Drug database can't take doctor's place
When North Vancouver resident Geoff Graham was prescribed the antidepressant Effexor to address possible depression from chronic back problems, he didn't stay on the drug for long. After suffering intermittent bouts of dizziness and agitation, the 45-year-old stopped taking the pills. Recently, out of curiosity, he logged on to Health Canada's Canadian Adverse Drug Reaction Monitoring Program Web site, or CADRMP ( www.hc-sc.gc.ca/dhp-mps/medeff/ad visories-avis/index_e.html), a database of suspected adverse reactions that has been available on-line since May, to see if other people had had similar experiences with Effexor.
Formerly available for a fee and only by request to the department's access-to- information office, the database was made free to the public to counter criticism that Health Canada wasn't transparent enough about adverse reactions. Some lauded the move as a step toward increased accountability, but critics say it's not enough and that for consumers it raises more questions than it answers.
What Graham found on the Web site confused and alarmed him: from 1995 to 2005, there were 282 adverse-reaction reports filed on Canadians taking Effexor, ranging from age 15 to 84. Of these, there were 18 deaths. One death, that of an 18-year-old, is listed as "due to adverse reaction". The 17 others are described as deaths in which the "drug may be contributory"; many of these people were in their 30s and 40s. Forty-eight of the 282 are listed as "not yet recovered", and for 90 people, the outcome was reported as "unknown".
"My instincts told me I was right to stay away from a drug even remotely associated with 18 deaths in 10 years," Graham told the Georgia Straight in an interview. Yet apart from stronger warnings on its packaging, beginning in 2004, of increased risk of self-harm or harm to others, Effexor remains on the market and is regularly prescribed for depression or menopausal symptoms.
According to Health Canada, CADRMP should not be used on its own to make decisions about a drug's safety. From her Ottawa office, Health Canada spokesperson Jirina Vlk said the adverse-reaction reports are meant to detect possible signs of troubling patterns and are just one weapon in the government's arsenal for assessing whether or not a drug poses a serious danger. "You can't use the data to make comparisons among different drugs," Vlk said, "because some of the drugs have been on the market longer than others, and some are more popular so there are higher numbers of users. Nor can you make a causal link between the drug and the adverse reaction," Vlk added, "because there are too many variables involved." The Health Canada Web site includes a whole section on the limitations of interpretation.
One of the biggest weaknesses, Vlk readily conceded, is the quality of the information. Although adverse-reaction reporting by drug manufacturers is compulsory under the Food and Drugs Act, health professionals and consumers are not mandated to report adverse events. (Vlk did point out, though, that Health Canada is looking into mandatory reporting for all parties, similar to what happens in France.) But for now, underreporting is a significant problem because of the time involved in filling out the 30-question form. In addition, those that are filed are sometimes incomplete. "No matter how comprehensive the questionnaire, you can't force people to fill in all the blanks," Vlk said.
So how does she recommend consumers like Graham use CADRMP? "First, they should read the product label, which lists the possible side effects," she said, "because what they're experiencing might be perfectly normal. What they shouldn't do is make a decision based on what they find on the database. What really needs to happen is a discussion between doctor and patient about risks and benefits, a point clearly stated on the Web site."
Alan Cassels, a drug-policy researcher at the University of Victoria and a board member of PharmaWatch, a Vancouver-based group advocating for consumer rights and safe medicines, believes the onus is on Health Canada to do a better job of collecting post-market data, particularly in light of the fact that clinical trials are routinely too short to be definitive. "The information we get about a drug when it's launched is very different from what happens after it's been on the market for a while and people excluded from the clinical trials start taking it," he told the Straight in a phone interview. He cited the Paxil example. Manufacturer GlaxoSmithKline Inc. reported on its label a two-percent incidence rate for what it called the "discontinuation effect" (a euphemism for serious withdrawal effects, Cassels said). After public pressure in Great Britain, the company rewrote the label, citing a 25-percent incidence rate, but it had already been on the market for five years.
The fact that Health Canada has made what data it does have available to the public is hardly reassuring, Cassels said. "It's better than nothing," he quipped, "but can you imagine a plane crash and Transport Canada not knowing how many people were killed and what went wrong? Yet when it comes to drug safety, these are exactly the things Health Canada doesn't know."
His suggestion: for at least two years after a drug is approved, it should be dispensed with a postcard questionnaire for consumers to return. "At least this would be an effective early-warning system," he said.
In the meantime, he agreed with Vlk that people who are looking at taking any drug long-term, including nonprescription and natural remedies, should make a point of finding out what the side effects are. "They need to ask their doctor which ones are the most common, the most serious, and how often they occur. They should also ask if the drug has been tested on people of similar age, gender, and health status," he said.
Having the information from the adverse-reaction database with them wouldn't hurt either, according to Cassels. "Flawed or not," he said, "at least it would be a point of discussion between doctor and patient."