Artificial intelligence in COVID-19

The COVID-19 pandemic has taken the world by storm, placing healthcare systems around the globe under immense pressure. The exceptional circumstance has made the scientific community turn to artificial intelligence (AI), with hopes that AI techniques can be used in all aspects of combating the pandemic, whether it is in using AI to uncover sequences in the genomic code of the severe acute respiratory syndrome coronavirus (SARS-CoV-2) virus for the purposes of developing therapeutics, such as antivirals, antibodies, or vaccines, or using AI to provide (near-) instantaneous clinical diagnosis techniques by way of analysis of chest X-ray (CXR) images, computed tomography (CT) scans or other useful modalities, or using AI for as a tool for mass population testing by analyzing patient audio recordings. In this chapter, we survey the AI research literature with respect to applications for COVID-19 and showcase and critique notable state of the art approaches. © 2023 Elsevier Inc. All rights reserved.

COVID-19 dynamics in an Ohio prison.

Incarcerated individuals are a highly vulnerable population for infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Understanding the transmission of respiratory infections within prisons and between prisons and surrounding communities is a crucial component of pandemic preparedness and response. Here, we use mathematical and statistical models to analyze publicly available data on the spread of SARS-CoV-2 reported by the Ohio Department of Rehabilitation and Corrections (ODRC). Results from mass testing conducted on April 16, 2020 were analyzed together with time of first reported SARS-CoV-2 infection among Marion Correctional Institution (MCI) inmates. Extremely rapid, widespread infection of MCI inmates was reported, with nearly 80% of inmates infected within 3 weeks of the first reported inmate case. The dynamical survival analysis (DSA) framework that we use allows the derivation of explicit likelihoods based on mathematical models of transmission. We find that these data are consistent with three non-exclusive possibilities: (i) a basic reproduction number >14 with a single initially infected inmate, (ii) an initial superspreading event resulting in several hundred initially infected inmates with a reproduction number of approximately three, or (iii) earlier undetected circulation of virus among inmates prior to April. All three scenarios attest to the vulnerabilities of prisoners to COVID-19, and the inability to distinguish among these possibilities highlights the need for improved infection surveillance and reporting in prisons.

Participatory management of the response to the Covid-19 emergency. The unique experience of the Autonomous Province of Bolzano

In Europe, several countries introduced mass screening but in Italy, the Province of Bolzano has been the first to use it, introducing a paradigm shift in management of the Covid-19 emergency. The "Rapid tests in South Tyrol” program (20-22 November 2020) tested about 65% of the population and was followed by the "Monitoring South Tyrol” program, so far unique in Europe, which monitored a representative sample of the population of over 20,000 people (December 2020-April 2021). The process was characterized by the close collaboration between public institutions and between these and private associations, by the strong investment in research and experimentation of innovative solutions and by individual and collective responsibility, accouting for the effects of the choices made not only in terms of health, but also in terms of economic and social impact. © 2022 FrancoAngeli. All rights reserved.

COVID-19 mass testing and sequencing: Experiences from a laboratory in Western Kenya.

Background: The Basic Science Laboratory (BSL) of the Kenya Medical Research Institute/Walter Reed Project in Kisumu, Kenya addressed mass testing challenges posed by the emergent coronavirus disease 2019 (COVID-19) in an environment of global supply shortages. Before COVID-19, the BSL had adequate resources for disease surveillance and was therefore designated as one of the testing centres for COVID-19. Intervention: By April 2020, the BSL had developed stringent safety procedures for receiving and mass testing potentially infectious nasal specimens. To accommodate increased demand, BSL personnel worked in units: nucleic acid extraction, polymerase chain reaction, and data and quality assurance checks. The BSL adopted procedures for tracking sample integrity and minimising cross-contamination. Lessons learnt: Between May 2020 and January 2022, the BSL tested 63 542 samples, of which 5375 (8.59%) were positive for COVID-19; 1034 genomes were generated by whole genome sequencing and deposited in the Global Initiative on Sharing All Influenza Data database to aid global tracking of viral lineages. At the height of the pandemic (August and November 2020, April and August 2021 and January 2022), the BSL was testing more than 500 samples daily, compared to 150 per month prior to COVID-19. An important lesson from the COVID-19 pandemic was the discovery of untapped resilience within BSL personnel that allowed adaptability when the situation demanded. Strict safety procedures and quality management that are often difficult to maintain became routine. Recommendations: A fundamental lesson to embrace is that there is no 'one-size-fits-all' approach and adaptability is the key to success.

How to design and implement a university-based COVID-19 testing programme? An evaluation of a novel RT-LAMP COVID-19 testing programme in a UK university.

BACKGROUND: Little is known about how asymptomatic testing as a method to control transmission of COVID-19 can be implemented, and the prevalence of asymptomatic infection within university populations. The objective of this study was to investigate how to effectively set-up and implement a COVID-19 testing programme using novel reverse transcriptase loop-mediated isothermal amplification (RT-LAMP) technology and to quantify the scale of asymptomatic infection on a university campus. METHODS: An observational study to describe the set-up and implementation of a novel COVID-19 testing programme on a UK university campus between September and December 2020. RT-LAMP testing was used to identify asymptomatic cases. RESULTS: A total of 1,673 tests were performed using RT-LAMP during the study period, of which 9 were positive for COVID-19, giving an overall positivity rate of 0.54%, equivalent to a rate in the tested population of 538 cases per 100,000 over the duration of testing. All positive tests were found to be positive on RT-PCR testing, giving a false positive rate of 0%. CONCLUSIONS: This study shows that it is possible to rapidly setup a universal university testing programme for COVID-19 in collaboration with local healthcare providers using RT-LAMP testing. Positive results were comparable to those in the local population, though with a different peak of infection. Further research to inform the design of the testing programme includes focus groups of those who underwent testing and further interrogation of the demographics of those opting to be tested to identify potential access problems or inequalities.

A Descriptive Comparison of Mass Testing During the COVID-19 Pandemic in Montreal, Paris, Bamako, and Recife.

Objective: The aim of this descriptive article was to compare mass testing for SARS-CoV-2 during the first wave of the COVID-19 pandemic in Montreal, Canada; Bamako, Mali; Paris, France; and Recife, Brazil. Methods: Data was collected through interviews with key informants involved in the testing response and a review of the grey literature. The TIDieR-PHP checklist was then used to provide the basis of the intervention descriptions and to compare the data between cities. Results: Descriptive comparisons revealed that the type of test, the testing process, and materials used were similar between the cities during the first wave of the pandemic. In addition, all cities experienced similar material and personnel resource shortages, directly affecting testing accessibility and capacity. The main differences were related to testing capacity and implementation timelines, which were dependent on the state of the health care systems, governance, and access to resources. Conclusion: Results of this study highlight the similarities and differences in testing between the cities and demonstrate the importance of comprehensive intervention descriptions to highlight lessons learned, increase knowledge sharing, and inform policy decisions.

Hope and trust: Public attitudes toward mass COVID-19 testing programs in Guangzhou, China.

Mass testing is one COVID-19 pandemic response strategy. The effect of population-wide testing programs is influenced by public attitudes toward COVID-19 viral tests. However, the public's attitudes toward mass testing and related factors in mainland China are not adequately understood. This study focuses on pandemic responses during the first wave of the Delta variant outbreak in southern China and explores how residents responded to population-wide mass COVID-19 testing programs. The research relies on data collected from short videos recording residents' experiences of being in lockdown, media reports, and semi-structured interviews. Thematic analysis was used to analyze the data, and four themes emerged: public attitudes toward mass testing, the technology of viral tests, policy and governance, and cultural practices. The study finds that residents actively participated in mass testing campaigns as mass viral tests are associated with hope and trust. The Reverse Transcription-Polymerase Chain Reaction (RT-PCR), negative test results, lockdown policies, waiting times, medical staff, and media representations are all actors that assemble and mobilize hope and trust. The research reveals some critical factors influencing people's attitudes toward mass testing policies in response to COVID-19 and provides practical suggestions for public health professionals in rolling out effective population-wide mass COVID-19 testing.

Cost-effectiveness of a whole-area testing pilot of asymptomatic SARS-CoV-2 infections with lateral flow devices: a modelling and economic analysis study.

BACKGROUND: Mass community testing for SARS-CoV-2 by lateral flow devices (LFDs) aims to reduce prevalence in the community. However its effectiveness as a public heath intervention is disputed. METHOD: Data from a mass testing pilot in the Borough of Merthyr Tydfil in late 2020 was used to model cases, hospitalisations, ICU admissions and deaths prevented. Further economic analysis with a healthcare perspective assessed cost-effectiveness in terms of healthcare costs avoided and QALYs gained. RESULTS: An initial conservative estimate of 360 (95% CI: 311-418) cases were prevented by the mass testing, representing a would-be reduction of 11% of all cases diagnosed in Merthyr Tydfil residents during the same period. Modelling healthcare burden estimates that 24 (16-36) hospitalizations, 5 (3-6) ICU admissions and 15 (11-20) deaths were prevented, representing 6.37%, 11.1% and 8.2%, respectively of the actual counts during the same period. A less conservative, best-case scenario predicts 2333 (1764-3115) cases prevented, representing 80% reduction in would-be cases. Cost -effectiveness analysis indicates 108 (80-143) QALYs gained, an incremental cost-effectiveness ratio of £2,143 (£860-£4,175) per QALY gained and net monetary benefit of £6.2 m (£4.5 m-£8.4 m). In the best-case scenario, this increases to £15.9 m (£12.3 m-£20.5 m). CONCLUSIONS: A non-negligible number of cases, hospitalisations and deaths were prevented by the mass testing pilot. Considering QALYs gained and healthcare costs avoided, the pilot was cost-effective. These findings suggest mass testing with LFDs in areas of high prevalence (> 2%) is likely to provide significant public health benefit. It is not yet clear whether similar benefits will be obtained in low prevalence settings or with vaccination rollout.

Adaptive group testing in a compartmental model of COVID-19.

Various measures have been implemented around the world to prevent the spread of SARS-CoV-2. A potential tool to reduce disease transmission is regular mass testing of a high percentage of the population, possibly with pooling (testing a compound of several samples with one single test). We develop a compartmental model to study the applicability of this method and compare different pooling strategies: regular and Dorfman pooling. The model includes isolated compartments as well, from where individuals rejoin the active population after some time delay. We develop a method to optimize Dorfman pooling depending on disease prevalence and establish an adaptive strategy to select variable pool sizes during the course of the epidemic. It is shown that optimizing the pool size can avert a significant number of infections. The adaptive strategy is much more efficient, and may prevent an epidemic outbreak even in situations when a fixed pool size strategy can not.

Universal Community Nucleic Acid Testing for Coronavirus Disease 2019 (COVID-19) in Hong Kong Reveals Insights Into Transmission Dynamics: A Cross-Sectional and Modeling Study.

BACKGROUND: Testing of an entire community has been used as an approach to control coronavirus disease 2019 (COVID-19). In Hong Kong, a universal community testing program (UCTP) was implemented at the fadeout phase of a community epidemic in July to September 2020. We described the utility of the UCTP in finding unrecognized infections and analyzed data from the UCTP and other sources to characterize transmission dynamics. METHODS: We described the characteristics of people participating in the UCTP and compared the clinical and epidemiological characteristics of COVID-19 cases detected by the UCTP versus those detected by clinical diagnosis and public health surveillance (CDPHS). We developed a Bayesian model to estimate the age-specific incidence of infection and the proportion of cases detected by CDPHS. RESULTS: In total, 1.77 million people, 24% of the Hong Kong population, participated in the UCTP from 1 to 14 September 2020. The UCTP identified 32 new infections (1.8 per 100000 samples tested), consisting of 29% of all local cases reported during the two-week UCTP period. Compared with the CDPHS, the UCTP detected a higher proportion of sporadic cases (62% vs 27%, P<.01) and identified 6 (out of 18) additional clusters during that period. We estimated that 27% (95% credible interval: 22%, 34%) of all infections were detected by the CDPHS in the third wave. CONCLUSIONS: We reported empirical evidence of the utility of population-wide COVID-19 testing in detecting unrecognized infections and clusters. Around three quarters of infections have not been identified through existing surveillance approaches including contact tracing.

Perception of testing for COVID-19 during the first wave of the pandemic in Slovakia with emphasis on population age groups.

OBJECTIVE: The aim of this study was to evaluate the perception of the respondents in selected areas of testing for COVID-19 during the first waves of this disease with an emphasis on the age categories. METHODS: The research sample consisted of 806 Slovak respondents and the collection of data took place in February 2021. The study examined six areas, of which five focused on the perception of testing and one area focused on examining the risk of population behaviour. All areas were examined in the first as well as in the last testing for COVID-19. In terms of age, we focused on 4 age categories: up to 24 years, 25-44 years, 45-59 years, and over 60 years. Data were obtained by online survey, using descriptive analysis and nonparametric analysis of differences. RESULTS: The results of the analysis show that age is a socioeconomic characteristic that is of great importance for the creation and implementation of epidemiological processes and programmes and therefore epidemiologists should pay increased attention to it when creating prevention programmes. Younger people perceived testing more positively than older ones, even though the course of the disease caused by COVID-19 is much more severe in older people. The difference in the perception of testing between age categories was confirmed in the first as well as in the last test study. Respondents perceived the last test more negatively and thus they showed slightly riskier behaviour compared to the first test. CONCLUSIONS: We perceive testing as an efficient way in the fight against pandemics generally, but we point out that testing should be well accomplished in a managerial way, otherwise testing can have a negative impact on society's confidence.

Crucial control measures to contain China's first Delta variant outbreak.

The SARS-CoV-2 B.1.617.2 (Delta) variant flared up in late May in Guangzhou, China. Transmission characteristics of Delta variant were analysed for 153 confirmed cases and two complete transmission chains with seven generations were fully presented. A rapid transmission occurred in five generations within 10 days. The basic reproduction number (R0) was 3.60 (95% confidence interval: 2.50-5.30). After redefining the concept of close contact, the proportion of confirmed cases discovered from close contacts increased from 43% to 100%. With the usage of a yellow health code, the potential exposed individuals were self-motivated to take a nucleic acid test and regained public access with a negative testing result. Facing the massive requirement of screening, novel facilities like makeshift inflatable laboratories were promptly set up as a vital supplement and 17 cases were found, with 1 pre-symptomatic. The dynamic adjustment of these three interventions resulted in the decline of Rt from 5.00 to 1.00 within 9 days. By breaking the transmission chain and eliminating the transmission source through extending the scope of the close-contact tracing, health-code usage and mass testing, the Guangzhou Delta epidemic was effectively contained.

Modelling the Impact of Mass Testing to Transition from Pandemic Mitigation to Endemic COVID-19.

As countries transition from pandemic mitigation to endemic COVID-19, mass testing may blunt the impact on the healthcare system of the liminal wave. We used GeoDEMOS-R, an agent-based model of Singapore's population with demographic distributions and vaccination status. A 250-day COVID-19 Delta variant model was run at varying maximal rapid antigen test sensitivities and frequencies. Without testing, the number of infections reached 1,021,000 (899,400-1,147,000) at 250 days. When conducting fortnightly and weekly mass routine rapid antigen testing 30 days into the outbreak at a maximal test sensitivity of 0.6, this was reduced by 12.8% (11.3-14.5%) and 25.2% (22.5-28.5%). An increase in maximal test sensitivity of 0.2 results a corresponding reduction of 17.5% (15.5-20.2%) and 34.4% (30.5-39.1%). Within the maximal test sensitivity range of 0.6-0.8, test frequency has a greater impact than maximal test sensitivity with an average reduction of 2.2% in infections for each day removed between tests in comparison to a 0.43% average reduction per 1% increase in test frequency. Our findings highlight that mass testing using rapid diagnostic tests can be used as an effective intervention for countries transitioning from pandemic mitigation to endemic COVID-19.

Analysis of a city-wide COVID-19 prevention strategy for aged-care facilities during third and fifth waves of COVID-19 in Kyoto City, Kyoto, Japan.

BACKGROUND: During the third wave of the COVID-19 pandemic at the end of 2020, clusters occurred frequently in aged-care facilities (ACFs), which put pressure on the medical field in Japan. Based on this experience, Kyoto University and Kyoto City collaborated to promote a citywide COVID-19 prevention strategy to prevent the spread of COVID-19 within ACFs. The aim of this study was to clarify the effect of the prevention strategy among ACFs in Kyoto City during the third and fifth waves of the pandemic. METHODS: During the study period, the following measures were adopted as the prevention strategy in all ACFs: (1) active polymerase chain reaction (PCR) mass testing and facility-wide testing when a single case was identified, (2) implementation of strategies to prevent transmission within a facility, and (3) vaccination program for ACFs. RESULTS: Of the 1,144 facilities subjected to the mass testing, 71.0% participated in the whole program including active PCR testing. The remainder participated in the rest of the programs. The prevalence of ACF-related COVID-19 cases among total COVID-19 cases in Kyoto City decreased from 7.9% in the third wave to 4.1% in the fourth wave and 2.1% in the fifth wave. The incidence of clusters and proportion of severe elderly cases also decreased during the study period. CONCLUSIONS: A city-wide multidisciplinary effort including PCR mass testing and a vaccination program in cooperation with a university and local administrative office successfully reduced the clusters and transmission in ACFs in Kyoto City, Japan.

Perception of COVID-19 Testing in the Entire Population.

In the Slovak Republic, a mass testing of the entire population was performed. Estimates show that this testing cost more than 400 million EUR and thousands of euros were paid for one positively identified case. Thus, it is possible to state a high cost for such a project, which has been criticized by many parties. On the other hand, from a public health point of view, mass testing has helped fight the pandemic. Both the health and economic perspectives are important in assessing the success of a pandemic strategy, but the social perspective is equally important. In fact, the situation is perceived from the position of public leaders who make decisions, but also from the position of the society that bears individual political decisions. It is not appropriate to forget about the society that is most affected by restrictions, testing, health status, but also the burden on the state budget. The objective of the presented research was to examine the perception of testing for coronavirus disease 2019 (COVID-19) in the Slovak population. Non-parametric difference tests and correspondence analysis were used for statistical processing. The research sample consisted of 806 respondents and data collection took place in February 2021. The main findings include significant differences in perceptions between the first and the last participation in testing in terms of gender, age, testing experience, and time aspect. The last participation in testing showed lower rates of positive aspects related to the internal motivation to test compared to the first participation. In contrast, external stimulation by government regulations related to restrictions in the absence of a negative result was higher in the last participation in testing. There were also differences between the first and the last test in the level of doubts about the accuracy of the test result, while a higher level was found at the last testing participation. It can be concluded that the frequency of testing and its requirements need to be approached very carefully over time, as it is likely that the positive perceptions may deteriorate. The recommendations include clear and timely government communication, trust building and health education.

Mass testing against COVID-19: a logistic simulation model to design and manage swab test clinics

COVID-19 pandemic caused several million deaths worldwide since the beginning of 2020. One of the most effective activities to contrast its diffusion is the execution of mass testing campaigns to track the virus spread. To design an efficient logistic system for such purpose, it is important to correctly plan the clinic layout and size the medical resources involved in the swab testing campaign, to avoid long patient queues or personnel underutilization. This paper describes the development of an original logistics simulation model to support the planning and design of clinics for the walk-in mass testing campaign against COVID-19 performed by South Tyrol Health Agency in just one weekend in November 2020 which involved more than 350’000 citizens. The developed model represents the targeted physical system considering all the different phase of such healthcare process. Furthermore, the duration of the multiple process phases is statistically distributed according to a large dataset collected during the COVID-19 testing campaign for touristic operators conducted in September and October 2020 in South Tyrol. The simulation model virtually evaluates the swab testing clinics with different parameters to determine the best scenario to be implemented. It concerns the number of medical resources allocated, the necessary clinic spaces and the time spent inside the clinic by each patient. The obtained results suggest that the so-defined configuration is distinguished by an average throughput of 8.8 minutes per patient. This clinic prototype has been replicated and set-up all over South-Tyrol territory to reach the targeted number of tested citizens. Indeed, 362’050 people were effectively tested from 20th to 22nd November 2020 leveraging 184 clinics and about 1400 healthcare co-workers, both medical and non-medical personnel. 3’615 Covid-positive people were detected and the virus transmission index of this Province fell from 1.22 to 0.74 just in the following 2 weeks. © 2021, AIDI - Italian Association of Industrial Operations Professors. All rights reserved.

Ultrasensitive assay for saliva-based SARS-CoV-2 antigen detection.

OBJECTIVES: Widespread SARS-CoV-2 testing is invaluable for identifying asymptomatic/pre-symptomatic individuals. There remains a technological gap for highly reliable, easy, and quick SARS-CoV-2 diagnostic tests suitable for frequent mass testing. Compared to nasopharyngeal (NP) swab-based tests, saliva-based methods are attractive due to easier and safer sampling. Current saliva-based SARS-CoV-2 rapid antigen tests (RATs) are hindered by limited analytical sensitivity. Here, we report one of the first ultrasensitive, saliva-based SARS-CoV-2 antigen assays with an analytical sensitivity of <0.32 pg/mL, corresponding to four viral RNA copies/µL, which is comparable to that of PCR-based tests. METHODS: Using the novel electrochemiluminescence (ECL)-based immunoassay, we measured the SARS-CoV-2 nucleocapsid (N) antigen concentration in 105 salivas, obtained from non-COVID-19 and COVID-19 patients. We then verified the results with a second, independent cohort of 689 patients (3.8% SARS-CoV-2 positivity rate). We also compared our method with a widely used point-of-care rapid test. RESULTS: In the first cohort, at 100% specificity, the sensitivity was 92%. Our assay correctly identified samples with viral loads up to 35 CT cycles by saliva-based PCR. Paired NP swab-based PCR results were obtained for 86 cases. Our assay showed high concordance with saliva-based and NP swab-based PCR in samples with negative (<0.32 pg/mL) and strongly positive (>2 pg/mL) N antigen concentrations. In the second cohort, at 100% specificity, sensitivity was also 92%. Our assay is about 700-fold more sensitive than the Abbott Panbio Rapid Test. CONCLUSIONS: We demonstrated the ultrasensitivity and specificity assay and its concordance with PCR. This novel assay is especially valuable when compliance to frequent swabbing may be problematic.

Multi-dimensional factors related to participation in a population-wide mass COVID-19 testing program among Hong Kong adults: A population-based randomized survey.

RATIONALE: Mass testing is considered as an important policy response to the COVID-19 pandemic, and high population coverage is pivotal to its effectiveness. A range of factors derived from health behaviour theories were hypothesized to be associated with public uptake of mass testing, including illness representations of COVID-19, perceived susceptibility to COVID-19, perceived efficacy of the testing program, and general trust toward governmental measures for controlling COVID-19. OBJECTIVE: This study aimed to investigate the multi-dimensional factors associated with participation in a free and voluntary population-wide mass COVID-19 testing program. METHODS: A cross-sectional study was conducted in Hong Kong within two weeks after the Universal Community Testing Program for COVID-19 concluded on September 14, 2020. A random population-based telephone survey interviewed 443 Hong Kong general adults who were aged ≥18 and had not joined other COVID-19 testing programs. The dependent variable was participation in the Universal Community Testing Program. Logistic regression analysis was conducted to test the associations of participation in the program with the proposed factors. RESULTS: The standardized participation rate of the testing program was estimated to be about 37.2% among the general adults (33.0% among males; 40.8% among females) in Hong Kong. The participation rates were significantly lower among males and younger adults. Adjusted for socio-demographics, significant factors included four dimensions of illness representations of COVID-19 (treatment control: adjusted odds ratio [AOR] = 1.41; illness identity: AOR = 1.10; concern: AOR = 1.14; emotions: AOR = 1.10), perceived susceptibility to COVID-19 (AOR = 1.40), perceived efficacy of the testing program (AOR = 2.73), and trust toward governmental control measures (AOR = 4.30). CONCLUSIONS: The participation rate of the population-wide mass testing program was not high among general adults in Hong Kong, evidence-based health promotion is necessary. The study informs some critical factors to be addressed to effectively boost public support for the mass testing policy in response to emerging infectious diseases.

COVID-19 Testing in Sweden During 2020-Split Responsibilities and Multi-Level Challenges.

Sweden's use of soft response measures early in the COVID-19 pandemic received a good deal of international attention. Within Sweden, one of the most debated aspects of the pandemic response has been COVID-19 testing and the time it took to increase testing capacity. In this article, the development of and the debate surrounding COVID-19 testing in Sweden during 2020 is described in detail, with a particular focus on the coordination between national and regional actors in the decentralised healthcare system. A qualitative case study was carried out based on qualitative document analysis with a chronological presentation. To understand COVID-19 testing in Sweden, two aspects of its public administration model emerged as particularly important: (i) the large and independent government agencies and (ii) self-governing regions and municipalities. In addition, the responsibility principle in Swedish crisis management was crucial. Overall, the results show that mass testing was a new area for coordination and involved a number of national and regional actors with partly different views on their respective roles, responsibilities and interpretations of the laws and regulations. The description shows the ambiguities in the purpose of testing and the shortcomings in communication and cooperation during the first half of 2020, but after that an increasing consistency among the crucial actors. During the first half of 2020, testing capacity in Sweden was limited and reserved to protect the most vulnerable in society. Because mass testing for viruses is not normally carried out by the 21 self-governing regions responsible for healthcare and communicable disease prevention, and the Public Health Agency of Sweden stated that there was no medical reason to test members of the public falling ill with COVID-like symptoms, the responsibility for mass testing fell through the cracks during the first few months of the pandemic. This article thus illustrates problems associated with multi-level governance in healthcare during a crisis and illustrates the discrepancy between the health service's focus on the individual and the public health-oriented work carried out within communicable disease control.