An algorithmic approach to chronic dyspnea

Abstract

Question: The objective of the study was to prospectively evaluate an algorithmic approach to the cause(s) of chronic dyspnea. MATERIALS/PATIENTS/METHODS: Prospective observational study. The study group consisted of 123 patients with a chief complaint of dyspnea of unknown cause present for >8 weeks. Dyspnea severity scores were documented at entry and after therapy. Patients underwent an algorithmic approach to dyspnea. Therapy could be instituted at any time that data supported a treatable diagnosis. Whenever possible, accuracy of diagnosis was confirmed with an improvement in dyspnea after therapy. Tests required, spectrum and frequency of diagnoses, and the values of individual tests were determined.

Results: Cause(s) was(were) diagnosed in 122/123 patients (99%); 97 patients had one diagnosis and 25 two diagnoses. Fifty-three percent of diagnoses were respiratory and 47% were non-respiratory. Following therapy, dyspnea improved in 63% of patients.

Conclusions: The prospective algorithmic approach led to diagnoses in 99% of cases. A third of patients were diagnosed with each tier of the algorithm, thus minimizing the need for invasive testing. Specific diagnoses led to improvement in dyspnea in the majority of cases. Based on the results of this study, the algorithm can be revised to further minimize unnecessary tests without loss of diagnostic accuracy.