摘要：Background: Preventive measures on the COVID-19 pandemic is an effective way to control its spread. We aimed to investigate the effect of control measures and holiday seasons on the incidence and mortality rate of COVID-19 in Iran. Study design: An observational study. Methods: The daily data of confirmed new cases and deaths in Iran were taken from the Johns Hopkins University COVID-19 database. We calculated weekly data from 19 Feb to 6 Oct 2020. To estimate the impact of control measures and holiday seasons on the incidence rate of new cases and deaths, an autoregressive hidden Markov model (ARHMM) with two hidden states fitted the data. The hidden states of the fitted model can distinguish the peak period from the non-peak period. Results: The control measures with a delay of one-week and two-week had a decreasing effect on the new cases in the peak and non-peak periods, respectively ( P =0.005). The holiday season with a two-week delay increased the total number of new cases in the peak periods ( P =0.031). The peak period for the occurrence of COVID-19 was estimated at 3 weeks. In the peak period of mortality, the control measures with a three-week delay decreased the COVID-19 mortality ( P =0.010). The expected duration of staying in the peak period of mortality was around 6 weeks. Conclusions: When an increasing trend was seen in the country, the control measures could decline the incidence and mortality related to COVID-19. Implementation of official restrictions on holiday seasons could prevent an upward trend of incidence for COVID-19 during the peak period.
关键词：Autoregressive Hidden Markov Model (ARHMM);Holiday seasons;Control measures;COVID-19
