Skip to main content

Table 2 Comparison of estimated time-varying Rt and prediction accuracy of the models under consideration

From: A comparison of five epidemiological models for transmission of SARS-CoV-2 in India

 

Model

Baselinea

eSIR

SAPHIREb

SEIR-fansy

ICMc

Estimated mean reproduction number R [95% CrI]

Lockdown 1.0 (March 25 – April 14)

2.12 [1.44, 2.16]

2.54 [2.41, 2.74]

5.03 [5.01, 5.04]

1.77 [1.58, 1.96]

Lockdown 2.0 (April 15 – May 3)

1.48 [1.00, 1.51]

1.60 [1.36, 2.17]

1.90 [1.89, 1.91]

1.22 [1.18, 1.27]

Lockdown 3.0 (May 4 – May 17)

0.87 [0.59, 0.89]

1.69 [1.46, 1.97]

2.71 [2.67, 2.73]

1.33 [1.28, 1.38]

Lockdown 4.0 (May 18 – May 31)

0.89 [0.61, 0.91]

1.54 [1.29, 2.00]

2.33 [2.30, 2.36]

1.41 [1.35, 1.47]

Unlock 1.0 (June 1 – June 30)

0.85 [0.58, 0.87]

1.27 [1.19, 1.32]

1.74 [1.73, 1.75]

1.05 [0.99, 1.10]

Unlock 2.0 (July 1 – July 31)

0.77 [0.52, 0.78]

1.31 [1.22, 1.36]

1.80 [1.79, 1.81]

1.11 [1.08, 1.14]

Unlock 3.0 (August 1 – August 31)

0.79 [0.54, 0.81]

1.16 [1.06, 1.31]

1.25 [1.24, 1.26]

1.05 [1.04, 1.07]

Unlock 4.0 (September 1 – September 30)

0.69 [0.47, 0.7]

1.12 [0.98, 1.49]

1.06 [1.05, 1.07]

0.89 [0.86, 0.91]

Unlock 5.0 (October 1 – October 15)

0.67 [0.45, 0.68]

1.09 [0.91, 1.69]

0.86 [0.85, 0.87]

0.83 [0.82, 0.84]

Prediction accuracy using %-SMAPE (MSRPE)d

Active reported cases

37.955 (2.283)

35.141 (1.114)

Cumulative reported cases

6.889 (0.173)

6.593 (0.198)

2.250 (0.056)

2.285 (0.048)

Cumulative reported deaths

8.943 (0.253)

4.737 (0.115)

0.771 (0.020)

  1. aThe baseline model does not return estimates of time-varying R(t) or projections of active reported cases or cumulative reported deaths
  2. bThe SAPHIRE model does not return projections of active reported cases or cumulative reported deaths
  3. cThe ICM model does not return projections of active or cumulative reported cases
  4. dWe compare model projections with observed reported data from October 16 till December 31, 2020