 a	1 b	67 a + b
	Negative	2 c	78 d	80 c + d
Totals		a + c 68	b + d 79	a + b + c + d 147

Sensitivity = a/(a+c)
= 97.1%

Specificity = d/(b+d)
= 98.7%

Likelihood Ratio for a positive test result (LR+) = sens/(1-spec)
= 75

Likelihood Ratio for a negative test result (LR-) = (1-sens)/spec
= 0.03

Positive Predictive Value = a/(a+b)
= 99%

Negative Predictive Value = d/(c+d)
= 98%

Pre-test Probability (prevalence) = (a+c)/(a+b+c+d)
= 46%

Pre-test-odds = prevalence/(1-prevalence)
= 0.85

Post-test odds for a negative result = Pre-test odds x Likelihood Ratio
= 0.85 x 0.03 = 0.0255

Post-test Probability for a negative result = Post-test odds/(Post-test odds + 1)
= 2.5%

Can you apply this valid, important evidence about a diagnostic test in caring for your patient?

Is the diagnostic test available, affordable, accurate, and precise in your setting?

Yes.

Can you generate a clinically sensible estimate of your patient's pre-test probability (from practice data, from personal experience, from the report itself, or from clinical speculation)?

Yes. Could audit own practice if don't feel 46% of babies with projectile vomiting and no tumour palpable is realistic.

Will the resulting post-test probabilities affect your management and help your patient? (Could it move you across a test-treatment threshold?; Would your patient be a willing partner in carrying it out?)

Depends on results. Negative test means post-test probability now < 5%, and you would be happy for baby to go home. Both +ve and -ve tests move patient across treatment thresholds.

Would the consequences of the test help your patient?

Yes. Earlier discharge if negative. earlier surgery if positive.

Additional Notes

• This is a SpPin (Specificity = 99% so positive USS rules in diagnosis)
• In fact, there were only 142 patients and 5 of USS were re-examinations. Only first USS should have been included in results. If read text, it is apparent that repeat scans were performed mostly on true positive or true negative cases, which means sensitivity and specificity will not be altered greatly.
• If the surgeon knew the result of the USS (ie. not blind), this might exaggerate sensitivity and specificity.
• Emphasise the importance of thinking about the confidence intervals around the likelihood ratios.

Pyloric stenosis - Ultrasound is diagnostic

Clinical Bottom Line

In young infants with projective vomiting but no palpable pyloric tumour, ultrasound is useful to rule in and rule out pyloric stenosis.

Citation

Neilson D, Hollman AS. The ultrasonic diagnosis of infantile hypertrophic pyloric stenosis: technique and accuracy. Clinical Radiology 1994;49:246-247

Clinical Question

In young infants with projectile vomiting and no palpable pyloric tumour, what is the probability of hypertrophic pyloric stenosis with a negative or a positive ultrasound of the pylorus?

Search Terms

'pyloric-stenosis' and 'infant' and 'ultrasound' and ('diagnosis' or 'sensitivity-and-specificity')

The Study

• Reference-standard - review of final diagnosis (time after test not stated) and operative findings applied to all.
• Test - Ultrasound scan (USS) of the pylorus: considered positive if pyloric canal length ≥ 16 mm, diameter of pylorus ≥ 11 mm, muscle;thickness ≥ 2.5 mm and/or dynamic appearance of pylorus.
• Study setting - retrospective audit of infants less than 5 months old who had projectile vomiting, no clearly palpable pyloric tumour and who were referred for ultrasound.

The Evidence

Reference standard - pyloric stenosis at follow-up

		+	-
Test	+	66	1	67
USS	-	2	78	80
		68	79	147

			95% CI
LR+		77	11 to 538
LR-		0.03	0.01 to 0.12
Pre-test probability		46%	38% to 54%
Post-test probability	Test +	99%	87% to 100%
	Test -	2.5%	0% to 12%

Comments

• Surgeons deciding diagnosis were not blind to the ultrasound result.
• Unclear whether positive test result based on one or all of above criteria.
• Test was always performed by consultant paediatric radiologists.