Six studies (n approximately 2,121) were included, some of which evaluated more than one diagnostic test.
Two studies scored 6 out of 8 for methodological quality and the rest scored 5. All studies had consecutive patient selection, prospective data collection and reported sufficient details of the test and control populations. Only one study compared the diagnostic test with a 'gold' standard test and one other study had blinded test interpretation. In all studies the diagnosis of IBS was made using only symptom-based criteria and before any diagnostic tests were applied.
Colonic examinations.
Four studies assessed flexible sigmoidoscopy, barium enema or colonoscopy with 0% (0 out of 89), 0% (0 out of 125), 1% (3 out of 196) and 1.3% (4 out of 306) of patients receiving alternative diagnoses to explain their GI based on the test results. One study assessed rectal biopsies and no patients from the sample of 89 suspected cases and 59 controls received an alternative diagnosis.
Laboratory evaluations.
Two studies assessed CBC, FOBT or serum chemistries, but neither study found that the laboratory test results led to an alternative diagnosis.
Stool analysis.
Two studies assessed stool analysis: one found that 1.7% (19 out of 1,154) of patients with suspected IBS had evidence of an intestinal pathogen; the other found no patients had enteric infection.
TSH testing.
Two studies assessed TSH: one found that 6% of patients (67 out of 1,200) had thyroid function abnormalities, but it was unclear if these were responsible for the IBS symptoms as there was no report of symptom relief after treatment (the gold standard); the other study found one patient (out of 171) with an abnormal TSH, similarly there was no report of symptom relief after treatment.
Hydrogen breath testing.
Two studies assessed hydrogen breath testing for lactose malabsorption in patients with suspected IBS. One found that 23% (256 out of 1,122) of patients with suspected IBS demonstrated impaired lactose absorption when given a 25-g dose of lactose. The other study found a similar prevalence of malabsorption with 25.8% of patients (48 out of 186) having abnormal results after a 50-g dose of lactose. Neither study reported symptom relief on treatment.
Imaging studies.
One study assessed abdominal ultrasound and found 20% (of 100 women) and 8% (of 25 men) had an abnormality following an abdominal scan, and 10% of women had pelvic abnormalities. The identification of an abnormality was not correlated to any GI symptoms and did not lead to additional treatment in any patient.