This page is to keep up to date with the latest literature on COVID19. Anyone is welcome to add new papers, particularly modelling papers or those that might inform modelling. For each paper please add a link and a couple of sentences summarising the paper. Since things are changing so fast, it's also helpful to put the date the paper was put up, rather than just the year. If you have come across papers that look interesting but haven't had time to read them, leave them in the list at the top so that others can read them and move them down with summarising sentences. You can also add new headings if your paper doesn't fit into the existing categories.
To save duplicated effort, if you start reading a paper from the "Papers to read" section, please change the label from "UNREAD" to "being read by xxx", before moving it down the page with your summary once you have finished.
Papers to read
- Pellis, Lorenzo, et al. (Manchester Maths) "Challenges in control of Covid-19: short doubling time and long delay to effect of interventions." arXiv (31/03/2020). UNREAD
- Ma et al "Epidemiological parameters of coronavirus disease 2019: a pooled analysis of publicly reported individual data of 1155 cases from seven countries" (24/03/2020) UNREAD
- Ganyani et al "Estimating the generation interval for COVID-19 based on symptom onset data" (medRxiv, 08/03/2020) UNREAD
- Dudel et al "Monitoring trends and differences in COVID-19 case fatality rates using decomposition methods: Contributions of age structure and age-specific fatality "(medRxiv, 02/04/2020) UNREAD
- Lopez and Rodo "A modified SEIR model to predict the COVID-19 outbreak in Spain: simulating control scenarios and multi-scale epidemics" (medRxiv, 30/03/2020) UNREAD
- Wenbao et al "Transmission dynamics of SARS-COV-2 in China: impact of public health interventions" (medRxiv, 27/03/2020) UNREAD
- Banerjee et al "Estimating excess 1- year mortality from COVID-19 according to underlying conditions and age in England: a rapid analysis using NHS health records in 3.8 million adults" (medRxiv 24/03/2020) UNREAD
- Davies et al. (LSHTM) "The effect of non-pharmaceutical interventions on COVID-19 cases, deaths and demand for hospital services in the UK: a modelling study" (01/04/2020) UNREAD
- Woelfel et al "Virological assessment of hospitalised patients with COVID-2019" (Nature, 01/04/2020) UNREAD
- Noriega and Samore "Increasing testing throughput and case detection with a pooled-sample Bayesian approach in the context of COVID-19" (BioRxiv, 05/04/2020) UNREAD
- Huang et al Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China (The Lancet, 24/01/2020) UNREAD
Lei et al Clinical characteristics and outcomes of patients undergoing surgeries during the incubation period of COVID-19 infection (The Lancet, 04/04/2020), UNREAD
Finding R0/Early spread
- Hilton, Joe, and Matt J. Keeling. "Estimation of country-level basic reproductive ratios for novel Coronavirus (COVID-19) using synthetic contact matrices." medRxiv (27/02/2020).
- Using contact matrices from different countries to infer how R0 will change between countries
- Danon, Leon, et al. "A spatial model of CoVID-19 transmission in England and Wales: early spread and peak timing." medRxiv (14/02/2020).
- Provides initial estimates of the potential course of CoVID-19 in England and Wales in the absence of control measures.
- Estimates peak to occur ~4 months after start of person-to-person transmission, with initial location of cases having limited impact on the timing of the epidemic.
Current patterns of transmission
- Abbott, Sam, et al. (LSHTM) "Temporal variation in transmission during the COVID-19 outbreak." CMMID repository (updated regularly).
- Developed real-time dashboards to analyse current transmission dynamics in different countries.
- Abbott, Sam, et al. (LSHTM) "Temporal variation in transmission during the COVID-19 outbreak in Italy." CMMID repository (updated regularly).
- Developed real-time dashboards to analyse current transmission dynamics in Italy.
Estimating case numbers
- Jombart, Thibaut, et al. (LSHTM) "Inferring the number of COVID-19 cases from recently reported deaths." medRxiv (13/03/2020).
- Estimate the number of COVID-19 cases from newly reported deaths in a population without previous reports. Approach implemented in a publicly available, user-friendly, online tool.
- Perkins, Alex, et al. "Estimating unobserved SARS-CoV-2 infections in the United States." medRxiv (18/03/2020).
Informed a stochastic model of local transmission with data on the number and timing of reported cases and deaths, both imported and local.
- Compared simulations of symptomatic infections with data on reported cases to estimate local case detection over time.
- SARS-CoV-2 Infection in Children (letter to the editor) NEJM (18/03/2020)
- Brief data regarding symptoms in children (aged 15 and under) treated at the Wuhan Children's hospital who have tested positive (they tested symptomatic and asymptomatic children with known contacts to confirmed or suspected cases). Diagnosis: 15.8% asymptomatic, 19.3% upper respiratory tract infection, 64.9% pneumonia. The paper has a list of symptoms including 48.5% with a cough and 71% with a fever.
- Mizumoto, Kenji, et al. "Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020." Eurosurveillance 25.10 (12/03/2020): 2000180.
- Estimated an asymptomatic proportion of 17.9% (95% credible interval: 15.5-20.2%)
- Quilty, Bill, et al. (LSHTM) "Effectiveness of airport screening at deteccting travellers infected with novel coronavirus (2019-nCoV)." The Lancet Global Health (06/02/2020).
- Analysed airport screening, finding that screening at arrival would likely miss the vast majority of incoming infected travellers
- Keeling, Matt J., T. Deirdre Hollingsworth, and Jonathan M. Read. "The Efficacy of Contact Tracing for the Containment of the 2019 Novel Coronavirus (COVID-19)." medRxiv (17/02/2020).
- With contact tracing less than 1 in 5 cases will generate any subsequent untraced cases, but there is an average of 36.1 individuals (95th percentiles 0-182) traced per case.
- Hellewell, Joel, et al. (LSHTM) "Feasibility of controlling COVID-19 outbreaks by isolation of cases and contacts." The Lancet Global Health (28/02/2020).
- Looked at feasibility of controlling COVID through continued contact tracing, estimating that if R0=2.5, then to control the majority of outbreaks, more than 70% of contacts would need to be traced.
- Clifford, Sam, et al. (LSHTM) "Interventions targeting air travellers early in the pandemic may delay local outbreak of SARS-COV-2" medRxiv (28/02/2020).
- Syndromic screening and traveller sensitisation in combination could delay outbreaks in yet unaffected countries and support local containment efforts, but only if infected traveller numbers are very low.
- Prem, Kiesha, et al. (LSHTM) "The effect of control strategies that reducce social mixing on outcomes of the COVID-19 epidemic in Wuhan, China" medRxiv (12/03/2020).
- Potential impact of social distancing intervention measures in Wuhan, and what might happen when the interventions stop.
- Changes in mixing patterns may have contributed to reducing the number of infections in mid-2020 by 92% (interquartile range: 66-97%). There are benefits to sustaining these measures until April in terms of reducing the height of the peak, overall epidemic size in mid-2020 and probability that a second peak may occur after return to work.
- Cowling, Benjamin J, et al. "Impact assessment of non-pharmaceutical interventions against COVID-19 and influenza in Hong Kong: an observational study" medRxiv (16/03/2020)
- Measuring the impact of social distancing in Hong Kong through two telephone surveys, comparing to estiamtes of the daily effective reproduction number for COVID and influenza A.
- Influenza reduced: 44% reduction in transmissibility in the community and 33% reduction in transmissibility based on paediatric hospitalisation rates. in the earlier survey 74.5% of adult population wore masks when going out, rising to 97.5% and 61.3% avoided going to crowded places, rising to 90.2%.
- Jombart, Thibaut, et al. (LSHTM) "Forecasting critical care bed requirements for COVID-19 patients in England". CMMID repository (22/03/2020).
- Estimating critical care bed needs up to 31st March 2020.
- "These results imply that unless transmissibility is strongly reduced in the coming days, ICU/HDU capacity for COVID-19 in England (in January 2020: 4,123 critical beds for adults, 312 in paediatrics) may be challenged by the end of March, without even considering capacity requirements for other conditions."
- Ferguson et al. (Imperial group) "Impact of non-pharmaceutical interventions (NPIs) to reduce COVID19 mortality and healthcare demand" (16/03/2020)
- The modelling provided directly to the SAGE group (other evidence can be found here), seems to have influenced decisions a lot
- Uses a detailed IBM originally made for flu, changing some parameters for COVID (I think - they don't put a lot of actual detail in). Considers whether we want to "mitigate" (reduce R0, but not below 1) or "suppress" (reduce R0 below 1) and concludes that all the mitigate strategies lead to running out of critical care beds. This model finds that only by using very severe interventions can we stay below the critical care beds required, and finishes with the suggestion that we could oscillate between more and less severe cases, based on something like the number of critical care beds available.
- Kissler, Stephen, et al. "Social distancing strategies for curbing the COVID-19 epidemic" medRxiv(24/03/2020)
- Assessed that one-time interventions will be insufficient to maintain COVID-19 prevalence within the critical care capacity of the United States. Intermittent distancing measures may be the only way to avoid overwhelming critical care capacity while building population immunity.
Model is an adapted SEIR model with three tracks, each of differing disease severity (asymptomatic & mild; hospitalised; critical care).
- Peak, Corey M., et al. "Modeling the Comparative Impact of Individual Quarantine vs. Active Monitoring of Contacts for the Mitigation of COVID-19." medRxiv (08/03/2020).
- Using a branching process model. Need to quarantine at least 3/4s of infected contacts.
- Ferretti, Luca, et al.( Christophe Fraser group) "Quantifying dynamics of SARS-CoV-2 transmission suggests that epidemic control and avoidance is feasible through instantaneous digital contact tracing". medRxiv (12/03/2020)
- "We show that first-degree instantaneous contact tracing, informing users when they can move safely or when to seek medical help and avoid vulnerable individuals, has the potential to stop the spread of the epidemic if used by a sufficiently large number of people with reasonable fidelity."
- Rossberg and Knell "How will this continue? Modelling interactions between the COVID-19 pandemic and policy responses" (medRxiv, 01/04/2020).
- A “management strategy evaluation” approach. Uses a matrix population model to simulate disease spread, takes key infection parameter estimates from Ferguson 2020 paper (e.g. 4.6 days to go from exposure to infectiousness; 20% assumed to be asymptomatic; 30% of hospitalized cases require critical care etc).
- Assumes that policy interventions reduce infection rates by d. d changes according to various rules (i.e. imposition of a disease control policy if the case numbers exceed a certain threshold/relaxation of policy when case numbers below a certain threshold). Variation in policy compliance is also incorporated in d.
- Key conclusions: “To be effective [at minimising overall mortality], the frequency of policy revisions should be comparable to the doubling time of the outbreak”, and “any decline of the population’s compliance through time has a comparatively weak effect on cumulative mortality” (but this is assuming that the government keeps changing policies, and this can make up for the declining compliance).
- Pepe, Emanuele et al. "COVID-19 outbreak response: first assessment of mobility changes in Italy following lockdown" (updating daily, accessed 25/03/2020) and on medRxiv
- Using mobility data from Cuebiq, which provides anonymous location data from smartphones. Various measurements of mobility between regions, and a loose estimate of networks of contacts by saying a contact is being within 50m of another person over a 1 hour period. "In some provinces (Milano, Piacenza, Rimini, and others), the average network degree has dropped about 30% with respect to the pre-outbreak period." In the summary they say "The restrictions in mobility, closure of public spaces and the enhancement of smart/remote working, led to an average reduction of potential encounters of 8% during week 2 and almost 19% during week 3" but I can't find this in the actual results.
- Kraemer, Moritz et al. ""The effect of human mobility and control measures on the COVID-19 epidemic in China"(25/03/2020)
- "use real-time mobility data, crowdsourced line-list data of cases with reported travel history, and timelines of reporting changes to identify early shifts in the epidemiological dynamics of the COVID-19 epidemic in China, from an epidemic driven by frequent importations to local transmission"
- "We find that the magnitude of the early epidemic (total number of cases until February 10, 2020) outside of Wuhan is remarkably well predicted by the volume of human movement out of Wuhan alone"
- Tests the contribution of the epidemic in Wuhan to seeding epidemics elsewhere in China through use of a naïve COVID-19 GLM model of daily case counts.
- Chris Jarvis, Kevin van Zandvoort et al. "Impact of physical distance measures on transmission in the UK" medRxiv (03/04/2020)
Survey of contact patterns of representative sample of UK adults (18+) post-lockdown. Found a 73% reduction in the average daily number of contacts observed per participant (from 10.8 to 2.9).
- Verhagen et al (Oxford and Denmark) "Mapping hospital demand: demographics, spatial variation, and the risk of "hospital deserts" during COVID-19 in England and Wales" OSF preprints (21/03/2020)
- They take the hospitalisation estimates from the Ferguson paper and put it together with age distributions in the various regions of England and Wales, and the number of beds and critical care beds available. It's unclear to me what number's they're actually using from the Ferguson paper - the peak hospitalisation, or total over the epidemic. It should also depend how long people are hospitalised for.
- The result is that London is much better off than more coastal regions, because there are generally more beds and a younger population.
- Yang Liu, Roz Eggo and Adam Kucharski. (LSHTM) "Secondary attack rate and superspreading events for SARS-CoV-2." The Lancet (27/02/2020).
- compiled and analysed data on early outbreaks among close contacts, used to assess risk during close-knit gatherings.
- Russell, Timothy W, et al. (LSHTM) "Estimating the infection and case fatality ratio for COVID-19 using age-adusted data from the outbreak on the DIamong Princess cruise ship." Eurosurveillance (26/03/2020).
- "We estimated that the all-age cIFR on the Diamond Princess was 1.3% (95% confidence interval (CI): 0.38–3.6) and the cCFR was 2.6% (95% CI: 0.89–6.7)"
- Comparing deaths onboard with expected deaths based on naive CFR estimates using China data, suggests the cCFR in China during that period was 1.2% (95% CI: 0.3–3.1) and the IFR was 0.6% (95% CI: 0.2–1.7)
- Zhanwei, Du et al. "Serial Interval of COVID-19 among Publicaly Reported Confirmed Cases." Emerging Infectious Diseases (19/03/2020).
- Distribution of serial intervals for 468 confirmed cases of 2019 novel coronavirus disease reported in China as of February 8, 2020.
- The mean interval was 3.96 days (95% CI 3.53–4.39 days), SD 4.75 days (95% CI 4.46–5.07 days); 12.6% of case reports indicated presymptomatic transmission.
- Davies, Nick et al. (LSHTM) "Age-dependent effects in the transmission and control of COVID-19 epidemics". medRxiv (27/03/2020).
- Find strong age dependence in the probability of developing clinical symptoms, rising from around 20% in under 10s to over 70% in older adult.
Sources of data
- Johns Hopkins University GitHub data (including list of original sources they got it from)
- Kaggle resource for COVID-19
- Klepac, Petra, et al. (LSHTM) "Contacts in context: large-scale setting-specific social mixing matrices from the BBC panemic project." medRxiv (05/03/2020).
- Social mixing data from the BBC Pandemic project, to help understand where transmission risk might be concentrated.
- Zhang, Juanjuan, et al. "Age profile of susceptibility, mixing, and social distancing shape the dynamics of the novel coronavirus disease 2019 outbreak in China" medRxiv (20/03/2020).
Analysed contact surveys data for Wuhan and Shanghai before and during the outbreak and contact tracing information from Hunan Province. Daily contacts were reduced 7-9 fold during the COVID-19 social distancing period, with most interactions restricted to the household.
- Amanat, Fatima "A serological assay to detect SARS-CoV-2 seroconversion in humans" medRxiv (18/03/2020)
- I haven't read in detail, but looks like they've successfully developed an ELISA assay for seroconversion. It would still need to be scaled up (massively) for large scale testing.
- Data from Italy
- Data from South Korea
- Xu, Bo et al. "Open access epidemiological data from the COVID-19 outbreak". The Lancet Infectious Diseases (19/02/2020).
- A centralised repository of individual-level information on patients with laboratory-confirmed COVID-19.
- GitHub repository: https://github.com/beoutbreakprepared/nCoV2019/tree/master/latest_data
Yang Yang et al. "Epidemiological and clinical features of the 2019 novel coronavirus outbreak in China Development of tests" medRxiv
- Unfortunately, the preprint is currently withdrawn!
Had some excellent data on early cases in Wuhan - age structure of 4000+ cases.
- Cereda,D, et al. "The early phase of the COVID-19 outbreak in Lombardy, Italy" arRxiv (22/03/2020)
- Some amount of data here for Lombardy, including number of cases by age or by province for three periods of time, finishing 05/03/2020. They do some fitting and look at how the effective reproduction number varies.