key: cord-266526-8csl9md0 authors: Li, Shuai; Xu, Yifang; Cai, Jiannan; Hu, Da; He, Qiang title: Integrated environment-occupant-pathogen information modeling to assess and communicate room-level outbreak risks of infectious diseases date: 2020-10-24 journal: Build Environ DOI: 10.1016/j.buildenv.2020.107394 sha: doc_id: 266526 cord_uid: 8csl9md0 Microbial pathogen transmission within built environments is a main public health concern. The pandemic of coronavirus disease 2019 (COVID-19) adds to the urgency of developing effective means to reduce pathogen transmission in mass-gathering public buildings such as schools, hospitals, and airports. To inform occupants and guide facility managers to prevent and respond to infectious disease outbreaks, this study proposed a framework to assess room-level outbreak risks in buildings by modeling built environment characteristics, occupancy information, and pathogen transmission. Building information modeling (BIM) is exploited to automatically retrieve building parameters and possible occupant interactions that are relevant to pathogen transmission. The extracted information is fed into an environment pathogen transmission model to derive the basic reproduction numbers for different pathogens, which serve as proxies of outbreak potentials in rooms. A web-based system is developed to provide timely information regarding outbreak risks to occupants and facility managers. The efficacy of the proposed method was demonstrated by a case study, in which building characteristics, occupancy schedules, pathogen parameters, as well as hygiene and cleaning practices are considered for outbreak risk assessment. This study contributes to the body of knowledge by computationally integrating building, occupant, and pathogen information modeling for infectious disease outbreak assessment, and communicating actionable information for built environment management. This study aims to develop a framework for room-level outbreak risk assessment based on 105 integrated building-occupancy-pathogen modeling to mitigate the spread of infectious disease in 106 buildings. The rationale is twofold. First, buildings are highly heterogeneous with a variety of 107 compartments of distinctive functionalities and characteristics, providing diverse habitats for 108 humans and various pathogens [17, 18] . Modeling the pathogen transmission and exposure 109 within a building at the room level will provide useful information at an unprecedented resolution 110 to implement appropriate disease control strategies. Second, the spread of infectious diseases 111 can be mitigated if occupants and facility managers have adequate and timely information 112 regarding the outbreak risks within their buildings. Communicating actionable information to 113 occupants and facility managers through an easily accessible interface will help occupants to 114 follow hygiene and social distancing practice, and help facility managers to schedule disinfection 115 for rooms with high outbreak risks. 116 117 To address the knowledge gaps, a novel environment-occupant-pathogen modeling framework 119 and a web-based information visualization system are developed to assess the outbreak risks 120 and mitigate the spread of infectious diseases in buildings ( Fig. 1) . First, to assess the outbreak 121 risks, the fomite-based pathogen transmission model proposed in [24] is adopted in this study. 122 The limitation of the model is that the environmental parameters and occupant characteristics 123 are not automatically extracted and incorporated in the model, hindering the computation of the 124 spatially-varying environmental infection risks in buildings. To overcome this limitation, BIM is 125 exploited to automatically retrieve venue-specific parameters including building characteristics 126 and occupancy information that are relevant to pathogen transmission and exposure. Then, the 127 extracted building and occupant parameters are used with pathogen-specific parameters in a 128 human-building-pathogen transmission model to compute the basic reproduction number R 0 for 129 each room in a building. R 0 is used as a proxy to assess the outbreak risks of different infectious 130 diseases. Second, a web-based system is developed to enable information visualization and 131 communication in an interactive manner to provide guidance for occupants and facility 132 managers. This study innovatively establishes the computational links among building, occupant, 133 and pathogen modeling to predict outbreak risks. The risk prediction for spatially and 134 functionally distributed rooms in a building provides useful information for end-users to combat 135 and respond to the spread of infectious diseases, including the seasonal flu and COVID-19. The 136 developed method and system add a health dimension to transform the current building 137 management to a user-centric and bio-informed paradigm. 138 139 Fig. 1 In this study, a computational tool is developed based on Dynamo [29] to extract the geometry 232 and properties of each room in a building, and to compute the corresponding venue-specific 233 parameters. Fig. 4 shows the workflow of the information retrieval process. Lines in Fig The workflow for information retrieval is detailed as follows. 241 242 The steps for extracting room parameters are: Thereafter, the total furniture area in each room (Named ) is calculated by summing 282 up the surface area of all furniture inside the room. In epidemiology literature, R 0 is one of the most widely used indicators of transmission intensity 384 to demonstrate the outbreak potential of an infectious disease in a population. Commonly, R 0 > 385 1 means the epidemic begins to spread in the population, R 0 < 1 means the disease will 386 gradually disappear, and R 0 = 1 means the disease will stay alive and reach a balance in the 387 population. With the increase of R 0 , the outbreak risk will increase, and more severe control 388 measures and policies will be needed [37] . In this study, we categorize the level of outbreak risk 389 into low, mild, moderate, and severe based on the range of R 0 . Specifically, the risk is low when 390 R 0 < 1; the risk is mild when 1 ≤ R 0 < 1.5 because there is a fair chance that the transmission 391 will fade out as , is not much larger than 1 [38]; the risk is moderate when 1.5 ≤ R 0 < 2, 392 indicating an epidemic can occur and is likely to do so [39, 40] ; and the risk is severe when R 0 > 393 2 and immediate actions should be taken by facility managers, such as cleaning the surfaces, to 394 reduce the risk. 395 396 To better communicate the infection risk to occupants and facility managers, a web-based 398 system was developed to visualize the outbreak risk of different pathogens in each room within 399 a building. Fig. 5 illustrates the architecture of the web-based system, which consists of four 400 modules, i.e., data management, model derivative, web application, and user. Three add-in functions were developed to help users visualize the interior layout of the building 441 and color-coded rooms with their corresponding risk levels, as well as search specific room-442 related disease outbreak risk information. The first add-in function is "vertical explode", which is 443 used to view each level of the building. This function can help the user visualize the interior and 444 room layout. The facility users can also use this function to visualize the outbreak risk of rooms 445 on each floor and take appropriate practices. For facility managers, the "vertical explode" 446 function enables them to obtain a holistic view of risk distribution at each level and take 447 informed actions, such as limiting the number of occupants and implementing cleaning and 448 disinfection protocols, to control the spread of the disease. This function is integrated with the 449 web-based system, and clicking buttons were created to activate and deactivate it. The second 450 function is "room filtering", which is used to highlight rooms at different risk levels for a specific 451 pathogen. The user needs to first select one of the three pathogens from the dropdown menu: 452 SARS-CoV-2, Influenza, and norovirus. Thereafter, the user can set a risk threshold to highlight 453 rooms with R 0 greater than a specific value. In addition, different highlighting colors are used to 454 represent different infection risk levels. Low, mild, moderate, and severe risks are represented 455 by color green, blue, celery, and red, respectively. The third function is "room query", which 456 enables the user to search for a specific room and retrieve infection risk for the three pathogens. 457 The "room query" function is displayed as a search box on the web-based system. The users 458 can easily find the potential risk of a specific room using this function. Finally, end users can 459 access the web-based information communication system and obtain information about 460 outbreak risk in each room of the building through various channels, including laptops, 461 smartphones, and tablets. 462 463 A hypothetical case study is used as an example to demonstrate the efficacy of the proposed 465 framework and the newly developed web-based system. The building information model of a 466 six-floor school building with 221,000 square feet is used. The building contains classrooms and 467 faculty and graduate assistant offices. 468 469 The room types considered in the case study include offices and classrooms. Five offices and 471 five classrooms were selected. The venue-specific parameters of the rooms are extracted and 472 listed in Table 3 , and the computed R 0 values of the three diseases are listed in Table 4. 473 474 Table 3 Venue-specific parameters in representative rooms 475 From Table 4 , the values of R 0 vary across different rooms and different diseases. R 0 values in 482 offices are smaller than the values in classrooms, which stems from the small occupancy and 483 the low rate of fomite touching in offices compared to those in classrooms. For influenza, the R 0 484 values in all the rooms are less than 1, indicating that influenza is unlikely to outbreak in the 485 building through the fomite-mediated transmission. This could be partially explained by the 486 relatively short infectious period, high inactivation rate in hands, low hand-to-fomite pathogen 487 transmission efficiency, and relatively low infectiousness with the same amount of pathogens. 488 For COVID-19, the R 0 values in all rooms are higher than those of influenza, and the risk in 489 Classroom 4 reaches a moderate level, indicating that COVID-19 has the potential to outbreak 490 in the classroom. COVID-19 has a relatively high outbreak risk in most cases because it has a 491 high shedding rate, small surface inactivation rate, and high transfer efficiency from fomites to 492 hands. For norovirus, the R 0 values are high in most classrooms, which might be because of its 493 high infectivity, long infection period, and high hand-to-fomite transmission efficiency compared 494 to the other two diseases. This finding also aligns with the trend obtained in [24]. The above 495 results prove that the outbreak risk of an infectious disease is influenced by both venue-specific 496 and pathogen-specific parameters, which highlights the significance of integrating BIM and the 497 pathogen transmission model in assessing spatial-varying disease outbreak risk. 498 499 Sensitivity analysis was further conducted to evaluate the influence of the rate of fomite 500 touching (+ e ) and the shedding rate (%) of SARS-COV-2 on R 0 based on the estimated ranges 501 of the two parameters (listed in Table 2 ). Fig. 6 illustrates the changes in R 0 with the increase of 502 + e for all three diseases in both classrooms and offices. From Fig. 6 , the disease outbreak risk 503 increases as the increase of + e . The values of R 0 for norovirus and COVID-19 in Classroom 1, 2, 504 and 4 may exceed 1 with the increase of + e . On the other hand, the infection risk in offices and 505 that for influenza in classrooms will remain low even occupants touch objects in the rooms more 506 frequently. Therefore, it is particularly important to educate students in classrooms with 507 relatively high occupancy to not touch the common areas frequently. Fig. 7 illustrates the 508 changes in R 0 of COVID-19 with varying shedding rates. From the figure, % has a significant 509 impact on the outbreak risk of COVID-19 in Classroom 1, 2, and 4. Therefore, for classrooms 510 with relatively large occupancy, control strategies should be taken to reduce pathogen shedding 511 from the occupants, such as using face masks, and covering the mouth when coughing. applied to different rooms to reduce the risks to an acceptable low level. Cleaning the surface 532 five times per day will decrease R 0 by over 50%, compared to no surface cleaning. Considering 533 the ongoing outbreak of COVID-19, classrooms with high occupancy (e.g., Classroom 4) should 534 be given particular attention on surface cleaning. Cleaning surfaces at least two times per day is 535 needed to achieve a low risk level. For norovirus, classrooms with relatively large occupancy 536 (e.g., Classroom 1, 2, and 4) will require more frequent surface cleaning to reduce the outbreak 537 risk to the low level. Other complementary strategies, such as increasing hand washing and 538 limiting occupancy, should be adopted to maintain a low level of outbreak risks. 539 540 As shown in Fig. 10 , room filtering and room query functions can help the user easily locate 556 rooms with high risk and query risk information for a specific room. Specifically, Fig. 10 (a) 557 shows an exemplary output of the room filtering function that highlights the rooms with R 0 value 558 greater than 1 for COVID-19. Fig. 10 (b) displays an example of the room query function in the 559 web system. The pathogen risk information for influenza, norovirus, and COVID-19 is retrieved 560 with corresponding recommendations. With the web-based information communication system, 561 facility managers can take important measures to control the spread of diseases, such as 562 designing appropriate cleaning and disinfection strategies, promoting hand hygiene, reducing 563 maximum occupancy, and accommodating facility usage schedule based on risk distribution 564 across rooms within the building. For instance, deep cleaning and disinfection are required for 565 rooms with severe outbreak risk. In addition, facility managers can post signs at these high-risk 566 areas to remind occupants to take essential practices such as social distancing and hand 567 hygiene. The web-based system will also keep facility users, including teachers, students, and 568 other staff, aware of up-to-date outbreak risk information within the building, and thus taking 569 informed actions to avoid further spread of diseases. For example, facility users can avoid 570 entering rooms with high outbreak risk. 571 572 4. Discussion 573 The results and insights derived from the analysis have important implications on adaptive built 574 environment management to prevent infectious disease outbreak and respond to on-going 575 pandemic. Due to varying building characteristics, occupancy levels, and pathogen parameters, 576 the microbial burdens and outbreak risks differ significantly even in the same building, 577 highlighting the need for spatially-adaptive management of the built environment. The proposed 578 method automates the batch process for simulation and prediction of outbreak risks for different 579 pathogens at the room level, and visualizes the risks for adaptive management. The results on 580 outbreak risks at room level enables the paradigm for spatially-adaptive management of the 581 built environment. With the new streams of risk information, customizable interventions can be 582 designed. For instance, in consistent with the practice during the COVID-19 pandemic, reducing 583 the accessible surfaces in rooms and restricting the occupancy in the room are some of the 584 effective strategies to reduce the outbreak risks. The spatially-varying risk information can also 585 guide the facility managers to pay close attention to high-risk areas by adopting more frequent 586 disinfection practices. 587 588 A BIM-based information system is developed to extract the necessary information for modeling 589 infection within buildings, and to visualize the derived information in an easy-to-understand and 590 convenient way through web pages. As such, the information-driven interventions could 591 alleviate the pathogenic burdens in the buildings to prevent the spread of infectious diseases. 592 Providing information to end-users is critically important for them to change behaviors. Human 593 behavior plays an important role in the transmission of pathogens such as the SARS-Cov-2. 594 Changing behaviors is critical to preventing transmission. Providing timely and contextual 595 information can be a promising option to motive the change of human behaviors. With the room-596 level outbreak risk information, the users could be motivated or persuaded by the visualized 597 risks to practice appropriate behaviors such as wearing a mask, social distancing, and hand-598 washing. The facility managers can use the information to conduct knowledge-based 599 management, such as limiting the occupancy in the room, managing crowd traffic, and 600 rearranging room layout. 601 602 This study has some limitations that deserve future research. First, the model does not consider 603 factors such as sunlight exposure, humidity, and airflow that may impact the persistence and 604 transmission of pathogens in built environments. This is mainly because the quantitative 605 impacts of these factors on pathogen persistence and transmission are largely ambiguous, if not 606 unknown. If these impacts can be quantified and the environmental parameters can be 607 monitored and modeled in BIM, our proposed framework can be extended to incorporate these 608 factors. Second, the computation of R 0 only considers the fomite-mediated transmission, and 609 does not consider the airborne and close contact transmission. Microbial pathogens may have 610 different transmission routes, including airborne, close-contact, and fomite-based transmission. 611 This study focused on fomite-based transmission to illustrate the modeling approach for 612 assessing the outbreak risks, and demonstrate the efficacy of the developed information system 613 to guide infection control practices and building operations. To fully assess the exposure risks 614 and outbreak potentials, all important routes need to be considered. In addition, the outbreak 615 potentials of a variety of pathogens can be considered together to develop an aggregate index, 616 which could be more intuitive for occupants and facility managers who are not public health 617 experts. Third, the system mainly relies on static models and does not make full use of dynamic 618 and real-time data regarding built environments and occupant behaviors such as presence and 619 interactions with objects. In future studies, the internet of things sensors can be installed in the 620 buildings and algorithms can be developed to retrieve dynamic data for integration with the 621 models for accurate and robust risk estimation. Fourth, the web-based system can be further 622 improved by connecting it with smart devices such as robots for automated cleaning and 623 disinfection and smartphones for precision notifications. 624 625 This study creates and tests a computational framework and tools to explore the connections 627 among built environment, occupant behavior, and pathogen transmission. Using BIM-based 628 simulations, building-occupant characteristics, such as occupancy and accessible surface, are 629 extracted as venue-specific parameters. The fomite-mediated transmission model is used to 630 predict the contamination risks in the built environment by calculating a room-by-room basic 631 reproductive number R 0 , based on which the level of infection risk at each room is characterized 632 into low, mild, moderate, and severe. A web-based system is then created to communicate the 633 infection risk and outbreak potential information within buildings to occupants and facility 634 managers. The case study demonstrated the efficacy of the proposed methods and developed 635 systems. Practically, the method and system can be used in a variety of built environments, 636 especially, schools, hospitals, and airports, where transmission of infectious pathogens is of 637 particular concern. The outbreak risks predicted at room resolutions can inform the facility 638 managers to determine room disinfection and cleaning frequency, schedule, and standard. In 639 addition, appropriate operational interventions including access control, occupancy limits, social 640 distancing, and room arrangement (e.g. reducing the number of tables and chairs) can be 641 designed based on the derived information. The occupants can access the useful information 642 via webpage to plan their visit and staying time in the facilities, and practice appropriate 643 personal hygiene and cleaning practice based on the information. Microbial Exchange via Fomites and Implications for Human Health How quickly viruses can contaminate buildings --from just a single 658 doorknob The occurrence of influenza A virus on household and day care 661 center fomites An interactive web-based dashboard to track COVID-19 in 663 real time Prolonged Infectivity of 666 SARS-CoV-2 in Fomites Exaggerated risk of transmission of COVID-19 by fomites Role of fomite contamination during an 671 outbreak of norovirus on houseboats Epidemiologic and molecular trends of "Norwalk-like viruses" 675 associated with outbreaks of gastroenteritis in the United States Microbiology of the built environment Model Analysis of Fomite 680 Mediated Influenza Transmission Informing 683 Optimal Environmental Influenza Interventions: How the Host, Agent, and Environment 684 Alter Dominant Routes of Transmission Dynamics and control of infections 687 transmitted from person to person through the environment Bacterial transfer to fingertips during 691 sequential surface contacts with and without gloves, Indoor Air. (2020) ina.12682 Evaluating a transfer gradient 695 assumption in a fomite-mediated microbial transmission model using an experimental and 696 Bayesian approach Physical factors that affect 699 microbial transfer during surface touch Architectural design drives the biogeography of indoor bacterial communities Architectural design influences the diversity and 708 structure of the built environment microbiome The diversity and distribution of fungi on 711 residential surfaces Microbiota of the indoor environment: 713 a meta-analysis Bacterial communities on classroom surfaces 716 vary with human contact What Have We Learned about the Microbiomes of Indoor Environments? BIM handbook: A guide to building 720 information modeling for owners, managers, designers, engineers and contractors A conceptual framework 723 for integrating building information modeling with augmented reality Fomite-mediated transmission as a sufficient pathway: A 727 comparative analysis across three viral pathogens 11 Medical and Health Sciences 1117 728 Public Health and Health Services Building Information Modeling (BIM) for existing 731 buildings -Literature review and future needs Determining the level of development for BIM implementation in 734 large-scale projects: A multi-case study Transmission of Influenza A in a Student Office Based on Realistic 737 Person-to-Person Contact and Surface Touch Behaviour Risk of fomite-mediated transmission of SARS-CoV-2 in child 741 daycares, schools, and offices: a modeling study Visual scripting environment for designers -Dynamo 747 Predicting infectious SARS-CoV-2 from diagnostic samples Deducing the Dose-response Relation for Coronaviruses from 750 COVID-19, SARS and MERS Meta-analysis Results Estimated Surface Decay of SARS-CoV-2 (virus that causes COVID-753 19) A study of the probable transmission routes of 756 MERS-CoV during the first hospital outbreak in the Republic of Korea, Indoor Air CoV-2: Clinical Presentation, Infectivity, and Immune Responses Pandemic potential of a strain of 769 influenza A (H1N1): Early findings, Science On the definition and the computation of 772 the basic reproduction ratio R0 in models for infectious diseases in heterogeneous 773 populations Unraveling R0: Considerations for public health 775 applications Assessing the pandemic potential of MERS-CoV Nuanced risk assessment for emerging infectious 780 diseases 782 Interventions to mitigate early spread of SARS-CoV-2 in Singapore: a modelling study Web development with MongoDB and NodeJs Hand hygiene and surface cleaning should be paired 789 for prevention of fomite transmission, Indoor Air