Id |
Subject |
Object |
Predicate |
Lexical cue |
T184 |
0-10 |
Sentence |
denotes |
Discussion |
T185 |
11-289 |
Sentence |
denotes |
Our results suggest that physical distancing measures were effective in British Columbia; we estimated that individuals practising physical distancing in British Columbia to be experiencing approximately 0.22 (0.11–0.34 90% CI) of their normal contact rate as of April 11, 2020. |
T186 |
290-419 |
Sentence |
denotes |
This was below the threshold of 0.55 at which prevalence was expected to grow, which left some room to relax distancing measures. |
T187 |
420-500 |
Sentence |
denotes |
These results were supported by declines in hospitalizations and ICU admissions. |
T188 |
501-597 |
Sentence |
denotes |
We estimated that there was varying room to relax measures in other locations as of May 7, 2020. |
T189 |
598-688 |
Sentence |
denotes |
Strong control in New Zealand suggested considerable scope for restrictions to be relaxed. |
T190 |
689-800 |
Sentence |
denotes |
However, we found that there was relatively little room to relax measures in New York, Florida, and Washington. |
T191 |
801-1019 |
Sentence |
denotes |
The overall picture in California, as of May 7, 2020, appeared to be that contacts were above the threshold that leads to increasing prevalence, and hence restrictions were not sufficient to curb spread of the disease. |
T192 |
1020-1159 |
Sentence |
denotes |
We note that in California, and all locations considered, it is likely that there was strong regional variation around our broad estimates. |
T193 |
1160-1281 |
Sentence |
denotes |
Our estimates for BC are consistent with local mobility data, and with contact patterns in other international locations. |
T194 |
1282-1466 |
Sentence |
denotes |
For example, survey data from the UK [39] suggested a 73% reduction in contacts, and a modelling study found that a 70–80% reduction in contacts is consistent with data in France [13]. |
T195 |
1467-1623 |
Sentence |
denotes |
Our estimate of the effect of distancing on contact patterns in BC is consistent with independent lines of evidence for the strength of distancing measures. |
T196 |
1624-1787 |
Sentence |
denotes |
Local rail (SkyTrain) station crowding data, provided by Metro Vancouver’s transportation authority TransLink, gives a proxy for reduction in public transport use. |
T197 |
1788-1918 |
Sentence |
denotes |
Overall daytime travel was reduced by 16% for the week of March 9, 64% for the week of March 16, and 73% for the week of March 23. |
T198 |
1919-2039 |
Sentence |
denotes |
Estimates on adhering to physical distancing are also available from a publicly available respondent-driven survey [20]. |
T199 |
2040-2211 |
Sentence |
denotes |
The survey found the rate of respondents stating that there was a serious threat of a coronavirus outbreak in Canada increased from 42% on March 5–6 to 88% on March 20–23. |
T200 |
2212-2393 |
Sentence |
denotes |
For individuals who stated there was a serious threat, 89% stated they were keeping personal distance compared to 66% for individuals who did not believe there was a serious threat. |
T201 |
2394-2677 |
Sentence |
denotes |
Mobile phone location data from BlueDot [39] suggested that the maximum and cumulative distance travelled from home fell by approximately 90%, the portion of mobile phone check-ins at home rose by over 10%, and the portion of devices for which every check-in was at home rose by 60%. |
T202 |
2678-2793 |
Sentence |
denotes |
These estimates are also consistent with Google mobility [21], Citymapper index [40], and Apple mobility data [41]. |
T203 |
2794-2931 |
Sentence |
denotes |
These are indirect reflections of the contact rate but are supportive of a dramatic change in contact patterns as reflected in our model. |
T204 |
2932-3002 |
Sentence |
denotes |
Furthermore, similar data are widely available for many jurisdictions. |
T205 |
3003-3092 |
Sentence |
denotes |
Our results provide a direct estimate of contact rates in conjunction with mobility data. |
T206 |
3093-3300 |
Sentence |
denotes |
Our results suggest that some relaxation of distancing measures may have been possible in BC and New York, considerably so in New Zealand, but that relaxation would have been risky in Washington and Florida. |
T207 |
3301-3451 |
Sentence |
denotes |
In BC, we simulated fixed and dynamic measures—less stringent than the measures in place at present—which would continue to maintain low case numbers. |
T208 |
3452-3559 |
Sentence |
denotes |
This is feasible either through continual strong distancing, or via well-monitored dynamic on/off measures. |
T209 |
3560-3806 |
Sentence |
denotes |
We have illustrated the model’s high case volumes and long time frames that would result from cessation of distancing and the absence of continued strong public health and behavioural intervention in BC; the dynamics would look similar elsewhere. |
T210 |
3807-4021 |
Sentence |
denotes |
In all jurisdictions, we found that immunity has not built up in the model; our estimates of the decline are not due to a natural peak in an epidemic curve, but are the direct effect of changes in contact patterns. |
T211 |
4022-4140 |
Sentence |
denotes |
Seasonal transmission could even amplify a peak in the winter season, if control and monitoring were ceased then [12]. |
T212 |
4141-4288 |
Sentence |
denotes |
The delays between exposure and reporting present obvious challenges for monitoring the success of control measures using reported case-count data. |
T213 |
4289-4596 |
Sentence |
denotes |
We suggest that two different kinds of monitoring will be important if distancing measures are to be relaxed: (1) monitoring cases through testing, contact tracing, and other case finding, and (2) monitoring contact patterns and distancing behaviour in the population, in a “distancing surveillance” effort. |
T214 |
4597-4913 |
Sentence |
denotes |
This latter form of monitoring, derived from mobile phones, surveys and apps, could be available rapidly, whereas the incubation period places an unavoidable delay between control measures and detecting their impact in reported cases—even if testing of symptomatic cases were widespread and reporting were immediate. |
T215 |
4914-5080 |
Sentence |
denotes |
There is considerable interest in real-time monitoring of mobile phone movements, population surveys on the uptake of physical distancing, and other behavioural data. |
T216 |
5081-5175 |
Sentence |
denotes |
While these are potentially promising avenues, the outcome of interest is incident infections. |
T217 |
5176-5282 |
Sentence |
denotes |
Locations of mobile phones, traffic patterns, and survey information are proxies for this outcome at best. |
T218 |
5283-5457 |
Sentence |
denotes |
The work we have presented here could help to calibrate distancing surveillance measurements, to understand how they relate to changes in contact rates for modelling efforts. |
T219 |
5458-5520 |
Sentence |
denotes |
Our modelling framework has a number of important limitations. |
T220 |
5521-5663 |
Sentence |
denotes |
We do not model age and contact structure explicitly, except to distinguish between two populations: those participating in distancing or not. |
T221 |
5664-5844 |
Sentence |
denotes |
This has the advantage that we do not require data on age-specific contact patterns, responses to distancing measures, or infectiousness; these data are not available at this time. |
T222 |
5845-5957 |
Sentence |
denotes |
It also limits our ability to provide guidance on where and how contact reduction measures could be implemented. |
T223 |
5958-6209 |
Sentence |
denotes |
It is a simplification of behaviour in many ways; true distancing responses are a continuum, and the measures in place (e.g., no mass gatherings or dine-in services) also mean that the whole population is experiencing some changes in contact patterns. |
T224 |
6210-6379 |
Sentence |
denotes |
Our model is deterministic, and so does not capture the possibility of extinction; in addition, we have not simulated introductions of COVID-19 from other jurisdictions. |
T225 |
6380-6557 |
Sentence |
denotes |
We have not accounted for geographic structure; differences in distancing behaviour, health care practices, and demographics in different jurisdictions could impact the results. |
T226 |
6558-6771 |
Sentence |
denotes |
We have also not modelled either conventional or automated contact tracing [42, 43]; in our model these would decrease the duration of the infectious period and change the transitions for some exposed individuals. |
T227 |
6772-6811 |
Sentence |
denotes |
There are also limitations in our data. |
T228 |
6812-6966 |
Sentence |
denotes |
We have used an observation model to link reported cases to the modelled prevalence, and we included variation in the portion of cases detected over time. |
T229 |
6967-7181 |
Sentence |
denotes |
Modelling and forecasting based on reported cases faces challenges when testing is driven by clinical needs, testing capacities, and other constraints (and in particular is not designed to test population samples). |
T230 |
7182-7306 |
Sentence |
denotes |
Cases in long-term care facilities (LTCF) represent a substantial fraction of the cases, and particularly the deaths, in BC. |
T231 |
7307-7411 |
Sentence |
denotes |
Along with the low number of deaths in total, this is one rationale for not modelling deaths explicitly. |
T232 |
7412-7517 |
Sentence |
denotes |
We included LTCF cases but also modelled a wide range of under-reporting to account for potential biases. |
T233 |
7518-7670 |
Sentence |
denotes |
If many cases in an LTCF cluster were all reported on the same day (or within a short time frame) this could increase the noise in reported case counts. |
T234 |
7671-7996 |
Sentence |
denotes |
We have modelled case counts as over-dispersed compared to a Poisson distribution to account for such variation; we have developed the R package to model a range of data types individually or in combination (e.g., reported cases, hospitalizations, ICU admissions), which could help to overcome limitations of particular data. |
T235 |
7997-8077 |
Sentence |
denotes |
The testing criteria include a number of categories that have changed over time. |
T236 |
8078-8182 |
Sentence |
denotes |
Testing volume increased sharply in mid March in BC and most jurisdictions have changed testing volumes. |
T237 |
8183-8233 |
Sentence |
denotes |
The base population being tested has also changed. |
T238 |
8234-8442 |
Sentence |
denotes |
In BC, testing first focused primarily on those hospitalized or likely to be hospitalized, health care workers, residents of long-term care facilities, and other cluster investigations (mid-March to April 9). |
T239 |
8443-8626 |
Sentence |
denotes |
Testing was then expanded to include residents of remote, isolated, or Indigenous communities, people who are homeless or have unstable housing, and by physicians’ clinical judgement. |
T240 |
8627-8794 |
Sentence |
denotes |
There is likely some inconsistency in the application of these guidelines across hospitals and facilities, and base populations differ in across jurisdictions as well. |
T241 |
8795-8925 |
Sentence |
denotes |
As a consequence, the base population being tested changes with time and contributes a changing portion of the force of infection. |
T242 |
8926-9031 |
Sentence |
denotes |
Underestimation is therefore complex and is comprised of varying under-ascertainment and under-reporting. |
T243 |
9032-9199 |
Sentence |
denotes |
There remain important unknowns about COVID-19 that give rise to additional limitations for modelling efforts; immunity and asymptomatic transmission are two of these. |
T244 |
9200-9425 |
Sentence |
denotes |
We have included pre-symptomatic transmission but we have not explicitly modelled asymptomatic individuals, who may have few or no symptoms but nonetheless be transmitting, and who may or may not be building lasting immunity. |
T245 |
9426-9553 |
Sentence |
denotes |
Recent work suggests that both pre- and asymptomatic individuals may be contributing considerably to transmission [27, 28, 44]. |
T246 |
9554-9701 |
Sentence |
denotes |
We have indirectly approached this uncertainty, exploring variable underestimation fractions and duration of the incubation and infectious periods. |
T247 |
9702-9790 |
Sentence |
denotes |
A wide range for underestimation and duration is consistent with the reported case data. |
T248 |
9791-10020 |
Sentence |
denotes |
Our conclusion about the impact of distancing measures appears to be robust to these uncertainties, although the basic reproductive number and the model prevalence vary according to assumptions about underestimation and duration. |
T249 |
10021-10249 |
Sentence |
denotes |
Model predictions for the peak timing and size of prevalence without strong public health interventions will depend strongly on the dynamics of immunity, including the numbers of asymptomatic individuals and their immunity [12]. |
T250 |
10250-10451 |
Sentence |
denotes |
Our model suggests that distancing measures were working well in British Columbia as of April, 2020, that some relaxation of these measures may have been possible, but that this must be done carefully. |
T251 |
10452-10614 |
Sentence |
denotes |
More broadly, we estimated a range of effectiveness of physical distancing in other locations (from very strong in New Zealand to weak in California as of May 7). |
T252 |
10615-10727 |
Sentence |
denotes |
Given the likely low levels of immunity, long-term public health measures will be necessary to control COVID-19. |
T253 |
10728-11071 |
Sentence |
denotes |
If data were available describing contact patterns and transmissibility by age, along with data describing the impact of specific activities on these contact patterns, then models could be effective tools to determine how to safely relax distancing measures—for example, restarting specific activities such as particular workplaces or schools. |
T254 |
11072-11335 |
Sentence |
denotes |
Without knowledge of prevalence and transmissibility in children, of the extent of asymptomatic infection, and of the contact patterns that would result from restarting specific activities, models aiming to simulate these activities will have large uncertainties. |
T255 |
11336-11532 |
Sentence |
denotes |
We therefore suggest that there is an urgent need for longitudinal measurements of population-level prevalence and immunity via viral testing and serological studies, even where prevalence is low. |
T256 |
11533-11798 |
Sentence |
denotes |
Furthermore, if distancing measures are to be relaxed, it will be crucial to have strong surveillance through widespread testing, contact tracing, and isolation of new cases, as well as strong compliance with potentially shifting public health policy and messaging. |