Abstract
Objective:
Determine the association of fragmented ambulatory healthcare with uncontrolled blood pressure (BP) and apparent treatment-resistant hypertension (aTRH) among older adults taking antihypertensive medication, overall and by race and gender.
Study Design:
Cross-sectional study using data from 2,868 REasons for Geographic and Racial Differences in Stroke (REGARDS) study participants ≥66 years of age who completed a study examination in 2013–2016, had Medicare fee-for-service coverage, and were taking antihypertensive medication.
Methods:
We used logistic regression to analyze the association of fragmented healthcare with uncontrolled BP and aTRH. Fragmented healthcare was operationalized as a reversed Bice-Boxerman Index ≥75th percentile, calculated using the number of ambulatory providers and healthcare visits in the year preceding the study examination. Uncontrolled BP was defined by systolic BP ≥140 mm Hg or diastolic BP ≥90 mm Hg. aTRH was defined by taking ≥3 classes of antihypertensive medication with uncontrolled BP or ≥4 classes with controlled BP.
Results:
The overall adjusted odds ratios (95%CI) for uncontrolled BP, aTRH with controlled BP, and aTRH with uncontrolled BP associated with fragmented healthcare were 1.10 (0.89–1.37), 1.08 (0.80–1.47), and 1.32 (0.96–1.81), respectively. Fragmented healthcare was not associated with uncontrolled BP or aTRH among white participants, women, or men. Among black participants, the odds ratio (95%CI) associated with fragmented healthcare was 1.21 (0.81–1.82) for uncontrolled BP, 1.22 (0.72–2.07) for aTRH with controlled BP, and 1.82 (1.07–3.11) for aTRH with uncontrolled BP.
Conclusions:
Fragmented healthcare may increase the likelihood of aTRH with uncontrolled BP among older black adults taking antihypertensive medication.
Keywords: hypertension, ambulatory care, Medicare, African Americans
Précis:
Among older black adults taking antihypertensive medication, fragmented ambulatory healthcare was associated with an increased likelihood of apparent treatment-resistant hypertension with uncontrolled blood pressure.
Introduction
Antihypertensive medication reduces the risk for cardiovascular disease (CVD) and all-cause mortality in adults with hypertension.1–3 However, 35.2% of US adults with hypertension taking antihypertensive medication have uncontrolled blood pressure (BP).4 In addition, 17.7% of US adults taking antihypertensive medication have apparent treatment-resistant hypertension (aTRH), defined as uncontrolled BP while taking ≥3 classes of antihypertensive medication or taking ≥4 classes of antihypertensive medication with controlled BP.5 Adults with aTRH have a higher risk of CVD and all-cause mortality compared to their counterparts with hypertension without aTRH.6,7
Several factors contribute to uncontrolled BP among adults taking antihypertensive medication, including medication non-adherence, an unhealthy diet, and low physical activity.8–10 Fragmented ambulatory healthcare, wherein a patient’s care is spread across many healthcare providers without a dominant provider, is common and may increase the likelihood of gaps in communication across providers caring for the same patient.11–13 Prior studies suggest that fragmented healthcare may increase the likelihood of uncontrolled BP among adults taking antihypertensive medication.14,15 Fragmented healthcare may also increase the likelihood that adults with hypertension are prescribed medicines that raise BP or counteract their antihypertensive medication, leading to aTRH.16,17 Among older US adults, uncontrolled BP is more frequent among those of black versus white race, and among women compared to men.18,19 Whether fragmented ambulatory healthcare differentially influences BP control by race and gender is unclear.
The aim of the current study was to determine the association of fragmented ambulatory healthcare with uncontrolled BP and aTRH among older US adults, overall and by race and gender. To accomplish this goal, we analyzed data from the Reasons for Geographic And Racial Differences in Stroke (REGARDS) study linked to Medicare claims.
Methods
Study population
The REGARDS study enrolled a population-based cohort of 30,239 black and white adults ≥45 years of age recruited from all 48 contiguous US states and the District of Columbia between January 2003 and October 2007. Black adults and adults residing in the Southeastern region of the US were oversampled by design.20 Participants completed a computer-assisted telephone interview (CATI) and an in-home study visit at baseline, and were followed through phone contacts to identify study outcomes.21–24 Study data from participants were linked to their claims in Medicare.25 Medicare is an insurance program that provides healthcare coverage to US adults ≥65 years of age and those <65 years of age who have disabilities or end-stage renal disease (ESRD). The REGARDS study was approved by Institutional Review Boards at the participating institutions, and all participants provided written informed consent.
We included REGARDS study participants who completed a second CATI and in-home study visit between May 2013 and November 2016 and:
Were ≥66 years of age.
Had continuous Medicare fee-for-service coverage for the 365 days preceding the date of their second in-home study visit, inclusive (i.e., the healthcare assessment period).
Did not have ESRD,26 defined by an ESRD indicator in Medicare.
Had ≥4 ambulatory care visits during their healthcare assessment period.
Had hypertension and were taking antihypertensive medication, defined by self-report in addition to having ≥1 antihypertensive medication class identified during a study medication inventory.
We excluded participants <66 years of age on their second in-home study visit (i.e., <65 years of age when the healthcare assessment period started) because this population may have qualified for Medicare on the basis of disability and have ambulatory care patterns that are sufficiently distinct to merit being considered separately.27 Participants with Medicare Advantage do not have fee-for-service coverage and were excluded because claims to assess ambulatory healthcare fragmentation were not available. We restricted the analysis to participants with ≥4 ambulatory visits as using fewer visits could produce unreliable estimates of healthcare fragmentation.12,28 Only 159 (5%) out of 3,027 participants with hypertension were excluded because they were not taking antihypertensive medication. After these criteria were applied, 2,868 REGARDS study participants taking antihypertensive medication were included in the current study (Figure 1).
Figure 1.
Flowchart of REGARDS study participants included the current analysis.
CATI: computer-assisted telephone interview; REGARDS: REasons for Geographic and Racial Differences in Stroke.
aMore black than white participants were excluded at this step. Of the 3,141 black participants with Medicare Parts A and B Coverage, 1,554 (49.5%) were excluded for having Medicare Advantage. Of the 6,268 white participants with Medicare Parts A and B Coverage, 1,989 (31.7%) were excluded for having Medicare Advantage.
bThe healthcare assessment period is defined by the 365 days prior to each participant’s second in-home study visit, inclusive.
cAmbulatory visits were defined using a modified version of the definition from the National Committee for Quality Assurance. Specifically, ambulatory visits were defined as an ambulatory or carrier file claim with a valid National Provider Identifier code and a Current Procedural Terminology code for an evaluation and management visit for adults that would take place in an office or clinic setting.
dAn outlier for the number of ambulatory visits or number of providers during the healthcare assessment period is defined by a value greater than 99.9th percentile of the distribution.
eHypertension was defined by a systolic blood pressure ≥140 mm Hg, or a diastolic blood pressure ≥90 mm Hg, or antihypertensive medication use. Antihypertensive medication use was defined by self-report of current use of medication to lower blood pressure during the second CATI and having ≥1 antihypertensive medication class in a medication inventory of all prescription and over-the-counter medications taken in the past 2 weeks conducted during the second in-home study visit.
Healthcare fragmentation
Healthcare fragmentation was assessed based on the number of ambulatory visits and ambulatory providers that each participant had during their healthcare assessment period (i.e., the year preceding each participant’s second in-home study visit) using the reversed Bice-Boxerman Index (rBBI; eAppendix 1).29 Ambulatory visits, including visits to federally qualified healthcare centers and rural clinics, were defined using a modified version of the definition by the National Committee for Quality Assurance, restricting the definition to evaluation-and-management visits in office or clinic settings.30 Ambulatory providers were defined by National Provider Identifier codes. The Bice-Boxerman Index captures the “dispersion” of care (the spread of care across multiple healthcare providers) and the “density” of care (the share of visits to each healthcare provider) received by patients.31 The rBBI ranges from 0, all visits with the same provider or no healthcare fragmentation, to 1, each visit with a different provider or maximum healthcare fragmentation. Fragmented ambulatory care was defined by an rBBI ≥75th percentile of the distribution in the overall study population, informed by prior analyses.32–34 All participants included in the analysis had at ≥4 ambulatory visits in their healthcare assessment period.
BP measurements and antihypertensive medication use
During each participant’s second in-home study visit, trained staff performed two BP measurements after a 5-minute rest using an aneroid sphygmomanometer, which were averaged for all analyses. Trained staff also conducted an inventory of medications taken in the previous two weeks, including antihypertensive medication classes (eAppendix 2). Following the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC7), we defined uncontrolled BP as a systolic BP (SBP) ≥140 mm Hg or a diastolic BP (DBP) ≥90 mm Hg. aTRH was defined as taking ≥3 classes of antihypertensive medication with uncontrolled BP (aTRH with uncontrolled BP), or taking ≥4 classes of antihypertensive medication with controlled BP (aTRH with controlled BP).
Participant characteristics
Participant characteristics defined using data collected at the REGARDS study baseline included gender, race, geographic region of residence, rural residence, and education. Participant characteristics defined using data collected at the second REGARDS study examination included age, annual household income, body mass index (BMI), current smoking, diabetes, chronic kidney disease (CKD), atrial fibrillation, total and high-density lipoprotein (HDL) cholesterol, statin use, and self-rated health. History of coronary heart disease (CHD), stroke, and heart failure were defined using data collected at baseline and the second in-home study visit supplemented with adjudicated outcome events. We provide definitions for participant characteristics in eAppendix 3 and a figure detailing when data used for the current analysis were collected in eAppendix 4.
Statistical analysis
We used descriptive statistics to estimate characteristics of participants overall and stratified by fragmented healthcare. We also calculated the prevalence of uncontrolled BP and aTRH overall and stratified by fragmented healthcare. We used two logistic regression models with progressive adjustment for covariates to calculate odds ratios (OR) and 95% confidence intervals (CI) for uncontrolled BP associated with fragmented healthcare. Model 1 included adjustment for age, gender, race, geographic region of residence, rural residence, annual household income, and education. Model 2 included adjustment for variables in Model 1 and BMI, smoking, diabetes, CKD, history of CHD, stroke or heart failure, atrial fibrillation, total and HDL cholesterol, statin use, and self-rated health. We also calculated the prevalence and ORs for uncontrolled BP associated with quartiles of the rBBI distribution. The analysis described above was repeated for the outcome of aTRH. In a separate analysis, we divided the study population into 3 mutually exclusive groups: those with no aTRH, aTRH with controlled BP, and aTRH with uncontrolled BP and used multinomial logistic regression models with adjustment for covariates as described above to calculate ORs and 95% CIs for for aTRH with controlled and uncontrolled BP, separately, versus no aTRH associated with fragmented healthcare.
We repeated the analysis described above by race and gender subgroups, separately. Because black adults have lower rBBI scores versus white adults, we repeated the analysis among black and white participants using race-specific 75th percentiles of the rBBI distribution to define fragmented healthcare.35 In exploratory analyses, we repeated the analyses described above excluding participants with an rBBI of 0 or 1, and examined effect modification by education, income and region of residence. We used chained equations to impute 10 datasets to keep participants with missing covariates in the regression models (eAppendix 5).36,37 For a secondary analysis, we repeated calculations described above using the definition of uncontrolled BP in the 2017 American College of Cardiology (ACC)/American Heart Association (AHA) BP guideline. Under this definition, uncontrolled BP was defined as an SBP ≥130 mm Hg or DBP ≥80 mm Hg.38 Analyses were conducted using SAS, Version 9.4 (SAS Institute, Inc, Cary, NC).
Results
Participant characteristics and healthcare utilization patterns
Participants with fragmented healthcare were more likely to be white and had higher education and income versus those without fragmented healthcare (Table 1). Compared to participants without fragmented healthcare, those with fragmented healthcare had more ambulatory visits, saw more ambulatory providers, and had a lower percentage of visits with the most commonly seen provider during the healthcare assessment period (Table 2).
Table 1.
Characteristics of REGARDS study participants included in the analysis.
| Characteristics | Overall (n=2,868) |
Without fragmented healthcarea (n=2,111) |
With fragmented healthcarea (n=757) |
P-value |
|---|---|---|---|---|
| Age, years, mean (SD) | 76.0 (6.5) | 76.1 (6.6) | 75.8 (6.2) | 0.23 |
| Women, n (%) | 1,524 (53.1) | 1,124 (53.2) | 400 (52.8) | 0.87 |
| Blacks, n (%) | 878 (30.6) | 688 (32.6) | 190 (25.1) | <0.01 |
| Geographic region of residence, n (%) | ||||
| Stroke buckleb | 770 (26.8) | 569 (27.0) | 201 (26.5) | 0.58 |
| Stroke beltc | 1,020 (35.6) | 744 (35.2) | 276 (36.5) | |
| Other US regions | 1,078 (37.6) | 798 (37.8) | 280 (37.0) | |
| Rural residence, n (%) | 359 (14.0) | 266 (14.0) | 93 (13.9) | 0.95 |
| Annual household income <$35,000, n (%) | 987 (40.6) | 749 (41.9) | 238 (36.8) | 0.02 |
| Less than high school education, n (%) | 218 (7.6) | 164 (7.8) | 54 (7.1) | 0.63 |
| Body mass index, n (%) | ||||
| <25 kg/m2 | 636 (22.3) | 474 (22.5) | 162 (21.5) | 0.28 |
| 25 to <30 kg/m2 | 1,080 (37.8) | 802 (38.2) | 278 (36.8) | |
| ≥30 kg/m2 | 1,141 (39.9) | 826 (39.3) | 315 (41.7) | |
| Current smoking, n (%) | 144 (5.1) | 111 (5.3) | 33 (4.4) | 0.38 |
| Diabetes, n (%) | 892 (32.2) | 677 (33.3) | 215 (29.3) | 0.05 |
| Chronic kidney disease, n (%) | 1,313 (49.6) | 981 (50.3) | 332 (47.7) | 0.25 |
| History of coronary heart disease, n (%) | 987 (34.4) | 711 (33.7) | 276 (36.6) | 0.15 |
| History of stroke, n (%) | 258 (9.0) | 195 (9.3) | 63 (8.4) | 0.51 |
| History of heart failure, n (%) | 296 (10.4) | 223 (10.7) | 73 (9.8) | 0.53 |
| Atrial fibrillation, n (%) | 534 (18.7) | 394 (18.7) | 140 (18.5) | 0.96 |
| Total cholesterol, mg/dL, mean (SD) | 171.7 (38.8) | 171.8 (39.0) | 171.4 (38.2) | 0.82 |
| HDL cholesterol, mg/dL, mean (SD) | 54.4 (17.3) | 54.6 (17.4) | 53.8 (17.1) | 0.28 |
| Statin use, n (%) | 1,721 (60.0) | 1,264 (59.9) | 457 (60.4) | 0.83 |
| Self-rated health, n (%) | ||||
| Excellent | 266 (9.4) | 186 (8.9) | 80 (10.6) | 0.09 |
| Very good/good | 2,065 (72.6) | 1,518 (72.5) | 547 (72.7) | |
| Fair/poor | 514 (18.1) | 389 (18.6) | 125 (16.6) |
HDL: high-density lipoprotein; REGARDS: REasons for Geographic and Racial Disparities in Stroke; SD: standard deviation.
Fragmented healthcare was defined by a reversed Bice-Boxerman Index score ≥75th percentile of the distribution in the overall study population. The 75th percentile of reverse Bice-Boxerman index scores among all participants included in the analysis was 0.89.
The stroke buckle includes the coastal plains region of North Carolina, South Carolina, and Georgia
The stroke belt includes the remaining parts of North Carolina, South Carolina and Georgia, and Alabama, Mississippi, Tennessee, Arkansas, and Louisiana.
Table 2.
Ambulatory healthcare utilization among REGARDS study participants included in the analysis.
| Overall (n=2,868) |
Without fragmented healthcarea (n=2,111) |
With fragmented healthcarea (n=757) |
P-value | |
|---|---|---|---|---|
| Number of ambulatory visits | ||||
| Mean (SD) | 12.8 (7.7) | 12.3 (7.6) | 14.3 (7.9) | <0.01 |
| Median (range) | 11.0 (4.0 – 59.0) | 10.0 (4.0 – 59.0) | 13.0 (4.0 – 48.0) | |
| 25th – 75th percentile | 7.0 – 16.0 | 7.0 – 16.0 | 8.0 – 19.0 | |
| Number of ambulatory providers | ||||
| Mean (SD) | 5.8 (2.9) | 4.9 (2.3) | 8.2 (2.9) | <0.01 |
| Median (range) | 5.0 (1.0 – 17.0) | 5.0 (1.0 – 15.0) | 8.0 (4.0 – 17.0) | |
| 25th – 75th percentile | 4.0 – 7.0 | 3.0 – 6.0 | 6.0 – 10.0 | |
| Proportion of visits with most commonly seen ambulatory provider | ||||
| Mean (SD) | 0.42 (0.17) | 0.47 (0.16) | 0.25 (0.06) | <0.01 |
| Median (range) | 0.38 (0.09 – 1.00) | 0.44 (0.20 – 1.00) | 0.25 (0.09 – 0.40) | |
| 25th – 75th percentile | 0.29 – 0.50 | 0.36 – 0.56 | 0.21 – 0.29 | |
| Reversed Bice-Boxerman Index | ||||
| Mean (SD) | 0.78 (0.16) | 0.73 (0.16) | 0.92 (0.03) | <0.01 |
| Median (range) | 0.82 (0.00 – 1.00) | 0.79 (0.00–0.88) | 0.91 (0.89 – 1.00) | |
| 25th – 75th percentile | 0.72 – 0.89 | 0.67 – 0.84 | 0.90 – 0.93 |
REGARDS: REasons for Geographic and Racial Disparities in Stroke; SD: standard deviation.
Fragmented healthcare was defined by a reversed Bice-Boxerman Index score ≥75th percentile of the distribution in the overall study population. The 75th percentile of reverse Bice-Boxerman index scores among all participants included in the analysis was 0.89.
Uncontrolled BP and aTRH
Overall, 18.4% of participants had uncontrolled BP (Table 3). Also, 16.4% of participants had aTRH, including 9.0% who had aTRH with controlled BP and 7.4% who had aTRH with uncontrolled BP. After adjustment, there was no evidence of a difference in the likelihood of uncontrolled BP (Model 2 OR 1.10 [95% CI 0.89–1.37]) or aTRH (Model 2 OR 1.18 [95% CI 0.93–1.49]) between participants with versus without fragmented healthcare. Also, there was no evidence of an association between fragmented healthcare and aTRH with controlled BP (Model 2 OR 1.08 [95% CI 0.80–1.47]) or aTRH with uncontrolled BP (Model 2 OR 1.32 [95% CI 0.96–1.81]). There was no evidence of a difference in the likelihood of uncontrolled BP, aTRH, aTRH with controlled BP, or aTRH with uncontrolled BP across quartiles of the rBBI distribution (eAppendix 6). After excluding participants with rBBI scores of 0 or 1, the OR for aTRH with uncontrolled BP associated with healthcare fragmentation after adjustment for the variables in Model 2 was 1.37 (95% CI 1.00–1.89, eAppendix 7).
Table 3.
Odds ratios for uncontrolled blood pressure and aTRH associated with healthcare fragmentation overall.
| Overall (n=2,868) |
Without fragmented healthcarea (n=2,111) |
With fragmented healthcarea (n=757) |
|
|---|---|---|---|
| Uncontrolled Blood pressure | |||
| Prevalence, N (%) | 528 (18.4) | 385 (18.2) | 143 (18.9) |
| Odds ratio (95% CI) | |||
| Model 1 | -- | 1 (ref) | 1.09 (0.88–1.35) |
| Model 2 | -- | 1 (ref) | 1.10 (0.89–1.37) |
| aTRH | |||
| Prevalence, N (%) | 469 (16.4) | 337 (16.0) | 132 (17.4) |
| Odds ratio (95% CI) | |||
| Model 1 | -- | 1 (ref) | 1.19 (0.95–1.49) |
| Model 2 | -- | 1 (ref) | 1.18 (0.93–1.49) |
| aTRH with controlled blood pressure | |||
| Prevalence, N (%) | 258 (9.0) | 189 (9.0) | 69 (9.1) |
| Odds ratio (95% CI) | |||
| Model 1 | -- | 1 (ref) | 1.10 (0.82–1.48) |
| Model 2 | -- | 1 (ref) | 1.08 (0.80–1.47) |
| aTRH with uncontrolled blood pressure | |||
| Prevalence, N (%) | 211 (7.4) | 148 (7.0) | 63 (8.3) |
| Odds ratio (95% CI) | |||
| Model 1 | -- | 1 (ref) | 1.32 (0.96–1.80) |
| Model 2 | -- | 1 (ref) | 1.32 (0.96–1.81) |
aTRH: apparent treatment-resistant hypertension; CI: confidence interval.
Uncontrolled blood pressure is defined as a systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg. Controlled blood pressure is defined as a systolic blood pressure <140 mm Hg and a diastolic blood pressure <90 mm Hg. aTRH is defined as taking ≥3 antihypertensive medications with uncontrolled blood pressure, or taking ≥4 antihypertensive medications regardless of blood pressure levels.
Model 1 adjusted for age, gender, race, geographic region of residence, rural residence, annual household income and education.
Model 2 adjusted for variables in Model 1 and body mass index, smoking, diabetes, chronic kidney disease, history of coronary heart disease, stroke or heart failure, atrial fibrillation, total and high-density lipoprotein cholesterol, statin use, and self-rated health.
Fragmented healthcare was defined by a reversed Bice-Boxerman Index score ≥75th percentile of the distribution in the overall study population. The 75th percentile of reverse Bice-Boxerman index scores among all participants included in the analysis was 0.89.
Subgroup analyses by race and gender
Among black participants, there was no evidence of an association of fragmented healthcare with uncontrolled BP (Model 2 OR 1.21 [95% CI 0.81–1.82]). Fragmented healthcare was associated with a higher likelihood of aTRH (Model 2 OR 1.51 [95% 1.01–2.26], Table 4) and aTRH with uncontrolled BP (Model 2 OR 1.82 [95% CI 1.07–3.11]) among black participants. There was no evidence of an association between fragmented healthcare and aTRH with controlled BP among black participants. Among white participants, there was no evidence of an association of fragmented healthcare with uncontrolled BP, aTRH, or aTRH with controlled or uncontrolled BP, separately (Table 4). Results among black and white participants using race-specific 75th percentiles of the rBBI distribution to define fragmented healthcare and excluding participants with rBBI scores of 0 or 1 are shown in eAppendices 8 and 9, respectively. Among black participants, fragmented healthcare was associated with aTRH with controlled BP among those with less than high school education or annual household income <$35,000, but not among those with a high school education or greater or annual household income ≥$35,000. (eAppendices 10 and 11). There was no evidence of an effect modification by region of residence among black participants (eAppendix 12) or by education, income, or region of residence among white participants (eAppendices 13–15).
Table 4.
Odds ratios for uncontrolled blood pressure and aTRH associated with healthcare fragmentation stratified by race.
| Black | White | |||
|---|---|---|---|---|
| Without fragmented healthcarea (n=688) |
With fragmented healthcarea (n=190) |
Without fragmented healthcarea (n=1,423) |
With fragmented healthcarea (n=567) |
|
| Uncontrolled Blood pressure | ||||
| Prevalence, N (%) | 138 (20.1) | 45 (23.7) | 247 (17.4) | 98 (17.3) |
| Odds ratio (95% CI) | ||||
| Model 1 | 1 (ref) | 1.28 (0.86–1.89) | 1 (ref) | 1.03 (0.79–1.33) |
| Model 2 | 1 (ref) | 1.21 (0.81–1.82) | 1 (ref) | 1.04 (0.80–1.35) |
| aTRH | ||||
| Prevalence, N (%) | 135 (19.6) | 51 (26.8) | 202 (14.2) | 81 (14.3) |
| Odds ratio (95% CI) | ||||
| Model 1 | 1 (ref) | 1.52 (1.04–2.23) | 1 (ref) | 1.07 (0.81–1.42) |
| Model 2 | 1 (ref) | 1.51 (1.01–2.26) | 1 (ref) | 1.08 (0.81–1.45) |
| aTRH with controlled blood pressure | ||||
| Prevalence, N (%) | 76 (11.1) | 25 (13.2) | 113 (7.9) | 44 (7.8) |
| Odds ratio (95% CI) | ||||
| Model 1 | 1 (ref) | 1.26 (0.77–2.09) | 1 (ref) | 1.02 (0.71–1.48) |
| Model 2 | 1 (ref) | 1.22 (0.72–2.07) | 1 (ref) | 1.07 (0.73–1.56) |
| aTRH with uncontrolled blood pressure | ||||
| Prevalence, N (%) | 59 (8.6) | 26 (13.7) | 89 (6.3) | 37 (6.5) |
| Odds ratio (95% CI) | ||||
| Model 1 | 1 (ref) | 1.84 (1.10–3.06) | 1 (ref) | 1.13 (0.75–1.69) |
| Model 2 | 1 (ref) | 1.82 (1.07–3.11) | 1 (ref) | 1.14 (0.75–1.72) |
aTRH: apparent treatment-resistant hypertension; CI: confidence interval.
Uncontrolled blood pressure is defined as a systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg. Controlled blood pressure is defined as a systolic blood pressure <140 mm Hg and a diastolic blood pressure <90 mm Hg. aTRH is defined as taking ≥3 antihypertensive medications with uncontrolled blood pressure, or taking ≥4 antihypertensive medications regardless of blood pressure levels.
Model 1 adjusted for age, gender, geographic region of residence, rural residence, annual household income and education.
Model 2 adjusted for variables in Model 1 and body mass index, smoking, diabetes, chronic kidney disease, history of coronary heart disease, stroke or heart failure, atrial fibrillation, total and high-density lipoprotein cholesterol, statin use, and self-rated health.
Fragmented healthcare was defined by a reversed Bice-Boxerman Index score ≥75th percentile of the distribution in the overall study population. The 75th percentile of reverse Bice-Boxerman index scores among all participants included in the analysis was 0.89.
There was no evidence of a difference in the association of fragmented healthcare with uncontrolled BP, aTRH, or aTRH with controlled or uncontrolled BP, separately, across gender subgroups (eAppendix 16). Results among men and women excluding participants with rBBI scores of 0 or 1 are shown in eAppendix 17. Fragmented healthcare was associated with aTRH and aTRH with controlled BP among women with less than high school education, but no association was present among those with a high school education or greater. Also, fragmented healthcare was associated with aTRH among women with annual household income <$35,000 but no association was present among those with annual household income ≥$35,000. There was no evidence of an effect modification by region of residence among women (eAppendices 18–20), or by education, income, or region of residence among men (eAppendices 21–23).
Secondary analysis following the 2017 ACC/AHA BP guideline
Overall, 45.6% of participants had uncontrolled BP as defined in the 2017 ACC/AHA BP guideline. According to this guideline, 22.8% of participants had aTRH, including 6.0% with aTRH and controlled BP and 16.8% with aTRH and uncontrolled BP. There was no evidence of a difference in the likelihood of uncontrolled BP, aTRH, or aTRH with controlled or uncontrolled BP, separately, among participants with versus without fragmented healthcare (eAppendix 24 and eAppendix 25). There was no evidence of an association of fragmented healthcare with uncontrolled BP, aTRH, or aTRH with controlled or uncontrolled BP, separately, among black or white participants (eAppendices 26 and 27), or men or women (eAppendix 28).
Discussion
In the current analysis of older adults taking antihypertensive medication, fragmented healthcare was not associated with uncontrolled BP or aTRH in the overall study population. Also, no association was present between fragmented healthcare and uncontrolled BP or aTRH, among men and women, and among white adults. However, fragmented healthcare was associated with a higher likelihood of aTRH with uncontrolled BP among black adults.
Prior studies have suggested an association between fragmented healthcare and BP control.14,15 In an analysis of 5,886 adults with hypertension in the Third National Health and Nutrition Examination Survey, participants who reported having a doctor or healthcare professional they usually saw were more likely to have controlled BP (multivariable-adjusted OR 2.29 [95% CI 1.74–3.02]).14 The higher likelihood of BP control associated with seeing one particular doctor or health professional was present among non-Hispanic white, non-Hispanic black and Mexican American participants. This prior analysis included participants who did not see a healthcare provider in the previous year. Therefore, some participants not seeing a particular healthcare provider may not have been receiving any healthcare, including antihypertensive medication, which may explain their lower likelihood of BP control. In the current study of 2,868 adults taking antihypertensive medication who had ≥4 ambulatory visits in the preceding 365 days, the multivariable-adjusted OR for uncontrolled BP associated with healthcare fragmentation was 1.10 (95% CI 0.89–1.37). Results from the current study suggest that fragmented healthcare is not associated with BP control among adults taking antihypertensive medication.
Black adults have lower healthcare fragmentation compared with white adults.35 Despite this, fragmented healthcare was associated with a higher likelihood of aTRH with uncontrolled BP among black participants, but not among white participants. Several factors could explain the association between fragmented healthcare and aTRH with uncontrolled BP. If providers caring for the same patient are unaware of the complete medication regimen the patient is taking, this could increase the likelihood that the patient is prescribed therapies that counteract their antihypertensive medication, necessitating the addition of more medications for BP control.11,16,17 Also, fragmented healthcare has been associated with lower medication adherence.39 Minimizing healthcare fragmentation may reduce the prevalence of aTRH among black adults.
In an exploratory analysis, fragmented care was associated with a higher likelihood of aTRH with controlled BP among black participants with less than high school education or annual household income <$35,000. Fragmented healthcare was also associated with a higher likelihood of aTRH and aTRH with controlled BP among women with less than high school education and with a higher likelihood of aTRH among women with annual household income <$35,000. Given the exploratory nature of these analyses, these results should be interpreted with caution, and need to be confirmed in future studies.
There are multiple reasons why patients may have fragmented care, including patient-, provider-, organization-, and environment-level factors. However, most causes of fragmented care are unrelated to medical needs (e.g., patient preference for geographic proximity, or the widespread availability of urgent care centers).11 Therefore, it may be possible to reduce healthcare fragmentation without affecting the care a patient is receiving. Among those with 12 ambulatory visits, to decrease fragmentation from an rBBI of 0.9 (the mean among participants with fragmented care) to an rBBI of 0.7 (the mean among participants without fragmented care), an intervention would need to increase the proportion of visits with the most frequently seen provider from 25% to 50%.40 Such a change may be feasible, but there have been very few interventions designed to change fragmentation of care. One previous pilot study decreased fragmentation by using alerts in the electronic health record to inform providers if they were seeing patients with fragmented care in real time; however, this study was conducted in a single clinic among pediatric patients and may not be generalizable to adults.41 More interventions need to be designed and evaluated to address patient-, provider-, organization-, and environment-level drivers of fragmentation.
Strengths and limitations
The current study has several strengths. BP was measured by trained staff following a standardized procedure. Participants’ healthcare fragmentation was assessed using a validated index and claims data.28,42 This study also has limitations. Results from the current study may not be generalizable to older adults without Medicare fee-for-service coverage. We used the definitions of uncontrolled BP and aTRH in the 2017 ACC/AHA BP guideline in the secondary analysis.38 However, this guideline was published after the data for the current study were collected. Future studies should examine the association of fragmented healthcare with BP control using data collected after this guideline was published. Lastly, we cannot exclude the possibility of reverse causation, i.e. that participants with uncontrolled BP were more likely to be referred to more doctors. However, we attempted to minimize that possibility by assessing participants’ ambulatory care in the 365 days preceding their BP measurement.
Conclusion
No association was present between fragmented healthcare and uncontrolled BP or aTRH in the overall study population of older adults taking antihypertensive medication. However, fragmented healthcare was associated with a higher likelihood of aTRH with uncontrolled BP among black participants. Results from the current study suggest that fragmented healthcare may be a risk factor for aTRH with uncontrolled BP among older black adults.
Supplementary Material
Take-Away Points:
In the current analysis including adults ≥66 years of age with Medicare fee-for-service coverage in a large population-based cohort who were taking antihypertensive medication in 2013–2016:
Fragmented healthcare was not associated with uncontrolled blood pressure (BP) or apparent treatment-resistant hypertension (aTRH) in the overall study population or among white participants, men, or women.
Among black participants, fragmented healthcare was associated with an increased odds of aTRH with uncontrolled BP (odds ratio 1.82, 95% CI: 1.07–3.11).
Among older black adults taking antihypertensive medication, fragmented healthcare may increase the likelihood of aTRH with uncontrolled BP.
Acknowledgements
The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org.
Source of Funding:
This Reasons for Geographic and Racial Differences in Stroke (REGARDS) study is supported by cooperative agreement U01 NS041588 co-funded by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institute on Aging (NIA), National Institutes of Health, Department of Health and Human Service. The ancillary study on healthcare fragmentation and cardiovascular outcomes was funded by the National Heart, Lung, and Blood Institute (NHLBI; R01 HL135199). The funding agencies did not have any role in the design and conduct of the study, the collection, management, analysis, and interpretation of the data, or the preparation or approval of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This manuscript was reviewed and approved by the REGARDS Executive Committee prior to submission, ensuring adherence to standards for describing the REGARDS study. The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at: https://www.uab.edu/soph/regardsstudy/.
Author Disclosures:
MMS, PM, and LDC receive research support from Amgen, Inc. PM also serves as a consultant for Amgen Inc. LMK serves as a consultant to Mathematica, Inc. CLC reports no disclosures.
References
- 1.Bundy JD, Li C, Stuchlik P, et al. Systolic Blood Pressure Reduction and Risk of Cardiovascular Disease and Mortality: A Systematic Review and Network Meta-analysis. JAMA Cardiol 2017; 2(7):775–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Brunström M, Carlberg B. Association of Blood Pressure Lowering With Mortality and Cardiovascular Disease Across Blood Pressure Levels: A Systematic Review and Meta-analysis. JAMA Intern Med 2018; 178(1):28–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ettehad D, Emdin CA, Kiran A, et al. Blood pressure lowering for prevention of cardiovascular disease and death: a systematic review and meta-analysis. Lancet 2016; 387(10022):957–967. [DOI] [PubMed] [Google Scholar]
- 4.Muntner P, Hardy ST, Fine LJ, et al. Trends in Blood Pressure Control Among US Adults With Hypertension, 1999–2000 to 2017–2018. JAMA 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Carey RM, Sakhuja S, Calhoun DA, Whelton PK, Muntner P. Prevalence of Apparent Treatment-Resistant Hypertension in the United States. Hypertension 2019; 73(2):424–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kaczmarski KR, Sozio SM, Chen J, Sang Y, Shafi T. Resistant hypertension and cardiovascular disease mortality in the US: results from the National Health and Nutrition Examination Survey (NHANES). BMC Nephrol 2019; 20(1):138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Irvin MR, Booth JN 3rd, Shimbo D, et al. Apparent treatment-resistant hypertension and risk for stroke, coronary heart disease, and all-cause mortality. J Am Soc Hypertens 2014; 8(6):405–413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Butler MJ, Tanner RM, Muntner P, et al. Adherence to antihypertensive medications and associations with blood pressure among African Americans with hypertension in the Jackson Heart Study. J Am Soc Hypertens 2017; 11(9):581–588.e585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cherfan M, Vallée A, Kab S, et al. Unhealthy behaviors and risk of uncontrolled hypertension among treated individuals-The CONSTANCES population-based study. Sci Rep 2020; 10(1):1925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fagard RH, Cornelissen VA. Effect of exercise on blood pressure control in hypertensive patients. Eur J Cardiovasc Prev Rehabil 2007; 14(1):12–17. [DOI] [PubMed] [Google Scholar]
- 11.Kern LM, Safford MM, Slavin MJ, et al. Patients’ and Providers’ Views on Causes and Consequences of Healthcare Fragmentation in the Ambulatory Setting: a Qualitative Study. J Gen Intern Med 2019; 34(6):899–907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kern LM, Seirup JK, Rajan M, Jawahar R, Miranda Y, Stuard SS. Extent of Health Care Fragmentation in Different Payer Populations: Evidence from the Hudson Valley of New York. Popul Health Manag 2018; 22(2):138–143. [DOI] [PubMed] [Google Scholar]
- 13.O’Malley AS, Reschovsky JD. Referral and consultation communication between primary care and specialist physicians: finding common ground. Arch Intern Med 2011; 171(1):56–65. [DOI] [PubMed] [Google Scholar]
- 14.He J, Muntner P, Chen J, Roccella EJ, Streiffer RH, Whelton PK. Factors associated with hypertension control in the general population of the United States. Arch Intern Med 2002; 162(9):1051–1058. [DOI] [PubMed] [Google Scholar]
- 15.Fisher M, Sloane P, Edwards L, Gamble G. Continuity of care and hypertension control in a university-based practice. Ethn Dis 2007; 17(4):693. [PubMed] [Google Scholar]
- 16.Elliott WJ. Drug interactions and drugs that affect blood pressure. J Clin Hypertens (Greenwich) 2006; 8(10):731–737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Guo J-Y, Chou Y-J, Pu C Effect of Continuity of Care on Drug-Drug Interactions. Med Care 2017; 55(8):744–751. [DOI] [PubMed] [Google Scholar]
- 18.Daugherty SL, Masoudi FA, Ellis JL, et al. Age-dependent gender differences in hypertension management. J Hypertens 2011; 29(5):1005–1011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Foti K, Wang D, Appel LJ, Selvin E. Hypertension Awareness, Treatment, and Control in US Adults: Trends in the Hypertension Control Cascade by Population Subgroup (National Health and Nutrition Examination Survey, 1999–2016). Am J Epidemiol 2019; 188(12):2165–2174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Howard VJ, Cushman M, Pulley L, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology 2005; 25(3):135–143. [DOI] [PubMed] [Google Scholar]
- 21.Luepker RV, Apple FS, Christenson RH, et al. Case Definitions for Acute Coronary Heart Disease in Epidemiology and Clinical Research Studies. Circulation 2003; 108(20):2543–2549. [DOI] [PubMed] [Google Scholar]
- 22.Thygesen K, Alpert JS, White HD. Universal Definition of Myocardial Infarction. Circulation 2007; 116(22):2634–2653. [DOI] [PubMed] [Google Scholar]
- 23.Prineas RJ, Crow RS, Zhang Z-M. The Minnesota code manual of electrocardiographic findings. Springer Science & Business Media; 2009. [Google Scholar]
- 24.Stroke−−1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989; 20(10):1407–1431. [DOI] [PubMed] [Google Scholar]
- 25.Xie F, Colantonio LD, Curtis JR, et al. Linkage of a Population-Based Cohort With Primary Data Collection to Medicare Claims: The Reasons for Geographic and Racial Differences in Stroke Study. Am J Epidemiol 2016; 184(7):532–544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.United States Renal Data System. 2017 USRDS annual data report: Epidemiology of kidney disease in the United States National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Chapter 9: Healthcare Expenditures for Persons with ESRD. 433–440. Bethesda, MD, 2017. Access date: March 25, 2020. [Google Scholar]
- 27.Cubanski J, Neuman T, Damico A. Medicare’s role for people under age 65 with disabilities. Kaiser Family Foundation, Aug 12, 2016. Access date: March 25, 2020 https://www.kff.org/medicare/issue-brief/medicares-role-for-people-under-age-65-with-disabilities/.
- 28.Nyweide DJ, Anthony DL, Bynum JPW, et al. Continuity of care and the risk of preventable hospitalization in older adults. JAMA Intern Med 2013; 173(20):1879–1885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bice TW, Boxerman SB. A quantitative measure of continuity of care. Med Care 1977; 15(4):347–349. [DOI] [PubMed] [Google Scholar]
- 30.Kern LM, Seirup JK, Casalino LP, Safford MM. Healthcare Fragmentation and the Frequency of Radiology and Other Diagnostic Tests: A Cross-Sectional Study. J Gen Intern Med 2017; 32(2):175–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jee SH, Cabana MD. Indices for continuity of care: a systematic review of the literature. Med Care Res Rev 2006; 63(2):158–188. [DOI] [PubMed] [Google Scholar]
- 32.Kern LM, Ringel JB, Rajan M, et al. Ambulatory Care Fragmentation and Subsequent Hospitalization: Evidence From the REGARDS Study. Med Care 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kern LM, Ringel JB, Rajan M, et al. Ambulatory Care Fragmentation and Incident Stroke. J Am Heart Assoc 2021; 10(9):e019036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kern LM, Rajan M, Ringel JB, et al. Healthcare Fragmentation and Incident Acute Coronary Heart Disease Events: a Cohort Study. J Gen Intern Med 2021; 36(2):422–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Kern LM, Rajan M, Colantonio LD, et al. Differences in ambulatory care fragmentation by race. BMC Health Serv Res 2021; 21(1):154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ 2009; 338:b2393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med 2011; 30(4):377–399. [DOI] [PubMed] [Google Scholar]
- 38.Whelton PK, Carey RM, Aronow WS, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension 2018; 71(6):e13–e115. [DOI] [PubMed] [Google Scholar]
- 39.Chen CC, Tseng CH, Cheng SH. Continuity of care, medication adherence, and health care outcomes among patients with newly diagnosed type 2 diabetes: a longitudinal analysis. Med Care 2013; 51(3):231–237. [DOI] [PubMed] [Google Scholar]
- 40.Kern LM, Seirup JK, Rajan M, Jawahar R, Stuard SS. Fragmented ambulatory care and subsequent emergency department visits and hospital admissions among Medicaid beneficiaries. Am J Manag Care 2019; 25(3):107–112. [PubMed] [Google Scholar]
- 41.Christakis DA, Wright JA. Can Continuity of Care Be Improved? Results From a Randomized Pilot Study. Ambul Pediatr 2004; 4(4):336–339. [DOI] [PubMed] [Google Scholar]
- 42.Pollack CE, Hussey PS, Rudin RS, Fox DS, Lai J, Schneider EC. Measuring Care Continuity: A Comparison of Claims-based Methods. Med Care 2016; 54(5):e30–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.