Objective To analyze the epidemiological characteristics of pneumonia in Gansu province from 2019 to 2024, and provide reference for the pilot reform of “medical treatment and prevention integration” in Gansu.

Methods The incidence data of pneumonia in Gansu during this period were collected from the Gansu Provincial National Health Information Platform, and spectral analysis and mutation point detection algorithms were used for signal analysis to explore the new epidemiological characteristics of pneumonia.

Results The signal wave formed by the number of pneumonia patients in Gansu from 2019 to 2024 was a typical unstable sine one. The amplitude fluctuated greatly, and the waveform showed no obvious periodicity. During the surveillance period, there were abnormal fluctuations or transient disturbances in the number of pneumonia patients seeking treatment in Gansu. The signal wave formed by the number of hospital visits of patients with pneumonia was similar to a Delta one, indicating the incidence of abnormal conditions. The number of hospital visits of the pneumonia patients showed a low-frequency dominant characteristic, with an amplitude of 150000 within the frequency range of 1.098−4.277 Hz. The number of the hospital visits was affected by resonance and periodicity, indicating the presence of commodities and regular influencing factors. The 10 mutation points detected and analyzed were almost consistent with the adjustment of public health policies and the incidences of key public health emergencies in China.

Conclusion Application of signal analysis and result interpretation in disease surveillance and early warning can benefit the rapid analysis on the periodicity of diseases and identification of more causal factors. The impacts of resonance and periodicity are the shortcomings of traditional epidemiological research, and signal analysis can intuitively reflect these two hidden phenomena of diseases. Importantly, the detection of mutation points has great public health significance, especially in the field of disease surveillance, by which key nodes of “important events” can be identified and health policy-making can be informed.