Study design
According to organising experts in epidemiology, statistics, and health management to conduct sufficient discussion, and combined with the actual situation in Zhejiang province, we developed a method to calculate the epidemic risk in county areas as well as the subsequent adjustment rules. We selected the total number of confirmed cases, the proportion of local cases, and the clustered outbreaks (non-family and family clustered outbreaks) as positive indicators, accounting for 50%, 30%, and 20%, respectively, to calculate the county-level epidemic risk.
Meanwhile, the absence of newly confirmed cases for three consecutive days (or seven days) was selected as the negative indicator, which was used to adjust the county-level epidemic risk score.
In the initial stage of assessment, the total county-level epidemic risk score ( Si ) = score of total confirmed cases ( αi + score of local case proportion ( βi) + clustered outbreak score ( γi) ( i = 1,2,3,4.90).
where αi is based on the number of confirmed cases on the day before the assessment; the county with the highest number of cases was given a full score of 50, and the rest of the counties’ score was calculated by the proportion of the number of confirmed cases.
βi is based on the data on the day before the assessment; the county with the highest proportion of local cases to the total number of confirmed cases was given a full score of 30, and the proportions of local cases in other counties were calculated proportionally.
γi refers to the number of clustered outbreaks and the total number of cases in clustered outbreaks, each accounting for 50%. Among them, the clustered outbreaks were divided into 30% of family clustered outbreaks (A1) and 70% of non-family clustered outbreaks (A2). The number of cases was divided into family cluster cases (B1) accounting for 30% and non-family cluster cases (B2) accounting for 70%. The total clustered outbreak score = (0.3 A1 + 0.7 A2) × 0.5+ (0.3 B1 + 0.7 B2) × 0.5. We assigned the highest clustered outbreak score in the county to a full score of 20 and calculated the score of the clustered outbreaks in other counties proportionally.
Then we used the negative indicator to adjust the county-level epidemic risk score. The rule was that if there were no case reports for three consecutive days, 5% was deducted from the total score, and if no case reports for seven consecutive days, 10% was deducted from the total score.
According to the adjusted Si , the epidemic risk level of each county was classified as follows: more than 90 points, very-high-risk areas; 60–90 points, high-risk areas; 20–60 points, medium- risk areas; 10–20 points, medium-risk areas; and below 10 points, low-risk areas.
Thematic maps are represented by red, orange, yellow, blue, and green colours, respectively, i.e., a five-colour epidemic chart21.
The county-level epidemic risk was adjusted every three days according to the subsequent adjustment rules. The rules for increasing risk levels were as follows: the risk was increased by one level if the total number of newly confirmed cases for three consecutive days in the five days before the assessment was 10 or more, or there was report of one non-family cluster outbreak with the number of cases greater than or equal to five, or two or more non-family cluster outbreaks within three days before the assessment. In areas that were initially at low risk, as soon as cases were reported within three days before the assessment, it was raised to mild risk. The risk level reduction rule was as follows: if no case was reported for six consecutive days before assessment, the risk was reduced by one level.
Data source
According to the latest version of the guideline on the diagnosis and treatment of novel coronavirus pneumonia released by the National Health Commission13, a confirmed case was defined as a suspected case with respiratory specimens that tested positive for the COVID-19 by real-time reverse-transcription polymerase chain reaction (RT-PCR) or a genetic sequence that is highly homologous with the known novel coronavirus. A cluster outbreak refers to two or more confirmed cases or asymptomatic infections that are found in a small area (such as a family, a construction site, a unit, etc.) within 14 days, with the possibility of human-to-human transmission due to close contact, or a possibility of infection due to co-exposure.
As a legal infectious disease, Chinese law required all COVID-19 cases to be immediately reported to China’s Infectious Disease Information System. Demographic information, consultation information, and epidemiological information related to each case were investigated and collected by local epidemiologists and public health workers and transmitted to the information system. Based primarily on epidemiological investigations, cases were categorised as imported cases if they had resided in or visited Wuhan and surrounding areas or other communities with cases reported within 14 days before the onset of illness, or if they had close contact with someone with fever or respiratory symptoms. Otherwise, it was a local case.
According to whether the epidemic outbreaks occurred in the same family, clustered outbreaks were divided into family and non-family clustered outbreaks.
In this study, data of all 1205 COVID-19 cases reported and 218 clustered outbreaks reported in Zhejiang province were extracted from China’s Infectious Disease Information System by the end of 27 February 2020.
Statistical analysis
The epidemic curve for confirmed cases was constructed based on the date of diagnosis, and key dates relating to control measures were overlaid to aid interpretation. The dates of diagnosis for imported cases and local cases were stacked to show total cases over time. The curve for clustered outbreaks was also overlaid with the number of cases versus the date of diagnosis.
MapInfo Pro (version 16.0; MapInfo Corporation, New York, NY, USA) was used to map the geographic distribution and variation.
Ethics approval
This study was in response to a public health emergency. As such, it was exempt from the requirement for ethical approval and informed consent.