Abstract
We used electronic health records (EHRs) data from 5658 ambulatory chronic kidney disease (CKD) patients with hypertension and prescribed antihypertensive therapy to examine antihypertensive drug prescribing patterns, blood pressure (BP) control, and risk factors for resistant hypertension (RHTN) in a real‐world setting. Two‐thirds of CKD patients and three‐fourths of those with proteinuria were prescribed guideline‐recommended renoprotective agents including an angiotensin‐converting enzyme inhibitor (ACEI) or an angiotensin receptor blocker (ARB); however, one‐third were not prescribed an ACEI or ARB. CKD patients, particularly those with stages 1‐2 CKD, who were prescribed regimens including beta‐blocker (BB) + diuretic or ACEI/ARB + BB + diuretic were more likely to have controlled BP (<140/90 mm Hg) compared to those prescribed other combinations. Risk factors for RHTN included African American race and major comorbidities. Clinicians may use these findings to tailor antihypertensive therapy to the needs of each patient, including providing CKD stage‐specific treatment, and better identify CKD patients at risk of RHTN.
Keywords: antihypertensive drugs, blood pressure control, chronic kidney disease, combination therapy, drug prescribing, electronic health records, hypertension
1. INTRODUCTION
Hypertension is the second leading risk factor of chronic kidney disease (CKD) after diabetes. It occurs in the majority of people with CKD (67%‐92%)1, 2 and accounts for a third of end‐stage renal disease (ESRD) cases in the United States (US).3 Antihypertensive treatment can slow the progression of CKD to renal failure4, 5; however, the treatment of hypertension can be challenging among CKD patients because of the bidirectional cause and effect relationship between hypertension and CKD.6, 7 Additionally, hypertensive CKD patients often require treatment with multiple antihypertensive drugs.8, 9, 10, 11, 12
Numerous patient and clinician‐related factors contribute to the difficulty of treating hypertension in CKD patients, one of which is variability in adopting guideline‐recommended hypertension treatment practices.1, 13 Major clinical guidelines, including the American College of Cardiology (ACC)/American Heart Association (AHA) hypertension treatment guidelines and the Kidney Disease Improving Global Outcomes (KDIGO) guidelines, provide consensus recommendations about optimal antihypertensive therapy for CKD patients with regard to both BP‐lowering and renal protection.2, 14 These recommendations include angiotensin‐converting enzyme inhibitor (ACEI) or angiotensin receptor blocker (ARB) as first‐line therapy for CKD patients, particularly for those with proteinuria and/or reduced renal function, defined as estimated glomerular filtration rate (eGFR) <60 mL/min2, 14, 15 (Table 1). However, it is not known to what extent and which clinicians adopt guideline‐recommended or preferred antihypertensive therapy among CKD patients in the real‐world clinical setting. A few electronic health record (EHR)‐based studies have reported that clinicians primarily prescribe ACEI or ARB (hereafter ACEI/ARB) for CKD patients, although the prescribing frequency is modest; estimates have ranged from less than one‐half to three‐quarters of CKD patients prescribed these agents.7, 16, 17
Table 1.
Guideline‐recommended antihypertensive therapy among CKD patients
| Hypertension treatment guideline | Recommended initial monotherapy or combination therapy (CKD patients) |
|---|---|
| 2017 ACC/AHA2 |
|
| 2012 KDIGO14 |
|
| 2014 JNC815 |
|
| 2003 JNC740 |
|
ACEI, angiotensin‐converting enzyme inhibitor; AHA/ACC, American College of Cardiology/American Heart Association; ARB, angiotensin receptor blocker; BB, beta‐blocker; BP, blood pressure; CCB, calcium‐channel blocker; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; JNC, Joint National Committee; KDIGO, Kidney Disease Improving Global Outcomes; TD, thiazide or thiazide‐like diuretic.
Also, avoid combining ACEI or ARB with direct renin inhibitor or aldosterone receptor antagonist.
“Other” acceptable dual therapy combinations in CKD patients include BB + diuretic, ACEI/ARB + BB, alpha‐blocker/centrally‐acting alpha agonist + diuretic.
There is also evidence to suggest that tailoring antihypertensive therapy according to stage of CKD matters. For instance, thiazide diuretics are considered to be ineffective in patients with eGFR <30 mL/min, and one contributing factor is that, in more advanced CKD, there is less sodium reaching the distal convoluted tubule where thiazide diuretics act to reduce sodium reabsorption.14 However, some recent studies have suggested that thiazide diuretics are still effective in stage 4 CKD.18, 19 Additionally, clinicians are cautious about prescribing ACEI/ARB among patients with more advanced stages of CKD because of the risk of hyperkalemia associated with these agents in more advanced CKD.20, 21, 22 The clinical guidelines do not provide further information about tailoring therapy according to stage of CKD, with the exception of KDIGO, which briefly mentions that loop diuretics may be preferred over a thiazide/thiazide‐like diuretic for patients with stage ≥4 CKD (eGFR <30 mL/min).14
In the present study, we sought to gain a more comprehensive understanding of real‐world antihypertensive drug prescribing patterns, BP control status, and BP‐lowering effectiveness of different drug regimens among hypertensive patients with CKD. We also aimed to stratify these analyses by stage of CKD to better understand how prescribing decisions and the BP‐lowering ability of various antihypertensive drug regimens may vary by stage of CKD. Lastly, we sought to identify significant socio‐demographic and clinical risk factors associated with resistant hypertension (RHTN), defined as uncontrolled BP on three or more antihypertensive drugs, one being a diuretic, or controlled BP on four or more drugs.23 Since hypertensive CKD patients are more likely to have RHTN,24, 25 these findings may assist clinicians to provide more targeted risk management and CKD stage‐specific antihypertensive treatment for patients at high risk of adverse outcomes.
2. METHODS
2.1. Study population and data source
We used EHR data for all analyses. The entire EHR data used for analyses were obtained from the University of Florida (UF) Health Integrated Data Repository (IDR) (idr.ufhealth.org). The UF Institutional Review Board approved the EHR data acquisition and analysis.
The present study included ambulatory patients aged ≥18 years with hypertension and prescribed any antihypertensive medication and seen in the outpatient clinic at UF Health between August 1, 2013, and August 1, 2016 (study period). Of the patients diagnosed with hypertension, we identified patients with a CKD diagnosis. The process of study population screening, inclusion, and analysis is shown in Figure 1.
Figure 1.
Flow diagram showing selection of CKD study population. *In the analysis of RHTN (Analysis 3), the antihypertensive therapy was selected based on the treatment period in which the patient was on the maximum number of drug classes. anti‐HTN, antihypertensive; BP, blood pressure; CHF, congestive heart failure; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate (mL/min/1.73m2); HTN, hypertension; RF, renal failure; RHTN, resistant hypertension
We only included patients who were diagnosed with hypertension and CKD (and other comorbidities, if any) prior to the prescription order (during the study period) of antihypertensive drugs to increase the likelihood that the drug prescribing decisions were only informed by the patients’ pre‐existing conditions. We excluded patients with eGFR <15 mL/min, as well as those with a diagnosis of stage 5 CKD, ESRD/renal failure, dialysis, history of kidney transplant, and congestive heart failure (CHF) because we considered that these conditions would significantly confound the assessment of BP response and/or attenuate any differences in BP response due to differences in drug regimen. Also, patients with a diagnosis of allergy to ACEI/ARB, angioedema, or hyperkalemia were excluded.
2.2. Research design
2.2.1. Descriptive analyses
First, we conducted a descriptive study in which we described the CKD population in terms of their demographic and clinical characteristics, proportion of patients prescribed different drug regimens, and proportion with controlled BP by drug regimen, stratified by stage of CKD (stages 1‐4). We also described the frequency of various drug regimens in a subset of CKD patients with proteinuria. In additional analysis, we examined the frequency of prescription of different drug regimens by race, comorbidities, and number of diuretics, as well as examined the association between ACEI/ARB prescription and potassium level. We also evaluated the proportion of patients with controlled BP by race, number of diuretics (one vs more than one), source of antihypertensive drug prescription (nephrology vs other clinic), and treatment with fixed‐dose combination (FDC) pills.
2.2.2. Multivariable regression analyses
Second, we conducted a case‐control study, in which we tested the association of different mono‐ and combination therapy regimens (explanatory variables) with controlled vs uncontrolled BP (dependent variable) among all patients and in subgroups stratified by stage of CKD (stages 1‐2 and stages 3‐4); we also stratified these analyses by diabetes and obesity status, as well as race. In sensitivity analysis, we re‐examined the main findings of the study by comparing patients with stages 1‐3 of CKD against those with stage 4 of CKD.
In supplementary analysis, we examined the associations of different drug regimens with yearly percent eGFR change, calculated as ((eGFR at nearest post‐prescription follow‐up minus pre‐prescription eGFR) divided by pre‐prescription eGFR*100) divided by follow‐up duration in years.
We also conducted a case‐control study in which we examined the socio‐demographic and clinical factors (explanatory variables) associated with RHTN vs non‐RHTN (dependent variable). RHTN was defined as uncontrolled BP on ≥3 drugs, one being a diuretic, or controlled BP on ≥4 drugs (cases) vs controlled BP on 1‐3 drugs (controls).23, 26 This analysis was conducted among all CKD patients (stages 1‐4) and in subgroups stratified by stage of CKD.
2.2.3. Definition of variables
BP control (dependent variable)
In the primary multivariable regression analyses examining BP control by drug regimen, we defined controlled BP as systolic BP (SBP) <140 and diastolic BP (DBP) <90 mm Hg, consistent with the definition of hypertension at the time at which the EHR data were recorded in the clinic. Only office/clinic BP measurements were used in our study.
Combination therapy (explanatory variable)
Patients were considered to be on combination therapy if the prescription order dates of two or more drug classes overlapped by at least one month (≥28 days); for the RHTN analysis, we selected the treatment period in which patients were on the maximum number of concomitant drugs. Antihypertensive drugs were classified as ACEI, ARB, beta‐blocker (BB), calcium‐channel blocker (CCB), diuretic (thiazide/thiazide‐like, loop, potassium‐sparing), vasodilator, alpha‐1 blocker, centrally‐acting agent, direct renin inhibitor, and neprilysin inhibitor.
Definition of diagnoses
Relevant diagnoses were identified using International Classification of Diseases (ICD) 9th and 10th Revision diagnosis codes provided in the US Renal Data System 2017 Annual Data Report.27 Since the majority of CKD‐diagnosed patients were missing stage‐specific CKD diagnosis, we instead used eGFR measurements to determine stages of CKD14 as follows: ≥90 mL/min (stage 1), 60 to <90 mL/min (stage 2), 30 to <60 mL/min (stage 3), and 15 to <30 mL/min (stage 4).
Of note, we used diagnoses obtained from multiple sources to determine diagnoses for past medical history (professional billing, ICD‐9 and ICD‐10 diagnosis codes from hospital encounters, problem list). Only one episode of diagnosis was used to determine past medical history. However, when applicable, information on the use of corresponding medications or relevant laboratory values were also considered for diagnosis. For example, CKD diagnosis was based on ICD codes and eGFR information; hypertension was based on ICD code for hypertension and use of antihypertensive medications.
2.3. Statistical methods
In the descriptive analytic portions of the study, we used the chi‐square test, Student’s t test, or simple univariate regression analysis to examine the association between drug regimens and BP control and other variables of interest. We presented descriptive analytic results as numbers and percentages, means ± standard deviations, or as beta coefficients or odds ratios. We also used the Cochran‐Armitage trend test to assess the significance of correlation between the frequency of a treatment characteristic and stage of CKD progression.
In the multivariable regression analyses of BP control, we tested the association of different mono‐ and combination therapy regimens with controlled vs uncontrolled BP among all patients and in subgroups stratified by stage of CKD (stages 1‐2 and stages 3‐4). In sensitivity analysis, we also stratified these primary analyses by stages 1‐3 vs stage 4 of CKD. We tested the association between BP control (controlled vs uncontrolled BP) and different drug regimens using multivariable logistic regression analysis. Covariates that were tested for potential inclusion in the final multivariable regression model are listed in Table 2. Final covariates were first selected using univariate analysis (P < 0.20) followed by stepwise logistic regression modeling, with P < 0.20 as the criterion for variable entry and P < 0.05 as the criterion for variable retention in the model. The final regression model adjusted for age, sex, race (African American vs other), history of diabetes, history of ischemic heart disease (IHD), method of CKD stage determination (laboratory‐provided [74%] vs calculated [26%] eGFR), pre‐prescription SBP, treatment duration (duration between drug prescription order date and subsequent BP measurement date during the study period), and number of concurrent antihypertensive drugs. Additionally, in the analysis of the association between combination therapy regimens and BP control among patients prescribed two or more drugs, the regression model further adjusted for body mass index (BMI).
Table 2.
Patient characteristics by stage of CKD
| Patient characteristics |
All n = 5658 |
Stage of CKD | |||
|---|---|---|---|---|---|
|
Stage 1 eGFR ≥90 n = 665 |
Stage 2 eGFR 60 to <90 n = 1682 |
Stage 3 eGFR 30 to <60 n = 2817 |
Stage 4 eGFR 15 to <30 n = 494 |
||
| Age, y | 63.0 ± 14.9 | 51.3 ± 15.8 | 61.5 ± 13.6 | 66.6 ± 13.8 | 63.6 ± 15.3 |
| Male | 2926 (52) | 335 (50) | 951 (56) | 1420 (50) | 220 (44) |
| BMI, kg/m2 | 30.7 ± 7.2 | 31.3 ± 8.2 | 30.8 ± 7.1 | 30.4 ± 7.0 | 30.4 ± 6.9 |
| BMI ≥ 30 | 2588 (47) | 325 (50) | 801 (49) | 1240 (45) | 222 (47) |
| Race | |||||
| White | 3857 (68) | 386 (58) | 1163 (69) | 1980 (70) | 328 (66) |
| Black | 1492 (26) | 241 (36) | 438 (26) | 675 (24) | 138 (28) |
| Other | 309 (6) | 38 (6) | 81 (5) | 162 (6) | 28 (6) |
| Hispanic | 193 (3) | 25 (4) | 50 (3) | 96 (3) | 22 (4) |
| Disease | |||||
| IHD | 2882 (51) | 266 (40) | 847 (50) | 1527 (54) | 242 (49) |
| MI | 1079 (19) | 88 (13) | 315 (19) | 580 (21) | 96 (19) |
| Diabetes | 2869 (51) | 319 (48) | 750 (45) | 1519 (54) | 281 (57) |
| Stroke/TIA | 1496 (26) | 136 (20) | 408 (24) | 813 (29) | 139 (28) |
| Insurance | |||||
| Private | 1346 (24) | 221 (33) | 475 (28) | 558 (20) | 92 (19) |
| Medicaid | 541 (9) | 150 (23) | 160 (10) | 194 (7) | 37 (8) |
| Medicare | 3548 (63) | 259 (39) | 966 (57) | 1980 (70) | 343 (69) |
| Other | 48 (1) | 6 (1) | 19 (1) | 16 (1) | 7 (1) |
| Uninsured | 175 (3) | 29 (4) | 62 (4) | 69 (2) | 15 (3) |
BMI, body mass index; BP, blood pressure; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; IHD, ischemic heart disease; MI, myocardial infarction; TIA, transient ischemic attack.
Number of patients with percentage (%) in parentheses or mean ± SD shown; eGFR values in mL/min/1.73m2
In supplementary analysis examining the associations of different drug regimens with yearly percent eGFR change, we used multivariable regression model adjusted for the aforementioned covariates, with the exception of adjustment for pre‐prescription order eGFR instead of pre‐prescription order SBP.
In total, 5658 CKD patients were available for these aforementioned analyses. In all analyses, P < 0.05 was defined as the threshold for statistical significance.
In the analysis of RHTN, we used multivariable logistic regression analysis to identify significant socio‐demographic and clinical risk factors associated with RHTN among all CKD patients (stages 1‐4) and by stage of CKD (stages 1‐2 and stages 3‐4). We defined RHTN as uncontrolled BP on ≥3 drugs, one being a diuretic, or controlled BP on ≥4 drugs (cases); non‐RHTN was defined as controlled BP on 1‐3 drugs (controls).23, 26 We screened covariates using stepwise logistic regression modeling, applying the same variable entry and retention criteria as mentioned previously. The list of tested covariates is provided in Supplementary Table S1. Age and BMI were entered into the model as categorical variables, with age in 10‐year increments and BMI in 5‐unit (kg/m2) increments. In total, 4459 CKD patients with RHTN or non‐RHTN were available for the RHTN analysis.
3. RESULTS
3.1. Results of descriptive analyses: study population and prescribing frequency and BP control by drug regimen
Overall, the average age of the sample population was 63.0 ± 14.9 years; one‐half was male. White, African American/Black, and other/unknown race groups comprised 68%, 26%, and 6% of patients, respectively. Diabetes, IHD, and stroke/transient ischemic attack (TIA) were a listed diagnosis in 51%, 51%, and 26% of CKD patients, respectively (Table 2).
CKD patients were, on average, prescribed 2.31 ± 1.08 drug classes, with a greater number of drugs prescribed among patients with more advanced stages of CKD. The most frequently prescribed drug class was ACEI/ARB (64%). With the exception of ACEI/ARB and thiazide and potassium‐sparing diuretics, the proportion of patients prescribed all other drug classes increased with progression of CKD (Table 3). Additionally, the frequency of treatment with more than one diuretic type was 10%; this frequency did not change across CKD stages (P = 0.115).
Table 3.
Antihypertensive therapy characteristics by stage of CKD
| Treatment characteristics | All | Stage of CKD | P‐trendb | |||
|---|---|---|---|---|---|---|
|
Stage 1 eGFR ≥90 |
Stage 2 eGFR 60 to <90 |
Stage 3 eGFR 30 to <60 |
Stage 4 eGFR 15 to <30 |
|||
| n = 5658 | n = 665 | n = 1682 | n = 2817 | n = 494 | ||
| No. drugs | ||||||
| 1 drug | 1440 (25) | 215 (32) | 490 (29) | 630 (22) | 105 (21) | <0.001 |
| 2 drugs | 2066 (36) | 273 (41) | 582 (35) | 1027 (36) | 184 (37) | <0.001 |
| 3 drugs | 1350 (24) | 128 (19) | 406 (24) | 718 (25) | 98 (20) | <0.001 |
| ≥4 drugs | 802 (14) | 49 (7) | 204 (12) | 442 (16) | 107 (22) | <0.001 |
| No. drugs, mean | 2.31 ± 1.08 | 2.04 ± 0.96 | 2.22 ± 1.05 | 2.39 ± 1.09 | 2.52 ± 1.24 | <0.001 |
| Drug classes | ||||||
| ACEI/ARB | 3642 (64) | 437 (66) | 1122 (67) | 1835 (65) | 248 (50) | <0.001 |
| BB | 3119 (55) | 304 (46) | 868 (52) | 1642 (58) | 305 (62) | <0.001 |
| CCB | 2337 (41) | 235 (35) | 661 (39) | 1199 (43) | 242 (49) | <0.001 |
| dCCB | 1990 (35) | 197 (30) | 565 (34) | 1023 (36) | 205 (41) | <0.001 |
| ndCCB | 345 (6) | 37 (5) | 95 (6) | 176 (6) | 37 (7) | 0.140 |
| Diuretic | 2959 (52) | 318 (48) | 863 (51) | 1506 (53) | 272 (55) | <0.001 |
| Thiazidea | 1371 (24) | 176 (26) | 484 (29) | 635 (22) | 76 (15) | <0.001 |
| Loop | 1720 (30) | 146 (22) | 407 (24) | 953 (34) | 214 (43) | <0.001 |
| K+ ‐sparing/MRA | 490 (9) | 65 (10) | 172 (10) | 230 (8) | 23 (5) | 0.001 |
| Other | 811 (14) | 53 (8) | 170 (10) | 448 (16) | 140 (28) | <0.001 |
ACEI, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, beta‐blocker; BP, blood pressure; CCB, calcium‐channel blocker; CKD, chronic kidney disease; dCCB, dihydropyridine CCB; eGFR, estimated glomerular filtration rate; MRA, mineralocorticoid/aldosterone receptor antagonist; ndCCB, non‐dihydropyridine CCB.
“Other” drugs included alpha‐blockers, vasodilators, centrally‐acting agents, direct renin inhibitors, and neprilysin inhibitors.
Number of patients with percentage (%) shown in parentheses or mean ± SD shown; eGFR values in mL/min/1.73m2.
Thiazide also includes thiazide‐like diuretics.
P‐values based on the Cochran‐Armitage trend test.
In additional analysis, we found that patients who were prescribed ACEI/ARB were more likely to have higher baseline or pre‐prescription potassium level compared to those who were not prescribed ACEI/ARB, regardless of concurrent diuretic use; after adjusting for diuretic therapy, ACEI/ARB therapy was associated with beta = 0.076 mEq/L higher potassium level (P < 0.0001). Also, there was no difference in the proportion of patients treated with ACEI/ARB between patients with hyperkalemia (>5.5 mEq/L) and those without hyperkalemia (≤5.5 mEq/L; 70% vs 64%, P = 0.450).
We also examined whether there are any racial differences in the frequency of treatment with different antihypertensive medications. We found that a higher proportion of African Americans were prescribed BB, CCB, and diuretic compared to non‐African Americans, with no difference by race in the prescribing of ACEI or ARB. With regard to the prescribing of two‐drug combination therapy, African Americans were less likely to be prescribed BB + diuretic compared to other race groups (11% vs 16%, P = 0.016); there was no difference by race in the prescribing of the other two‐drug combinations. Moreover, African Americans were less likely to be prescribed ACEI/ARB + BB + diuretic combination compared to others (37% vs 43%, P = 0.047) but more likely to be prescribed ACEI/ARB + CCB + diuretic compared to other race groups (24% vs 17%, P = 0.006). It is also worth noting that, in general, African Americans were more likely to be treated with three or more medications than others; African Americans were prescribed on average 2.6 medications while others were prescribed 2.2 medications (P < 0.0001).
We next examined the proportion of patients with controlled BP according to the number of concomitant drugs and stage of CKD. Among all CKD patients, 65% had SBP <140 and DBP <90 mm Hg. Patients with more advanced stages of CKD were less likely to have controlled BP than those with earlier stages of CKD (Table 4).
Table 4.
Proportion of patients with controlled BP by number of antihypertensive drugs and stage of CKD
| Proportion of patients with controlled BP | All | BP control by stage of CKD | P‐trenda | |||
|---|---|---|---|---|---|---|
|
Stage 1 eGFR ≥90 |
Stage 2 eGFR 60 to <90 |
Stage 3 eGFR 30 to <60 |
Stage 4 eGFR 15 to <30 |
|||
| Overall | ||||||
| <140/90 mm Hg | 3691 (65) | 459 (69) | 1128 (67) | 1799 (64) | 305 (62) | 0.008 |
| No. of drugs | ||||||
| 1 drug | 989 (69) | 154 (72) | 337 (69) | 427 (68) | 71 (68) | 0.330 |
| 2 drugs | 1335 (65) | 189 (69) | 385 (66) | 646 (63) | 115 (62) | 0.035 |
| 3 drugs | 903 (67) | 87 (68) | 272 (67) | 486 (68) | 58 (59) | 0.414 |
| ≥4 drugs | 464 (58) | 29 (59) | 134 (66) | 240 (54) | 61 (57) | 0.086 |
| b P‐trend | <0.0001 | 0.099 | 0.406 | <0.0001 | 0.039 | |
BP, blood pressure; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.
Number of patients with percentage (%) shown in parentheses; eGFR values in mL/min/1.73m2.
Sample sizes provided in Table 3.
P‐values based on the Cochran‐Armitage trend test across stages of CKD.
P‐values based on the Cochran‐Armitage trend test across number of drugs (within each stage of CKD).
In analysis of BP control stratified by race, among non‐African Americans, patients who were prescribed diuretic were more likely to have controlled BP than those not on a diuretic (69% vs 66%, P = 0.022); there was no difference by BB therapy. Additionally, among non‐African Americans patients, those treated with BB + diuretic were more likely to have controlled BP than those not on this combination (71% vs 66%, P = 0.006). Among African Americans, treatment with a BB and/or diuretic was not associated with BP control.
We also found that there was no significant difference in the proportion of patients with controlled BP according to treatment with one vs more than one diuretic as part of combination therapy (64% vs 65%, P = 0.918). In analysis stratified by race (African Americans vs other), comorbidities (diabetes, IHD) as well as CKD stage, there was no difference in BP control between the aforementioned two diuretic groups (African Americans: 58% vs 56%, P = 0.683; Non‐African Americans: 67% vs 68%, P = 0.710; Diabetic: 62% vs 64%, P = 0.722; Non‐diabetics: 66% vs 66%, P = 0.865; IHD: 65% vs 72%, P = 0.088; Non‐IHD: 63% vs 55%, P = 0.086; CKD stages 1‐2:68% vs 64%, P = 0.403; CKD stages 3‐4:62% vs 65%, P = 0.456).
With regard to treatment with a FDC pill, during the study period, 3% of patients were prescribed a FDC; there was no difference in BP control between patients who were prescribed FDC vs those not prescribed FDC (63% vs 65%, P = 0.546).
3.2. Results of multivariable regression analyses: blood pressure control by drug regimen
We found that among patients with stages 1‐2 CKD prescribed ≥2 drugs, those who were prescribed the combination BB + diuretic were more likely to have controlled BP compared to those who were not prescribed this combination (OR: 1.43, 95% CI: 1.03‐2.00, P = 0.034) (Figure 2, panel A); this finding largely remained after further adjusting for type of BB and diuretic (alpha/beta‐blockers vs beta‐blockers, and thiazide vs other diuretics, P = 0.056). Moreover, although the analysis adjusted for history of diabetes and BMI, in additional analysis stratified by diabetes and obesity status, the aforementioned finding was primarily observed among non‐diabetic (P = 0.018) and non‐obese (P = 0.007) patients. Lastly, among all CKD patients (stages 1‐4), we found a slight trend toward association between controlled BP and BB + diuretic therapy (P = 0.103; Supplementary Figure S1, panel A).
Figure 2.
Association between combination therapy regimens and BP control among patients with stages 1‐2 CKD. A, patients on ≥2 drugs, B, patients on ≥3 drugs. Reference group in each regression model is all other combinations excluding the combination indicated; models adjusted for age, sex, race, treatment duration, method of CKD stage determination, history of diabetes, history of IHD, pre‐prescription SBP, and number of antihypertensive medications; model in panel A further adjusted for BMI; Sample sizes: Panel A (n = 2132), Panel B (n = 1859). ACEI, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BB, beta‐blocker; BMI, body mass index; BP, blood pressure; CCB, calcium‐channel blocker; CKD, chronic kidney disease; DIUR, diuretic; IHD, ischemic heart disease
Among patients with stages 1‐2 CKD prescribed ≥3 drugs, those who were prescribed the combination ACEI/ARB + BB + diuretic showed a trend toward association with controlled BP compared to those who were not prescribed this combination (OR: 1.36, 95% CI: 0.94‐1.98, P = 0.100; Figure 2, panel B). We did not find significant associations between antihypertensive therapy regimens and BP control among patients with stages 3‐4 CKD (Supplementary Figure S2) and patients on monotherapy (Supplementary Figure S3).
In analysis examining the association between different drug regimens and eGFR change, we found that among patients on ≥2 drugs, those prescribed ACEI/ARB + diuretic or BB + CCB were more likely to have a higher eGFR compared to those prescribed other combinations. The ACEI/ARB + diuretic and BB + CCB combinations were associated with a yearly average increase in eGFR of 31% (P = 0.018) and 33% (P = 0.019), respectively; this finding was primarily observed among patients with stages 3‐4 CKD. We did not find significant associations between different drug regimens and eGFR change among patients on monotherapy or among patients on ≥3 drugs.
3.3. Results of sensitivity analysis
We conducted sensitivity analysis in which we re‐examined the primary findings of the study with analysis stratified by stages 1‐3 vs stage 4 CKD. Among patients on ≥2 drugs with stages 1‐3 CKD, treatment with BB + diuretic was not associated with controlled BP (OR: 0.90, 95% CI: 0.68‐1.19, P = 0.465). Among patients on ≥2 drugs with stage 4 CKD, those on BB + diuretic showed a slight toward having controlled BP than those not on BB + diuretic (OR: 2.44, 95% CI: 0.96‐6.25, P = 0.062). Among patients on ≥3 drugs, treatment with ACEI/ARB + BB + diuretic was not associated with BP control (among patients with stages 1‐3 CKD, OR: 0.91, 95% CI: 0.66‐1.26, P = 0.583; and stage 4 CKD, OR: 1.14, 95% CI: 0.37‐3.44, P = 0.831).
3.4. Results of resistant hypertension analyses
In the analysis of RHTN, among all CKD patients (stages 1‐4), the risk factors that were significantly associated with RHTN included African American/Black race, older age, higher BMI, and history of diabetes, IHD, left ventricular hypertrophy (LVH), ischemic stroke, and obstructive sleep apnea (OSA; Figure 3). In stage‐specific analyses, we identified similar risk factors among patients with stages 1‐2 and stages 3‐4 CKD (Supplementary Figure S4, panels A and B, respectively) with a few exceptions; notably, peripheral vascular disease (PVD) and angina were only significant among patients with stages 1‐2 CKD, and TIA was only significant among patients with stages 3‐4 CKD. Overall, African American race, BMI, and history of LVH were consistently associated with RHTN regardless of stage of CKD.
Figure 3.
Risk factors significantly associated with RHTN among CKD patients: results of multivariable logistic regression analysis. All covariates tested in univariate and multivariate analyses are listed in Supplementary Table S1. Sample sizes: n = 4459 (1241 RHTN cases, 3218 non‐RHTN controls). BMI, body mass index; CKD, chronic kidney disease; IHD, ischemic heart disease; LVH, left ventricular hypertrophy; RHTN, resistant hypertension
4. DISCUSSION
In this study, we evaluated antihypertensive therapy prescribing patterns and the frequency of controlled BP among real‐world hypertensive CKD patients using EHR data. We also explored the associations of different antihypertensive drug regimens with BP control and found that some drug regimens are associated with a greater likelihood of controlled BP compared to other drug regimens, with differences by stage of CKD. Lastly, we identified significant socio‐demographic and clinical risk factors of RHTN according to stage of CKD.
Overall, the majority of CKD patients were prescribed the preferred antihypertensive drug class, namely an ACEI/ARB (64%), with a higher proportion of those with proteinuria prescribed these agents (73%). This finding is consistent with major clinical guidelines, including the ACC/AHA hypertension treatment guideline and KDIGO, which recommend ACEI/ARB for hypertensive CKD patients, particularly those with proteinuria and/or stage ≥3 CKD (eGFR < 60 mL/min) based on the well‐established cardiovascular and renal benefits of these agents.2, 14 Conversely, these findings also suggest that one‐third of CKD patients were not prescribed an ACEI/ARB, including those with proteinuria.2, 14 In addition, patients with more advanced CKD were less likely to have been prescribed ACEI/ARB compared to those with earlier stage of CKD. This may reflect clinicians’ concerns about the risk of hyperkalemia associated with these agents among advanced CKD patients20, 21, 22; however, most experts recommend that the benefits of ACEI/ARB therapy outweigh the risks.28 Indeed, its discontinuation in hypertensive CKD patients may lead to worse outcomes, as reported by one EHR‐based study.29 Collectively, our findings highlight a potential need to increase awareness among clinicians about the importance of ACEI/ARB therapy among CKD patients, coupled with close electrolyte monitoring in those with more advanced CKD. Moreover, thiazide/thiazide‐like diuretic therapy was less frequent among patients with more advanced stages of CKD, consistent with general thoughts that thiazides are less effective as eGFR declines below 30 mL/min.14 However, some experts argue that thiazide diuretics can still be useful in advanced CKD by providing additional diuresis and long‐term BP‐lowering benefits18; indeed, combining thiazide and loop diuretics may be the most optimal diuretic strategy in this subgroup.30 In our study, we did not find a significant difference between treatment with one vs more than one diuretic with regard to BP control.
In supplemental analysis on the association between potassium level and prescription of ACEI/ARB, we found that patients who were prescribed ACEI/ARB were more likely to have higher baseline or pre‐prescription potassium level compared to those who were not prescribed ACEI/ARB, regardless of concurrent diuretic use. However, it must be noted that, since our study is essentially cross‐sectional, we do not have a true baseline or pre‐prescription potassium level because not all patients with a prescription for ACEI/ARB were ACEI/ARB naïve at the time that the baseline potassium level was measured; thus, caution is warranted in interpreting the role of potassium level in guiding the prescription of ACEI/ARB.
Our study also explored the association of different antihypertensive drug regimens with BP control in multivariable logistic regression analysis. Among patients with stages 1‐2 CKD prescribed ≥2 drugs, those who were prescribed BB + diuretic were more likely to have controlled BP compared to those who were prescribed other combinations. Similarly, among patients with stages 1‐2 CKD prescribed ≥3 drugs, those who were prescribed ACEI/ARB + BB + diuretic showed a trend toward controlled BP compared to those who were prescribed other combinations. Thus, while BB + diuretic has fallen out of favor in recent years,2, 15 our findings suggest that this combination could be effective at controlling BP in hypertensive CKD patients, though use of ACEI/ARB remains critical for its renoprotective effects. Given the importance of the sympathetic nervous system in hypertension and CKD pathogenesis and given the increased frequency of adverse cardiovascular outcomes observed in CKD patients, the use of BB, particularly the newer vasodilating BB, could be an important adjunct therapy for hypertension that provides additional cardiorenal protection in CKD patients.31, 32 Furthermore, the above findings may be especially useful for CKD patients who need ≥3 drugs to achieve BP control. The clinical guidelines recommend ACEI/ARB in nearly all CKD patients; however, among patients who need ≥3 drugs, adding BB + diuretic instead of BB + CCB or CCB + diuretic to an existing ACEI/ARB regimen may lead to the most BP‐lowering, particularly for patients with stages 1‐2 CKD. Lastly, the lack of significant associations between drug regimens and BP control among stages 3‐4 patients suggests that controlling high BP in advanced CKD is more difficult, regardless of regimen, and will require more personalized treatment strategies.
We also examined BP control by number of drugs and stage of CKD. Overall, more than two‐thirds of CKD patients had controlled BP. This proportion was slightly reduced in more advanced CKD, consistent with the conjecture that controlling high BP becomes more difficult with CKD progression. It is possible that involving specialist‐provided care (nephrology or hypertension specialist) may lead to greater BP control,7, 33, 34 particularly among those with advanced CKD. In our study, we did not find significant differences in BP control proportions by source of antihypertensive drug prescription (nephrology vs other clinic).
In supplementary analysis, we also examined the association between different antihypertensive drug regimens and eGFR change. Among patients on ≥2 drugs, those prescribed ACEI/ARB + diuretic or BB + CCB were more likely to have improved eGFR compared to those prescribed other combinations. This finding corroborates the standing of ACEI/ARB + diuretic as a guideline‐recommended combination for CKD patients and provides new evidence that BB + CCB may be a renoprotective combination; thus, it could potentially be used as an add‐on to an ACEI/ARB, which already has well‐demonstrated renoprotective benefits.
Our study results also indicate that there are racial differences in the prescribing of antihypertensive therapy. A higher proportion of African Americans were prescribed BB, CCB, and diuretic compared to non‐African Americans; there was no difference by race in the prescribing of ACEI or ARB. African Americans were less likely to use BB + diuretic or ACEI/ARB + BB + diuretic combination and more likely to use ACEI/ARB + CCB + diuretic compared to other race groups. Also, they were more likely to be treated with three or more medications than others. These increased use of CCB and need for more medications to treat hypertension among African Americans are consistent with the literature and hypertension treatment guidelines.2, 15 Demographic, clinical, patient and prescriber related factors all contribute to racial differences in prescribing patterns and BP control in real‐world hypertensive patients.2, 35, 36
We also examined the socio‐demographic and clinical determinants of RHTN according to stage of CKD. In all patients, the strongest determinant of RHTN was race, where African Americans had a nearly twofold greater odds of RHTN compared to other race groups. African American race, higher BMI, and history of LVH were consistently associated with RHTN regardless of stage of CKD and level of BP control. These and the other significant risk factors that we identified are consistent with those reported in previous studies of RHTN among people with various cardiovascular risk factors,23 including renal impairment.12, 37 Previous studies of RHTN among CKD patients have also identified male sex and PVD as important correlates of RHTN.37 In our study, PVD was associated with RHTN among stages 1‐2 CKD patients only.
The strengths of our study include the use of a rich EHR dataset that contained information on a variety of patient demographic and clinical characteristics. We were able to conduct associations with BP, which is not possible with other observational data, for example, claims data. We also were able to examine the association of BP control with not only major drug classes, as is typically the case in RCTs, but also with third/fourth‐line agents. Moreover, the wide variety of patient characteristics available for analysis allowed us to study the real‐world correlates of RHTN in greater detail.
The limitations of the study are acknowledged. We did not have pharmacy refill data to assess adherence. We utilized BP values recorded in the EHR, which are not necessarily taken according to accepted BP measurement protocols, and thus, they may not accurately reflect actual BP values; in addition, we do not know what the home BP of the individuals were. Also, although we excluded BP data from emergency department visits and only included BP data from outpatient clinics, we did not distinguish by type of outpatient clinic. Thus, the analyses regarding the BP‐lowering effectiveness of different drug regimens should be interpreted with caution. We attempted to mitigate this limitation partly through multivariable regression models adjusted for various demographic and clinical variables, including comorbidities and number of antihypertensive medications, which may be correlated with adherence.38, 39 Moreover, we did not validate our findings through manual chart reviews; however, this was beyond the scope of the study, which was to gain clinical insights through large biomedical data analysis. Thus, while EHR data have limitations, the resulting findings may more closely reflect the reality of treating patients in a routine clinical setting and thus more directly inform clinical decisions. Additionally, the achievement of BP control should ultimately translate into improved cardiovascular and renal outcomes, particularly in high‐risk patients such as CKD patients; however, we were not able to address this issue and make this important link in the present study.
In summary, we found that guideline‐recommended ACEI/ARB‐based therapy was prescribed in two‐thirds of CKD patients and three‐fourths of CKD patients with proteinuria. Although this is encouraging, there is still a substantial proportion of patients who are not being afforded the benefits of these renoprotective agents. However, since this is a retrospective study, we do not know if individuals were previously on an ACEI/ARB and were taken off it due to side effects. Moreover, patients who were prescribed combination therapy regimens that included BB + diuretic or ACEI/ARB + BB + diuretic were more likely to have controlled BP compared to those prescribed other combinations. These findings may be especially useful for early stage patients who need ≥3 drugs to control BP and ultimately reduce adverse outcomes. Among these patients, clinicians may select BB + diuretic as an add‐on to ACEI/ARB to better prevent the risk of cardiovascular and renal complications, including the progression of CKD. We also identified significant real‐world socio‐demographic and clinical risk factors associated with RHTN among CKD patients, which clinicians could use to more adequately manage cardiovascular risk factors and prevent adverse outcomes among CKD patients. Furthermore, our study findings may help fill knowledge gaps in the hypertension/CKD treatment guidelines about optimal stage‐specific antihypertensive therapy for CKD patients. Future hypertension treatment guidelines may incorporate more real‐world studies such as those based on EHRs to inform recommendations concerning the treatment of different patient subgroups, particularly those with CKD. Lastly, similar studies are needed using EHR data from other major health centers across the country to corroborate our findings.
CONFLICT OF INTEREST
The authors have no conflict of interest to disclose.
Supporting information
ACKNOWLEDGMENTS
We thank the UF Health IDR/i2b2 informatics team and the UF Clinical and Translational Science Institute (CTSI).
Magvanjav O, Cooper‐DeHoff RM, McDonough CW, et al. Antihypertensive therapy prescribing patterns and correlates of blood pressure control among hypertensive patients with chronic kidney disease. J Clin Hypertens. 2019;21:91–101. 10.1111/jch.13429
Funding information
This research was supported by the UF CTSI, which is funded in part by the NIH National Center for Advancing Translational Sciences under award number UL1TR001427. Additional funding for this study comes from NIH training grants T32 DK104721 (PI: Segal; Magvanjav) and KL2 TR001429 (McDonough).
REFERENCES
- 1. Muntner P, Anderson A, Charleston J, et al. Hypertension awareness, treatment, and control in adults with CKD: results from the Chronic Renal Insufficiency Cohort (CRIC) Study. Am J Kidney Dis. 2010;55:441‐451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. 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. J Am Coll Cardiol. 2018;71:e127‐e248. [DOI] [PubMed] [Google Scholar]
- 3. Gargiulo R, Suhail F, Lerma EV. Cardiovascular disease and chronic kidney disease. Dis Mon. 2015;61:403‐413. [DOI] [PubMed] [Google Scholar]
- 4. Hamrahian SM, Falkner B. Hypertension in chronic kidney disease. Adv Exp Med Biol. 2017;956:307‐325. [DOI] [PubMed] [Google Scholar]
- 5. Lv J, Ehteshami P, Sarnak MJ, et al. Effects of intensive blood pressure lowering on the progression of chronic kidney disease: a systematic review and meta‐analysis. CMAJ. 2013;185:949‐957. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Becker GJ, Wheeler DC. Blood pressure control in CKD patients: why do we fail to implement the guidelines? Am J Kidney Dis. 2010;55:415‐418. [DOI] [PubMed] [Google Scholar]
- 7. Allen AS, Forman JP, Orav EJ, Bates DW, Denker BM, Sequist TD. Primary care management of chronic kidney disease. J Gen Intern Med. 2011;26:386‐392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Rossignol P, Massy ZA, Azizi M, et al. The double challenge of resistant hypertension and chronic kidney disease. Lancet. 2015;386:1588‐1598. [DOI] [PubMed] [Google Scholar]
- 9. Wolley MJ, Stowasser M. Resistant hypertension and chronic kidney disease: a dangerous liaison. Curr Hypertens Rep. 2016;18(36):1‐9. [DOI] [PubMed] [Google Scholar]
- 10. De Nicola L, Gabbai FB, Agarwal R, et al. Prevalence and prognostic role of resistant hypertension in chronic kidney disease patients. J Am Coll Cardiol. 2013;61:2461‐2467. [DOI] [PubMed] [Google Scholar]
- 11. de Beus E, Bots ML, van Zuilen AD, Wetzels J, Blankestijn PJ. Prevalence of apparent therapy‐resistant hypertension and its effect on outcome in patients with chronic kidney disease. Hypertension. 2015;66:998‐1005. [DOI] [PubMed] [Google Scholar]
- 12. De Nicola L, Borrelli S, Gabbai FB, et al. Burden of resistant hypertension in hypertensive patients with non‐dialysis chronic kidney disease. Kidney Blood Press Res. 2011;34:58‐67. [DOI] [PubMed] [Google Scholar]
- 13. Fraser SD, Roderick PJ, McIntyre NJ, et al. Suboptimal blood pressure control in chronic kidney disease stage 3: baseline data from a cohort study in primary care. BMC Fam Pract. 2013;14:88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. CKD Kdigok , Group W . KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2013;3:1‐150. [DOI] [PubMed] [Google Scholar]
- 15. James PA, Oparil S, Carter BL, et al. 2014 evidence‐based guideline for the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507‐520. [DOI] [PubMed] [Google Scholar]
- 16. Pilotto LS, Ball PA, Smithard JM, Kennedy DR. Electronic records suggest suboptimal management of chronic kidney disease in general practice. Aust J Rural Health. 2012;20:195‐199. [DOI] [PubMed] [Google Scholar]
- 17. Nash DM, Brimble S, Markle‐Reid M, et al. Quality of care for patients with chronic kidney disease in the primary care setting: a retrospective cohort study from Ontario, Canada. Can J Kidney Health Dis. 2017;4:2054358117703059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Sinha AD, Agarwal R. Thiazide diuretics in chronic kidney disease. Curr Hypertens Rep. 2015;17:13. [DOI] [PubMed] [Google Scholar]
- 19. Agarwal R, Sinha AD. Thiazide diuretics in advanced chronic kidney disease. J Am Soc Hypertens. 2012;6:299‐308. [DOI] [PubMed] [Google Scholar]
- 20. Yildirim T, Arici M, Piskinpasa S, et al. Major barriers against renin‐angiotensin‐aldosterone system blocker use in chronic kidney disease stages 3–5 in clinical practice: a safety concern? Ren Fail. 2012;34:1095‐1099. [DOI] [PubMed] [Google Scholar]
- 21. Raebel MA. Hyperkalemia associated with use of angiotensin‐converting enzyme inhibitors and angiotensin receptor blockers. Cardiovasc Ther. 2012;30:e156‐e166. [DOI] [PubMed] [Google Scholar]
- 22. Weir MR, Rolfe M. Potassium homeostasis and renin‐angiotensin‐aldosterone system inhibitors. Clin J Am Soc Nephrol. 2010;5:531‐548. [DOI] [PubMed] [Google Scholar]
- 23. Calhoun DA, Jones D, Textor S, et al. Resistant hypertension: diagnosis, evaluation, and treatment: a scientific statement from the American Heart Association Professional Education Committee of the Council for High Blood Pressure Research. Circulation. 2008;117:e510‐e526. [DOI] [PubMed] [Google Scholar]
- 24. Tanner RM, Calhoun DA, Bell EK, et al. Prevalence of apparent treatment‐resistant hypertension among individuals with CKD. Clin J Am Soc Nephrol. 2013;8:1583‐1590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Drexler YR, Bomback AS. Definition, identification and treatment of resistant hypertension in chronic kidney disease patients. Nephrol Dial Transplant. 2014;29:1327‐1335. [DOI] [PubMed] [Google Scholar]
- 26. Benjamin EJ, Blaha MJ, Chiuve SE, et al. Heart disease and stroke statistics‐2017 update: a report from the American Heart Association. Circulation. 2017;135:e146‐e603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Saran R, Robinson B, Abbott KC, et al. US Renal Data System 2017 Annual Data Report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2018;71:A7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Talreja H, Ruzicka M, McCormick BB. Pharmacologic treatment of hypertension in patients with chronic kidney disease. Am J Cardiovasc Drugs. 2013;13:177‐188. [DOI] [PubMed] [Google Scholar]
- 29. Epstein M, Reaven NL, Funk SE, McGaughey KJ, Oestreicher N, Knispel J. Evaluation of the treatment gap between clinical guidelines and the utilization of renin‐angiotensin‐aldosterone system inhibitors. Am J Manag Care. 2015;21:S212‐S220. [PubMed] [Google Scholar]
- 30. Karadsheh F, Weir MR. Thiazide and thiazide‐like diuretics: an opportunity to reduce blood pressure in patients with advanced kidney disease. Curr Hypertens Rep. 2012;14:416‐420. [DOI] [PubMed] [Google Scholar]
- 31. Kaur J, Young BE, Fadel PJ. Sympathetic overactivity in chronic kidney disease: consequences and mechanisms. Int J Mol Sci. 2017;18:1‐18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Ptinopoulou AG, Pikilidou MI, Lasaridis AN. The effect of antihypertensive drugs on chronic kidney disease: a comprehensive review. Hypertens Res. 2013;36:91‐101. [DOI] [PubMed] [Google Scholar]
- 33. Scherpbier‐de Haan ND, Vervoort GM, van Weel C, et al. Effect of shared care on blood pressure in patients with chronic kidney disease: a cluster randomised controlled trial. Br J Gen Pract. 2013;63:e798‐e806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Wang LY, Yin DX, Zhang DL, Xu R, Cui WY, Liu WH. Improvement and influencing factors of blood pressure control by nephrologist referral in chronic kidney disease patients in China: a cohort study. Int Urol Nephrol. 2013;45:1345‐1353. [DOI] [PubMed] [Google Scholar]
- 35. Gu A, Yue Y, Desai RP, Argulian E. Racial and ethnic differences in antihypertensive medication use and blood pressure control among US adults with hypertension: The National Health and Nutrition Examination Survey, 2003 to 2012. Circ Cardiovasc Qual Outcomes. 2017;10:e003166. [DOI] [PubMed] [Google Scholar]
- 36. Musemwa N, Gadegbeku CA. Hypertension in African Americans. Curr Cardiol Rep. 2017;19:129. [DOI] [PubMed] [Google Scholar]
- 37. Thomas G, Xie D, Chen HY, et al. Prevalence and prognostic significance of apparent treatment resistant hypertension in chronic kidney disease: report from the chronic renal insufficiency cohort study. Hypertension. 2016;67:387‐396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Gupta P, Patel P, Strauch B, et al. Risk factors for nonadherence to antihypertensive treatment. Hypertension. 2017;69:1113‐1120. [DOI] [PubMed] [Google Scholar]
- 39. Kurdi AI, Chen LC, Elliott RA. Exploring factors associated with patients' adherence to antihypertensive drugs among people with primary hypertension in the United Kingdom. J Hypertens. 2017;35:1881‐1890. [DOI] [PubMed] [Google Scholar]
- 40. Chobanian AV, Bakris GL, Black HR, et al. The seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure: the JNC 7 report. JAMA. 2003;289:2560‐2572. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials