Table 1. Base case parameters and sources.
| Parameter | Mean and 95% CI | Reference |
|---|---|---|
| Contact structure (contacts per day) | ||
| General population | 12.60 | [10] |
| General population | 12.12 | Calculated [10] |
| University students | 0.38 | Calculated [10] |
| Long-term care residents | 0.10 | Calculated |
| University students | 23.70 | |
| General population (includes faculty, staff, and graduate students) | 9.48 (5.0, 15.0) | Calculated [11, 14] |
| University students | 14.22 (10.0, 28.4) | Calculated [11, 14] |
| Long-term care residents | 0 | Assumed |
| Long-term care residents | 34.1 | [13, 15] |
| General population (includes LTC staff) | 14.2 (11.4, 17.0) | [13, 15] |
| University students | 0 | Assumed |
| Long-term care residents | 19.9 (11.3, 28.5) | [13] |
| Infectiousness and COVID-19 prevention behaviours | ||
| R0: Average number of new infections per infection | 3.0 (2.85, 3.3) | Empirically estimated a |
| Reduction in contacts when aware of infected status and in-home isolation | 90% (80%, 95%) | Assumed |
| Reduction in contacts when in hospital | 100% | Assumed |
| Effectiveness of mask wearing, reduction in transmission during a close contact between a susceptible and an infected person | 40% | [16] |
| General population | ||
| Initial proportion who are ‘high-intensity physical distancers’ | 40% | [17] |
| High-intensity physical distancer | ||
| Reduction in contacts | 75% | [17] |
| Mask wearing | 86% | [17] |
| Low-intensity physical distancers / Unable to reduce contacts | ||
| Reduction in contacts | 30% | Calculated g |
| Mask wearing | 38% | [17] |
| University students | ||
| Initial reduction in contacts | 40% | Calculated h |
| Mask wearing | 57% | [17] |
| Response to COVID-19 community outcomes | ||
| General population increase participation in high-intensity physical distancing | ||
| COVID-19 patients in critical care exceeds 15 | 0.5% per day | Assumed |
| COVID-19 deaths in the past 10 days exceeds 10 | 1% per day | Assumed |
| Maximum level of participation in high-intensity physical distancing | 80% | Assumed |
| University students increase reduction in contacts | ||
| COVID-19 patients in critical care exceeds 15 | 0.5% per day | Assumed |
| COVID-19 deaths in the past 10 days exceeds 10 | 1% per day | Assumed |
| Maximum level of contact reduction | 50% | Assumed |
| Time to diagnosis | ||
| Minimum time from symptom onset to clinical presentation (average days) | 2.1 (1, 3) | |
| Daily probability of diagnosis by symptom-based surveillance and contact tracing, general population and student population | ||
| Symptomatic cases | 15.8% | Calculated b |
| Asymptomatic cases | 4.1% | Calculated b |
| Sensitivity of nasopharyngeal swab PCR test for COVID-19 | 72.1% | [18, 19] |
| Disease severity distribution | ||
| Long-term care residents | ||
| Asymptomatic | 12% (1.2%, 22.6%) | [20, 21] |
| Symptomatic, cared for in long-term care | 76.2% | Calculated |
| Hospitalized, no critical care resources | 11.4% (9%, 14%) | [22] |
| Critical, requires mechanical ventilation (MV) | 0.3% (0, 0.7%) | [22] |
| Critical, requires renal replacement therapy (RRT) | 0.1% (0, 0.2%) | Estimated c |
| University students | ||
| Asymptomatic | 31% (18%, 80%) | [23] |
| Mild or Moderate | 67.8% | Calculated |
| Severe | 1.0% (0.5%, 1.5%) | Estimated d |
| Critical, requires MV | 0.18% (0, 0.4%) | [24] |
| Critical, requires RRT | 0.06% (0, 0.1%) | Estimated c |
| General population | ||
| Asymptomatic | 31% (26%, 37%) | [23] |
| Mild or Moderate | 60.4% | Calculated |
| Severe | 3.75% (2.0%, 8.0%) | Calibrated e |
| Critical, requires MV | 1.25% (1.0%, 1.8%) | [24] |
| Critical, requires RRT | 0.45% (0.2%, 0.7%) | Estimated c |
| Time to health event transition (Mean f , days) | ||
| Average duration of infectiousness | 10 (6.3, 16.0) | [25–28] |
| Incubation period: Exposure → Symptom onset | 5.6 (5.1, 6.1) | [29] |
| Infectiousness prior to symptom onset | 2.5 (2.0, 3.0) | [25, 27, 28] |
| Diagnosis: Symptom onset → First opportunity for diagnosis | 2.1 (1.1, 3.1) | [30] |
| Symptom onset → Progression to severe or critical symptoms | 5.8 (4, 8) | [31] |
| Severe symptoms: In hospital → Recovery or Death | 8.3 (6, 12) | [31] |
| Critical care: MV in ICU → Post-ICU in hospital or Death | 15.5 (10, 32) | [32] |
| Critical care: Post-ICU → Recovery | 10.1 (6, 18) | [32] |
| Critical: RRT → Discharge or Death | 25.0 (12, 44) | [32] |
| Symptomatic in LTC: Symptom onset → Recovery or Death | 18.0 (14, 24) | Estimated in calibration |
| Clinical improvement in patients receiving lower level of care than is medically indicated | ||
| Severe symptoms: Home isolation → Recovery | 18.0 (14, 24) | Assumed |
| Mortality | ||
| Long-term care residents | ||
| Symptomatic, cared for in long-term care | 25.5% (21%, 30%) | [22] |
| Hospitalized, no critical care resources | 47.4% (34%, 60%) | [22] |
| Critical, requires MV | 70.8% (66%, 75%) | Based on outcomes in ≥ 70 year olds [32] |
| Critical, requires RRT | 74.9% (67%, 83%) | Based on outcomes in ≥ 70 year olds [32] |
| University students | ||
| Mild or Moderate | 0% | |
| Severe (In hospital) | 0.43% (0.1%, 0.7%) | Estimated non-ICU mortality for < 55 year olds [31] |
| Critical, requires MV | 21.5% (17%, 25%) | Based on outcomes in 16 to 39 year olds [32] |
| Critical, requires RRT | 35.9% (26%, 46%) | Based on outcomes in 16 to 39 year olds [32] |
| General population | ||
| Mild or Moderate | 0% | Assumed |
| Severe (In hospital) | 14.4% (4%, 33%) | Estimated non-ICU mortality for < 75 year olds [31] |
| Critical, requires MV | 42.9% (41%, 45%) | Based on outcomes in < 70 year olds [32] |
| Critical, requires RRT | 53.4% (50%, 57%) | Based on outcomes in < 70 year olds [32–34] |
| Mortality in patients unable to receive medically indicated care | ||
| Case fatality rate, Severe patient requiring hospitalization, In home isolation | 25% (16%, 35%) | Assumed |
| Daily rate, Patients who need MV or RRT, In hospital | 40% (21%, 60%) | Assumed, 2-day life expectancy |
| Daily rate, Patients who need MV or RRT, In home isolation | 60% (41%, 80%) | Assumed, 1-day life expectancy |
Mean and 95% confidence interval representing the uncertainty in the mean used in sensitivity analysis.
a. Using exponential regression, we empirically estimated the basic reproduction number, R0, the average number of secondary infections produced by one infected individual during the infected individual’s entire infectious period assuming a fully susceptible population, is 3.0 based on Ontario’s reported cases between March 7 to March 22 [35].
b. The observed median time to diagnosis through symptom-based surveillance alone of 4.6 days (95%CI: 4.2, 5.0) and symptom-based surveillance in combination with contact tracing efforts of 2.9 days (95%CI 2.4, 3.4) in Shenzhen, China [36]. From this, we estimated that symptom-based surveillance and contact tracing results in a daily probability of diagnosis of 15.8% and the daily probability of detection from contact tracing of 4.1% in asymptomatic infections.
c. Among critical care patients, we estimate the ratio of patients requiring renal replacement therapy (RRT) to mechanical ventilation (MV) based on the UK Intensive Care National Audit and Research Centre (ICNARC) report describing the care and outcomes of 10,118 critical care COVID-19 patients in the UK. In this report, 7,277 patients required MV and 2,673 required RRT, resulting in a ratio of 0.37 RRT patients per mechanical ventilation patient [32].
d. In Canada, based on 63,800 COVID cases in people who were not residents of long-term care facilities reported between February 23 and June 21, 20.3% of hospitalized patients received critical care [22]; this is also consistent with rates of critical care observed in the UK (22% overall hospitalized patients go to ICU) [31]. Therefore, we estimate the ratio of 3.92 hospitalized without critical care patients per critical care patient.
e. Initially estimated using the same process as is described in footnote d. Adjusted in calibration process to better fit the observed data (see Supplemental methods in S1 File).
f. Median and IQR presented in the cited primary work were transformed to Mean (95%CI range) assuming a gamma distribution.
g. Due to reduced density in public spaces and reduced availability of their usual contacts, low-intensity physical distancers also experience an overall contact reduction calculated at each time to be equal to the reduction in overall contacts imposed by the ‘high intensity physical distancers’ [Proportion of the population that are ‘high intensity physical distancers’ × 75% reduction in contacts].
h. Calculated as 32% self-reporting high-intensity physical distancing × 75% reduction in contacts and the remainder having reduced contacts due to the reduced access to their usual contacts and reduced population density in public spaces. So, 32% × 75% + (1–32%) × (32% × 75%) = 40.3%.