Although Kearney and colleagues1
estimated that the prevalence of hypertension would increase by 24% between 2000 and 2025 in developed countries, we found that the age-and sex-adjusted prevalence of hypertension among adults aged 20 years and older in Ontario increased by 60.0% from 1995 to 2005, and by 20.9% from 2000 to 2005. The increase in the most recent 5-year period has almost reached what Kearney and colleagues predicted would occur over 25 years. We also found a 25.7% increase in the age-and sex-adjusted incidence of hypertension from 1997 to 2004; however, this increase in incidence alone did not appear to be sufficient to explain the overall increase in prevalence. Thus, in a separate article in this issue of CMAJ
we examine changes in mortality among patients with hypertension over the same period.
Overall, we found that, similar to estimates of the prevalence of diabetes, the prevalence of hypertension has been underestimated.2
Although not to the same magnitude as with diabetes, the increase in prevalence of hypertension was slightly higher among younger adults than among older adults. Rising obesity rates were suggested to have contributed to the increase in diabetes, and obesity has also likely led to an increase in hypertension.3
We found that diabetes was increasingly prevalent among people with a diagnosis of hypertension. We found a more rapid increase in incidence of hypertension between 1999 and 2001. This period coincided with the start of the Canadian Hypertension Education Program.21,22
The intent of this program is to improve awareness of hypertension guidelines among health care providers, and the program may have contributed to a decrease in undiagnosed hypertension. Although the ICD diagnostic codes for hypertension changed in 2003 in the hospital admission database, a sensitivity analysis using an algorithm of 2 physician billing codes in 3 years showed a similar pattern of increase, and there have been no changes in physician billing practices or policies, such as incentives for hypertension, that could account for this increase. An increasing number of visits to the physician did not appear to increase the possibility of hypertension being diagnosed. This was similar in all 3 years assessed, which also suggests that there were no major changes in physician coding practices for hypertension during the study period.
The prevalence rates of physician-diagnosed hypertension in our study are about 3%–4% higher than those found in national self-report surveys during a similar period.23
They are a little lower than the estimated 5% of patients taking antihypertensive medications who do not report that they have hypertension, perhaps because they erroneously believe that they are cured if their blood pressure is under control.8
Our finding of a higher prevalence of hypertension among older women than among older men was similar to sex-related differences in the national self-report surveys;23
however, the sex-related differences in our study were slightly greater than the global predicted estimates.1
By using a validated case-definition algorithm for the administrative data, we were able to examine hypertension prevalence and incidence over a large, ethnically diverse population, and to follow our population over time. However, the validation of the algorithm that we used applied to a population aged 35 years and older. The algorithm was likely as accurate in our population, given that younger adults probably have fewer competing comorbidities than older adults.
We chose a case-definition algorithm of 2 physician billing codes or 1 hospital admission code in 2 years in part because of its high level of accuracy and because of constraints on data availability. Case-definition algorithms that do not use a hospital admission database may miss patients with more comorbidities because the outpatient physician billing database typically records only the primary diagnosis.
There are other limitations inherent to the use of administrative data that may have led to an underestimation of the occurrence of hypertension, which further strengthens our argument that current projections are too low. First, administrative data do not capture actual blood pressure measurements. Therefore, we rely on physician recognition of patients with hypertension. Second, our algorithm requires that patients use the health care system either as an outpatient or as an in-hospital patient. As a result, we are unable to capture undiagnosed hypertension. In addition, our algorithm has been found to miss up to 28% of cases when compared with audits of primary care physician charts, which may also be responsible for the underestimation of the true prevalence of hypertension.12
Third, our findings of a greater prevalence of hypertension among women than among men may, in part, be related to women visiting physicians more often then men24
and, therefore, having more opportunities to receive a diagnosis of hypertension.25
The extent to which this factor affects the sex-based differences in prevalence in our study cannot be measured with our administrative data. Finally, about 2% of physicians are paid salaries under alternate funding plans, and their billings do not appear in the Ontario Health Insurance Plan database. Nonetheless, it is estimated that less than 6% of the population report not having a family physician, and more than 75% of the population report visits to a primary care physician at least once a year.26
Thus, within a 2- or 3-year period, it is likely that almost all residents have at least 1 visit to a physician. Furthermore, in the 2001 Canadian Community Health Survey, 73% of respondents reported having their blood pressure measured within the past year and 85% reported having a blood pressure measured in the 2 years before the survey.7
The prevalence of hypertension is steadily increasing, and in fewer than 10 years it will likely exceed the prevalence projected for 2025. This anticipated increase, in conjunction with a comparable increase in diabetes,2
will invariably lead to dramatic rises in the incidence of cardiovascular diseases and their consequences, which has the potential to overwhelm a health care system that is modelled on underestimates. The increasing prevalence of hypertension will also have financial implications for provincial drug plans because there is increasing evidence that the majority of patients with hypertension will require 2 or more drugs to achieve blood pressure control.27
Our study highlights the need for strategies to improve the prevention of hypertension.
@ See related articles, pages 1436, 1441 and 1458