James W. Kinn, MD, Midwest Heart Specialists, 3825 Highland Avenue, Suite 400, Downers Grove, IL firstname.lastname@example.org
Clinical studies suggest that hypertension is often undiagnosed, undertreated, and poorly controlled. In 1997, the authors developed a comprehensive electronic medical record that interfaces with physicians during each outpatient visit and provides real-time feedback about patient care management, including the management of hypertension. The purpose of this study was to determine whether this interactive electronic medical record results in better detection and control of hypertension. During a 12-month study period, consecutive outpatients (n=1076) were seen for routine follow-up; patient care documentation relied solely on the electronic medical record. Quality indicators for hypertension included: 1) documentation of the diagnosis of hypertension; 2) use of blood pressure-lowering drugs; and 3) successful blood pressure lowering to ≤140/90 mm Hg. The authors compared the hypertension management of these patients to a control group of similar patients (n=723) with medical records consisting solely of traditional “pen and paper” charts. Baseline characteristics were similar between the two groups, including the prevalence of hypertension (73 % vs. 70%; p=NS). However, the electronic medical record resulted in higher documentation rates of hypertension (90% vs. 77%; p<0.001), greater use of antihypertensive therapy (94% vs. 90%; p<0.01), and more successful blood pressure lowering to ≤140/90 mm Hg (54% vs. 28%; p<0.001). In conclusion, the electronic medical record with real-time feedback improves the physician's ability to detect, treat, and control hypertension.
An estimated 50 million Americans have hypertension.1 Patients with hypertension are at increased risk for myocardial infarction, stroke, death, congestive heart failure, and end-stage renal disease. Unfortunately, hypertension is often underdiagnosed and undertreated. The Third National Health and Nutrition Examination Survey (NHANES III, phase 2), reported that only 68% of patients with hypertension are diagnosed, only 53% are being treated, and only 27% are adequately controlled to a blood pressure of less than 140/90 mm Hg. This “treatment gap” has been evident since the 1970s.1,2 There has been significant progress in developing new and effective antihypertensive drugs over this same time period. Nevertheless, hypertension remains underdiagnosed and undertreated.3,4 New approaches to the management of patients with hypertension are needed.
Recently, we developed an electronic medical record (EMR) to help manage our outpatient cardiovascular practice. This EMR provides real-time documentation of patient visits, including all of the elements of a complete medical record. This “paperless chart” was developed and implemented in our practice in 1997 and has become an integral part of the standard of care for our patients. It is widely accepted and used by all of our physicians as the primary source of documentation. This EMR recognizes data elements, including blood pressure measurements, medical history, and diagnoses. It detects undiagnosed and undertreated hypertension and provides real-time feedback, which alerts the physician to potential deficiencies in patient care. We hypothesized that this process may potentially modify and improve physician behavior and improve patient care.
The purpose of this study was to evaluate this computer-assisted hypertension management process, which includes an EMR with an integrated blood pressure management tool. To achieve this, we compared the hypertension management of patients with office visits documented in the EMR to similar patients with documentation that relied on traditional “pen and paper” charts.
To evaluate the effectiveness of the EMR, three physician volunteers from our single-specialty cardiology practice agreed to fully implement the EMR for 100% of their outpatient office visits. Consecutive patients (n=1076) were scheduled for routine outpatient longitudinal visits and the EMR was effectively utilized for 100% of these outpatient visits over a 12-month study period (November, 1998 through October, 1999). For the purposes of this study, patients were defined as having hypertension if they had a systolic blood pressure ≥140 mm Hg, a diastolic blood pressure ≥90 mm Hg, or a previously established diagnosis of hypertension. Quality of care parameters included: 1) documentation of the diagnosis of hypertension (awareness); 2) use of antihypertensive medication (treatment); and 3) successful lowering of blood pressure to ≤140/90 mm Hg (control).
Electronic Medical Record
The EMR was co-developed by Midwest Heart Specialists, Ltd. (Downers Grove, IL) and Cardio-works (Westman, IL). It is a MicroSoft Windows-based system designed to allow physicians and other health care professionals to document and retrieve clinical information. The EMR stores clinical information in the form of data elements, allowing for multiple data queries. It organizes, stores, and displays relevant clinical information by extracting key data elements from current and prior EMR documentation. In addition, it also provides physicians with timely reminders and online alerts when patients fail to meet established guidelines.
The EMR is designed to record patient data during outpatient visits as the sole source of patient care documentation, and functions as a “paperless chart.” The health care provider uses a hand-held device, which transmits data to and from a main central computer. Providers can enter and retrieve patient information during office visits in real-time. Health care workers are electronically provided with complete lists of diagnoses (including hypertension) and medications at the start of each office visit. Clinical information is updated during patient care documentation. Alerts are designed to remind physicians to diagnose and treat hypertension as well as other common disorders. Clinical information is displayed in the form of physician reminders. In response to these prompts, physicians may choose to establish and document new diagnoses (diagnose hypertension) or modify treatment plans (adjust pharmacologic therapy).
Computer-assisted patient management using the EMR was compared to the management of a control group comprising similar patients chosen at random from the same group practice (19 cardiologists) who had office visits documented by traditional “pen and paper” charts (n=723). Patients in the control group were evaluated by retrospective chart review (ACCESS Medical Ltd., Arlington Heights, IL., under contract by Merck and Co., Inc.). To establish the control group, all active patients in the practice for less than 12 months were excluded. Remaining established patients were then sorted according to physician, and 40 patients for each of the 19 physicians were chosen at random to form the control group. Analysis of the control group was performed on 723 of the anticipated 760 patient charts. Thirty-seven charts in the control group were excluded because the charts were inaccessible or data appeared incomplete or unreliable. Note that this control group included active “pen and paper” charts from outpatient visits done in 1996, as this time period predated the transition period that began in 1997 when the EMR was introduced into practice. Quality of care was determined by measuring the same parameters designed to measure the awareness, treatment, and control of hypertension.
We used quality-of-care measurements that have previously been well established in the medical community.2,3 Discrete variables were compared using the chi-square statistic. Continuous variables were compared using the t test for unpaired data. Results are reported as the means±SEM unless otherwise indicated.
Of the 1076 consecutive patients in the EMR group, 783 patients (73%) were identified as having hypertension. Of the 723 patients in the “pen and paper” control group, 502 patients (70%) were identified as having hypertension (p=NS). The Table shows that baseline characteristics of each group were also similar with respect to age, gender, and other cardiovascular risk factors. The prevalence of myocardial infarction, prior coronary bypass surgery, or prior angioplasty was higher in the control group.
Table Table. Baseline Patient Characteristics
CABG=history of coronary artery bypass surgery; MI=history of myocardial infarction; PTCA=history of percutaneous coronary intervention
Figure 1 shows that patients with hypertension in the EMR group were more likely to be given a diagnosis of hypertension (90% vs. 77%; p<0.001). In addition, patients in this group were also more likely to receive at least one antihypertensive drug (94% vs. 90%; p<0.01). The EMR was also associated with a greater number of blood pressure medications per patient (2.63±0.03 vs. 1.67±0.04; p<0.001). Figure 2 illustrates the increased use of blood pressure drugs associated with the EMR. The EMR group utilized combination therapy (two or more blood pressure drugs) in 81% of hypertensive patients, compared with 51% of the hypertensive patients in the “pen and paper” group (p<0.001). The increased use of medications in the EMR group suggests more aggressive upward drug titration to improve blood pressure control. As a result, the EMR was associated with a lower systolic blood pressure (138.9±0.4 vs. 147.1±0.4 mm Hg; p<0.001) and a lower diastolic blood pressure (76.8±0.2 vs. 80.6±0.2 mm Hg; p<0.001) after treatment. In addition, more patients in the EMR group achieved the blood pressure goal of ≤140/90 mm Hg (54% vs. 28%; p<0.001). Similar data regarding hypertension diagnosis, treatment, and control from NHANES III are also shown in Figure 1, to serve as a community benchmark.
Figure 3 demonstrates that patients followed in the EMR were more likely to receive angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers, β blockers, and diuretics. The utilization of calcium channel blockers and other antihypertensives did not significantly differ between the two study groups. Of note, the control group was more likely to have patients with prior myocardial infarction (39% vs. 18%; p<0.01) or prior coronary bypass surgery (51% vs. 34%; p<0.01). Thus, despite a seemingly greater relative indication for pharmacologic therapy in the control patients, these high-risk control patients were less likely to receive β blockers or ACE inhibitors than patients in the EMR group. This suggests that the EMR may also have benefits that extend beyond the management of hypertension alone, and that the disease management of coronary artery disease may also be enhanced by the EMR.
Individual Physician Performance
Subset analysis was done for each of the three physician volunteers to determine their performance in hypertension management using the traditional pen and paper charts (1996) and the EMR (1998–1999). After the EMR was implemented, patients treated by these three physicians were more likely to obtain a diagnosis of hypertension (Figure 4A) and achieve successful blood pressure lowering to ≤140/90 mm Hg (Figure 4B). In addition, use of the EMR was associated with an increase in the number of antihypertensive drugs per patient for each of the three individual physicians (Figure 4C). Furthermore, quality-of-care measurements for each of the three individual physicians using pen and paper charts did not significantly differ from the pen and paper charts in the rest of the control group. This demonstrates that prior to using the EMR, each of the three physician volunteers in the EMR group had similar (and suboptimal) practice patterns as the physicians in the control group. Then, after using the EMR for 12 months, each of the three physicians demonstrated improved hypertension management. These findings further support our hypothesis that the EMR is associated with improved patient care.
This study demonstrates that in comparison to traditional methods, the EMR is associated with more consistent and successful management of patients with hypertension. The EMR is associated with more patients being diagnosed with hypertension, more patients being treated for hypertension, and more patients receiving combination therapy. As a result, the EMR was associated with a two-fold increase in patients achieving the blood pressure goal of ≤140/90 mm Hg.
The reason that the EMR improves the management of hypertension is that it uses timely reminders that jog physicians' memories to recognize and treat hypertension. These reminders also serve as a means to establish common and well-communicated treatment goals so that every health care provider on the team knows when the goals of therapy have not been met. In addition, although this study involved only physician office visits, this process may also help guide and reinforce outpatient longitudinal clinics that utilize advance practice nurses and other nonphysician health care providers who may also direct patient care. In addition, we observed that the benefits of the EMR were not limited to just a fraction of those physicians who used it. Instead, the EMR was associated with improved hypertension management for each of the individual physicians who used it. This also supports the idea that the EMR may potentially improve patient care for other providers who may use it. Further studies including a larger number of physicians as well as other nonphysician health care providers are necessary to more accurately evaluate the full impact of the EMR on the management of hypertension.
It should be emphasized that the two-fold increase in successful blood pressure reduction occurred in conjunction with an increased use of combination therapy. More than one half of the patients in the EMR group were on three or more antihypertensive drugs, and the greatest change was in the increased use of diuretics, β blockers, and ACE inhibitors. This supports the hypothesis that the EMR prompts physicians to modify treatment regimens, but these findings may also be indicative of a general trend toward increased use of combination therapy between 1996 and 1999.
Several limitations from the current study require comment. First, defining the hypertensive population and success of blood pressure lowering in these study patients relied solely on data queries of the most recent office blood pressure reading. This study did not account well for “white coat” hypertension. Ambulatory blood pressure monitoring and home blood pressure readings were not available for review. This may cause some patients to be included as hypertensive who are in fact normotensive. This may also cause some patients to seem inadequately treated when in fact their blood pressure is well controlled outside the office. Fortunately, these limitations should affect both study groups similarly and should not introduce any bias into the comparisons made between them. In addition, this limitation may also lower the measured rates of hypertension diagnosis, treatment, and control. This may, in turn, falsely decrease the success rates of the EMR in comparison to other community benchmarks. Despite this obstacle, the EMR was still associated with a higher rate of diagnosing hypertension and a two-fold increase in achieving adequate blood pressure control (≤140/90 mm Hg) compared with the general population, as previously reported by NHANES III.1
A second limitation of the current study is the potential for time bias. The control group data were obtained by reviewing the charts of established patients who made office visits in 1996, while the EMR data were obtained between 1998 and 1999; theoretically, physicians may have been able to more successfully treat hypertension in the later time period. However, data from NHANES III suggest a paradoxical decline in the diagnosis, treatment, and control of patients with hypertension in the early part of that same decade,1,2 and there has been no convincing evidence since then to suggest otherwise. Other recent reports from Minnesota have shown a similar decrease in awareness, treatment, and control of hypertension.5,6 Therefore, it is unlikely that these small differences in time are responsible for any of the observed differences between the EMR and control group documentation. One additional concern might include a potentially greater selection of antihypertensive medications available to the EMR group as a result of new medications being developed and made available in the United States. However, in response to this concern, we see that the greatest increase in physician prescribing habits occurred in the class of drugs that included β blockers, ACE inhibitors, and diuretics, and these drugs had been available for many years before the interval of our study (1996–1999).
Successful treatment of hypertension requires one to overcome many barriers, which include physician and patient education and compliance with guidelines, access to adequate pharmacologic therapy and medical care, and establishing hypertension as a priority for physicians and patients. Improving physician compliance with hypertension management guidelines has been a difficult task over the past 30 years. Recommendations from the Joint National Committee (JNC VI) are widely accepted, yet they have not been fully implemented. We believe that the EMR, with real-time prompts, can have widespread applications to improve physician understanding and implementation of these guidelines, enhance the management of hypertension, and significantly improve patient care.