Treating hypertension: the drugs and the patients. Method
Hypertension is among the most frequently encountered health problems in primary care. The many clinical practice guidelines for hypertension have been remarkably consistent in recommending initial therapeutic choices: diuretics and β-blockers have been recommended as first-choice therapy by expert panels around the world.
Some studies, however, suggest that physicians, particularly in North America, do not apply these guidelines. Diuretics are the most frequently prescribed drugs, but angiotensin-converting enzyme (ACE) inhibitors and calcium channel blockers, which are more expensive, are prescribed more often than is recommended as first-choice therapy. For example, Psaty et al, in a study of prescribing trends, observed that newly treated hypertensive patients were about half as likely as previously treated patients to receive diuretics or β-blockers and about twice as likely to receive calcium channel blockers or ACE inhibitors.
Although hypertension treatment guidelines are remarkably consistent, they are also remarkably complex. First-choice drugs might be contraindicated because of associated medical conditions. Because hypertension often is associated with myriad other chronic diseases, it is possible that only a few patients actually are able to take diuretics or β-blockers, making it difficult to ascertain how widely clinical practice strays from the guidelines.
Few studies have reported on factors associated with the choice of antihypertensive medication. Many of those studies are surveys of reported prescribing practices rather than descriptive studies based on chart review or database analysis. Some authors have reported that younger patients tend to receive ACE inhibitors more frequently than older ones, but this has not been observed consistendy. In a survey conducted in Sweden, female and older physicians chose diuretics as first-line therapy more often than younger male physicians did. Some authors suggest that general practitioners choose less expensive treatment strategies more frequently than specialists, an observation that does not take into consideration the difference in case severity between the two groups of physicians. We found only two studies where comorbidity was ascertained. One was conducted in a health management organization; the other was based on a study of a community of residents older than 65 years. In neither study was comorbidity associated with choice of therapy.
Our study aims to evaluate the appropriateness of prescribing antihypertensive drugs according to the latest recommendations of the Canadian Hypertension Society, in a cohort of patients followed for at least 2 years in a family medicine unit in Montreal.
Method
Setting
The study took place in a teaching unit affiliated with the Family Medicine Program at the University of Montreal. The clinic, which opened in 1988, serviced a mostly French-speaking population and was located in a poor socioeconomic area of Montreal. (At that time, the Quebec provincial government paid the full cost of prescription drugs for disadvantaged and elderly people.) The unit had 13 regular staff physicians, nine of whom practiced full time at the teaching unit, and 13 family medicine residents on rotation for 2 years. In 1994, the clinic entered all its patient records into a computerized register. Two fourth-year medical students (functioning as clinical clerks at the externship level) effected the transfer under the supervision of one of the authors. We reviewed records of the previous 2 years’ visits to establish an active problem list, which we captured in the computerized database.
Study population
To be eligible, patients registered as having hypertension had to be regular patients of the clinic (not walk-in patients), who had consulted at least once a year between 1993 and 1995 and who had been treated for hypertension by a family physician at the clinic rather than by a specialist at the hospital. Charts were excluded if diagnosis of hypertension was not supported by chart review; if charts contained insufficient information; if patients were pregnant or younger than 18 years; or if patients were suffering from secondary hypertension and presenting with complex medical conditions, such as severe cardiac failure, a trial fibrillation with anticoagulant therapy, and metastatic cancer. Of the 350 patients registered since 1988,183 met the eligibility criteria. Reasons for exclusion appear in Table 1.
| Table 1. Reasons for exclusion from the study: 167 of 350 patients were excluded. | |
| REASON FOR EXCLUSION | N (%) |
| Only one or no consultation between 1993 and 1995 | 122 (73.1) |
| Not followed at the clinic for hypertension | 20 (12.0) |
| Diagnosis not supported by chart review | 11 (6.6) |
| Multiple complex medical conditions | 10 (6.0) |
| Insufficient information in the chart | 3 (1.8) |
| Secondary high blood pressure (renal artery stenosis) | 1 (0.6) |
| Total | 167 |
Variables
Variables were extracted from the charts by the same person, who had been trained by the two other authors. Interobserver agreement was not measured formally, but periodic checks of the data abstraction process were conducted during the study. As previously mentioned, to be included in the analysis, charts had to contain information on patients’ problems, blood pressure readings, and medications. Only three charts were excluded, which illustrates the high quality of the clinic’s charts.
All visits were reviewed to ascertain treatment modifications and occurrence of new health problems. The dependent variable was the antihypertensive medication. Mean blood pressure reading was considered an outcome variable and readings at each visit as independent variables that could explain treatment modifications. Other independent variables were age and sex of patients, duration of hypertension, total number of visits and number of visits for hypertension during the study period, number of physicians consulted at the clinic, number of consultations with specialists, associated medical conditions, diagnosis of target organ damage, and associated medications.
Analysis
χ2 and Student’s t tests were used for nominal and continuous variables, respectively. Logistic regression analysis was performed to evaluate factors associated with the probability of receiving each major class of medication (diuretics, β-blockers, calcium channel blockers, and ACE inhibitors). Control variables were medical conditions considered indications or contraindications, age and sex of patients, presence of target organ damage, duration of hypertension, and systolic and diastolic blood pressure.
This post has been viewed 4341 times.
Comments are closed.
