Data analysis and Forecasts of COVID-19 in Pune

Department of Scientific Computing, Modeling & Simulation, in collaboration with Pune Knowledge Cluster has put together these charts to help monitor the progress of disease, thereby enabling the appropriate response from policymakers. The analyzed data is obtained from Pune Municipal Corporation and is curated by the team of Pune Knowledge Cluster. The forecasting is done using INDSCI-SIM epidemiological model.

Serological Surveillance

Epidemiological and Serological Surveillance of COVID-19 in Pune City

Rationale: Global data suggests that a significant proportion of SARS-CoV-2 infections are asymptomatic and remain undetected unless populations are actively screened. Understanding the extent of infection in the population permits better understanding of the severity of COVID-19 and transmission dynamics.

Serological testing for antibodies against SARS-CoV-2 in a representative sample population can estimate the cumulative incidence of infection in the population. This study was undertaken to initiate generation of unbiased data important for gaining insights into the spread of the pandemic in Pune city and estimated the seroprevalence of antibodies against SARS-CoV-2 in five high-incidence prabhags (subwards)

The selected prabhags were Yerwada, Lohiyanagar-Kasewadi, Rastapeth-Ravivarpeth, Kasbapeth - Somwarpeth, Navipeth-Parvati.

SARS-CoV-2 IgG antibody seroprevalence in five high-incidence prabhags

Number Prevalence of Seropositivity (%) 95% CI
Prabhag 6 Yerwada 367 56.6 51.5-61.7
Prabhag 16 Kasbapeth-Somwarpeth 352 36.1 31.1-41.1
Prabhag 17 Rastapeth-Raviwarpeth 335 45.7 40.4-51.0
Prabhag 19 Lohiya Nagar-Kasewadi 312 65.4 60.1-70.9
Prabhag 29 Navipeth-Parvati 298 56.7 51.1-62.3
Overall Average 1664 51.5 49.1-53.9

Sero-Surveillance Team
Principal Investigators:Aarti Nagarkar (SPPU, Pune); Aurnab Ghose (IISER Pune).
Co-Investigators:Abhay Kudale (SPPU, Pune), LS Shashidhara (Ashoka University, on lien from IISER Pune); Susmita Chaudhuri and Guruprasad Medigeshi(THSTI, Faridabad)


Forecast

Forecast of number of critical cases

Using state of the art INDSCI-SIM model critical cases are forecasted using the data on 1st April. Reported cases are also shown in black dots for the reference. Best case scenario is shown on green, while worst case scenario is in red. The likely scenarios are in the shaded region. Crossing the shaded region may require additional measures.


Ventilators: Typically, 20% of critical patients require ventilators.


Forecast made on June 10th 2020, based on the reported cases on 8th April.


City level analysis

Cumulative case count

All the cases reported till date. This includes new as well as recovered/deceased cases.

Testing

All the tests carried out till date.

Daily new cases

New positive cases detected on a given day.

Active cases

Cases that still active on a given day

Doubling time

Days required for cases to double. Higher the number, for cases, better. Doubling time for testing should be lower than that for cases

Effective reproduction number R(t)

R(t) indicates the number of people a single infected person infects before recovering. R(t) less than 1 for sustained period of time is desired.

Active fraction

We plot the ratio of Active cases on the date, and Total number of cases till the date. At the beginning of the epidemic this ratio is 1, while at the end of the epidemic it is 0. Thus this ratio can be taken to be a rough measure of the stage of the epidemic.

Test positivity

Test positivity is defined as the ratio of the number of detected positive cases to the total number of tests that have been done. It depends on (a) the prevalence of the infection in the region (b) The testing strategy. If most tests are done on cases which are clinically judged to have a high probability (e.g. symptomatic contacts of confirmed cases) the test positivity will be high. On the other hand if the tests are done more extensively it will be low.

Case Fatality Rate

Case fatality rate (CFR) tells us the percentage of people admitted with SARS CoV2 who unfortunately die in the days after admission to the hospital. If the CFR reduces over time, this means that more people are recovering from the infection instead of passing away. In more recent days, there are still active cases and we don't know whether they will be discharged or will die. Active cases are called "unresolved cases" and those whose outcome is known (discharged or death) are called "resolved cases".

In the above plots, there are three CFR curves for those dates when there are unresolved cases Blue line: CFR based on only resolved cases. This is likely an overestimate Green line: CFR based on both resolved and unresolved cases. This is likely an underestimate Red line: The mean of the Blue and Green curves. This is likely a better estimate then either the Blue or Green for the reasons stated above.

We calculate CFR for a particular day using all the previous data available. This makes the plot smoother and easier to interpret. Because the CFR for any given day depends on all deaths prior to this date, it may not be the best estimate for the CFR for the latest period. To account for this issue, we provide three CFR curves which use data from different time periods -- 40, 30 and 20 days previous to the day for which the latest data is available. The curve based on 40 day data represents the long-term trends better, whereas the curve based on 20 day data represents the short-term trends better. The curve based on 30 day data represents a balance between long and short-term trends.


Ward level analysis


Prabhag level analysis