'Catching the spike and tracking the flow': Holter-temperature monitoring in patients admitted in a general internal medicine ward.

OBJECTIVE

To study if a 24-h continuous monitoring of temperature reveals information not accessible through conventional care. This included omitted fever peaks and circadian and complexity characteristics that may correlate with specific aetiologies.

DESIGN

Ours was a prospective, observational study. A total of 62 patients, admitted to a general internal medicine ward, in whom a temperature > 38 °C had been observed the day before inclusion underwent a 24-h long continuous monitoring of both central and peripheral temperatures. The time series were recorded in a file, while they otherwise followed conventional care. Time series were analysed for standard statistics, chronobiological analysis (amplitude, mesor, acrophase, intra-daily variability) and complexity analysis (Approximate Entropy of both central and peripheral temperature, cross-ApEn). A month after discharge, the clinical reports were reviewed and a definitive diagnosis of the febrile syndrome was established.

RESULTS

A total of 62 patients were initially included. In six cases, no time series could be obtained because of technical problems, leaving 56 patients accessible for analysis. In 10 cases, no definitive diagnosis was established. Continuous monitoring detected a mean of 0.7 (CI = 0.27-1.33) peaks of fever (central temperature > 38.0 °C) unobserved by conventional care per patient. A proportion of 16% (CI = 6-26) of patients considered afebrile by conventional care had at least one fever peak detected by continuous monitoring. Circadian rhythm persisted or was exacerbated in febrile patients. Circadian amplitude was increased in patients with tuberculosis. Complexity analysis did not differ among different diagnostic groups, although in subgroup analysis, viral infections had a higher complexity than other infectious diseases.

CONCLUSIONS

Temperature Holter monitoring reveals fever peaks that pass otherwise unobserved. Furthermore, chronobiological and complexity analysis of the temperature profile may provide quick and easy 'hidden information', not available to conventional care.