Table 1 Study characteristics

Ahmed

2018

Review

Workplace social distancing

Modelling studies estimated workplace social distancing reduced the cumulative influenza attack rate 23%.

Armbruster

2007

Modelling

Contact tracing

Screening

Contact tracing is cost-effective only when population prevalence is low (e.g. under 8% for tuberculosis)

Barnes

2007

Modelling

Case isolation

Personal protection equipment

Antiviral prophylaxis

Antiviral prophylaxis can contain an influenza strain with R0 = 2; healthcare workers contribute disproportionately to the transmission of the infection when not protected against infection.

Becker

2005

Modelling

Face masks

Social distancing

Hand hygiene

Case isolation

School closures

Quarantining affected households

Contact tracing

Combinations more effective than single interventions; household quarantines and contact tracing reduced reproduction number from 6 to below 1.

Bell

2006

Review

Travel restrictions

Screening and quarantining entering travellers were ineffective in past pandemics; WHO recommends providing information to international travellers and screening travellers departing infected countries.

Bell

2006

Review

School closures

Isolation of patients

Quarantine of contacts

Antiviral therapy

Travel restrictions

Hand hygiene

Respiratory hygiene

Face masks

Disinfected household surfaces

At the start of the outbreak, detect and isolate cases, quarantine contacts, restrict travel in and out of affected communities, and consider targeted antiviral therapy. If sustained transmission in the general population, ill persons are advised to remain at home; increase social distance; promote hand-washing and respiratory hygiene/cough etiquette; wear face masks; and disinfect contaminated household surfaces.

Bin Nafisah

2018

Review

School closures

School closures reduced the peak of epidemics by an average of 29.65%.

Bolton

2012

Modelling

School closures

Travel restrictions

Generalized social distancing

Quarantining of close contacts

Treatment of cases with antivirals

Prophylaxis of contacts

A combination of non-pharmaceutical interventions proved as effective as the targeted use of antivirals and reduced the mean attack rate from approximately 23 to 21% (severe pandemic scenario).

Bootsma

2007

Observational

School closures

Church closures

Mass gatherings

Face masks

Case isolation

Disinfection/hygiene measures

The timing of interventions during the 1918–1919 flu pandemic (specifically early intervention) was the most strongly correlated factor with total mortality. Population size and density were not significant factors in overall mortality.

Buonomo

2012

Modelling

Health promotion campaign

Vaccination

Early vaccination is more cost-effective; interventions before vaccines are available need to be balanced with the potential gains in the cost-effectiveness of future vaccines.

Caley

2007

Modelling

Border screening

Face masks during transit

Immediate presentation following symptom onset

Flight-based quarantining

The most effective strategy to prevent spread is control in the source country; targeting travellers is not effective.

Chen

2020

Modelling

School closure

Case isolation

Antiviral treatment

Antiviral prophylaxis

Vaccination

Case isolation was the most effective single intervention, adding antiviral therapeutics, antiviral prophylaxis, vaccination prior to the outbreak, and school closure decreased the attack rate only slightly and shortened outbreak duration by only 9 days

Chinazzi

2020

Modelling

Travel restrictions

90% travel restrictions to and from China only modestly affect the epidemic trajectory unless combined with a 50% or higher reduction of transmission in the community.

Chong

2012

Modelling

Travel restrictions (air, land, sea)

Antiviral treatment

Restricting air travel from infected regions 99% delays epidemic peak up to 2 weeks. Restricting air and land travel delays peak ~3.5 weeks. Neither 90% nor 99% travel restrictions reduced the epidemic magnitude more than 10%. Antiviral treatment and hospitalization of infectious subjects are more effective.

Dan

2009

Modelling

Personal protection equipment

Personal protection equipment plus restricting visitors and cancelling elective procedures

Personal protection equipment cost-effective for H1N1 ($23,600 per death prevented)

Personal protection equipment plus restricting visitors and cancelling elective procedures less cost-effective ($2500,000 per death prevented)

Day

2006

Modelling

Quarantine

Quarantine of all contacts is beneficial only when case isolation is ineffective, significant asymptomatic transmission, and asymptomatic period is neither very long nor short.

Figueroa

2017

Observational

Mass gatherings

Outbreaks at mass gatherings were uncommon, even during the 2009 H1N1 pandemic

Gostic

2020

Modelling

Traveller screening

Even under best-case assumptions, screening (at border or locally) will miss more than half of people infected with SARS-CoV-2.

Halder

2011

Modelling

School closures

Workplace closure

Antiviral treatment

Household antiviral prophylaxis

Extended antiviral prophylaxis

Combinations were more cost-effective than single interventions; best combination included treatment and household prophylaxis using antiviral drugs and limited duration school closure ($632 to $777 per case prevented)

Halton

2013

Review

Surveillance

Contact tracing

Isolation and quarantine

Contact tracing and progressively earlier isolation of probable SARS cases were associated with control of the SARS outbreak.

Handel

2006

Modelling

Hypothetical interventions

Interventions before vaccines are available must be balanced with potential gains of future vaccines or potential multiple outbreaks.

Hellewell

2020

Modelling

Contact tracing

Highly effective contact tracing and case isolation enough to control a COVID-19 outbreak within 3 months. Transmissibility is an important factor (when Ro = 2.5, 80%+ contacts needed to be traced and isolated). Timing is also important (5 initial cases, 50%+ chance of achieving control, even at modest contact-tracing levels; however, 40 initial cases, control much less likely). Delays from symptom onset to isolation decreased the probability of control.

Herrera-Diestra

2019

Modelling

Vaccination

Vaccinating (or self isolating) based on the number of infected acquaintances is expected to prevent the most infections while requiring the fewest intervention resources.

Ishola

2011

Review

Mass gatherings

Some evidence to indicate that mass gatherings may be associated with an increased risk of influenza transmission

Jackson

2014

Review

School closures

School closures are most effective when they cause large reductions in contact, when the basic reproduction number is below 2, and when attack rates are higher in children than in adults

Jefferson

2011

Review

Screening at entry ports

Isolation

Quarantine

Social distancing

Barriers

Personal protection

Hand hygiene

Overall masks were the best performing intervention across populations, settings, and threats

Lee

2009

Review

Antivirals

Antibiotics

Vaccination

Case isolation

Quarantine

Personal hygiene measures

Social distancing

Travel restrictions

Combinations delayed spread, reduced the overall number of cases, and delayed and reduced peak attack rate more than individual strategies.

Lee

2011

Modelling

Hand hygiene

Disinfectant measures

Patient isolation

Personal protection equipment

Staff exclusion policies

Ward closures

Implementing increased hand hygiene, using protective apparel, staff exclusion policies, or increased disinfection separately or in bundles provided net cost-savings, even when the intervention was only 10% effective in preventing further norovirus transmission. Patient isolation or ward closure was cost-saving only when transmission prevention efficacy was very high (≥90%), and their economic value decreased as the number of beds per room and the number of empty beds per ward increased.

Lee

2010

Modelling

Vaccination

Vaccination priority should be given to at-risk individuals and to children within high-risk groups

Li

2013

Modelling

Quarantine of close contacts

Quarantine in Beijing during 2009 H1N1 reduced confirmed cases by a factor of 5.6; given that H1N1 was mild, “not economically effective”.

Lin

2010

Modelling

Social distancing

Case isolation

Supports early containment; the best strategy depends on the transmission characteristics of virus, the state of the pandemic, and the cost and implementation levels of intervention.

MacIntyre

2019

Modelling

Case isolation

Contact tracing

Outbreak controlled in 100 days when 95% of cases isolated and 50% of contacts traced.

MacIntyre

2015

Review

Face masks

Face masks provide protection against infection in various community settings

Markel

2007

Observational

School closures

Isolation or quarantine

Public gathering ban

Cities that implemented interventions earlier had greater delays in reaching peak mortality (Spearman r=−0.74, P<0.001), lower peak mortality rates (Spearman r=0.31, P=.02), and lower total mortality (Spearman r=0.37, P=.008). A significant association between increased duration of interventions and a reduced total mortality burden (Spearman r=−0.39, P=.005).

Martinez

2014

Modelling

School closures

Workplace closures

Case isolation

Household quarantine

School closure was the single most effective intervention; combination of all interventions was most effective.

Mateus

2014

Review

Travel restrictions

Evidence did not support travel restrictions as an isolated intervention for the containment of influenza.

Nguyen

2018

Modelling

Vaccination

Vaccination should be administered 5 months before to 1 week after the start of an epidemic.

Pan

2020

Observational

Traffic restrictions

Cancellation of social gatherings

Home quarantines

Designated hospitals and wards

Personal protective equipment

Increased testing capacity

Quarantine of presumptive cases

Quarantine of confirmed cases and of their close contacts

Traffic restrictions, cancellation of social gatherings, and home quarantines are associated with reduced transmission, but not sufficient to prevent increases in confirmed cases. Ro reduced below 1 only when all interventions are implemented.

Pasquini-Descomps

2017

Review

School closures

Disease surveillance networks

Contact tracing and case isolation

Face masks

Preventive measures in hospitals

Antiviral treatment

Antiviral prophylaxis

Vaccination

Stockpiling antiviral medicine

quarantining confirmed cases at home

Self-isolation at home

The most cost-effective interventions were disease surveillance networks and contact tracing and case isolation; the least cost-effective intervention was school closure.

Perlroth

2010

Modelling

School closures

Quarantine of infected individuals

Child social distancing

Adult social distancing

Antiviral treatment

Antiviral prophylaxis

Combinations were more cost-effective than single interventions; the best combination included adult and child social distancing, school closure, and antiviral treatment and prophylaxis ($2700 per case).

Prosser

2011

Modelling

Vaccination

Incremental cost-effectiveness ratios ranged from $8000 to $52,000 per quality-adjusted life year for persons aged 6 months to 64 years without high-risk conditions

Rainey

2016

Review

Mass gatherings

Mass gathering-related respiratory disease outbreaks were relatively rare between 2005 and 2014 in the US, suggesting low transmission at most types of gatherings, even during pandemics

Rashid

2015

Review

School closures

School closures moderately effective in reducing influenza transmission and delaying epidemic peak; associated with very high costs

Ryu

2020

Review

Travel restrictions

Evidence does not support entry screening as an effective measure.

Sang

2012

Modelling

Quarantine

Isolation

Entry travel screening

Isolation was the best strategy; entry screening delays the peaks but does not prevent the epidemic.

Saunders-Hastings

2017

Modelling

School closure

Community-contract reduction

Hang hygiene

Face mask

Voluntary isolation

Quarantine

Vaccination

Antiviral prophylaxis

Antiviral treatment

Vaccination plus antiviral treatment most cost-effective (cost per life-year saved: $2581); however, it still led to 3026 life-years lost. Only 1607 life-years lost at a marginally higher cost ($6752) with combination including community-contact reduction, hand hygiene, face masks, voluntary isolation, and antiviral therapy. Combining all interventions saved most lives (267 life-years lost), but very costly ($199,888 per life-year saved).

Saunders-Hastings

2017

Review

Hand hygiene

Face masks

Hand hygiene significant protective effect (OR = 0.62; 95% CI 0.52–0.73).

Face masks non-significant protective effect (OR = 0.53; 95% CI 0.16–1.71) (randomized control trials and cohort studies)

Face masks significant protective effect (OR = 0.41; 95% CI 0.18–0.92) (randomized control trials and cohort studies pooled with case–control studies)

Schiavo

2014

Review

Communicating health risk

Promoting disease control measures

Evidence not conclusive

Shi

2010

Modelling

Mass gatherings

Mass gatherings that occur within 10 days before the epidemic peak can result in a 10% relative increase in peak prevalence and total attack rate; little effect when occurring more than 40 days earlier or 20 days after the infection peak (when initial Ro = 1.5)

Shiell

1998

Modelling

Vaccination

Vaccinating all unvaccinated school-aged children was the most cost-effective strategy ($32.90 marginal cost per case prevented).

Smith

2015

Review

School closure

Quarantine

Social distancing

Oral hygiene

Hand hygiene

Face masks

Social gatherings

Positive significant interventions included professional oral hygiene intervention in the elderly and hand washing.

Suphanchaimat

2020

Modelling

Vaccination

Incremental cost-effectiveness ratio of vaccination (compared with routine outbreak control) $1282–$1990/DALY.

Townsend

2017

Modelling

Hand hygiene

National behaviour change programme would net $5.6 billion (3.4–8.6) in India and $2.64 billion (2.08–5.57) in China

Tracht

2012

Modelling

Face masks

$573 billion saved if 50% of the US population used masks in an unmitigated H1N1 epidemic

Tuncer

2018

Modelling

Isolation

Quarantine

Education

Safe burial

Social distancing

Social distancing had the most impact on the 2014 Ebola epidemic in Liberia, followed by isolation and quarantine.

Van Genugten

2003

Modelling

Vaccination

Antiviral treatment

Similar results from vaccinating the entire population vs. only at-risk groups; best strategy combined pneumococcal vaccination of at-risk groups and antiviral treatment.

Velasco

2012

Review

School closure

Antiviral prophylaxis

Social distancing

Vaccination

Quarantine

Combinations were more cost-effective than vaccines and/or antivirals alone; reducing non-essential contacts, using pharmaceutical prophylaxis, and closing schools was the most cost-effective combination.

Viner

2020

Review

School closures

School closures did not help the control of the 2003 SARS epidemic in China, Hong Kong, and Singapore and would prevent only 2–4% of COVID-19 deaths

Young

2019

Modelling

Isolations and quarantines

Case isolation is likely ineffective when the identification of infected hosts is not sufficiently thorough or delayed.

Zhang

2015

Modelling

Voluntary self-isolation

Antivirals

Voluntary self-isolation at symptom onset can achieve the same level of effectiveness as starting antiviral prophylaxis; when delayed 2 days, strategy has a limited effect on reducing transmission.

Zhang

2012

Observational

Border screening

Close contact tracing (and quarantine)

Medical follow-up of international travellers

Influenza-like illness monitoring

Border screening: 132/600,000 (0.02%) people infected; contact tracing: 120/4768 (2.5%) infected; medical follow-up of international travellers: 18/346, 847 (0.005%) infected; influenza-like illness monitoring: 339/180,495 (0.2%) infected.

Zhao

2020

Observational

Domestic travel

Each increase of 100 in daily new cases and daily passengers departing from Wuhan was associated with an increase of 16.25% (95% CI: 14.86–17.66%) in daily new cases outside of Wuhan.