Risk assessment and visualisation of COVID-19 in Zhejiang Province of China: a ‘five- colour chart’ directing disease control and work resumption

Fan He, Xiaopeng Shang, Ling Feng, Zhiping Chen, Tianying Fu, Junfen Lin, Zhen WangResearch 2020 · DOI

Abstract

Since the outbreak of COVID-19 in December 2019 in Wuhan, Zhejiang has become the province with the largest number of cases. The aim of this article is to present Zhejiang province’s experience of establishing an accurate and smart control mechanism for epidemic prevention and control and resumption of work and production using a ‘five-colour epidemic chart’. The number of confirmed cases, proportion of local cases, and occurrence of clustered outbreaks were used as evaluation indicators to calculate the county-level epidemic risk and were assigned different weight coefficients; the absence of cases for 3 and 7 consecutive days was used as the adjustment index. When the first chart was published on February 9, there were 1 very-high-risk, 12 high-risk, and 12 low-risk counties. Under the five-colour chart, Zhejiang began to adopt precise measures to prevent and control the epidemic and resume work and production. By February 24, the low-risk counties had expanded to 82, with no high-risk and very-high-risk counties. The epidemic situation in Zhejiang province has been effectively controlled. The experience of epidemic prevention and control in Zhejiang is worthy to be emulated and learned by other countries and regions.

Introduction

In the early stages of the COVID-19 pandemic, governments worldwide were forced into a difficult trade-off: implement sweeping lockdowns to curb viral spread or maintain economic activity at the risk of public health. Zhejiang Province, a major economic hub in China, faced this dilemma acutely. Outside of Hubei, it initially reported the highest number of confirmed cases in the country. To navigate this crisis, Zhejiang developed a "third way"—a data-driven, granular approach to risk assessment known as the "five-colour epidemic chart." This mechanism allowed for "classified guidance," enabling precise lockdowns where necessary while facilitating the safe resumption of industrial production elsewhere.

Methodology

The core of the Zhejiang strategy was a standardized scoring system applied to 90 individual counties. This system moved away from "blanket" provincial policies to a high-resolution map of risk. The scoring was based on three primary "positive indicators" and a "negative adjustment" index to reflect real-time improvements.

Risk Indicators and Weighting

  1. Total Confirmed Cases (50% weight): This provided a baseline measure of the epidemic's scale within a specific jurisdiction.
  2. Proportion of Local Cases (30% weight): This metric distinguished between imported cases (travelers) and local transmission. Local cases were viewed as a higher threat, signifying active and potentially undetected transmission chains within the community.
  3. Clustered Outbreaks (20% weight): This indicator was further subdivided to prioritize public health risk. Family clusters accounted for 30% of this sub-score, while non-family clusters (e.g., in workplaces or construction sites) accounted for 70%, as they indicated a higher potential for super-spreading events.

Dynamic Adjustment

To incentivize and reflect successful containment, the model included a "negative adjustment" index. Counties that reported zero new cases for 3 consecutive days received a 5% score deduction, while those with zero cases for 7 consecutive days received a 10% deduction.

The Five-Colour Classification

The resulting scores categorized counties into five risk levels: Red (Very-high-risk), Orange (High-risk), Yellow (Medium-risk), Blue (Mild-risk), and Green (Low-risk). These charts were updated every three days, ensuring that the provincial response remained dynamic and responsive to the latest epidemiological data.

Results and Implementation

The implementation of the five-colour chart facilitated a rapid transition from crisis management to economic recovery. The first chart, published on February 9, 2020, revealed a stark landscape with 1 very-high-risk and 12 high-risk counties. However, the precision afforded by the chart allowed for targeted interventions.

Evolution of the five-colour epidemic chart in Zhejiang Province from February 9 to February 24, 2020.

By February 24, the landscape had shifted dramatically. There were zero high-risk or very-high-risk counties remaining, and 82 of the 90 counties were classified as green (low-risk). This shift was not merely a result of the passage of time but a reflection of the "surgical" application of control measures.

The economic impact was significant. By February 27, Zhejiang achieved a production capacity recovery rate of over 75%. Remarkably, despite the return of over 500,000 workers from outside the province, no new COVID-19 cases were reported among this population during the study period. The province achieved zero new confirmed cases by February 22, just 33 days after the first imported case was detected.

Discussion and Insights

The success of the Zhejiang model rests on several key principles:

  • Granularity: By analyzing data at the county level, the government prevented the economic paralysis of low-risk rural areas due to outbreaks in high-density urban centers.
  • Transmission Weighting: Prioritizing local cases and non-family clusters recognized that not all infections pose the same systemic risk to the healthcare infrastructure.
  • Transparency: The chart served as a public "barometer," fostering public awareness in high-risk zones and providing a clear, evidence-based roadmap for business reopening.
  • Digital Integration: The chart was part of a broader "one chart, one code, one index" system, integrating individual health codes and logistics indices into a holistic digital response.

Limitations and Future Work

The authors note that the system's efficacy depends entirely on the authenticity and timeliness of data. Furthermore, while the 50/30/20 weighting was effective for the ancestral strain of SARS-CoV-2, these parameters must be adjusted for different viral variants or epidemic stages. Future iterations could benefit from machine learning to automate these weight adjustments rather than relying solely on expert consensus.

Conclusion

The Zhejiang "five-colour chart" demonstrates how data visualization and granular risk assessment can balance public health with economic stability. This model provides a valuable framework for other regions seeking to implement precise, evidence-based pandemic responses.

Summary updated

Abstract

Abstract Since the outbreak of COVID-19 in December 2019 in Wuhan, Zhejiang has become the province with the largest number of cases. The aim of this article is to present Zhejiang province’s experience of establishing an accurate and smart control mechanism for epidemic prevention and control and resumption of work and production using a ‘five-colour epidemic chart’. The number of confirmed cases, proportion of local cases, and occurrence of clustered outbreaks were used as evaluation indicators to calculate the county-level epidemic risk and were assigned different weight coefficients; the absence of cases for 3 and 7 consecutive days was used as the adjustment index. When the first chart was published on February 9, there were 1 very-high-risk, 12 high-risk, and 12 low-risk counties. Under the five-colour chart, Zhejiang began to adopt precise measures to prevent and control the epidemic and resume work and production. By February 24, the low-risk counties had expanded to 82, with no high-risk and very-high-risk counties. The epidemic situation in Zhejiang province has been effectively controlled. The experience of epidemic prevention and control in Zhejiang is worthy to be emulated and learned by other countries and regions
BibTeX
@inproceedings{he2020risk,
  author = {Fan He and Xiaopeng Shang and Ling Feng and Zhiping Chen and Tianying Fu and Junfen Lin and Zhen Wang},
  title = {{Risk assessment and visualisation of COVID-19 in Zhejiang Province of China: a ‘five- colour chart’ directing disease control and work resumption}},
  booktitle = {Research},
  year = {2020},
  doi = {10.21203/rs.3.rs-24313/v1},
}
Figure