Clostridium difficile infection is the leading cause of nosocomial diarrhea in developed countries and may progress to pseudomembranous colitis, sepsis and death. The risk factors are antibiotics use, advanced age and prolonged hospitalization. The diagnosis of Clostridium difficile infection is based on clinical history in combination with laboratory tests, which detect the Clostridium difficile presence or toxins. Clostridium difficile remains in spore form contaminating the environment and requiring measures to prevent hospital transmission. Tests with more accurate results to identify true carriers of Clostridium difficile allow the clinician to determine a safer treatment. This study evaluated accuracy and cost-effectiveness of the real-time polymerase chain reaction compared with the enzyme-linked immunosorbent assay from the perspective of a Brazilian public cardiology hospital.

A study diagram was constructed by type of test, linking the data of prevalence in hospital, accuracy and direct costs of tests. The costs were based on a hypothetical population comparing two strategies to identify the incremental expenditure between technologies. The analysis included comparisons for each test versus no test, and with each other. The prices were converted to the American currency taking into account the date of purchase of each product and respective price.

For real-time polymerase chain reaction test versus no test, 214 patients would have tested to justify one empirical treatment suspension, at a cost of USD90,926.46. For enzyme-linked immunosorbent assay test, to prevent one unnecessary treatment, 375 patients would have to be tested at a cost of USD6,603.75. In the comparative analysis, only a single false-positive patient would have the treatment suspended after performing 375 real-time polymerase chain reaction tests at USD424.89 each one (USD159,333.75 in total). An incremental cost of USD152,730.00 may be necessary to benefit a single patient by discontinuing empirical treatment.

The Real-time polymerase chain reaction test has restrictions as a test of choice for the diagnosis of Clostridium difficile infection, in services with low disease prevalence. It undergoes a significant change in its positive predictive value and does not offer a great impact in the clinical diagnosis.