HEARTBURN AND DYSPEPSIA - THE UTILITY OF SYMPTOM ANALYSIS

Patients presenting with dyspepsia and heartburn are a heterogeneous g roup and a more precise identification of the cause of the symptoms is a prerequisite for a rational treatment. The patient's history is the basic diagnostic tool. Therefore it is important to evaluate the util ity of symptoms as a predictor of any organic disease, like peptic ulc er, or as a predictor of a favourable symptomatic response to a specif ic drug treatment, such as omeprazole. When the history alone was used to discriminate between peptic ulcer and non-ulcer dyspepsia half of the patients with endoscopically-confirmed active ulcer were misdiagno sed as not having ulcers. Furthermore, in two-thirds of the patients i n whom a clinical diagnosis of peptic ulcer was made, this was not con firmed at endoscopy. Heartburn is recognized as a symptom of reflux oe sophagitis, but most patients complaining of heartburn have no visible mucosal lesion at endoscopy. The predictive value of heartburn depend s on its severity; one-half of patients with severe heartburn, in fact , have oesophagitis. When symptoms of reflux oesophagitis are used in a scoring system, the diagnostic sensitivity is about 50%, while the s pecificity is 34-85%, depending on the severity of symptoms. When a co mplete patient history is used as a predictor, the diagnostic value in creases significantly. Patients with endoscopy-negative dyspepsia and heartburn are a common problem and acid inhibitory drugs are widely us ed to manage the disease. The rationale for prescribing an acid inhibi tor in this situation is the assumption that the symptoms are acid-rel ated, but this is only the case for a subgroup of these patients. Iden tification of those patients who have acid-related symptoms is possibl e using omeprazole as a diagnostic tool, either in an open design or i n a placebo-controlled single-subject trial. When heartburn is a predo minant symptom, 50% of the patients respond to the acid inhibitory tre atment, while this is the case for only one-third of patients when hea rtburn is not the dominating symptom. It is important to identify symp toms that, when present, significantly reduce the likelihood of a spec ific diagnosis or condition. This is the case with the symptom of loos e stools in patients with dyspepsia, where it reduces the likelihood o f a symptomatic improvement during treatment with omeprazole. In concl usion, symptoms do not reliably predict the underlying disorder, wheth er this is an organic disease or endoscopy-negative dyspepsia. Cluster s of symptoms or the more global clinical judgement seem to have a hig her discriminative value compared with a single symptom.