In February 2003, a 78-year-old Canadian woman returned home after a 10-day trip to China. In the days that followed, she spiked a fever, then developed muscle aches and a persistent cough. She deteriorated rapidly, and in early March she died at her Toronto apartment.

Doctors didn’t know it yet, but she had severe acute respiratory syndrome, or SARS. Her son soon developed symptoms and sought care at a hospital. While in the emergency room, he transmitted the illness to two patients in nearby beds, who then passed it to more than two dozen others, including family members, hospital staffers and other patients. Those people then infected an additional 34 people, who sickened 17 more in an ever-growing chain of disease.

One irony of an infectious-disease outbreak is that it turns hospitals, which should be a refuge for the sick, into hot zones, putting everyone inside at risk. But a silent culprit tends to be overlooked. It is not the patients or the staff or the visitors. It is the building itself. Thoughtful modifications to the built environment — to how health-care facilities are designed, operated and maintained — could help curb the spread of infectious disease, reducing the toll of future outbreaks as well as the covid-19 pandemic raging today.

Seventeen years after the Toronto woman’s son set foot in that Canadian hospital, we’re watching history repeat itself with covid-19, which, like SARS, is caused by a coronavirus. When doctors in Wuhan, China, where the new virus first emerged, studied 138 early cases, they concluded that 41 percent of patients had most likely contracted the disease in the hospital.

Several urgent steps must be taken to keep hospital occupants safe, including significantly expanding diagnostic testing and ensuring that health-care workers have access to personal protective equipment such as gloves and masks. But hospitals should also consider some environmental tweaks. One simple change would be to rethink where we provide hand sanitizer. Many U.S. hospitals post dispensers in the hallway or at the door to patient rooms, but several studies suggest that making these dispensers even more conspicuous and accessible, and placing them closer to the point of care, could improve staff hand hygiene.

“Ninety-nine percent of alcohol hand rub is way at the door entry or outside the door,” says Eli Perencevich, an infectious-disease doctor and epidemiologist at the University of Iowa, who recommends that hospitals reconsider this practice. If doctors, nurses and technicians have to leave a patient’s bedside every time they touch something, they’re far less likely to comply with standard hand hygiene protocols, he notes. Putting hand sanitizer by the bed sounds obvious, but alcohol-based hand rubs are flammable, and fire safety codes have prompted many U.S. hospitals to keep dispensers away from patients, Perencevich says. However, sanitizer-fueled-fires are exceedingly rare, and the benefits of relocating the sanitizer are likely to far outweigh the risks, especially during a pandemic.

Adding lids to toilet seats is another easy fix, Perencevich says. Flushing an uncovered toilet can spray aerosolized droplets of water and waste around the room; although covid-19 appears to spread mainly via respiratory droplets, scientists have also found the virus in toilet bowls and stool samples.

Some hospitals may be able to bump up their ventilation rates or bring more fresh air in from outside, diluting the concentration of the virus in indoor air. Since opening hospital windows is rarely an option, many hospitals could do that with their current air-handling infrastructure, simply by opening the damper to outside air, says Kevin Van Den Wymelenberg, who directs the Institute for Health in the Built Environment at the University of Oregon and is a co-author of a forthcoming paper on covid-19 and the built environment.

Van Den Wymelenberg and his colleagues suggest that hospitals consider humidifying their typically dry air. There’s some evidence that higher air humidity can reduce the viability and airborne transmission of certain kinds of viruses, including coronaviruses. Though more research is needed, aiming for a sweet spot of between 40 and 60 percent humidity might help curb covid-19 without fostering mold growth.

Then there’s sunlight, which can inactivate certain kinds of microbes (and boost patient mood). Though its effects on this specific virus are unknown, keeping the blinds open couldn’t hurt. Neither could reducing clutter; tucking patient belongings and infrequently used supplies into drawers and cabinets will give virus-containing droplets fewer surfaces on which to settle and make it easier to thoroughly clean patient rooms.

In the longer term, thinking carefully about the floor plans of new hospitals could help us contain future disease outbreaks. We should build hospitals that give all patients private rooms, which can dramatically reduce the risk of hospital-acquired infections. Of course, during major pandemics, overwhelmed hospitals may have no choice but to put patients wherever there’s space, but for smaller outbreaks, as well as the everyday treatment of infectious disease, single-patient rooms have huge payoffs. The less that patients have to share space with others, the safer everyone will be.

We could take this idea even further by emulating Sweden’s Skane University Hospital, which opened an infectious-disease building in 2010. In the emergency department and the outpatient clinic, doors lead directly from the outdoors to several private isolation rooms so patients suspected of being contagious can bypass the communal waiting areas. The inpatient rooms, on the upper floors, have doors that open onto balconies that wrap around the circular building. When patients are admitted, they can be taken to their rooms via these outdoor walkways. Another set of doors connects each patient room to an indoor corridor used by hospital staff.

Though hospitals in the midst of the current outbreak can’t change their layouts now, they can rethink patient flows, designating separate entrances and waiting areas for suspected covid-19 patients, says Torsten Holmdahl, a doctor in the Swedish hospital’s infectious-disease department. As of Friday, the hospital had admitted six covid-19 patients, Holmdahl said, and he was feeling good about how the facility was performing so far. When he and his colleagues began planning the building 15 years ago, the prospect of a future pandemic was exactly what they had in mind. “At this hospital, we are very prepared,” he said. “We feel confident that we have done our very best.”

Still, like others around the world, Holmdahl was bracing for more patients, and if the case numbers surge too fast, they could push even the most thoughtfully designed hospital to its limits. Changes to the built environment won’t cure covid-19, but the right design decisions, today and tomorrow, absolutely could save lives.