Temporal risk of coronavirus disease 2019 in Gansu province

  Objective  To describe the temporal risk characteristics of coronavirus disease 2019 (COVID-19) in Gansu province.

  Methods  The information about confirmed COVID-19 cases reported in Gansu as of February 16, 2020 and population data of local counties (districts) were collected to calculate temporal risk of frequency index (α), duration frequency (β) and intensity index (γ) of COVID-19 in Gansu with spatial statistic.

  Results  The Moran's I coefficient of α of COVID-19 in Gansu was 0.115 by global autocorrelation analysis, suggesting that the probability of incidence frequency increase was low. The Moran's I coefficient of γ of COVID-19 in Gansu was 0.070, which showed that the confirmed cases would have no clustering and the epidemic intensity was low. Frequency index (α) and duration frequency (β) had global spatial autocorrelation (P<0.05), while intensity index (γ) had no global spatial autocorrelation (P>0.05). Frequency index (α) showed that there was high-high clustering in local spatial autocorrelation analysis (mainly in Chengguan, Qilihe, Anning, Honggu districts, and Gaolan and Yuzhong counties). There was high-high clustering in Chengguan, Qilihe, Xigu, Anning districts, and Gaolan and Yuzhong countries indicated by duration frequency (β) in local spatial autocorrelation analysis.

  Conclusion  The majority of counties in Lanzhou showed high risk of clustering of COVID-19, but the intensity was low. The prevention and control measures were timely and effective during COVID-19 epidemic period.