Background: The mouth is the most common site for measuring temperature. Oral mucositis may affect up to 40% of patients who receive chemotherapy. Thus far, mucositis has not been studied with regard to accurate thermometry.
Methods: One hundred consecutive patients (25 per group) were self-referred on the basis of perception of mucositis, temperature measurement at home, or malaise or were referred by a nurse on the basis of an assessment of a potential infectious process and/or mucositis. Oral and tympanic temperature were assessed simultaneously in patients with no fever and no mucositis (group A), those with mucositis (group B), those with fever and no mucositis (group C), and those with neutropenia but not fever or mucositis (group D).
Results: The 4 groups of patients had higher oral temperatures than tympanic temperatures (group A, 36.9 degrees C vs. 36.8 degrees C [delta = 0.12 degrees C; P = .062]; group B, 38.0 degrees C vs. 37.1 degrees C [delta = 0.89 degrees C; P < .001]; group C, 38.7 degrees C vs. 38.4 degrees C [delta = 0.27 degrees C; P < .001]; and group D, 37.0 degrees C vs. 36.7 degrees C [delta = 0.27 degrees C; P < .001]). Scheffe a-posteriori test revealed that only the delta temperature in group B differed significantly (95% confidence limits, -2.471, -1.584; P < .001). A linear regression model that examined the effect of other variables on the delta temperature found that only mucositis was a significant factor (95% confidence limits, 0.582, 0.820; P < .001).
Conclusions: Mucositis causes an increase in oral temperature but does not elevate systemic body temperature, thereby casting doubt on the diagnosis of infection. Conceivably, mucositis may provide an "inflammation bias" that could lead to the overuse of antibiotics and growth factors in 20%-40% of patients with cancer. When one considers issues of antimicrobial resistance and cost, this concern should be tested and clarified in a prospective study based on accurate temperature measurement.