Innovative new website warns of weather's effects on health

A new website created by a Canadian doctor promises to warn patients by email of potentially dangerous interactions between their medical conditions and the weather, but the science behind the innovative idea may not be as well-established as advertised.

The site, MediClim.com, opened for business in early January. Patients who register to use its services receive email alerts tailored to specific medical conditions, such as asthma or migraines, based on the site’s twice-daily calculations of weather patterns across Canada. In just over two months, more than 5,000 users have signed up for the alerts.

Created by Toronto family physician John Bart and Environment Canada meteorologist Denis Bourque after years of study and original research, MediClim is a first in Canada. While existing provincial public-health projects do warn patients of weather conditions that could exacerbate their medical conditions, those projects have so far largely been limited to smog warnings disseminated to the general population via the media. A system like MediClim’s that allows personalized warnings -- if they can be shown to be based on statistically sound weather-to-health correlations -- is something of a holy grail in environmental public health. But that’s a big if and questions have been raised about the validity of the science behind MediClim’s claims.

DEVELOPING A THEORY
This year’s launch is the culmination of what has not been an easy process: MediClim was 24 years in the making. The idea arose when Dr Bart saw a new patient at his office named Denis Bourque, who is a meteorologist with Environment Canada. “We were just yakking,” recalls Dr Bart. “I told him, ‘You know, you’re a weatherman. I think that the weather affects my patients because I seem to get them in clumps.’” Mr Bourque called a few days later. “You’re right,” he told Dr Bart. “There is a ton of information on the subject.”

The two of them began reading the literature on biometeorology, the study of the effects weather has on health. Eventually, in 1996, based on European research, they designed their own mathematical algorithm to describe the properties of the weather -- the MediClim Index, they called it -- that they theorized would allow them to model the connection between certain weather conditions and specific medical conditions. Their research efforts, however, ran into a chicken-or-egg dilemma. “We tried to ask the federal government to take over the program,” says Dr Bart. “They weren’t interested and nobody else was either because it was statistically unproven.”

“We tried every which way to get money to do the research,” Dr Bart says. “We got some money but because I’m a GP we couldn’t get much.” Nearly 10 years after Dr Bart and Mr Bourque created the MediClim Index, they finally received some funding to put it to the test, through a federal government grant to study the connection between the environment and human health. Their work was supported by a team of researchers including Gordon McBean, a University of Western Ontario geography and political science professor who was the lead author of the 1990 and 1995 assessment reports issued by the Intergovernmental Panel on Climate Change, the organization that shared the 2007 Nobel Peace Prize with Al Gore.

Dr Bart, who has a family practice, attended research meetings and decided, with the help of another physician, what medical information the study should collect -- all in his spare time. “I’ve been carrying this idea, this understanding, for so long now that when the opportunity arose, I just couldn’t pass it up,” he explains. “I just made the time.”

The 18-month-long study tested the MediClim Index by measuring its correlation with emergency department admissions at five Ontario hospitals. The 14 weather conditions in the Index’s formula -- temperature, pressure, humidity, wind, etc. -- didn’t correlate with emergency admissions individually, but when the data was run through the Index and amalgamated, the results were encouraging: “We proved that MediClim showed statistical relationships with [emergency department] appearances at four out of the five hospitals,” says Dr Bart. The results were presented on a poster at the 2007 Canadian Meteorological and Oceanographic Society meeting but have not been published in a peer-reviewed journal.

From that research, the new MediClim.com website eventually emerged. Once again, though, funding problems arose. Pharmaceutical companies and other private-sector investors didn’t bite when Dr Bart pitched the idea to them. “And so Denis and I floated the money ourselves,” he says. “We’re tens of thousands of dollars out of pocket and, let me tell you, our wives are very understanding.”

The site opened at the beginning of January, becoming the first service of its kind available to the Canadian public. What makes it unique is that, as opposed to existing public health warning systems like the provinces’ smog alerts, MediClim issues personalized health warnings to patients with five medical conditions: migraine, arthritis, asthma, cardiovascular disease, and diabetes. A patient can go to www.mediclim.com and register anonymously by entering her medical condition, her email address and her postal code, and the website will automatically notify her by email when an air mass is approaching that is correlated with exacerbations of her condition.

The site, which also features a blog by Dr Bart and message boards for users, has ambitious plans to solicit advertisements and expand. It’s already also available in the US, the UK and Ireland, and versions are in the works for France and Spain. Another future addition will enable users to report the symptoms they experience at a given time and location, to give MediClim a giant store of data to analyze to see if any weather-health trends emerge for medical conditions they haven’t studied yet, such as fibromyalgia, allergies and Alzheimer’s.

The January launch was a success. Within just one week, without any marketing or even a public announcement, the site registered 1,300 visits and 500 people signed up to receive email alerts. “We were quite surprised,” says Dr Bart. “We ring true with patients.” By March, the number of registered users had increased ten-fold.

QUESTIONING THE EVIDENCE
MediClim has already achieved a fair amount of popularity among patients, but there remains some uncertainty about the strength of the research that the site is built upon. That doubt persists even for Dr Bart, to some extent, despite his confidence about the basic premise of the research. “Maybe our index isn’t perfect,” he hedges. “I would be really happy for someone to come along and improve it. All I want is interest. Once we get the interest everything else will follow, and the subject of biometeorology will attain its position in public health where it should be.”

Less confident is Dr Pierre Gosselin, a physician who studies the effects of weather and climate change on human health at the National Public Health Institute of Quebec and at Laval University. “I do think it is a valid hypothesis,” he prefaces his opinion of MediClim. “I like it, personally.” But to Dr Gosselin, who is also the director of the World Health Organization Collaborating Centre for Environmental and Occupational Health Impact Assessment and Surveillance, none of the research performed so far on weather-heath interactions -- much of it has come from Europe so far -- has been particularly convincing from a clinical standpoint. “The strength of the evidence is rather weak, I would say. So it’s not that science cannot be helpful, but the scientific background is relatively nonexistent or really weak in terms of being what normally is used as standards for clinical practice nowadays.” (Dr Gosselin was invited to participate in the federally funded MediClim study but declined due to time constraints.)

Most of the existing research on the subject does not meet the rigorous standards that epidemiologists demand in order to make clinical recommendations. The same goes for the MediClim research, which was not been published in a peer-reviewed journal and was only presented at a meteorology conference, says Dr Gosselin. “That is not place where lots of doctors attend at that kind of association, so you cannot expect criticism on your work on epidemiology at a meteorological meeting, so that is not strong support.”

He doesn’t doubt that some of the correlations might be true. “We will probably find some existing relationships [between health and weather] and will prove them, that is for sure,” says Dr Gosselin. “But exactly what? We don’t know yet.”

Dr Gosselin’s concerns are tempered by a belief that the site may be providing a valuable service for some patients. “That’s possible that people could be misled, but it’s the same in many other areas of medical practice,” he says. “I mean, about half of what we do is art. So let’s say I would classify this website on the art side of medicine rather than on the science side.”

FROM THEORY TO PRACTICE
A major unanswered question about MediClim is what the site’s users and their family physicians are supposed to do with the information that arrives in their email inboxes.

Dr Bart is aware of that conundrum. When the site opened in January, he asked a colleague, Dr Jon Gladstone, a Toronto migraine specialist, his opinion of it. “His email said, ‘Very nice, John. What do I do with the information when they come in and ask me about it?’ At which point I said, ‘I refer to you. I take your advice.’” Dr Gladstone was kidding around, Dr Bart hastens to explain, but there’s a chasm between the MediClim warning and what family doctors can do for patients worried about weather-health interactions.

“We hope to invite the medical community, when they are approached by their patients, to follow this,” says Dr Bart. “If you have migraines, what we’d like the physicians to say is, ‘This is one of the triggers and now be careful of all the other triggers, like don’t drink red wine, don’t book yourself a really heavy day,’ stuff like that, and maybe we can avoid people turning up in the emergency department.”

Dr Gosselin worries that patients who receive warnings from MediClim might expect something from her family doctor that the medical profession can’t really provide yet.

Part of the problem is that biometeorology is a relatively new and underdeveloped field of study. Dr Gosselin cites a recent example: his department is currently conducting research on what effects high temperatures have on prescription drugs. “The science behind it is far from being foolproof, and when you check a little more deeply you realize not much is known actually,” he says. “And you cannot make a strong recommendation to family doctors to reduce, modify or change the drug that is prescribed. Can we be more specific about who is at risk and who is not? But we cannot at this time for most prescription drugs...” He chuckles. “So nobody knows actually.”

“Very often we refer to the easy ‘Go see your doctor,’” he said about the problem in a separate interview with Canadian Medicine last year. “The poor doctor very often doesn’t know what to do exactly.”

PERSONALIZED MEDICINE
Despite his reservations, Dr Gosselin is enthusiastic about one aspect of MediClim. “The concept of a personalized warning system is a major improvement on the general public [warnings],” he says. In fact, the National Public Health Institute of Quebec (INSPQ), where he works, has a similar program of its own, funded by the federal and provincial health ministries. Using data from what’s known as the Air Quality Health Index, the INSPQ set up an automated telephone messaging service to notify employees with respiratory conditions of potentially harmful weather, with different messages delivered depending on the severity of the employees’ problems. The INSPQ is now in the process of making their system available to Quebec City residents this spring, and then potentially expanding it to other cities in the province.

The idea of a competing project doesn’t faze Dr Bart. “I am so keen to get this idea started,” he says. “I have no doubt that smarter guys than me will do a better job than me. I don’t care. I just want to get this idea out that we are not homeostatic, as we thought, without costs to us maintaining it.”

Dr Bart’s hope -- both by forging ahead with his warning system and by asking MediClim users to submit subjective weather-health data in the future -- is to provide the basis for future biometeorology researchers to improve on his work. “I just want to give them the raw material to go ahead and do it. I want to die happy one day and know that a certain amount of public health information is backed up by some research by guys who are coming out of medical school when I am in my dotage.”

The site has attracted interest from the lay media and from bloggers. It has also received a warm reception from some influential figures in the field of environmental health.

Dr Warren Bell, a former president of the Canadian Association of Physicians for the Environment (CAPE) who practises family medicine in Salmon Arm, BC, and sits on the CAPE board, was impressed by MediClim’s innovation. “I certainly know of no other service that supplies that info,” he says, though he expressed concerns about the site’s interest in accepting advertising.

“I think it’s an intriguing concept,” says Gideon Forman, the executive director of CAPE. He dismissed concerns about insufficiently solid research, saying that the site’s warnings may be a huge help to some patients and at worst they will be harmless. “First, I think it would be very useful to people who suffer from things like migraine, asthma, arthritis to have something like this -- a lot of practical value. The other thing is I think it will cause people to think more about the environmental determinants of health generally.”

Get Canadian Medicine news by email or in an RSS reader