Doctors, patients and influenza-like illness: clinicians or patients at risk?

BACKGROUND: Healthcare providers can transmit influenza and influenza-like illness (ILI) to patients and vice versa. However, the magnitude of this problem in the healthcare system as a whole is unknown. Using population-based administrative health data, we tested the hypothesis of a temporal association of ILI diagnosis among clinicians and their patients. METHODS: Healthcare providers under study included physicians (85%) and a variety of other medical professionals (such as chiropractors, dentists and optometrists). Cases of ILI were defined as having an ICD-9 code of influenza (487), pneumonia (480-486) or bronchitis (466 and 490) in a province-wide healthcare-provider billing system. Rates of ILI among persons who saw a sick (case) and non-sick (control) doctor were calculated and compared. RESULTS: The rate of ILI was lower among providers than among patients for every year of the study. The mean number of exposures to patients diagnosed with ILI was higher among case providers than among control providers (P = 0.044). However, exposure to case providers did not significantly increase the risk of ILI diagnosis among patients (OR=1.11; 95% confidence intervals 0.85-1.36). INTERPRETATION: Our findings are consistent with the hypothesis of patient-to-provider transmission. However, we may have underestimated the association and magnitude of this effect. Our results do not support the hypothesis that infectious doctors transmit disease to their patients, although this may be partly accounted for by the lower rates of ILI among the provider population in general.