High above the heads of staff and residents at Boston’s Pine Street Inn, a well-known homeless shelter in the city’s South End, ultraviolet light fixtures hum and glow a soft blue as they have for more than 15 years. Back in the early 1980s, Dr. Edward Nardell, then an influential figure on the American Lung Association’s Board of Directors, encouraged the organization to fund installing the fixtures as a preventive measure to stem the rising tide of tuberculosis. The School’s Professor Melvin First, also on the board back then, objected because UV’s effectiveness against the disease had never been proven. Nardell recalls, "I raised the idea, and of course Mel got up and said, ‘Won’t work. No data.’" But facing an urgent TB situation, the board ended up voting with Nardell, and the UV lights became a permanent feature of the Pine Street Inn. And while First persuaded Nardell to take advantage of the situation and do an epidemiologic study, it quickly became clear that such an experiment wasn’t feasible, with only one shelter and too many variables. "So it was never studied," says First, who adds with a laugh: "But on the other hand, Nardell’s the first to point out that they’ve never had another case of tuberculosis among the staff since they put the lights in."

Today Nardell, associate professor of medicine at Harvard Medical School and the Massachusetts State TB control officer, and First, professor emeritus of environmental health engineering, are on the same side. They are working on a large-scale, multidisciplinary project designed to prove once and for all whether or not upper-roomultraviolet radiation prevents the spread of tuberculosis. At the heart of this $6 million undertaking is a series of double-blinded, controlled experiments comparing the incidence of tuberculosis infection in U.S. homeless shelters equipped with active and inactive ultraviolet light fixtures. The project also has an engineering component that is testing how various kinds of UV fixtures work and how their killing power is influenced by factors like room ventilation and humidity.

Why UV?

Scientists have known for a long time that ultraviolet light kills microorganisms; in tuberculosis, UV’s deadly effects are achieved by disrupting the bacterium’s dna. Much of the early work on the practical use of ultraviolet light was initiated by the School’s Professor William Wells in the 1930s and ’40s. From studies of Massachusetts textile mills, Wells concluded that many respiratory infections are transmitted by organisms that are so small they remain suspended in the air in the form of "droplet nuclei." Tuberculosis is the classic example of this kind of airborne pathogen. Wells began to ponder how such floating organisms might be controlled. "And the most effective way he could think of, or that anybody’s thought of, really, was to irradiate the air with ultraviolet radiation," says Richard Riley, who worked as a researcher under Wells while attending Harvard Medical School in the 1930s .

Riley would eventually become the chair of environmental medicine at Johns Hopkins and with Wells, whom he enticed down to Baltimore, would perform some of the most famous experiments in airborne infection and air disinfection. Using colonies of guinea pigs connected by ventilation ducts to a TB ward in the Baltimore VA Hospital, Riley and Wells carried out a landmark study that proved TB was airborne. Part of this four-year experiment included the use of ultraviolet in one of the two ducts linked to the ward so that the guinea pigs supplied with air from this duct would serve as controls for those receiving unirradiated air. The result? Not a single guinea pig connected to the UV-associated duct contracted TB while the others continued to get infected. Unfortunately, plans for studying upper-room ultraviolet were abandoned when both Wells and his staunch supporter, Dr. John Barnwell, director of research and education for the VA, died: "Without our fairy godfather in Washington and without Wells, the money dried up," sighs Riley. Add in the declining number of TB cases in the U.S. and the rising belief that antibiotics had tuberculosis licked, and the inclination to study ultraviolet disinfection began to fade.

But in the early ’80s, tuberculosis would stage a comeback; in 1985 the first upswing in cases was recorded in almost a century. Soon people began to look for a quick fix and rediscovered Wells and Riley’s original research. And while UV lights were installed in hospitals and prisons all over the country, they have never been fully tested to show that they actually work, how much TB is prevented, and at what cost. "The people who put up and designed these installations were sort of doing it on the back of a brown paper bag," says Jonathan Freeman, assistant professor in the School’s epidemiology department and one of the researchers.

Interestingly, the original idea for this new UV study came from neither Nardell nor First but from Dr. Philip Brickner, chairman of the Department of Community Medicine at St. Vincent’s Hospital and Medical Center of New York. Known for his work with the homeless, Brickner was intrigued with the idea of using UV to fight TB, a notion he had first encountered as a medical intern. Brickner approached both Nardell and Freeman with his concept: the first real-world trial of the efficacy of upper-room ultraviolet germicidal irradiation against TB. Nardell was skeptical: "I told him it couldn’t be done. I know, here I am taking Mel’s role now, but I said there are just too many variables." Freeman, however, was more optimistic and created a blueprint for a study in a number of homeless shelters across the country. Nardell was won over.

The epidemiologic arm of the project involves installing ultraviolet light fixtures at near-ceiling height in homeless shelters in nine U.S. cities/areas: New York; Birmingham, Ala.; New Orleans; Houston; four smaller cities in south Texas; and Los Angeles. As of today, the first two cities are up and running, the third is in installation, and the last six are in the fund-raising stage. Half of these shelters will have UV bulbs, while the other half will have non-UV placebo bulbs. Which shelters have which type of bulbs will change yearly, and even the epidemiologists and the engineers will have no idea where the ultraviolet is active. The investigators will ultimately examine each shelter for the number of "conversions"--subjects who test positive on a tuberculin skin test but who may not yet have developed the disease. If fewer conversions occur when the UV lights are active than during the placebo period, it will prove that the ultraviolet is effective on some level against the spread of the disease.

So why homeless shelters? TB transmission in this country takes place primarily in prisons, inner-city hospitals, and homeless shelters. Although experiments were done on prison volunteers in the 1950s and ’60s, it was later deemed unethical to study inmates. The problem with studying TB in hospitals is that most people don’t stay long enough and better infection control practices of all kinds have greatly reduced transmission. Homeless shelters, on the other hand, are good places to study TB transmission; because they have become, in effect, an amalgam of educational, housing, job-search, and detox programs, residents will often stay for six months to a year.

Homeless shelters also have the advantage of providing their own health care, including the physicals required to enter one of these programs. "So what we’re doing is piggybacking this study on a health care system that already exists," says Freeman. When people join a shelter program, for which they are going to need a TB test anyway, the epidemiologists explain the purpose of the study and ask them if they would like to take part; so far, 95 percent of those asked have volunteered.

Simply comparing the number of conversions in the UV-lighted shelters with those in the placebo-lighted shelters "is fine as far as the epidemiologist is concerned," notes First. "But from our standpoint--and we’re the engineering component--what we want to know is, how much ultraviolet irradiance do you need to get an effect, and is the effect directly related to how much UV you put into the spaces plus the integration of ventilation."

First’s engineering work has two parts. For one, he is measuring the irradiance--the measure of radiant energy over a certain area--of individual UV fixtures. Using specialized UV detectors, he has already found that these fixtures can vary widely and are often way out of sync with the manufacturers’ specifications. The second and more complicated issue is to figure out the relationship between ventilation and irradiance--and how engineers can best combine both factors to kill airborne bacteria without causing other health problems. Although the kind of ultraviolet light being used here is not the same as the UV in sunlight that causes skin cancer, even this UV, with its shorter wavelength, can cause superficial skin and eye irritations. "As far as killing the organism, we could flood the whole room--that would be better," says First. But not for the person trying to live or work there. For this reason, the lamps must be placed near the ceiling. But if there are no air currents to bring the expectorated bacteria up into this "killing zone," nothing will happen.

To make sense of all these particulars, First has had a small structure containing a room-sized exposure chamber built on top of the School’s Building Two. The chamber, which First’s colleagues have jokingly dubbed "Building Five," is equipped with a typical UV fixture. He and his fellow engineers, Stephen Rudnick, S.D.’78, a lecturer in the Department of Environmental Health, and Thomas Dumyahn, SM’94, an engineer for the department, will conduct experiments by aerosolizing bacillus Calmette-Gúerin (BCG), a tuberculosis vaccine, which is noninfective but otherwise a good surrogate for the tiny floating mycobacterium that causes TB. The aerosol will be propelled into the room through a valve that looks like an ordinary shower head but that can be made to "breathe" or even "sneeze" on demand. Factors like UV intensity, ventilation, temperature, humidity, and amount of the organism can be carefully manipulated and measured from a separate glassed-in "control room."

According to Brickner and Freeman, it took them almost ten years actually to embark on their UV project after the seed of the idea was planted. "Meanwhile," notes Freeman ruefully, "everyone’s jumped on the bandwagon, and they’re now doing a four-plus job on running all of these [tuberculosis] prevention programs that they used to run and abandoned. So we’re actually doing this in an era where TB’s declining again." According to the Centers for Disease Control and Prevention, U.S. tuberculosis rates have been falling since 1992 (19,855 new cases were reported in 1997, a 26 percent decline from five years prior). This decrease could have enormous implications for the outcome of these UV experiments: it is, after all, hard to study a prevention measure against something that isn’t there.

But while TB rates in this country have been falling, in developing countries tuberculosis is spiraling out of control, and resources for treatment are limited. UV could offer a very cost-effective intervention. The lights need only be bought once and, in the long run, prevention is much less expensive than treatment. Additionally, the proliferation of multidrug-resistant TB is further hampering therapeutic options, both here and abroad. The beauty of ultraviolet is that it doesn’t discriminate between regular and drug-resistant TB, killing both with equal abandon. Nardell wants to recreate Wells and Riley’s landmark guinea pig experiment on a TB ward in South Africa, a country where tuberculosis is running rampant, but so far his plans have been hampered by ongoing funding and logistical problems.

So will success elude this band of ultraviolet researchers? For now, they remain guardedly optimistic. Nardell proudly notes, "This is a very nice multidisciplinary effort involving a number of institutions and a number of disciplines with a lot of history behind it." But he cautions that the use of ultraviolet is still "a very controversial area." Nardell also wonders whether the study needs to factor in not only how many people get infected, but also the likelihood that an infectious case will walk into a shelter. "That may be the Achilles heel of our study to this day: that this is such a powerful variable and it’s very difficult to control." Riley mulls over whether the same homeless people really do stay in the same shelters long enough to see results. "We had control of things," he says of his original UV light experiments. "When you come to the real world, you don’t have that kind of control, and this is what makes it so very, very difficult to reach good conclusions."

As for First, he is still the same man who openly aired his concerns as a member of the American Lung Association Board more than 15 years ago. "As you know, I’m a skeptic, so I don’t know whether this is going to work or not," he grins. "But we’re going to find out."

-- Alexandra Benis