Abstract
Background
While the ambulatory setting is recognized as the best arena for optimizing antihypertensive drug treatment after a stroke, little is known about recent office-based antihypertensive drug treatment patterns in the United States. We assessed national trends in antihypertensive treatment of stroke patients in office-based medical practice.
Methods
Datafrom the 2000-2009 National Ambulatory Medical Care Surveys were analyzed comprising outpatient visits to physicians in office-based practice by patientsaged ≥ 40 yearswith a diagnosis of stroke(weighted estimate = 46,317,269). The main outcome measure was visits with a prescription of antihypertensive medication(s).
Results
The proportion of total visits that included a prescription of antihypertensive medicationwas 35.6% in 2000-2002, 29.5% in 2003-2005, and 49.3% in 2006-2009 (p=0.002);50.9% were primary care physician (PCP) visits vs.26.2% neurologist-visits (<0.0001).Age-adjusted logistic regression analyses confirmed a higher prescription rate in 2006-2009 vs. 2000-2002 (1.81, 95% CI=1.10-2.96) and PCP vs. neurologists (2.82, 95% CI=1.86-4.27). Use of two or more agent classes was 31.6% in 2000-2002, 44.2% in 2003-2005, and 56.7% in 2006-2009 (p=0.014). Age-adjusted logistic regression analyses confirmed a higher prescription rate of ≥ 2 agent classesin 2006-2009 vs. 2000-2002 (2.96, 95% CI=1.40-6.24). There were no significant differences in agent class type or number between neurologists vs. PCPs.
Conclusion
Over the last decade, there was a significant rise in use of antihypertensive drugs and combination of agent classes for patients aged≥ 40 years seen in an ambulatory setting with a diagnosis of stroke. PCPs were more likely than neurologists to prescribe these agents.
Keywords: Comparative Effectiveness; Outcomes; Stroke, Ischemic; Blood Pressure; Hypertension; Prognosis; Health services; Antihypertensive therapy; Target goals; Intensive; Guidelines
INTRODUCTION
Despite significant advances in the prevention and treatment of strokes in the last few decades, stroke remains a leading cause of disability, dementia and death in the United States.1Mitigating the immense societal and personal burden of stroke will require a greater emphasis on prevention through enhanced identification and control of established risk factors.2Of the 795 000 new strokes in the United States that occur each year, up to a quarter are recurrent strokes,1and the risk of dying is 41% after a recurrent stroke vs. 22% after a first-time stroke.3 Recurrent stroke prevention tends to be most effective when implemented early, monitored frequently, and maintained long-term after an index stroke event.4,5 Among all the established risk factors for stroke, none may be more important than elevated blood pressure, given its strong pathophysiologic contribution to virtually all types of stroke, and the plethora of therapies available to treat it. Yet, a substantial proportion of patients with, or at risk of stroke, are either not receiving antihypertensive treatments, or do not have their blood pressure under control.6,7Clinical trial evidence published over a decade ago indicated that antihypertensive drugs substantially lower recurrent stroke risk, even among stroke patients without hypertension.
Although hypertension is the premier treatable risk factor for recurrent stroke, and the most frequent principal diagnosis for ambulatory patients by US office-based physicians,8 research on factors that influence outpatient prescribing patterns and the extent of change produced by clinical trial findings among stroke patients is limited. Identification of variations in ambulatory practice among the clinicians who care for stroke patients may help direct specific efforts geared at improving overall stroke patient outcomes. 6The objective of this study was to assess trends in antihypertensive drug prescription rates among patients with a history of stroke in the United States over the last decade.
METHODS
This study involved an analysis of the National Ambulatory Medical Care Survey (NAMCS) data, an administrative dataset, which can be found at http://www.cdc.gov/nchs/ahcd.htm.NAMCS is a national, probability-sample survey conducted by the National Center for Health Statistics, Centers for Disease Control and Prevention. 9 It involves a continuing, in-person physician office survey of the non–federally employed US physicians providing office-based patient care conducted annually by the National Center for Health Statistics (NCHS). 9 NAMCS has been used previously to assess the characteristics of US office-based physicians, physician practice trends, and trends in ambulatory utilization.8,10-13
The NAMCS sampling frame consists of all physicians (excluding anesthesiology, pathology, and radiology) classified by the American Medical Association (AMA) or the American Osteopathic Association (AOA) as providing office-based patient care at a point approximately 6 months before the start of the survey year. It uses robust multistage probability sampling with the first stage sample, including counties, groups of counties, county equivalents, or towns and townships within the United States and the District of Columbia as the primary sampling units (PSUs). In the second stage, a probability sample of practicing physicians from the AMA and AOA master files are selected. In the final stage, patient visits to the selected sample of physicians are selected.
Each observation in the dataset represents a physician–patient encounter/visit. The physician aided by his/her office staff conducted the data collection. The NAMCS protocol instructs each physician’s office to record the primary diagnosis and 2 other secondary conditions related to the primary diagnosis. The NAMCS also collects data on patient demographics, services provided during the visit, chief complaint for the visit, and region of physician office.
For this study, to assess recent overall and physician specialty trends in antihypertensive drug therapy, NAMCS survey years over a 10-year period from 2000 to 2009 were evaluated.14To exclude strokes with very unusual causes that are almost always seen in individuals less than the age of 40 years, we selected all adults aged ≥40 years with a diagnosis of stroke (430.xxx-438.xxx) using valid International Classification of Diseases, 9th edition, Clinical Modification (ICD-9-CM) codes.15-17Up to three diagnoses were recorded for each visit. Up to six medications were recorded for each visit from 2000 to 2002. Starting in 2003, up to eight medications were recorded. To make years comparable in the present analysis, we limited the maximum number of medications to the first six listed in all years. These medications were used to identify visits including an antihypertensive drug. On the NAMCS data collection forms, physicians could record the antihypertensive drugs prescribed during the visit by using either generic drug names or brand names. Where brand names were listed, each generic name (active ingredient) component of the drug product was coded separately.
To examine patterns across years, the sample was grouped in years of three or four (2000-2002, 2003-2005, 2006-2009). To examine co-existing diagnosis we used ICD-9-CM; Hypertension (401.xxx), Myocardial infarction (410.xxx), Diabetes (250.xxx), Congestive heart failure (428.xxx) and Chronic kidney disease (585.xxx). Blood pressure was first collected in 2003 therefore years 2000-2002 were excluded from this part of the analysis. In additional analyses, we set out to compare antihypertensive drug treatment patterns between primary care physicians (PCPs) and neurologists, the latter group being the physician specialists recognized to have expertise in the acute and chronic management of patients with stroke. Thus visits were additionally characterized as follows - Visits where the physician marked physician specialty as “Neurology”; Primary care physicians (PCP): Visits where the physician indicated that they were the primary care physician and/or marked physician specialty as “General and family practice ” or “Internal medicine”. Of note, there were a few records with a physician indicating that they were the primary care physician and specialty as “Neurology”; these were classified as “Neurologists”.Finally, we set out to compare antihypertensive prescription patterns by class and number (one vs. two or more) of agents prescribed across time and between physician type.
Statistics
The data were analyzed by overall study population demographics, study period (2000-2002, 2003-2005, 2006-2009), and physician type (primary care physician vs. neurologist). Descriptive statistics were computed as frequency tables (row, column percentages) and statistical comparisons performed using chi-square tests. To examine antihypertensive drug prescription (yes/no) trends for stroke patients over the study period (2000-2002, 2003-2005, 2006-2009), unadjusted analysis using chi-square test were performed. Wherever the overall test was found to be statistically significant, pair-wise chi-square tests were performed reporting Holm’s adjusted p-values. We did not report any pairwise p-values, only the overall p-value. Furthermore, a logistic regression model adjusting for age (since age correlates very strongly with presence and degree of high blood pressure, as well presence of major vascular disease entities) was fit. To examine antihypertensive drug prescription (yes/no) patterns for stroke patients between primary care physicians and neurologists, unadjusted analysis using chi-square test were performed, then a logistic regression model adjusted for age was fit. To examine level of blood pressure for stroke patients across years (2003-2005, 2006-2009), the cutoff values listed below were used:
systolic blood pressure <140 mm Hg
diastolic blood pressure <90 mm Hg
systolic blood pressure <120 mm Hg
diastolic blood pressure <80 mm Hg
An unadjusted analysis using chi-square test was performed, and a logistic regression model adjusted for age was fit. Finally, to examine level of blood pressure for stroke patients between primary care physicians and neurologists, unadjusted analysis using chi-square test were performed, and subsequently, a logistic regression model adjusted for age was fit. Finally, unadjusted analyses for examining the relationship between class and number of antihypertensive agents versus 1) time period and 2) physician type among antihypertensive visits were performed using chi-square analyses. Additional analyses adjusted for age were performed using logistic regression. A sensitivity analyses in the latest time period with the most years (2006-2009) was conducted to see if time trends and physician specialty trends within this epoch were in accord with the overall trends across the decade. The determination of statistical significance was based on a 0.05 level of significance. All analyses were adjusted for the NAMCS complex survey design, including weights, clustering and stratification.
Ethics
Since this was an analysis of a publicly available de-identified administrative online database, formal review by the Institutional Review Board at our respective institutions was not required.
RESULTS
There were 18,197,254 weighted visits with stroke and antihypertensive treatment over the 9 year study period. Table 1 shows the descriptive data by time period studied over the decade. Several of the descriptive characteristics were fairly comparable across the time periods. However, the most notable difference across periods was for number of overall medications (of any type) listed, for which it was observed that there was a substantial increase in the number of drugs in 2006-2009 compared to 2000-2002 (≥ 6 drugs: 37.2 vs. 19.0%).Patients seen by PCPs were significantly more likely to be older, non-White, encountered in the Midwest/West, administered preventive care, been to the practice before, educated about diet/nutrition or exercise, prescribed multiple drugs, diabetic, and hypertensive (all p<0.05).Unadjusted data in table 2 show that antihypertensive drug prescription rates were significantly higher in 2006-2009 (49.3%) vs. 2000-2002 (35.6%) and 2003-2005 (29.5%), and significantly more so among PCPs (50.9%) compared to neurologists (26.2%). Age-adjusted logistic regression analyses (table 3)showed that antihypertensive drugs were prescribed more frequently in 2006-2009 compared to 2000-2002 (OR 1.81, 95% CI: 1.10-2.96, p=0.0188); likewise antihypertensive drugs were prescribed more frequently for PCP visits compared to neurology visits (OR 2.82, 95% CI: 1.86-4.27, P<0.0001).
Table 1.
Descriptive Summary by Time Period among Patients with Stroke in the National Ambulatory Medical Care Survey, 2000 to 2009
| Years 2000-2002 | Years 2003-2005 | Years 2006-2009 | |||||
|---|---|---|---|---|---|---|---|
| Variable | Description |
Weighted number of visits
(11,166,668) |
Weighted number of visits
(15,672,790) |
Weighted number of visits
(19,477,811) |
|||
| Proportion | Standard error | Proportion | Standard error | Proportion | Standard error | ||
| Age | |||||||
| 40-64 years | 27.7% | 3.5% | 25.7% | 3.3% | 34.4% | 2.4% | |
| 65-74 years | 32.1% | 3.6% | 25.5% | 2.6% | 25.5% | 2.0% | |
| >= 75 years | 40.2% | 3.7% | 48.7% | 3.4% | 40.1% | 2.4% | |
| Sex | |||||||
| Female | 52.7% | 3.9% | 47.5% | 3.2% | 53.5% | 2.7% | |
| Male | 47.3% | 3.9% | 52.5% | 3.2% | 46.5% | 2.7% | |
| Race | |||||||
| White | 84.0% | 3.4% | 85.2% | 2.8% | 87.2% | 2.0% | |
| Black | 10.4% | 2.6% | 10.9% | 2.3% | 9.8% | 1.8% | |
| Other | 5.6% | 2.3% | 3.8% | 1.8% | 3.0% | 0.8% | |
| Region | |||||||
| North East | 18.0% | 4.7% | 15.6% | 3.5% | 14.8% | 2.8% | |
| Mid West | 23.2% | 4.5% | 18.8% | 4.5% | 25.3% | 4.3% | |
| South | 33.2% | 5.4% | 46.9% | 7.0% | 44.9% | 5.1% | |
| West | 25.6% | 5.0% | 18.7% | 4.1% | 15.0% | 3.5% | |
|
Major
reason for visit |
|||||||
| Acute/new problem |
29.8% | 4.1% | 28.7% | 3.0% | 31.1% | 2.4% | |
| Chronic problem/routin e |
51.5% | 4.9% | 52.8% | 3.6% | 48.3% | 3.2% | |
| Chronic problem/flare up |
5.6% | 1.7% | 7.7% | 1.6% | 6.3% | 1.2% | |
| Pre or post surgery, injury follow up |
3.6% | 1.4% | 7.6% | 3.4% | 3.6% | 1.0% | |
| Preventive/non -illness care |
9.6% | 2.6% | 3.2% | 1.1% | 10.6% | 2.2% | |
|
Patient seen
before? |
|||||||
| yes | 89.8% | 2.3% | 87.5% | 2.2% | 87.8% | 1.5% | |
| no | 10.2% | 2.3% | 12.5% | 2.2% | 12.2% | 1.5% | |
|
Solo
Practice |
|||||||
| yes | 43.1% | 5.0% | 43.7% | 6.2% | 32.8% | 2.9% | |
| no | 56.9% | 5.0% | 56.3% | 6.2% | 67.2% | 2.9% | |
|
Diet
education |
|||||||
| no | 83.6% | 3.5% | 84.1% | 3.1% | 84.0% | 2.5% | |
| yes | 16.4% | 3.5% | 15.9% | 3.1% | 16.0% | 2.5% | |
|
Exercise
education |
|||||||
| no | 87.8% | 2.7% | 86.4% | 2.6% | 85.3% | 2.5% | |
| yes | 12.2% | 2.7% | 13.6% | 2.6% | 14.7% | 2.5% | |
|
Number of
medications |
|||||||
| 0-2 | 57.2% | 4.2% | 57.8% | 4.9% | 38.2% | 2.9% | |
| 3-5 | 23.8% | 4.0% | 21.5% | 3.4% | 24.6% | 2.3% | |
| >=6 | 19.0% | 3.7% | 20.7% | 3.4% | 37.2% | 2.6% | |
Table 2.
Unadjusted antihypertensive drug prescription rates by time period and provider type, NAMCS 2000-2009
| Analysis | Variable | No Prescription Percent (Standard Error) |
Yes Prescription Percent (Standard Error) |
p-value |
|---|---|---|---|---|
| #1: By time | ||||
| 2000-2002 | 64.4 (4.9) | 35.6 (4.9) | 0.0022 | |
| 2003-2005 | 70.5 (3.9) | 29.5 (3.9) | ||
| 2006-2009 | 50.7 (3.1) | 49.3 (3.1) | ||
| #2: By Provider | ||||
| Neurologist | 73.8 (3.1) | 26.2 (3.1) | <0.0001 | |
| Primary care provider | 49.1 (2.9) | 50.9 (2.9) |
Table 3.
Age-adjusted logistic regression analyses of antihypertensive drug prescription rates by time period and physician type, NAMCS 2000-2009
| Analysis | Variable | Odds Ratio | Lower 95% Confidence Limit |
Upper 95% Confidence Limit |
p-value |
|---|---|---|---|---|---|
| #1: By time | |||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 0.75 | 0.45 | 1.25 | 0.2685 | |
| 2006-2009 | 1.81 | 1.10 | 2.96 | 0.0188 | |
| Age | 1.01 | 1.00 | 1.03 | 0.0453 | |
| #2: By Provider | |||||
| Neurologist | Reference | Reference | --- | --- | |
| Primary care provider | 2.82 | 1.86 | 4.27 | <0.0001 | |
| Age | 1.01 | 0.99 | 1.02 | 0.3002 |
Unadjusted analyses of antihypertensive drug prescription rates by agent class and number of agent classes (supplementary table 1) showed generally similar rates by time period except for the use of thiazide diuretics which were significantly higher in 2006-2009 (19.0%) vs. 2000-2002 (12.4%), and the use of two or more agent classes in 2006-2009 (56.7%) vs. 2000-2002 (31.6%). There were no significant differences in antihypertensive drug prescription rates by agent class or number of agent classes between neurologists vs. PCPs. Age-adjusted logistic regression (table 4) only confirmed a higher use of two or more agent classes in 2006-2009 compared to 2000-2002 (OR 2.96, 95% CI 1.40-6.24, p=0.0044), and consistent withthe unadjusted analyses revealed no significant differences by provider type (table 5).
Table 4.
Age-adjusted logistic regression analyses of antihypertensive drug class prescription rates by time period, NAMCS 2000-2009
| Agent Class | Variable | Odds Ratio | Lower 95% Confidence Limit |
Upper 95% Confidence Limit |
p-value |
|---|---|---|---|---|---|
| Angiotensin II receptor blockers |
|||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.18 | 0.34 | 4.08 | 0.7936 | |
| 2006-2009 | 2.44 | 0.85 | 7.06 | 0.0988 | |
| Age | 0.99 | 0.97 | 1.01 | 0.4175 | |
| ACE Inhibitors | |||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.74 | 0.89 | 3.40 | 0.1028 | |
| 2006-2009 | 1.74 | 0.90 | 3.37 | 0.0995 | |
| Age | 0.99 | 0.98 | 1.01 | 0.5518 | |
| Beta blockers | |||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.36 | 0.62 | 2.98 | 0.4449 | |
| 2006-2009 | 1.52 | 0.73 | 3.15 | 0.2600 | |
| Age | 1.00 | 0.98 | 1.01 | 0.6920 | |
| Calcium channel blockers |
|||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 0.80 | 0.33 | 1.97 | 0.6333 | |
| 2006-2009 | 0.72 | 0.33 | 1.55 | 0.3983 | |
| Age | 1.01 | 0.99 | 1.03 | 0.4746 | |
| Non-thiazide diuretics |
|||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.66 | 0.69 | 4.01 | 0.2557 | |
| 2006-2009 | 1.43 | 0.69 | 2.94 | 0.3331 | |
| Age | 1.07 | 1.03 | 1.11 | 0.0002 | |
| Thiazide diuretics | |||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 0.41 | 0.13 | 1.34 | 0.1406 | |
| 2006-2009 | 1.64 | 0.59 | 4.53 | 0.3431 | |
| Age | 1.00 | 0.97 | 1.03 | 0.8606 | |
| Miscellaneous | |||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.51 | 0.24 | 9.44 | 0.6590 | |
| 2006-2009 | 1.08 | 0.28 | 4.11 | 0.9111 | |
| Age | 0.99 | 0.93 | 1.06 | 0.7531 | |
| Two or more classes (vs. one class) |
|||||
| 2000-2002 | Reference | --- | --- | --- | |
| 2003-2005 | 1.69 | 0.74 | 3.86 | 0.2114 | |
| 2006-2009 | 2.96 | 1.40 | 6.24 | 0.0044 | |
| Age | 1.01 | 1.00 | 1.03 | 0.1455 |
Table 5.
Age-adjusted logistic regression analyses of antihypertensive drug class prescription rates by provider type, NAMCS 2000-2009
| Agent Class | Variable | Odds Ratio | Lower 95% Confidence Limit |
Upper 95% Confidence Limit |
p-value |
|---|---|---|---|---|---|
| Angiotensin II receptor blockers |
|||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 0.75 | 0.41 | 1.38 | 0.3563 | |
| Age | 0.98 | 0.96 | 1.00 | 0.1158 | |
| ACE Inhibitors | |||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 1.26 | 0.77 | 2.04 | 0.3560 | |
| Age | 0.99 | 0.97 | 1.01 | 0.4827 | |
| Beta blockers | |||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 1.30 | 0.79 | 2.14 | 0.2989 | |
| Age | 0.99 | 0.97 | 1.01 | 0.5065 | |
| Calcium channel blockers |
|||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 0.97 | 0.55 | 1.71 | 0.9261 | |
| Age | 1.01 | 0.98 | 1.03 | 0.6248 | |
| Non-thiazide diuretics |
|||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 0.59 | 0.35 | 1.01 | 0.0540 | |
| Age | 1.08 | 1.03 | 1.14 | 0.0030 | |
| Thiazide diuretics | |||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 0.98 | 0.56 | 1.71 | 0.9308 | |
| Age | 0.99 | 0.96 | 1.02 | 0.4316 | |
| Miscellaneous | |||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 0.80 | 0.22 | 2.93 | 0.7375 | |
| Age | 1.00 | 0.92 | 1.09 | 0.9944 | |
| Two or more classes (vs. one class) |
|||||
| Neurologist | Reference | --- | --- | --- | |
| Primary care provider | 1.17 | 0.71 | 1.94 | 0.5371 | |
| Age | 1.00 | 0.98 | 1.02 | 0.8944 |
Results of additional analyses examining blood pressure levels in 2003-2005 vs. 2006-2009 showed no significant differences in the blood pressure levels measured during the visit by time period or physician type (supplementary table 2). Sensitivity analyses that explored time trends and physician specialty trends within just the time period of 2006-2009 show that there was no significant difference in antihypertensive drugs prescription rates across the5 year period, but similar to the overall 2000-2009 analysis, PCPs still had a significantly higher antihypertensive drug prescription rate than neurologists (supplementary table 3).
DISCUSSION
In this analysis of an administrative dataset comprising a sample of practice patterns among physicians based in the United States, it was observed that over the last decade, forvisits by patients aged≥ 40 years with a diagnosis of stroke there wasa significant rise in use of antihypertensive drugs from approximately one-third to one-half.This temporal increasecould have been due to progressive diffusion of clinical trial evidence and expert consensus guidelines on the use of antihypertensive drugs for recurrent stroke prevention into routine practice within ambulatory settings. However, we also noted that the number of medications of any type being prescribed in ambulatory setting significantly rose over the decade as well, so it could well be that the rise in antihypertensive drug prescriptions simply reflected secular trends. In support of this latter notion was the finding in the sensitivity analysis of a lack of increase in antihypertensive medications over the most recent 4 years in our study period. Still, it must be pointed out that the analysis of antihypertensive drug use among patients encountered across 2006 to 2009 may also have been limited by a lack of adequate statistical power. Even if one were to accept these data as reflecting at least in part evidence-practice diffusion, it would seem that there remains room for improvement in ensuring that virtually all known stroke patients without contraindications receive antihypertensive agents.
We also observed that visits by patients with a diagnosis of stroke to the neurologist were significantly less likely to prescription of antihypertensive drugs, compared to visits to the primary care physician. This result persisted after adjusting age, and was similarly noted in sensitivity analysis of pre-specified time period. Although, patients who visited PCPs were older and had more co-morbid conditions, it would still be expected that based on clinical trial evidence, the overwhelming majority of stroke patients aged≥40 years would benefit from being on an antihypertensive drug. When office blood pressure values were examined no significant differences were observed between PCP encounters and neurologist encounters suggesting that either blood pressure level during was not the main driver of the higher antihypertensive drug use seen among PCPs or patient blood pressure levels were higher in the PCP group thereby stimulating a higher treatment rate. While several studies have reported that neurologist’s management, compared with generalist’s management, is associated with improved patient outcomes among hospitalized stroke patients,18-22 and there are data on the influence of recent clinical trials on inpatient treatment rates,23 few comparisons have been conducted in the ambulatory care setting.24 Patients and neurologists seem to generally perceive that neurologic outpatient involvement adds value to care,25 but more objective data to support this notion are lacking.
It must also be noted that other strategies for mitigating future stroke risk like diet/nutrition education and counseling on exercise occurred more frequently during PCP visits. Again, it was apparent that patients seen by the PCPs had more comorbidities than those seen by neurologists by expert consensus guidelines mainly written academic neurologists advocate lifestyle modification counseling for patients with a diagnosis of stroke. Overall, the present data are in accord with results of a nationwide survey of neurologists in the United States that revealed that a majority of practicing neurologists defers traditional stroke risk factor control to primary care physicians.26 It is not immediately clear why neurologists may prefer to have PCPs manage these risk factors, and this retrospective analysis cannot shed more light on the reasons for this disparity in care. However, it is not inconceivable that neurologists who undergo only one year of general medicine training may feel less comfortable than PCPs in selecting, titrating, and monitoring antihypertensive drugs with their several classes, combinations, and potential systemic side effects.
The use of combinations of antihypertensive drug classes also rose significantly over the decade without any differences between PCPs and neurologists. This pattern may reflect greater acceptance among clinicians of the need to use combinations agents for more effective blood pressure control and minimization of side effects, and may be seen as a welcome trend.
This study has limitations. First, the survey data were not collected on a patient level therefore no patient level analysis could be performed. Second, there were slight differences in data collection across years (2000-2005 vs. 2006-2009), due to a change in overall classification being used to code drugs reported in NAMCS,which may result in a bias. However, we focused only on generic drug classifications, which would be expected to be consistent across years, and when we conducted a sensitivity analysis by examining individual years within the later epoch 2006-2009 we still found the same significant difference in antihypertensive drug visits between primary care physicians and neurologists.Third, NAMCS data are dependent on complete, accurate entry of clinical information by physicians, clinic staff, hospital staff, and Census Bureau representatives; and diagnoses in the NAMCS are based on the independent judgment of the prescribing physician rather than on research diagnostic interviews. Finally, no information was available in the NAMCS dataset regarding dosages, duration, and contraindications of medications.
Stroke is highly preventable and there is widespread agreement that the greatest single strategy for reducing the risk of recurrent stroke lies in blood pressure reduction via the use of antihypertensive therapy. Antihypertensive drug prescriptions for patient with a diagnosis of stroke are up, but there may be room for further improvement. In particular, neurologists might need to play a greater role when it comes to the use of antihypertensive drugs for their patients with a history of stroke. Future work should endeavor to pinpoint the reasons for this variation in physician practice patterns among patients with a history of stroke encountered in the ambulatory setting, and develop strategies to overcome it.
Supplementary Material
Acknowledgements
NIH-NINDS (U01 NS079179).
Role of the Sponsor:Not applicable.
Funding: Not applicable.
Footnotes
Author Contributions: Rema Raman PhD, Karin Ernstrom, MS and Daniela Markovic, MD had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors were involved in the final decision to submit the manuscript.
- Study concept and design: Ovbiagele.
- Acquisition of data:Not applicable.
- Analysis and interpretation of data: Ovbiagele, Raman, Ernstrom, Markovic
- Drafting of the manuscript: Ovbiagele
- Critical revision of the manuscript for important intellectual content:Ovbiagele, Raman, Ernstrom, Markovic
- Statistical analysis:Ernstrom, Markovic
- Obtained funding:Not applicable.
- Administrative, technical, or material support:Ovbiagele, Raman, Ernstrom, Markovic
- Study supervision:Ovbiagele, Raman
Financial Disclosures:
- Bruce Ovbiagele (None)
- Rema Raman (None)
- Karin Ernstrom (None)
- Daniela Markovic (None)
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statistics--2012 update: a report from the American Heart Association. Circulation. Jan 3;125(1):e2–e220. doi: 10.1161/CIR.0b013e31823ac046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Furie KL, Kasner SE, Adams RJ, et al. Guidelines for the prevention of stroke in patients with stroke or transient ischemic attack: a guideline for healthcare professionals from the american heart association/american stroke association. Stroke. Jan;42(1):227–276. doi: 10.1161/STR.0b013e3181f7d043. [DOI] [PubMed] [Google Scholar]
- 3.Hardie K, Hankey GJ, Jamrozik K, Broadhurst RJ, Anderson C. Ten-year survival after first-ever stroke in the perth community stroke study. Stroke. 2003 Aug;34(8):1842–1846. doi: 10.1161/01.STR.0000082382.42061.EE. [DOI] [PubMed] [Google Scholar]
- 4.Hardie K, Hankey GJ, Jamrozik K, Broadhurst RJ, Anderson C. Ten-year risk of first recurrent stroke and disability after first-ever stroke in the Perth Community Stroke Study. Stroke. 2004 Mar;35(3):731–735. doi: 10.1161/01.STR.0000116183.50167.D9. [DOI] [PubMed] [Google Scholar]
- 5.Moroney JT, Bagiella E, Paik MC, Sacco RL, Desmond DW. Risk Factors for Early Recurrence After Ischemic Stroke : The Role of Stroke Syndrome and Subtype. Stroke. 1998 Oct 1;29(10):2118–2124. doi: 10.1161/01.str.29.10.2118. 1998. [DOI] [PubMed] [Google Scholar]
- 6.Ovbiagele B. Pessin award lecture 2008: lessons from the Stroke PROTECT program. J Neurol Sci. 2008 Dec 15;275(1-2):1–6. doi: 10.1016/j.jns.2008.07.014. [DOI] [PubMed] [Google Scholar]
- 7.Wolf PA, Clagett GP, Easton JD, et al. Preventing ischemic stroke in patients with prior stroke and transient ischemic attack : a statement for healthcare professionals from the Stroke Council of the American Heart Association. Stroke. 1999;30(9):1991–1994. doi: 10.1161/01.str.30.9.1991. [DOI] [PubMed] [Google Scholar]
- 8.Hing E, Hall MJ, Xu J. National Hospital Ambulatory Medical Care Survey: 2006 outpatient department summary. Natl Health Stat Report. 2008 Aug 6;(4):1–31. [PubMed] [Google Scholar]
- 9.Centers for Disease Control and Prevention Web site . National Ambulatory Medical Care Survey (NAMCS) Centers for Disease Control and Prevention: National Center for Health Statistics; [Accessed September 2012]. Available from: http://www.cdc.gov/nchs/about/major/ahcd/ahcd1.htm#National%20Ambulatory%20Medical%20Care%20Survey%20(NAMCS) [Google Scholar]
- 10.Fang J, Alderman MH, Keenan NL, Ayala C, Croft JB. Hypertension control at physicians’ offices in the United States. Am J Hypertens. 2008 Feb;21(2):136–142. doi: 10.1038/ajh.2007.35. [DOI] [PubMed] [Google Scholar]
- 11.Hing E, Burt CW. Characteristics of office-based physicians and their practices: United States, 2003-04. Vital Health Stat. 2007 Jan;13(164):1–34. [PubMed] [Google Scholar]
- 12.Hing E, Burt CW. Office-based medical practices: methods and estimates from the national ambulatory medical care survey. Adv Data. 2007 Mar 12;(383):1–15. [PubMed] [Google Scholar]
- 13.Burt CW, Hing E. Making patient-level estimates from medical encounter records using a multiplicity estimator. Stat Med. 2007 Apr 15;26(8):1762–1774. doi: 10.1002/sim.2797. [DOI] [PubMed] [Google Scholar]
- 14.Roth PL, BeVier C. Response rates in HRM/OB survey research: Norms and correlates, 1990-1994. Journal of Management. 1998;24:97–117. [Google Scholar]
- 15.Goldstein LB. Accuracy of ICD-9-CM coding for the identification of patients with acute ischemic stroke: effect of modifier codes. Stroke. 1998 Aug;29(8):1602–1604. doi: 10.1161/01.str.29.8.1602. [DOI] [PubMed] [Google Scholar]
- 16.Roumie CL, Mitchel E, Gideon PS, Varas-Lorenzo C, Castellsague J, Griffin MR. Validation of ICD-9 codes with a high positive predictive value for incident strokes resulting in hospitalization using Medicaid health data. Pharmacoepidemiol Drug Saf. 2008 Jan;17(1):20–26. doi: 10.1002/pds.1518. [DOI] [PubMed] [Google Scholar]
- 17.Tirschwell DL, Longstreth WT., Jr Validating administrative data in stroke research. Stroke. 2002 Oct;33(10):2465–2470. doi: 10.1161/01.str.0000032240.28636.bd. [DOI] [PubMed] [Google Scholar]
- 18.Goldstein LB, Matchar DB, Hoff-Lindquist J, Samsa GP, Horner RD. VA Stroke Study: neurologist care is associated with increased testing but improved outcomes. Neurology. 2003 Sep 23;61(6):792–796. doi: 10.1212/01.wnl.0000082724.77447.3a. [DOI] [PubMed] [Google Scholar]
- 19.Gillum LA, Johnston SC. Influence of physician specialty on outcomes after acute ischemic stroke. J Hosp Med. 2008 May;3(3):184–192. doi: 10.1002/jhm.313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Horner RD, Matchar DB, Divine GW, Feussner JR. Relationship between physician specialty and the selection and outcome of ischemic stroke patients. Health Serv Res. 1995 Jun;30(2):275–287. [PMC free article] [PubMed] [Google Scholar]
- 21.Smith MA, Liou JI, Frytak JR, Finch MD. 30-day survival and rehospitalization for stroke patients according to physician specialty. Cerebrovasc Dis. 2006;22(1):21–26. doi: 10.1159/000092333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Mitchell JB, Ballard DJ, Whisnant JP, Ammering CJ, Samsa GP, Matchar DB. What role do neurologists play in determining the costs and outcomes of stroke patients? Stroke. 1996 Nov;27(11):1937–1943. doi: 10.1161/01.str.27.11.1937. [DOI] [PubMed] [Google Scholar]
- 23.Ovbiagele B, Schwamm LH, Smith EE, et al. Recent nationwide trends in discharge statin treatment of hospitalized patients with stroke. Stroke. Jul;41(7):1508–1513. doi: 10.1161/STROKEAHA.109.573618. [DOI] [PubMed] [Google Scholar]
- 24.Karve S, Balkrishnan R, Seiber E, Nahata M, Levine DA. Population Trends and Disparities in Outpatient Utilization of Neurologists for Ischemic Stroke. J Stroke Cerebrovasc Dis. Dec 22; doi: 10.1016/j.jstrokecerebrovasdis.2011.11.004. [DOI] [PubMed] [Google Scholar]
- 25.Narayanaswami P, Hahne JC. Outcome of outpatient neurology consultations: physician’s evaluations compared to patients’ perceptions. J Neurol Sci. 2009 Jan 15;276(1-2):175–178. doi: 10.1016/j.jns.2008.09.036. [DOI] [PubMed] [Google Scholar]
- 26.Ovbiagele B, Drogan O, Koroshetz WJ, Fayad P, Saver JL. Outpatient practice patterns after stroke hospitalization among neurologists. Stroke. 2008 Jun;39(6):1850–1854. doi: 10.1161/STROKEAHA.107.504860. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.