Description of the IDTM

The IDTM consists of three identical low-pressure inflatable pods: one staff pod (low-risk zone) and two patient-care pods (high-risk zone). Each patient-care pod contains a patient cabin, divided into two identical single-patient rooms, resulting in a total of four single-patient rooms per IDTM (Figure 1). The IDTM can be deployed either in a longitudinal configuration, with the staff pod positioned centrally between the two patient-care pods, or in an L-shaped configuration, with the staff pod located at the junction of the two patient pods.

Figure 1.

Structural layout and key features of the Infectious Diseases Treatment Module (IDTM). The IDTM consists of three identical inflatable modules: one staff area (low-risk zone) positioned centrally and two patient-care modules (high-risk zone), each containing two individual 10 m² patient rooms. The IDTM can be deployed in a longitudinal configuration (left) or in an L-shaped configuration (center), while maintaining staff visibility and workflow through the central low-risk zone. Each patient room includes five openable windows and a mosquito-net-equipped door, and each module features a patient porch area. Right: external view of the patient cabins and example of transparent monitoring screens that allow continuous patient observation from the staff area.

Figure 1. Long description

Panel A: A diagram of the longitudinal configuration of the Infectious Diseases Treatment Module (IDTM). The layout includes three identical inflatable modules: one central staff area (low-risk zone) and two patient-care modules (high-risk zone) on either side. Each patient-care module contains two individual patient rooms. Panel B: A diagram of the L-shaped configuration of the IDTM. The layout includes one central staff area (low-risk zone) and two patient-care modules (high-risk zone) arranged in an L-shape. Each patient-care module contains two individual patient rooms. Panel C: An external view of the patient cabins. Each patient room includes five openable windows and a mosquito-net-equipped door. The patient cabins feature transparent monitoring screens that allow continuous patient observation from the staff area.

Each patient room has a floor area of 10 m² (with a volume of 25 m³) and includes a transparent screen positioned behind the patient’s head, allowing continuous visual monitoring from the staff area without requiring staff entry into the high-risk zone (Figure 1). Each patient-care pod also incorporates a dedicated porch area, enabling patients to sit outdoors while remaining within the high-risk perimeter.

Each patient room is equipped with five openable windows and one door, all fitted with mosquito nets. These openings can be operated independently to regulate natural ventilation while maintaining vector protection. Together, these openings represent the primary adjustable ventilation pathways within the IDTM.

Environmental characterization

A full-scale deployment of the IDTM was carried out during a training simulation event in Brindisi, Italy. Although the primary purpose of the exercise was operational training, the deployment provided an opportunity to characterize the spatial configuration and ventilation behavior of the IDTM under realistic field conditions. The structure was fully assembled in accordance with standard operating procedures.

Air velocity and CO₂ tracer gas decay measurements

Ventilation performance and computational fluid dynamics (CFD) validity were assessed using a combined approach based on experimental air-velocity measurements (primary data set for CFD validation) and CO₂ tracer-gas decay tests (room-level ventilation rate quantification). Experiments were conducted in the IDTM patient cabin (right-side pod) under controlled conditions. To establish a repeatable boundary condition, all openings were sealed except window 2 (Figure 2), where a fan generated a steady inflow. The inlet boundary condition was characterized by nine pointwise velocity measurements distributed on a 3 × 3 grid across the window opening; these values were imposed in the CFD model as a spatially resolved (non-uniform) inlet velocity profile. Indoor velocities were then measured at five predefined locations inside the cabin at 1.55 m height using a vane anemometer as detailed in the Supplementary material.

Figure 2.

Numerical model setup. (a) computational domain and boundary conditions, (b) geometry of patient room highlighting the location of openable door and windows modeled as porous media to simulate mosquito nets, (c) detail of the computational mesh, selected for the CFD analysis, (d) detail of the three adopted computational meshes at an extrapolation plane height of 1.5 m.

In parallel, ACH was estimated using a CO₂ release-and-decay method. CO₂ was released in a single-patient room until concentrations reached approximately 2,000 ppm above ambient, then the decay was recorded using portable CO₂ sensors positioned centrally at 1.2 m height with 1 s logging. ACH was derived from the exponential decay.^{Reference Cui, Cohen, Stabat and Marchio10} Three replicate tests were conducted per room and averaged. ACH was converted to volumetric flow (L/s) using the fixed room volume to support comparison with WHO ventilation benchmarks.

Engineering analysis of ventilation scenarios

To extend the field measurements and evaluate ventilation performance across a range of environmental and architectural conditions, a CFD analysis was conducted. The analysis focused on wind-driven natural ventilation within individual patient rooms and was used to estimate ventilation rates under varying wind speeds and structural configurations.

The computational domain included two adjacent patient-care pods arranged in an L-shape, the internal volume of one patient room per pod, the porch area when present, and the surrounding external environment (Figure 2a). Doors and windows were modeled as ventilation openings, with mosquito nets represented as porous media whose resistance was derived from the manufacturer’s technical specifications (Figure 2b).

Wind was imposed to impinge on the L-shaped assembly at a 45° angle relative to the main door openings of both patient rooms (Figure 2a). This wind orientation was selected to reflect typical deployment practice, where patient-care pods are arranged side-by-side, and to generate a symmetric airflow pattern across the two analyzed pods. This symmetry ensures that ventilation performance in the simulated room pair is representative of conditions experienced by individual rooms within longer, linear multi-module deployments.

Simulations were conducted for wind speeds ranging from 1 to 4 m/s, representing light to gentle breezes, common in many outdoor settings.^{Reference Chen, Birkelund, Anfinsen, Staupe-Delgado and Yuan11} Airflow was modeled using the numerical solution of the incompressible Reynolds-averaged Navier–Stokes equations with a k-ω SST turbulence model, which is appropriate for low-speed, wind-driven ventilation flows around and within enclosed structures.^{Reference Chipongo, Khiadani and Sookhak Lari12} Uniform inlet velocity conditions were applied at the upwind boundary, with an outflow condition at the downstream boundary. External domain boundaries were treated as symmetry planes, while the IDTM envelope was modeled as no-slip walls.

Model development included mesh sensitivity testing (Figure 2c and d) and validation against experimental velocity measurements obtained during the deployment in Brindisi. Detailed descriptions of the numerical formulation, turbulence modeling, porous-media parameters for mosquito nets, mesh refinement, convergence, and validation are provided in Supplementary Material.

Following validation, four structural configurations were analyzed for each patient room: with both a porch and mosquito nets in place; without a porch but with mosquito nets; with a porch but without mosquito nets; and without either a porch or mosquito nets. For each configuration and wind speed, steady-state airflow rates through the patient room were extracted and converted into ACH based on the known internal volume, enabling direct comparison of ventilation performance across scenarios.

Infection risk assessment

Infection risk, defined as the probability that a susceptible occupant becomes infected through inhalation of infectious particles, was estimated using the quanta-based approach previously described ^{Reference Buonanno, Stabile and Morawska13} and validated,^{Reference Buonanno, Morawska and Stabile14,Reference Miller, Nazaroff and Jimenez15} applying the Wells–Riley framework ^{Reference Riley, Murphy and Riley16} to a single IDTM patient room. Two exposure scenarios were modeled: (i) one infectious patient present for 60 minutes before a HCW entered the room without respiratory protection for a 15-minute exposure, and (ii) continuous 24-hour co-exposure between one infectious patient and a second susceptible patient, also without respiratory protection. Infection risk was assessed across ventilation rates from 2 to 210 L/s (representing 0.3–30 ACH) spanning near-stagnant conditions up to and beyond the WHO benchmarks (160 L/s; 12 ACH). Risk was estimated for seven pathogens with established quanta-generation rates: MERS-CoV, SARS-CoV-2 (Ancestral), SARS-CoV-2 (Omicron), SARS-CoV-1, influenza, tuberculosis (TB), and measles.^{Reference Mikszewski, Stabile, Buonanno and Morawska17} Although WHO defines no explicit infection-risk threshold, the recommended 12 ACH standard corresponds to infection probabilities ranging from <1% to approximately 8% depending on quanta-emission rates.^{Reference Atkinson, Chartier and Pessoa-Silva8}