Global strategies and effectiveness for COVID-19 prevention through contact tracing, screening, quarantine, and isolation: a systematic review

Background COVID-19 is an emerging disease caused by highly contagious virus called SARS-CoV-2. It caused an extensive health and economic burden around the globe. There is no proven effective treatment yet, except certain preventive mechanisms. Some studies assessing the effects of different preventive strategies have been published. However, there is no conclusive evidence. Therefore, this study aimed to review evidences related to COVID-19 prevention strategies achieved through contact tracing, screening, quarantine, and isolation to determine best practices. Methods We conducted a systematic review in accordance with the PRISMA and Cochrane guidelines by searching articles from major medical databases such as PubMed/Medline, Global Health Database, Embase, CINAHL, Google Scholar, and clinical trial registries. Non-randomized and modeling articles published to date in areas of COVID prevention with contact tracing, screening, quarantine, and isolation were included. Two experts screened the articles and assessed risk of bias with ROBINS-I tool and certainty of evidence with GRADE approach. The findings were presented narratively and in tabular form. Results We included 22 (9 observational and 13 modeling) studies. The studies consistently reported the benefit of quarantine, contact tracing, screening, and isolation in different settings. Model estimates indicated that quarantine of exposed people averted 44 to 81% of incident cases and 31 to 63% of deaths. Quarantine along with others can also halve the reproductive number and reduce the incidence, thus, shortening the epidemic period effectively. Early initiation of quarantine, operating large-scale screenings, strong contact tracing systems, and isolation of cases can effectively reduce the epidemic. However, adhering only to screening and isolation with lower coverage can miss more than 75% of asymptomatic cases; hence, it is not effective. Conclusion Quarantine, contact tracing, screening, and isolation are effective measures of COVID-19 prevention, particularly when integrated together. In order to be more effective, quarantine should be implemented early and should cover a larger community.

The full spectrum of COVID-19 infection ranges from subclinical self-limiting respiratory tract illness to severe progressive pneumonia with multi-organ failure and death. As evidenced from studies and reports, more than 80% of cases remained asymptomatic and 15% of cases appeared as mild cases with common symptoms like fever, cough, fatigue, and loss of smell and taste [2][3][4][5][6]. Severe disease onset that needs intensive care might result in death due to massive alveolar damage and progressive respiratory failure [1,[4][5][6][7][8].
The virus transmits through direct and indirect contacts. Person-to-person transmissions primarily occur during close contact, droplets produced through coughing, sneezing, and talking. Indirect transmission occurs through touching contaminated surfaces or objects and then touching the face. It is more contagious during the first few days after the onset of symptoms, but asymptomatic cases can also spread the disease [5][6][7][8].
Recommended prevention measures was designed based on overcoming the mode of transmissions including frequent hand washing, maintaining physical distance, quarantine, covering the mouth and nose during coughs, and avoiding contamination of face with unwashed hands. In addition, use of mask is recommended particularly for suspected individuals and their caregivers. There is limited evidence against the community wide use of masks in healthy individuals. However, most of these preventive measures are recommended and were not researched well [4][5][6][7][8].
To the extent of our search, there is no systematic review on the preventive aspects and effectiveness of COVID-19 infection through contact tracing, screening, quarantine, and isolation. The findings were inconclusive; in some studies, certain preventive mechanisms were shown to have minimal effects, while in others different preventive mechanisms have better effect than expected. On the other hand, some studies have reported that integration of interventions is more effective than specific interventions [2,6,8].
Therefore, we aimed to conduct a comprehensive systematic review through reviewing globally published studies on the strategies and effectiveness of different preventive mechanisms (contact tracing, screening, quarantine, and isolation) developed to prevent and control COVID-19. This synthesized measure will be important to bring conclusive evidence, so that policy makers and other stakeholders could have clear evidence to rely on during decision making.

Objectives
To support the existing local and national COVID-19 prevention program with tangible evidence, we conducted a systematic review on global strategies for COVID-19 prevention through contact tracing, screening, quarantine, and isolation. We aimed to answer issues related to alternative strategic implementation and effectiveness in the prevention of the disease or death. The following key questions were considered: 1 Is contact tracing, screening, quarantine, and isolation effective to control the COVID-19 outbreak? 2 Is there difference in the effectiveness of contact tracing, screening, quarantine, and isolation in different settings? 3 How and when these strategies should be applied to control the COVID-19 outbreak?

Methods
We conducted the review in accordance with the PRIS MA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidance for reporting of systematic reviews and meta-analyses [9] and the Cochrane Handbook of Systematic Review [10] through systematic literature search of articles published to date (June 02/ 2020) containing information on COVID-19 prevention by contact tracing, screening, quarantine, and isolation. First, a working protocol was developed (but unpublished) and followed in the process.

Eligibility (inclusion and exclusion) criteria for the review
Based on the relevance of the reported evidence for decision making at local, national, and international levels, the papers were selected and prioritized for the review. The relevant outcomes observed in the review were reduction in incidence, transmission, adverse outcome, and cost-effectiveness of COVID-19 prevention through contact tracing, screening, quarantine, and isolation.

Types of studies
Due to the infancy of the epidemic, lack of researches, and ethical concerns, randomized controlled trials were not included. Therefore, we considered non-randomized observational studies and modeling (mathematical and/ or epidemiological) studies to supplement the existing evidences.
We included cohort studies, case-control studies, time series, case series, and mathematical modeling studies conducted anywhere, in any area, and in any setting reported in the English language. Whereas, commentaries, letter to editor, case reports, and governmental reports were excluded.

Types of participants
Depending on the type of the research, for each preventive methods, different participants were included. These includes individuals who have had contacts with confirmed or suspected case of COVID-19, or individuals who lived in areas with COVID-19 outbreak, or individuals considered to be at high risk for COVID-19/suspected cases or cases of COVID-19 infection. The number of participants varies according to the individual researches. Individuals who have confirmed other symptomatic respiratory diseases were excluded.

Types of interventions
We included different types of interventions applied specifically or in combination, either voluntary or mandatory and in different settings (facility or community). In comparative studies, the interventions were compared with the non-applied groups or other comparison groups. We excluded interventions other than the aforementioned strategies.

Types of outcome measures
To identify the extent to which these interventions were applied globally and to measure their effectiveness in COVID-19 prevention, we used the following outcome measures: incidence of COVID-19, onward transmission, mortality or other adverse outcomes, and cost-effectiveness. We did not address secondary outcomes such as psychological impacts, economic impacts, and social impacts.

Literature search strategy
A systematic literature search of articles was done by information system professionals and the researchers. Articles published between January 1, 2020, and June 2, 2020, containing information on different prevention strategies such as contact tracing, screening, quarantine, and isolation, and studies assessing their effectiveness were retained for the review. Electronic bibliographic databases and libraries such as PubMed/Medline, Global Health Database, Embase, CINAHL (Cumulative Index to Nursing and Allied Health Literature; Ebsco), the Cochrane Library, and African Index Medicus were used.
In addition, we searched gray literatures, pre-prints, and resource centers of The Lancet, JAMA, and N Engl J Med. Lastly, we screened the reference lists of systematic reviews for additional source. Combination of the following search terms were used with (AND, OR, NOT) Boolean (Search) Operators.

Data collection and analysis Study selection process
The team screened all the titles and abstracts based on predefined eligibility criteria. Two authors independently screened the titles and abstracts and reached consensus by discussion or by involving a third author. After that, the review author team retrieved the full texts of all included abstracts. Two review authors screened all the full-text publications independently, and disagreements were resolved with consensus or by a third person involvement.

Data extraction and management
Titles and abstracts found through primary electronic search were thoroughly assessed for the possibility of reporting the intended outcome and filtered for potential eligibility. One of the review authors who have experience extracted data from the included studies into standardized tables, and the second author checked completeness. From each eligible research, the following information was extracted based on the preformed format: author information, title, study participants, study design, study setting, type of intervention, length of intervention, year of publication, study duration, eligibility criteria, rate, and effect of intervention measures. For modeling studies, the data extraction items also included the type of model and the data source.

Assessment of risk of bias in included studies
Risk of bias was assessed through evaluating reliability and validity of data in included studies based on the Risk-Of-Bias In Non-randomized Studies -of Interventions (ROBINS-I) tool [11]. The first author rated the risk of bias, the second author checked the ratings, and the third author was involved in the disagreements. For each studies, the study design, participants, outcome, and presence of bias were assessed based on the eligibility criteria and quality assessment check list. Moreover, all studies with the same participants and outcome were measured using the same standard.
On the other hand, modeling studies were assessed by the International Society for Pharmaco-economics and Outcomes (ISPOR) and the Society for Medical Decision making (SMDM) for dynamic mathematical transmission model tools [12]. Modeling studies that fulfilled all the three criteria were rated as "no concerns to minor concerns, " and if one or more categories were unclear, it is rated as "moderate concerns," and if one or more categories were not fulfilled, we had it rated as "major concerns."

Data synthesis and analysis
The qualitative data was systematically reviewed and presented in accordance with the Cochrane guide line. We synthesized results from quantitative measures narratively and reported in tabular form. Because of the heterogeneity of the primary studies, quantitative analyses (meta-analysis) were not conducted.

Assessment of the certainty of the evidence
By using the GRADE approach [13], we graded the certainty of evidence for the main outcomes, reported in standard terms using tables. One of the authors conducted the certainty assessment which consists of assessments of risk of bias, indirectness, inconsistency, imprecision, and publication bias, and then, classified to one of the four categories: a high certainty (estimated effect lies close to the true effect), a moderate certainty (estimated effect is probably close to the true effect), a low certainty (estimated effect might substantially differ), and very low certainty (estimated effect is probably markedly different) from the true effect.

Studies included
The PRISMA flow diagram for the selected studies in the search process and the eligibility assessment are summarized in (Fig. 1). The initial electronic database search led to 1542 potentially relevant citations in the form of a title, abstract, bibliography, and full-text research. After removal of duplicates and initial screening, 125 articles were selected for further evaluation via fulltext articles. Of these full-text articles, 103 articles were excluded due to the following reasons: 38 studies reported the prevention of SARS other than COVID-19; 36 have measured prevention measures other than contact tracing, screening, quarantine, and isolation; 19 had inappropriate study designs (commentaries, letters and case reports); and 10 were reviews or protocols. Thus, 22 studies  met the inclusion criteria and were included in the systematic review.

Quality (risk of bias) assessment within included studies
Summaries of the risk of bias assessment of nonrandomized studies and quality rating of the modeling studies are presented in Tables 2 and 3, respectively. Two studies [14,19] have low bias due to confounding, eight studies have low bias in selection of participants into the study, and all studies have low bias in classification of interventions. The overall risk of bias is moderate for eight studies and serious for one study. On the other hand, we have no concern for nine modeling studies, and two studies have major concerns.

COVID-19 prevention strategies and effectiveness
The summary result is presented in Table 1. Among the nine observational studies, three of them assessed COVID-19 transmission with the existing prevention measures at a community level in Taiwan, China, and Hong Kong [18][19][20]. The other two studies assessed the effect of escalating prevention measures at health facilities in China and Hong Kong [21,22], and three studies [15][16][17] assessed national-and metropolitan-based quarantine strategies and the effect of laboratory-based quarantine in the prevention of COVID-19. The last study evaluated the effect of community-based contact tracing in UK [14]. The three studies [18][19][20] that assessed the overall prevention strategies found out that integration of interventions need to be applied instead of adhering to a single intervention. Cheng [18] reported that isolating symptomatic patients alone may not be sufficient enough to contain the epidemic. Wang [19] and Law [20] also concluded that in intimate contacts the transmission is 40-60%. Preventing contact through different strategies and integration is very important.
Studies conducted on the effect of quarantine [15][16][17] found that it can have a massive preventive effect. One of the studies [15] that assessed the effect of quarantine in different populations and quarantine strategies found that it should be integrated with input population reduction (travel restriction), and the other study [16] that assessed the effects of metropolitan-wide quarantine on the Spread of COVID-19 in China found that quarantine would prevent 79.27% (75.10-83.45%) of deaths and 87.08% (84.68-89.49%) of infections. Also, the other researcher [17] evidenced that laboratory-based screenings accomplished within hours can enhance the efficiency of quarantine.
Two studies described infection control preparedness measures in health care settings of Hong Kong and China [21,22]. One of these studies [21] reported that infection transmission is highly increased within a short period of time and multiplicity of infection prevention strategies were recommended for prevention in health care setups. The other study [22] also concluded that practicing working shift among professionals working in facilities can be used as strategy to prevent thetransmission of COVID infection.
A study conducted by Keeling et al. [14] assessed the efficacy of contact tracing for the containment of COVID-19 in the UK. The study evaluated the contact pattern of the community and concluded that rapid contact tracing to reduce the basic reproduction number (R 0 ) from 3.11 to 0.21 enables the outbreak to be contained. Additionally, it was found that each new case requires an average of 36 individuals to be traced, with 8.7% of cases having more than 100 close traceable contacts.
Three of these researches [25][26][27] particularly emphasized on the way how the R 0 can be reduced and the epidemic would be reduced. The simulation by Tang et al. [25] aimed to estimate the R 0 of SARS-CoV-2 and infer the required effectiveness of isolation and quarantine to contain the outbreak. Their susceptible-exposed-infected-recovered (SEIR) model estimated R 0 of 6.47 and generalized that 50% reduction of contact rate achieved by isolation and quarantine would decrease the confirmed cases by 44%; reducing contacts by 90% also can decrease the number of cases by 65%. The other researcher, Rocklov (27), by using data from the Diamond Princess Cruise ship, concluded that quarantine of passengers prevented 67% of cases and lowered the R 0 from 14.8 to 1.78. Similarly, the reduction of R 0 was achieved from quarantine [28]. To simulate and predict the disease variations of Guangdong province and to explore the impacts of the input population and quarantine strategies.

Study detail The impact of input population was evaluated with simulation
Interventions Quarantine Results ➢ The simulated peak value of the confirmed cases is 1002 at Feb 10, 2020 ➢ The disease will become extinction with peak value of 1397 at May 11, 2020. ➢ The increased numbers of the input population can mainly shorten the disease extinction days and the increased percentages of the exposed individuals

Results
➢ If social distancing reduces the reproductive number to 1.25 (e.g., 50% of person-to-person contact is removed in a setting where R 0 = 2.5), active monitoring of 50% of contacts can result in overall outbreak control (i.e., Re < 1). ➢ Tracing 10%, 50%, or 90% of contacts on top of social distancing resulted in a median reduction in R e of 3.2%, 15%, and 33%, respectively, for active monitoring and 5.8%, 32%, and 66%, for individual quarantine. ➢ Individual quarantine may contain an outbreak of COVID-19 with a short serial interval (4.8 days) only in settings with high intervention performance where at least three-quarters of infected contacts are individually quarantined.

18
Author/s and title [31] Biao Tang   S.N Study characteristics and summary report ✓ The strong measures implemented have reduced the effective reproduction number. These interventions may take a longer time to be effective as the second and third generations of infected people are exposed in succession.

19
Author/s and title [32] Can Study detail A mathematical model was employed to assess the feasibility of contact tracing and case isolation to control outbreaks of using simulated new outbreaks starting from 5, 20, or 40 introduced cases.

Interventions Contact tracing and case isolation
Results ✓ When R 0 was 2.5 or 3.5, the probability of controlling an outbreak decreased with the number of initial cases ✓ The majority of scenarios with an R 0 of 1.5 were controllable with less than 50% of contacts successfully traced. ✓ To control the majority of outbreaks, for R 0 of 2.5, more than 70% of contacts had to be traced, and for an R 0 of 3.5, more than 90% of contacts had to be traced. ✓ The delay between symptom onset and isolation had the largest role in determining whether an outbreak was controllable when R 0 was 1.5. ✓ For R 0 values of 2.5 or 3.5, if there were 40 initial cases, contact tracing and isolation were only potentially feasible when less than 1% of transmission occurred before symptom onset. ✓ Contact tracing and isolation might not contain outbreaks of COVID-19 unless very high levels of contact tracing are achieved. In addition to these, five studies [24,28,30,31,35] which modeled the effectiveness of different public interventions consistently reported that integrated intervention is better than a single intervention. One of these research conducted in the UK [24] found that combined isolation and tracing strategies would reduce transmission more than mass testing or self-isolation alone (50-60% compared to 2-30%). The other study [28] also reported that with R 0 of 2.4, a combination of case isolation and voluntary quarantine for 3 months could prevent 31% of deaths. The others also concluded that quarantine should be strict and integrated with contact tracing, screening, and other interventions [30,31,35].
Five modeling studies also assessed the effect of quarantine [23,29,32], contact tracing [33], and screening [34]. All of the studies [23,29,32] reported that quarantine has reduced the incidence of infection and shortened the duration of the epidemic. However, the effectiveness depends on the level of integration with other strategies. Similarly, model simulations that assessed the effect of contact tracing and screening reported that the strategies are effective. However, as the report of Hellewell [33] stated, contact tracing and isolation might not contain outbreaks of COVID-19 unless very high levels of contact tracing are achieved. Similarly, the other researcher [34] reported that in a stable epidemic, under the assumption that 25% of cases are subclinical, it is estimated that arrival screening alone would detect roughly one-third of infected travelers.

Interventions
Isolation and contact tracing in populations with social distancing

Results
✓ If the proportion of asymptomatic infections is larger than 30%, contact tracing and isolation cannot achieve containment for an R 0 of 2.5 ✓ To achieve containment by social distancing requires a reduction of numbers of non-household contacts by around 90%. ✓ Social distancing reduces non-household contacts only by 50%, tracing and isolation also of non-household contacts is needed for containment.

Discussion
This study aimed to assess the effectiveness of contact tracing, screening, and quarantine and isolation to prevent COVID-19 infection by reviewing existing literatures. The review identified and systematically synthesized the findings of 22 studies (9 observational and 13 modeling studies)  to bring the best available evidence that policy makers and implementers can use in the process of infection prevention interventions.
The studies consistently reported the benefit of contact tracing, screening, quarantine, and isolation in the prevention of COVID-19. The effectiveness of quarantine in particular is very high. Compared to individuals without any intervention quarantined people exposed to a confirmed case highly averted infections and deaths [15,[23][24][25][26][27][28][29][30][31]. Also, the effectiveness of quarantine increases whenever it is implemented along with other prevention measures such as isolation, contact tracing, and travel ban [23][24][25][26][27][28][29][30][31]. Although, screening and contact tracing are very important to control the epidemic, early initiation, larger coverage, and integration with other programs are very important. Unless the level of contact tracing and screening is high, prevention through isolation only is very limited, as the screening programs misses 75% of cases [3,24].
Quarantine measures applied alone or integrated with other measures were reported to be the most effective measures [25][26][27][28][29][30][31]. However, integration of quarantine with other public health measures increases the effectiveness and efficiency of the program [36]. Implementation of early quarantine measures makes the strategy a more cost effective one [28,30]. Quarantine implemented as self-quarantine and group quarantine is effective at varying levels once effectively implemented [28,32]. Total lockdown measures enhance the effectiveness of quarantine measures [15][16][17][18][19]. When laboratory tests are very fast, laboratory-based quarantine could be an effective in health care setups [17].
This evidence is in line with the finding of other reviews and modeling studies conducted to assess the effectiveness of these measures in the prevention of SARS, MERS, and COVID-19 [28,[35][36][37]. As reported before, combination of case isolation and voluntary quarantine for 3 months could prevent 31% of deaths compared to any single intervention. And adding social distancing on the previous interventions on people aged 70 years or older for 4 months increases the prevention proportion of deaths to 49%. It can also reduce the reproductive number by half; hence, it can tremendously reduce the incidence of infection, reduce the period of epidemic, and enhance effectiveness of control [28,36].
Our findings also witnessed the effectiveness of contact tracing measures used for pandemic response efforts at multiple levels of health care systems. Isolation of suspected and confirmed patients and their contact is at the heart of the prevention strategy. However, for the contact tracing to be an effective measure, it has to be integrated with other measures such as quarantine and screening. Because larger shares of individuals are asymptomatic, contact tracing may be difficult in areas where contact recording is unachievable. According to world health organization, contact tracing is also one of the most essential and effective strategies to control the epidemic [14,24,25,33,35]. Other studies also evidenced the importance of contact tracing and isolation in different settings [36,37]. The finding of our review revealed that screening and isolation are important measures of disease prevention [17,25,26,31,33,35]. Most of the researches recommend high-risk group screening and contact cases screening in a resource-limited setting. However, these programs are effective when the screening capacity is higher and contact tracing is effective. Otherwise, screening and isolation programs miss more than half of cases and may not be implemented alone [25,33,35]. Also evidences from different countries indicated that screening and isolation measures are implemented along with other measures, yet their role in the prevention of the epidemic is high [2,3,8,36,37].

Limitation
This review included a wide variety of study designs (observational and model studies); hence, it failed to include meta-analysis (statistical measures). Modeled studies also assume different scenarios, where it may not be true in the general cases. Also, the review has included only publications reported in the English language and open access resources.

Conclusion and recommendation
Quarantine, contact tracing, screening, and isolation are effective measures of COVID-19 prevention, particularly whenever integrated together. In order to be more effective, quarantine should be implemented early and covers larger community. Controlling population travel will enhance the effectiveness of quarantine. Screening, contact tracing, and isolation are effective particularly in areas where contact tracing is easily attainable. Although screening is the effective measure recommended by the WHO, since the disease is asymptomatic, it may miss a larger share of the population. Therefore, this should be integrated with other preventive measures. In order to control the COVID-19 epidemic, the health care system should consider high level of contact tracing, early initiation of nationwide quarantine measures, increasing coverage of screening service, and preparing effective isolation centers.