Physician Familiarity With Diagnosis and Management of Hypertension According to JNC 7 Guidelines

Abstract

Physician knowledge of the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) guidelines is unknown and may contribute to the prevalence of uncontrolled hypertension. Our objective was to determine physician knowledge of JNC 7 guidelines and whether online instruction could improve knowledge. A pretest served as baseline knowledge, and comparison with a post‐test after completing an online didactic demonstrated improvement in knowledge. Participants included 1280 physicians at 45 internal medicine residency training programs. Average baseline knowledge of six concepts of hypertension was 51.2%. Attending physicians performed better than trainees on some but not all concepts (p<0.05). Third‐year post‐graduate trainees performed better than first‐year trainees on some but not all concepts (p<0.05). Knowledge increased significantly on all concepts after completing the curriculum (p<0.05). The authors demonstrated that physician knowledge of JNC 7 guidelines is poor but can be improved by an online curriculum. Further study is needed to determine the impact of physician education on clinical outcomes in individuals with hypertension.

Hypertension is an important modifiable risk factor for cardiovascular disease. ¹ The US Department of Health and Human Services has set a goal that by 2010, 50% of hypertensive Americans have their blood pressure (BP) controlled. ² However, in 1999–2000, two thirds of patients being treated for hypertension were not at goal BP, ² even those with good access to health care. ³ , ⁴ Physician characteristics are therefore thought to play an important role in the prevalence of uncontrolled hypertension. ³

In 2003, the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) was published. ⁵ Recognizing that past JNC reports were not as effective as hoped, JNC 7 was published in part because of the “need for a new clear and concise guideline that would be useful for clinicians” and to simplify the classification of BP. ⁵ The goal of any clinical practice guideline is to improve the quality of care, ⁶ but despite the publication of JNC guidelines, two thirds of hypertensive patients have uncontrolled BP. ⁷

Since physician characteristics play a role in uncontrolled hypertension, it is important to understand factors that influence treatment decisions. ⁸ Barriers to physician implementation of guidelines include lack of familiarity with specific guidelines, as well as lack of agreement with guidelines, physician expectation of outcomes, and therapeutic inertia. ⁶ Little is known about physician knowledge of JNC 7 guidelines and the impact of medical residency training on this knowledge. We hypothesized that physicians, including those undergoing medical residency training, are unfamiliar with specific recommendations of JNC 7 guidelines and that computer‐assisted instruction could be used to effectively teach these guidelines.

Content Development

A didactic module on hypertension was developed using a six‐step approach to curriculum development. ⁹ JNC 7 guidelines and 39 additional references were used to support content of the didactic module and questions. Six key concepts detailed in JNC 7 guidelines were identified for instruction, which included staging of BP, laboratory evaluation of the newly diagnosed hypertensive, role of lifestyle modification in controlling BP, identifiable causes of hypertension, disease‐specific BP goals, and initial choice of an antihypertensive agent. Twelve clinical cases were written that described patients presenting with elevated BP, and respondents were asked to correctly stage and/or manage the patient based on JNC 7 guidelines. Multiple answer choices were provided, requesting the respondent to select a diagnosis or to make an initial decision on medical management. No time limitations were made on completion of the questions.

Face validity and content validity of questions were obtained by having diagnosis and management questions reviewed by a group of 10 self‐identified hypertension specialists and medical house officers; questions were revised until agreement was reached that the questions were clear and that sufficient information was provided to correctly diagnose and manage each case. Content validity was further verified by having five self‐identified hypertension specialists answer the multiple‐choice questions; the average score among these experts was 92%. One set of cases and questions on diagnosis and management was used as the module pretest, and a different set was used as the post‐test. To educate learners on diagnosis and management of patients according to JNC 7 guidelines, additional cases with didactic text were written to illustrate key concepts. Registered users began the module by completing the pretest, which enabled them to access the didactic section. The post‐test could not be accessed unless the didactic section had been completed. For both the pretest and post‐test, the learner was informed if their answer choice was correct and if incorrect, which answer choice was correct.

Study Population

The hypertension module was used by 45 internal medicine residency training programs in 21 states and Washington, DC as part of the Johns Hopkins Ambulatory Care Curriculum. ¹⁰ At participating residency training programs (internal medicine housestaff and faculty), 3667 had registered and been approved to use the curriculum Web site (www.hopkinsilc.org). The study sample consisted of 1280 physicians who completed the module (participation rate, 34.9%). Of these 1280 physicians, 456 (35.6%) were in their first postgraduate year (PGY)‐1, 373 (29.1%) were in PGY‐2, 369 (28.8%) were in PGY‐3, 48 (3.7%) were attending physicians, and 10 (0.8%) did not state their level of training. Participating residency training programs included primary affiliates of medical schools with extensive National Institutes of Health funding, as well as several community hospitals. ¹⁰ The Johns Hopkins Institutional Review Board approved this study. All personal identifiers were removed from group results before analysis.

Data Collection and Analyses

Performance data were tabulated by the Web site from July 1, 2004, through June 30, 2005, at which time the module was removed for updating for the new academic year. Results of partially completed modules were not included. Responses to the questions were tabulated electronically by the Web site and then analyzed based on year of training or attending status. Statistical analyses were performed with Stata software version 8.2 (StataCorp LP, College Station, TX). Comparisons between pretest and post‐test scores and comparisons of pretest scores at different training levels were performed using the chi‐square test. All p values were two‐sided, and a p value <0.05 was considered statistically significant.

Knowledge of Key Concepts in JNC 7. Average scores for medical residents and attending physicians on baseline knowledge of key concepts in diagnosis and management of patients presenting with elevated BP are shown in Figure 1.

Figure 1.

Performance of medical residents and attending physicians on the pretest of key concepts described in the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure is shown. PGY=post‐graduate year; *those concepts for which a significant difference (p<0.05) existed in performance between groups

All physicians performed best on the concept of choosing an initial agent for pharmacotherapy in the hypertensive patient who had no compelling indications. All physicians performed worst at identifying appropriate disease‐specific BP‐lowering goals according to JNC 7 guidelines. Performance was also poor for identifying a patient most likely to have a treatable secondary cause of hypertension. Performance was better on appropriate testing in the newly diagnosed hypertensive as well as the role of lifestyle modification in management of BP.

Attending physicians performed significantly better than medical residents on staging BP, the role of lifestyle modification in management of BP, and at choosing an appropriate initial agent for pharmacotherapy (Figure 1). Attending physicians did not perform significantly better than medical residents on choosing appropriate testing in the newly diagnosed hypertensive, identifying a patient most likely to have a treatable secondary cause of hypertension, or on identifying appropriate disease‐specific BP‐lowering goals.

Misconceptions in Diagnosis and Management of Hypertension. The concept of staging BP was tested by describing five hypothetical patients with elevated BP; other medical history was also provided. Each patient's BP was included in the description, along with the stage of BP, which was correct for only one of the choices (Table). Although a plurality of physicians (42.1%) correctly staged a 49‐year‐old man with a history of myocardial infarction and BP of 138/84 mm Hg as prehypertensive, 33% misclassified a patient whose BP was 156/96 mm Hg as having stage 2 hypertension. More than two thirds of physicians (68.4%) agreed with delaying pharmacotherapy to allow lifestyle modification to control BP in a patient with no other major cardiovascular risks, but the remainder disagreed and had misconceptions about sodium restriction, weight loss, and dietary potassium. Although JNC 7 identifies the most common age range of onset of hypertension as between ages 30 and 55 years, a plurality of physicians (33.3%) wanted to evaluate a 33‐year‐old woman with new‐onset hypertension for a treatable cause. Only one quarter (26.5%) identified a 52‐year‐old patient with uncontrolled hypertension despite treatment with maximal doses of metoprolol, enalapril, and hydrochlorothiazide as most likely among the choices to have a treatable cause of hypertension. When asked to identify the appropriate disease‐specific BP‐lowering goals according to JNC 7, physicians were incorrect on BP goals for diabetes (36.2%), congestive heart failure (26.0%), cerebrovascular disease (8.8%), and renal disease (8.0%). Of all the questions asked, agreement was strongest (76.2%) on choosing a diuretic as initial pharmacotherapy in a hypertensive patient with no compelling indications. A β blocker was a distant second choice (9.4%), followed by an angiotensin‐converting enzyme inhibitor (7.3%).

Table.

Detailed Answer Selections on Key Concepts

Concept: Question Synopsis	Answer Selections*
Staging: Which of the five patients listed has blood pressure (BP) (verified at multiple visits) appropriately staged?	49‐Year‐old man with a history of myocardial infarction and BP of 138/84 mm Hg is diagnosed with prehypertension (42.1)** 54‐Year‐old man with elevated cholesterol and BP of 156/96 mm Hg is diagnosed with stage 2 hypertension (33.0) 39‐Year‐old woman with diabetes and BP of 128/86 mm Hg is diagnosed with stage 1 hypertension (17.0) 36‐Year‐old man with polycystic kidney disease/proteinuria and BP of 128/78 mm Hg is diagnosed with stage 1 hypertension (4.8) 43‐Year‐old man with no significant medical history and BP of 118/72 mm Hg is diagnosed with prehypertension (3.2)
Testing: Of the following tests, in which group is all that is listed part of the standard evaluation of the patient with diagnosed hypertension?	Twelve‐lead electrocardiogram; urinalysis; potassium; calcium (72.7)** Echocardiogram; low‐density lipoprotein cholesterol; potassium; creatinine (13.2) Sodium; calcium; 24‐h urine protein; glucose (6.6) Spot cortisol; potassium; high‐density lipoprotein cholesterol; 12‐lead electrocardiogram (4.0) Echocardiogram; hematocrit; glucose; urinalysis (3.5)
Role of lifestyle modification: Choose the correct statement for a 43‐year‐old female patient with stage 1 hypertension who is overweight and sedentary but has no other cardiovascular risk.	Pharmacotherapy may be delayed to allow lifestyle modification to lower BP (68.4)** For diet to lower BP, sodium intake must be restricted to <2 g/d (13.6) Diet will not impact BP unless weight loss accompanies dietary change (10.2) For weight loss to impact BP, enough weight must be lost to correct body mass index to normal range (4.0) Dietary potassium, similar to dietary sodium, should be reduced to control BP (3.8)
Identification of a patient with a treatable cause of hypertension: Select the patient who should be evaluated for a secondary cause of hypertension.	A 33‐year‐old woman presenting with new‐onset hypertension confirmed over three visits during the past 6 months (33.3) A 52‐year‐old woman with a 6‐year history of hypertension that is uncontrolled despite treatment with maximal doses of metoprolol, enalapril, and hydrochlorothiazide (HCTZ) (26.5)** A 47‐year‐old man presenting for an initial health assessment noted to have BP of 210/116 mm Hg (22.4) A 56‐year‐old man with a 10‐year history of hypertension that is uncontrolled despite treatment with maximal doses of atenolol, doxazosin, and lisinopril (9.6) A 42‐year‐old man with a 2‐year history of hypertension who is noted to have optic disc cupping on physical examination (8.2)
Disease‐specific BP goals: Which one of the following disease‐specific BPlowering goals is recommended by JNC 7?	Diabetes: goal BP <135/85 mm Hg (36.2) Congestive heart failure: goal BP <120/80 mm Hg (26.0) Ischemic cardiovascular disease: goal BP <140/90 mm Hg (21.1)** Cerebrovascular disease: goal BP <135/85 mm Hg (8.8) Renal insufficiency with proteinuria: goal BP <135/85 mm Hg (8.0)
Initial pharmacotherapy: Choose the initial pharmacotherapy for a 43‐year‐old overweight sedentary male smoker who presents with stage 1 hypertension and elevated cholesterol. There is a family history of hypertension.	HCTZ daily (76.2)** Atenolol daily (9.4) Lisinopril daily (7.3) Lisinopril/HCTZ combination daily (6.2) Sustained‐release nifedipine daily (0.8)
*Percentages are given in parentheses and may not total 100% because of rounding. **=correct diagnosis; JNC 7=Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure

Performance Based on Year of Training. To assess the impact of year of training on knowledge, performance on the key concepts was compared between the PGY‐1 and PGY‐3 groups (Figure 2).

Figure 2.

Performance on key concepts is compared between post‐graduate year (PGY)‐1 and PGY‐3 medical residents. *Those concepts for which a significant difference (p<0.05) existed in performance between groups.

There was a significant increase in knowledge from PGY‐1 to PGY‐3 participants on appropriate testing in the newly diagnosed hypertensive, identification of a patient most likely to have a treatable cause of hypertension, and choice of an initial agent in a patient without a compelling indication. There were no statistically significant differences between PGY‐1 and PGY‐3 medical residents in knowledge of staging of BP, the role of lifestyle modification, and disease‐specific BP‐lowering goals.

Effect of Educational Module on Knowledge. After completing the didactic module, physicians completed a post‐test on the six key concepts; results are shown in Figure 3. Overall performance improved from 51.2% to 72.0%. Physician scores increased on each of the six concepts studied; however, physicians continued to have difficulty in selecting disease‐specific BP goals. Only one third (36.2%) were able to correctly do this after completion of the didactic module.

Figure 3.

Performance by all physicians on key concepts is compared between the pretest (Pre) and post‐test (Post). Improvement in all scores from the pretest to the post‐test were statistically significant for all concepts (p<0.005).

DISCUSSION

We found that physician knowledge of several key concepts of diagnosis and management of hypertension as described in JNC 7 is poor. In particular, physician knowledge of appropriate staging, clues to an underlying treatable cause, and disease‐specific BP goals was poor. Attending physicians perform better than trainees in some areas (e.g., staging BP, the role of lifestyle modification in managing BP, choosing a diuretic as initial pharmacotherapy in patients with no compelling indications), but not in others (e.g., identifying clues to a treatable cause of hypertension, disease‐specific BP goals). Similarly, PGY‐3 physicians were not more knowledgeable than PGY‐1 physicians in staging BP, the role of lifestyle modification, and disease‐specific BP goals. Knowledge of all concepts studied improved after completion of an online didactic module on the diagnosis and management of hypertension.

Staging of BP is important because it determines management. Appropriate clinical follow‐up (e.g., 1 month, 2 months, 1 year) and pharmacotherapy (e.g., no medication vs. single‐agent therapy vs. initiation of two agents) are determined by stage. ⁷ In our study, mistakes were made in classifying prehypertension, stage 1 hypertension, and stage 2 hypertension. While attending physicians did perform better than trainees, the lack of improvement in this area from PGY‐1 trainees to PGY‐3 trainees suggests inadequate education in this area during medical residency. There was also no improvement in upper‐level trainee knowledge on the role of lifestyle modification and disease‐specific BP goals.

Physician knowledge of disease‐specific BP goals trailed all other areas studied. There is a wealth of data demonstrating the benefit of more aggressive BP‐lowering goals in diabetes and renal disease, ¹¹ , ¹² , ¹³ yet even after completing the didactic in this area, confusion remains as to which diseases have specific BP‐lowering goals and what those goals are. In JNC 7, disease‐specific BP‐lowering goals were simplified from past guidelines, since the Committee felt that prior JNC reports were not being used to the best benefit of patients. ¹⁰ This study suggests that a significant proportion of physicians remain uncertain about these simplified guidelines, which may contribute to the inadequate BP control documented in patients with diabetes ¹⁴ , ¹⁵ and renal disease. ¹⁶

The results shown here are similar to those found with other guidelines. When a group of physicians (including 20 asthma specialists) was tested on knowledge of National Heart, Lung, and Blood Institute guidelines for the diagnosis and management of asthma, the mean total correct score was 60%. ¹⁷ Knowledge of American Cancer Society guidelines for colorectal cancer screening among a group of medical residents and attending physicians was also poor, with correct scores averaging 49% among residents and 56% among attendings. ¹⁸ Both of these studies are similar to performance of the physicians we studied, who had an average score of 51.2%.

Despite overall poor knowledge of JNC 7 guidelines, the role of diuretics as initial pharmacotherapy in hypertension has been widely accepted by the physicians in this study. As recently as 1999, a study of practicing physicians showed that 65% selected either an angiotensin‐converting enzyme inhibitor or a calcium channel blocker as monotherapy for a hypertensive patient with no compelling indications. ¹⁹ In our study, 75.9% of trainees and 93.8% of attending physicians chose a diuretic as mono‐therapy for a similar patient, perhaps as a result of the extensive publicity following publication of the Antihypertensive and Lipid‐Lowering Treatment to Prevent Heart Attack Trial (ALLHAT).

The interactive, case‐based curriculum used in this study, distributed via the Internet, effectively improved physician knowledge on specific concepts in the diagnosis and management of hypertension. Interactive, case‐based methods have been shown to be more effective than passive, didactic sessions in educating physicians. ²⁰ Computer‐assisted instruction, including that made possible by the Internet, allows for this interactivity that may improve educational outcomes. A study of passive techniques used to improve knowledge of JNC guidelines, specifically chart reminders, did not improve knowledge. ²¹

There are several limitations to this study. The impact of physician knowledge of JNC 7 guidelines on clinical outcomes is unknown; however, education has been shown to be effective in altering prescribing practices in management of hypertension. For example, an educational module on hypertension has been shown to increase prescribing of thiazide diuretics. ²² As shown for asthma guidelines, additional interventions such as feedback and peer review may be needed to improve clinical outcomes. ²³ The role of other physician factors, including lack of agreement with JNC guidelines and therapeutic inertia, was not studied. Finally, patient factors in diagnosis and management of hypertension were not measured.

This study demonstrates that physician knowledge of diagnosis and management of hypertension is poor and that residency education is not addressing many of these gaps in knowledge. The Internet can be used to increase knowledge of JNC 7 guidelines. It has been said that “undiagnosed hypertension and treated but uncontrolled hypertension occur largely under the watchful eye of the health care system.” ³ Addressing the educational needs of practicing physicians and medical trainees may reduce the role of physician factors in patients with hypertension that is undiagnosed or poorly controlled. Randomized controlled trials of educational interventions on the impact of clinical outcomes are needed to better define strategies to meet national goals on BP control.

Acknowledgment: The authors would like to thank Drs. Frederick Brancati and Lawrence Appel, Professors of Medicine, Johns Hopkins University, Baltimore, MD, for their helpful comments and review of this manuscript.