June 2001
The Riddle of Ritalin® Prescribing in British Columbia Resolved (Almost)
by John R. Graham
Ritalin® (methylphenidate) is a drug which is never far out of the public eye. Prescribed to treat Attention Deficit Hyperactivity Disorder (ADHD), it has generated opposition from those who believe it is abused. A few weeks ago, an American court dismissed a class action lawsuit against Novartis, the drug's manufacturer, and the American Psychiatric Association, which had claimed that those two organizations had conspired to broaden the definition of ADHD, thereby increasing sales of the drug.
Three years ago, British Columbia's provincial health officer noted significant differences in methylphenidate prescribing across the province's 20 health regions. "Regional variations in methylphenidate prescribing indicate that treatments for ADHD are not being uniformly provided. Children in the Thompson, North Okanagan, and Fraser Valley regions are up to three times as likely to receive a methylphenidate prescription as are children in Vancouver" (Millar, p. 95). While not explaining why this was occurring, Dr. Millar recommended: "diagnostic and treatment guidelines... are needed to ensure that children are receiving the best possible care for emotional and behavioral problems, including ADHD" (Millar, p. 95).
The Vancouver Province newspaper independently discovered similar regional variances (1999a). The Province also noted that boys were significantly more likely to be prescribed than girls (1999b). A trio of provincial health ministry officials charged that the newspaper's math was bad because it used an inappropriate population estimate as the denominator in its measurement of use of methylphenidate per capita (Dormuth, Anderson, and Warren). The newspaper strongly defended its case (Rees, 2000a).
The College of Physicians and Surgeons of BC, which licenses the province's physicians, sent a questionnaire to physicians who prescribed methylphenidate during the six months from May 1, 1998. The College determined that the drug was being responsibly prescribed in BC (Taylor, 2000). However, Ann Rees of the Vancouver Province reported further analyses of wide regional disparity in methylphenidate prescribing, including many more prescriptions written for boys, and expressed dissatisfaction with the College's finding (Rees, 2000b; 2000c). The College's Registrar questioned Rees' analysis (VanAndel, 2000).
Statistical analysis of methylphenidate prescribing versus certain population and socio-economic indicators indicates that two issues that Rees addressed, prescriptions for boys and regional differences, are related. The proportion of boys in the population of children explains a lot of the regional variance. For readers with a technical bent, table 1 shows a regression model that explains almost half of the variance in prescribing across BC's 20 health regions. The indicators are: proportion of boys in the population of 5- to 14-year-olds (1996), proportion of 5-to 14-year-olds in the entire population (1996), per capita income (1995), and population density (population per square kilometre, 1996). The variable to be explained is the proportion of people, aged 0 to 19, prescribed methylphenidate between 1990 and 1996. (This is not a cumulative figure. Each child was only counted once during the period, even if he received a prescription in more than one year.) The model's inputs use a narrower age range than the output does because there is far more methylphenidate prescribing for 5- to 14-year-olds than for toddlers or older teenagers (Millar 1998: 95).
The four independent variables explain about half the story. Most significantly, the proportion of boys in the children's population in a health region is very important. A 1 percent relative change in the proportion of boys (e.g. from 51.0 percent to 51.5 percent) explains a 19 percent relative change in the prescription of methylphenidate to children in the region (e.g. from 1.5 percent to 1.8 percent). This would not be apparent to the naked eye, because the proportion of boys in the children's population varies only from 50.5 percent to 51.8 percent across the 20 health regions. The regression shows that those who worry about the number of boys on methylphenidate and regional differences in prescribing are, to some degree, worrying about the same thing.
The other variables are harder to interpret. A 1 percent relative increase in children as a proportion of the overall population explains a 1.3 percent decrease in prescriptions. Could the occurrence of fewer children in a region be related to more scholastic or medical attention on each child, thereby increasing the chance of diagnosis and prescription? A 1 percent relative increase in per capita income explains a 1 percent relative decrease in prescriptions. Perhaps those with higher incomes are more likely to use more expensive therapy, such as counselling, rather than drugs for their ADHD children?
An absolute increase in population density explains a tiny, but significant, relative decrease in prescriptions. This apparently contradicts Rees' claim that urban kids are more likely to see a doctor than rural kids (2000a). However, this coefficient may be confounded because health regions can have pockets of high and low density. Analyses of subsets of the data (not reported) indicate that the direction of the relationship may change from high through medium to low density areas, but data for smaller areas than health regions are required to make a definitive statement.
So, the regional variance of methylphenidate prescribing in BC is almost explained. I anticipate that if doctors increase their prescribing of methyl- phenidate to girls, a significant amount of the variance will disappear.
|
Table 1: Regression Explaining the Proportion of Children ln(R) = 15.6081 + 19.3053*ln(M) – 1.3153*ln(P) – 0.9630*ln(I) –0.0002*D where M = proportion of children aged 0 to 19 prescribed methylphenidate in 1990-1996 (each child counted only once), B = proportion of boys in population of 5- to 14-year-olds in 1996, P = proportion of 5- to 14-year-olds in entire population in 1996, I = per capita income in 1995, D = population per square kilometre in 1996. Method: ordinary least squares, R-squared = 0.55, R-squared adjusted = 0.43, all coefficients' t-statistic and equation's F-statistic significant at 5 percent.
Sources: BC Stats (various); Millar; author's calculations. |
References
Dormuth, Colin, John F. Anderson, and Leanne Warren (2000). "Caveat Lector: Be Wary of Media Reports about Excessive Ritalin Use in BC." Canadian Medicine Association Journal 162, 3 (February 8): 313.
Millar, John S. (1998) A Report on the Health of British Columbians: Provincial Health Officer's Annual Report, 1997. Victoria, BC: Ministry of Health and Ministry Responsible for Seniors.
Rees, Ann (2000a). "Ritalin Use in BC." Canadian Medical Association Journal 162, 6 (March 21): 753-754.
Rees, Ann (2000b). "Hot Spots for Ritalin: In Parts of BC, Almost 4 Percent of Kids Aged 10 to 14 Are on the Drug." Vancouver Province (December 21): A6.
Rees, Ann (2000c). College Study Minimizes Numbers: After Looking at Just One Drug, It Said Over-prescribing Is not a Problem." Vancouver Province (December 21): A7.
Taylor, B.T.B. (2000). A Review of Ritalin Prescribing. Vancouver, BC: College of Physicians and Surgeons of British Columbia (August 30).
VanAndel, M. (2000). "Letter to the Editor." Vancouver Province (December 29): A35.
Vancouver Province (1999a). "When it Comes to Kids Taking Ritalin, BC's Breaking Records." (August 9): A15.
Vancouver Province (1999b). "Boys will be Boys." (August 9): A14.
info@fraserinstitute.ca
4th Floor, 1770 Burrard Street, Vancouver, BC, Canada, V6J 3G7
Tel: (604) 688-0221 Fax: (604) 688-8539 Book Orders: 1-800-665-3558 ext. 580
You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.