Abstract
Background/Objective
Given heightened cardiovascular (CVD) risk in rheumatoid arthritis (RA) and that higher blood pressure (BP) represents greater CVD risk, we hypothesized that higher BP would predict more BP-related communication in rheumatology visits. We examined predictors of documented BP communication during RA clinic visits.
Methods
Retrospective cohort study of RA patients identified in EHR records with uncontrolled hypertension receiving both primary and rheumatology care. Trained abstractors reviewed RA visit notes for “BP-communication” using a standardized tool to elicit documentation about BP or hypertension beyond recording vital signs. We used multivariate logistic regression to examine the impact of BP category (American Heart Association: ideal normotension, prehypertension, stage I and stage II hypertension) on odds ratios (OR, 95% confidence intervals) of BP-communication.
Results
Among 1,267 RA patients, 40% experienced BP elevations meeting the definition of uncontrolled hypertension. Of 2,677 eligible RA visits, 22% contained any documented BP communication. After adjustment, models predicted only 31% of visits with markedly high BPs ≥160/100 would contain BP-communication. Compared to stage I, stage II elevation did not significantly increase communication (OR 2.0, 1.4–2.8 vs. 1.5, 1.2–2.2), although both groups’ odds exceeded prehypertension and normotension. Less than 10% of eligible visits resulted in documented action steps recommending follow-up of high BP.
Conclusions
Regardless of BP magnitude, most RA clinic visits lacked documented communication about blood pressure despite compounded CVD risk. Future work should study how rheumatology clinics can facilitate follow-up of high BPs to address hypertension as the most common and reversible CVD risk factor.
Indexing Terms: Rheumatoid arthritis, cardiovascular disease, hypertension, blood pressure, quality of healthcare
INTRODUCTION
Compared to peers, patients with rheumatoid arthritis (RA) have higher rates of cardiovascular disease (CVD) events including myocardial infarction, stroke, and heart failure, [1] conditions causally linked to hypertension. Hypertension (HTN) affects 50% of people over age 55, yet nearly half of adults with known hypertension do not have it under control [2]. Hypertension is the most frequent comorbidity in RA [3, 4] and is a major determinant of organ damage and mortality [5]. Studies describe CVD event or CVD mortality risk ratios for hypertension between 2.8 and 4.1 in RA, notably higher than the 1.3–1.7 independent RA risk for CVD in many studies [1, 6–8]. Nevertheless, we previously reported that among patients meeting criteria for incident hypertension (i.e. newly developed, without prior diagnosis or treatment), RA patients were 29% less likely than peers without RA to be diagnosed with hypertension, despite more frequent visits [9]. Others have also reported high prevalence of uncontrolled hypertension in RA [5–7, 10–12].
Blood pressures (BP) are routinely measured in rheumatology clinics where RA patients may receive most of their care. Nationally, 73% of 2.6 million annual provider visits for RA occurred in specialty clinics [13]. Rheumatology visits equaled or exceeded primary care visits for half of the Medicare patients with RA [14]. Despite routine BP measurement at rheumatology visits, studies show that many rheumatologists are unwilling to treat hypertension [15, 16]. Recent quality measures [17] call not only for BP assessment but also “timely BP follow-up” across group practices defined as re-measurement within four weeks. A recent international rheumatology expert panel nearly reached consensus recommending that rheumatologists communicate findings of high BPs to ensure such primary care follow up [18].
Recognizing hypertension as a highly prevalent comorbidity and a highly reversible risk factor for CVD morbidity and mortality, we sought to examine how rheumatologists documented communication about high BP or hypertension during RA clinic visits. We termed this “BP-communication.” We hypothesized that higher BP measurements would predict more BP-related communication with RA patients in rheumatology clinic visits. Given that increased BP magnitude increases CVD risk, our objectives were to examine the relationship between the magnitude of BP elevation (none = ideal normotension, mild=prehypertension, moderate=stage I, or severe=stage II hypertension per clinical definitions at the time) [19] on the likelihood of documented BP-communication in RA visit notes and to examine other predictors of BP-communication.
MATERIALS AND METHODS
Sample
Using electronic health record (EHR) searches, we first identified potential RA patients by the presence of two or more International Classification of Diseases–Ninth edition (ICD-9) codes for RA (714.0–714.33, 714.4, 714.80, 714.81, 714.89) [20, 21] in a large academic multispecialty health system between 2008 and 2011. We then identified an RA cohort receiving regular primary and rheumatology care (Figure 1). Regular care was defined as having at least one ambulatory visit in primary care and one in rheumatology in 24 months, including at least one visit in the most recent 12 months adapting a published Wisconsin Collaborative for Healthcare Quality metric [22]. This definition guaranteed that an individual included in the study was receiving both specialty and primary care, which allowed us to review continuous BP trends over time in both contexts. Visits with patients who were pregnant during the observation period and visits in the emergency department were excluded to avoid secondary causes of hypertension from pregnancy, pain, or trauma. Visits with patients who did not have RA per rheumatologist notes were also excluded through a later manual abstraction item.
Figure 1.
Flow chart for RA hypertension control eligible cohort selection and visit inclusion and exclusion.
Next, we restricted eligibility to include only RA patients who met diagnostic criteria for uncontrolled hypertension defined by EHR algorithms using the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC-7) criteria, per clinical standards at the time [19]. “Uncontrolled hypertension” included those meeting incident (undiagnosed) hypertension criteria, and those with prior diagnosis or treatment whose BPs remained elevated. An individual was classified as having incident hypertension if they had two dates of having an ambulatory BP greater than 160/100 mmHg or three greater than or equal to 140/90 mmHg (or 130/80 for those with co-morbid diabetes) in the absence of a prior hypertension diagnosis code or medication [19, 23]. Elevated BP readings had to be at least 30 days apart and within 2 years. Likewise, those with prevalent hypertension identified by diagnosis codes for hypertension [23] or by the use of any baseline antihypertensive medications were deemed uncontrolled (control eligible) if BPs were not less than or equal to 140/90 mmHg, or 130/80 mmHg for those with comorbid diabetes.
Given that hypertension interpretation requires observation of BP trends over time, RA visits were reviewed starting from the study entry date defined as one year before the first date when all three of these criteria were met: (1.) RA defined by two ICD-9 codes, (2.) regular care definition, and (3.) uncontrolled hypertension. Patients were excluded if blood pressure normalized before their study entry date. The date of “normalized BP” was defined as the first date of three consecutive normal BPs <130/80 mmHg with diabetes, otherwise three BPs <140/90 mmHg [19]. Subjects were followed longitudinally and censored at the earliest date of the following: discontinuity of regular care (no visits in >12 mos.), death, BP normalization, or January 2012. We identified all rheumatology clinic RA visits for eligible patients from their entry date through their end date to determine whether BP-communication occurred at each rheumatology visit.
Ethics and approval
We received approval for this medical record review study (UW Health Science Institutional Review Board 2012–0053) from the Minimal Risk Institutional Review Board with a Health Insurance Portability and Accountability Act waiver of consent.
Electronic health record review for primary outcome
We manually abstracted each face-to-face rheumatology visit that occurred after an RA patient met entry criteria for uncontrolled hypertension until the date of control or censorship. Trained medical abstractors used a pre-defined electronic abstraction tool to review all eligible RA visit notes by rheumatology providers.
Abstractors reviewed the clinician’s encounter visit note, all electronic orders, and patient after-visit instructions for any written documentation about BP or hypertension. The abstraction tool contained 13 items detailing communication about BP or hypertension to the patient or primary care provider (PCP) (Table 1). The tool asked about the presence of any text beyond vital sign documentation regarding four areas: (1.) current or prior high blood pressure or hypertension, (2.) blood pressure interpretation (elevated/high), rechecks, or reviewed for additional or previous values, (3.) recommendation of patient or PCP follow up, (4.) review of numeric BP treatment goals or medications with the patient or PCP. “BP Communication” was met by any of these 13 items. On 5% audits, abstractors were required to demonstrate inter-observer agreement for BP-communication for quality assurance [24]. After every 200 charts were reviewed by an individual abstractor, quality assurance audits were performed by a second reviewer who re-abstracted ten charts to ensure agreement by consensus. This approach resulted in interrater agreement of >90% across BP communication items.
Table 1.
“BP communication” abstraction items.
| In this RA visit encounter note or after visit summary documentation: | |
|---|---|
| 1. | Was there any text regarding prior hypertension or high blood pressure? |
| 2. | Was this visit’s blood pressure value interpreted or described as “elevated/high”? |
| 3. | Was it documented that the blood pressure was repeated or rechecked during this visit? |
| 4. | Was there any text indicating that the rheumatologist reviewed additional blood pressures (not including this visit)? |
| 5. | Was there any text regarding blood pressure (other than the value in the vital signs)? |
| 6. | Was there any text recommending that the patient follow up with PCP or another clinician regarding blood pressure? |
| 7. | Was there any text recommending that the patient follow up with repeated blood pressures at home, or anywhere other than a visit to their PCP (i.e. community setting, nurse visit)? |
| 8. | Was there any text recommending that the PCP follow up on blood pressure? |
| 9. | Was there any text documenting that a numeric BP goal was given to the patient? |
| 10. | Was there any text documenting that a number BP goal was given to the PCP? |
| 11. | Was there any text documenting that current BP medication instructions were reviewed with the patient during the encounter? |
| 12. | Was there any text advising the PCP to consider BP medication changes? |
| 13. | Was there any text documenting that a change in BP medication instructions or a new prescription was given to the patient during the encounter? |
Abbreviations: BP = Blood Pressure; PCP = Primary Care Provider
Explanatory variable and covariates
Blood pressure magnitude, as defined by JNC-VII [19] and American Heart Association blood pressure stages, was assessed through EHR queries of longitudinal BP trends. The categorical magnitude of the BP at a given visit was the main explanatory variable [19]. Using this definition, a visit with BP <120/80 was considered ideal normotensive, BP ≥120/80 and <140/90 was prehypertensive, BP ≥140/90 and <160/100 was stage I hypertensive, and BP ≥160/100 was stage II hypertensive.
Covariates for the study included patient sociodemographics during the baseline year, such as age, gender, race, language, marital status, and ever receiving Medicaid, as a socioeconomic proxy. Behavioral health traits included tobacco history and quartiles of body mass index (BMI). Comorbidities included prior CVD (which included history of ischemic heart disease, congestive heart failure, peripheral vascular disease, transient ischemic attack or stroke), chronic kidney disease (CKD), end stage renal disease (ESRD), diabetes, and hyperlipidemia extrapolated using validated algorithms [25–27]. We also calculated a composite score using the Johns Hopkins Adjusted Clinical Group (ACG) System (Version 10) [28]. ACG calculates a score based on common morbidity patterns determined by 224 different clinical groupings. In this system, an individual scoring 1 is predicted to have average, while an individual scoring >1 would be predicted to have greater than average, composite health needs or illness burden. We created dummy variables to account for calendar year and clinic.
Analysis
For descriptive statistics comparing visits with normotension, prehypertension, and stage I and II BP elevations we used the chi-squared test for categorical variable comparisons and analysis of variance (ANOVA) for continuous variables. A p-value <0.05 was considered significant. Correlation matrices showed no evidence of collinearity for covariates. Multivariate logistic regression was used to analyze the relationship between explanatory variables and BP-communication. We used the Stata margins command to estimate adjusted predicted probabilities (APP) based on the recycled predictions approach. Odds ratios (OR) and 95% confidence intervals (CI) for BP-communication by BP magnitude were calculated using robust estimates of the variance that account for clustering. Sensitivity testing models re-analyzed to account for clustering by provider, and predictors did not differ.
Patient sample selection and data variable creation were conducted using SAS version 9.1.3 (SAS Institute, Inc., Cary, NC); statistical analysis was conducted using Stata version 12.0 (StataCorp, College Station).
RESULTS
Among 1,267 potential RA patients receiving both primary and specialty care in our health system, 40 % experienced visits with high blood pressures. Among 426 rheumatologist confirmed RA patients with uncontrolled hypertension, 232 (18%) also met incident hypertension criteria; by definition all had uncontrolled hypertension (Figure 1). Among visits of RA patients with uncontrolled hypertension, we observed a mean patient age of 61.6 years (SD 12.3), and female patients comprised 76% of the visits (Table 2). Overall, 11% were current smokers and >20% had a prior CVD diagnosis. In total, these 426 RA patients had 2,677 visits abstracted to examine BP-communication.
Table 2.
Visit Level Cohort Patient Characteristics (n=2,677 visits; n=436 patients)
| TOTAL | Ideal BP | Pre HTN | St 1 HTN | St II HTN | p | |
|---|---|---|---|---|---|---|
|
| ||||||
| N=2,677 n (%) |
N=461 n (%) |
N=931 n (%) |
N=952 n (%) |
N=333 n (%) |
||
| Age (mean±SD) | (61.8±12) | (61.6±12) | (62.3±13) | (61.2±13) | (62.4±14) | 0.27 |
| 20–39 | 71 (2.7) | 12 (2.6) | 17 (1.8) | 32 (3.4) | 10 (3.0) | <0.01 |
| 40–59 | 1115 (42) | 179 (39) | 396 (43) | 409 (43) | 131 (39) | |
| 60–79 | 1222 (46) | 230 (50) | 432 (46) | 421 (44) | 139 (42) | |
| >80 | 269 (10) | 40 (8.7) | 86 (9) | 90 (9.5) | 53 (16) | |
| Gender (female) | 2101 (78) | 363 (79) | 729 (78) | 743 (78) | 266 (80) | 0.91 |
| Race White | 2464 (92) | 427 (93) | 861 (92) | 872 (92) | 304 (91) | 0.03 |
| Black | 106 (4.0) | 14 (3) | 25 (2.7) | 50 (5.3) | 17 (5.1) | |
| Other | 107 (4.0) | 20 (4) | 45 (4.8) | 30 (3.2) | 12 (3.6) | |
| Language English | 2528 (94) | 439 (95) | 872 (94) | 909 (96) | 303 (93) | 0.13 |
| Non-English | 149 (5.6) | 22 (4.8) | 59 (6) | 43 (4.5) | 30 (9.0) | |
| Married/Partnered | 1561 (58) | 254 (55) | 575 (62) | 569 (60) | 163 (49) | <0.01 |
| Single | 465 (17) | 99 (21) | 164 (18) | 140 (15) | 62 (19) | |
| Separated/divorced | 651 (24) | 108 (23) | 192 (21) | 243 (26) | 108 (32) | |
| Medicaid (Ever) | 235 (8.8) | 37 (8) | 79 (8.5) | 90 (9.5) | 29 (8.7) | 0.81 |
| Tobacco Never Used | 1171 (44) | 195 (42) | 418 (45) | 399 (42) | 159 (48) | 0.02 |
| Current User | 312 (12) | 38 (8) | 98 (110) | 133 (14) | 43 (13) | |
| Quit | 984 (37) | 185 (40) | 350 (38) | 343 (36) | 106 (32) | |
| Missing/Passive | 210 (7.8) | 43 (9) | 65 (7.0) | 77 (8.1) | 25 (7.5) | |
| BMI (mean±SD) | 30.5±7.5 | 29.8±7.3 | 30±7.1 | 31.2±7.5 | 31.1±8.7 | <0.01 |
| ACG (mean±SD) | 1.4±0.68 | 1.5±0.75 | 1.4±0.63 | 1.3±0.66 | 1.4±0.72 | <0.01 |
| Cardiovascular disease | 795 (30) | 184 (40) | 315 (34) | 226 (24) | 70 (21) | <0.01 |
| Chronic kidney/ESRD | 134 (5.0) | 49 (11) | 40 (4.3) | 34 (3.6) | 11 (3.3) | <0.01 |
| Diabetes mellitus | 434 (16) | 53 (12) | 147 (16) | 173 (18) | 61 (18) | 0.01 |
| Hyperlipidemia | 0.42 | 0.48 | 0.43 | 0.4 | 0.38 | 0.01 |
Cardiovascular disease included ischemic heart disease, congestive heart failure, peripheral vascular disease, or transient ischemic attack or stroke.
Abbreviations: ACG = Adjusted Clinical Group a composite comorbidity and utilization measure with general age adjusted mean of 1; BMI = Body Mass Index; BP = Blood Pressure; ESRD = End Stage Renal Disease; HTN = Hypertension; SD = Standard Deviation; St = Stage.
Visit-level blood pressures in this cohort varied widely. Among the 2,677 abstracted visits, participants had a normal blood pressure at 20% of these visits, prehypertension in 45% of visits, and blood pressure was stage I in 32%, stage II in 11% of visits. Compared to normotensive patient encounters, visits with BPs meeting criteria for stage I or stage II hypertension occurred significantly more often in visits with patients who were older, black, tobacco users and in a higher BMI quartile (Table 2).
Primary outcome: BP-communication
Overall 23% of RA visits contained any BP-communication. After adjustment, only 31% of visit notes in which BP equaled or exceeded 160/100 mm/Hg contained BP-communication (APP 31%, CI 26–37). Compared to stage I elevations, stage II BP elevations did not significantly increase the probability of documented BP-communication (Figure 2), although rates were higher than for visits in which BP was normal or indicated prehypertension [stage I APP=25% (CI 22–28); stage II APP=31% (CI 26–37); normotension APP 19% (CI 15–22); prehypertension APP 19% (CI 17–22)]. Compared to visits in which BP was normal, the adjusted ORs for BP-communication in visits with stage I and stage II hypertension were 1.5 (1.2–2.2) and 2.0 (1.4–2.8) respectively with overlapping confidence intervals (Table 3); although there was a positive test for linear trend (P<0.001). Results were similar when analyzed clustering on provider (data not shown).
Figure 2. Probability of BP-communication by visit BP-category representing magnitude.
The x-axis represents visits by American Heart Association BP group and the y-axis demonstrates the percent predicted probability that BP-Communication occurred at each visit within that BP category.
Table 3.
Odds ratios for visit-level prediction of BP communication (n=2,677 visits)
| Unadjusted OR (95% CI) |
Adjusted OR (95% CI) |
|
|---|---|---|
| Blood Pressure Categories | ||
| Normotensive | Ref | Ref |
| Pre HTN | 1.0 (0.75–1.3) | 1.1 (0.83–1.5) |
| Stage I HTN | 1.4 (1.1–1.8) | 1.5 (1.2–2.2) |
| Stage II HTN | 1.9 (1.3–2.6) | 2.0 (1.4–2.8) |
| Age 20–39 | Ref | Ref |
| 40–59 | 0.72 (0.41–1.2) | 1.2 (0.67–2.2) |
| 60–79 | 0.88 (0.51–1.5) | 1.5 (0.80–2.7) |
| >80 | 1.3 (0.74–2.4) | 2.1 (1.1–4.2) |
| Gender (female) | 1.3 (1.1–1.7) | 1.1 (0.88–1.5) |
| Race White | Ref | Ref |
| Black | 0.95 (0.59–1.5) | 0.98 (0.56–1.7) |
| Other | 0.89 (0.55–1.4) | 0.86 (0.47–1.6) |
| Language English | Ref | Ref |
| Non-English | 0.7 (0.88–3.4) | 1.6 (0.70–3.6) |
| Married/Partnered | Ref | Ref |
| Single | 1.4 (1.1–1.8) | 1.4 (1.1–1.9) |
| Separated/divorced | 1.4 (1.2–1.8) | 1.3 (1.02–1.7) |
| Medicaid (Ever) | 1.2 (0.89–1.6) | 1.6 (1.1–2.3) |
| Tobacco Never | Ref | Ref |
| Current | 0.47 (0.33–0.67) | 0.40 (0.26–0.63) |
| Quit | 0.89 (0.72–1.1) | 0.91 (0.72–1.1) |
| Missing/Passive | 1.2 (0.83–1.6) | 1.2 (0.81–1.7) |
| ACG quartile Lowest | Ref | Ref |
| Second | 0.73 (0.57–0.94) | 0.61 (0.46–0.81) |
| Third | 0.68 (0.53–0.87) | 0.64 (0.48–0.84) |
| Highest | 0.66 (0.51–0.85) | 0.43 (0.32–0.57) |
| BMI quartile Lowest | Ref | Ref |
| Second | 1.3 (1.02–1.7) | 1.6 (1.2–2.2) |
| Third | 1.1 (0.86–1.5) | 1.2 (0.86–1.6) |
| Highest | 1.6 (1.2–2.1) | 1.9 (1.4–2.6) |
| Cardiovascular disease | 1.4 (1.2–1.7) | 1.5 (1.2–1.9) |
| Chronic kidney/ESRD | 1.3 (0.87–1.9) | 1.2 (0.76–1.8) |
| Diabetes mellitus | 1.2 (0.95–1.5) | 1.2 (0.92–1.6) |
| Hyperlipidemia | 1.2 (1.005–1.4) | 1.2 (0.97–1.5) |
Model also included dummy variables for year and clinic group (data not shown).
Abbreviations: ACG = Adjusted Clinical Group a composite comorbidity and utilization measure with general age adjusted mean of 1; BMI = Body Mass Index; BP = Blood Pressure; CI = Confidence Interval; ESRD = End Stage Renal Disease; HTN = Hypertension; OR = Odds Ratio.
Other predictors of BP-communication
In the adjusted OR model, active tobacco users were least likely to have BP-communication (See Table 3 OR 0.4, CI 0.26–0.63). Those in the highest ACG quartile for ACG co-morbidity scores had lower odds of BP-communication compared to the lowest ACG quartile (OR 0.43, 0.32–0.57). Patients who were over age 80, unmarried, ever on Medicaid, in the highest BMI quartile, or had prevalent CVD were more likely to have BP-communication documented.
When examining an adjusted model on visits with Stage I or II BP elevation as shown in Table 4, current tobacco use and increasing complexity indicated by ACG quartile predicted less likely BP-communication. In contrast, the highest BMI quartile and presence of baseline cardiovascular disease both predicted more BP-communication.
Table 4.
Odds ratios for visit-level BP communication in Stage I & II BP (n=1,285)
| Unadjusted OR (95% CI) |
Adjusted OR (95% CI) |
|
|---|---|---|
| Stage I HTN | Ref | Ref |
| Stage II HTN | 1.4 (1.03–1.8) | 1.3 (0.98–1.8) |
| Age 20–39 | Ref | Ref |
| 40–59 | 0.59 (0.31–1.2) | 0.98 (0.46–2.1) |
| 60–79 | 0.67 (0.35–1.3) | 1.1 (0.49–2.3) |
| >80 | 0.58 (0.28–1.2) | 0.82 (0.33–2.1) |
| Gender (female) | 1.1 (0.78–1.4) | 1.1 (0.74–1.5) |
| Race White | Ref | Ref |
| Black | 0.79 (0.44–1.4) | 0.70 (0.34–1.5) |
| Other | 0.55 (0.24–1.2) | 0.43 (0.16–1.2) |
| Language English | Ref | Ref |
| Non-English | 0.7 (0.57–5.4) | 2.9 (0.81–10) |
| Married/Partnered | ||
| Single | 1.3 (0.94–1.9) | 1.3 (0.87–2.0) |
| Separated/divorced | 1.2 (0.92–1.6) | 1.6 (1.1–2.3) |
| Medicaid (Ever) | 1.5 (0.995–2.2) | 1.8 (1.1–2.9) |
| Tobacco Never | Ref | Ref |
| Current | 0.61 (0.40–0.94) | 0.48 (0.27–0.86) |
| Quit | 1.1 (0.84–1.5) | 1.1 (0.79–1.5) |
| Missing/passive | 1.4 (0.91–2.2) | 1.5 (0.88–2.5) |
| ACG quartile Lowest | Ref | Ref |
| Second | 0.66 (0.47–0.91) | 0.46 (0.32–0.68) |
| Third | 0.64 (0.46–0.90) | 0.56 (0.38–0.81) |
| Highest | 0.39 (0.26–0.57) | 0.20 (0.13–0.31) |
| BMI quartile Lowest | Ref | Ref |
| Second | 2.2 (1.4–3.9) | 2.7 (1.7–4.4) |
| Third | 1.6 (1.04–2.3) | 1.6 (0.99–2.6) |
| Highest | 2.4 (1.6–3.5) | 2.8 (1.7–4.7) |
| Cardiovascular disease | 1.5 (1.1–2.0) | 2.1 (1.5–3.0) |
| Chronic kidney/ESRD | 1.4 (0.76–2.7) | 1.4 (0.66–2.8) |
| Diabetes mellitus | 1.08 (0.78–1.5) | 1.3 (0.83–1.9) |
| Hyperlipidemia | 0.65 (0.58–0.73) | 0.94 (0.68–1.3) |
Model also controlled for calendar year and clinic group.
Abbreviations: ACG = Adjusted Clinical Group a composite comorbidity and utilization measure with general age adjusted mean of 1; BMI = Body Mass Index; BP = Blood Pressure; CI = Confidence Interval; ESRD = End Stage Renal Disease; HTN = Hypertension; OR = Odds Ratio.
Describing BP-communication types
In total, 601 of 2,677 observed visits (22%) contained any BP-communication. Figure 3 shows that even when scaling only to the 601 visits with BP-communication, 70% of the time providers listed hypertension as a pre-existing comorbidity, and 25% simply interpreted that the current visit blood pressure as elevated. Only 102 visit notes included recommendations for patients or their primary care providers (PCPs) to follow up elevated BPs, <10% of the 1285 total eligible visits with high BPs. Again, scaled to the 601 visits with BP-communication, just 8% documented reviewing prior BP trends. Six percent of visit notes clarified blood pressure medication instructions, and less than 1% (n=4 each) documented discussing numeric blood pressure targets or discussed a new blood pressure medication prescription.
Figure 3. Visits with BP-communication by type.
This bar graph shows BP-communication content types and visit-level frequencies scaled to n=601 visits with BP-communication; among 1285 visits with elevated blood pressures eligible for such communication. Sums greater than 100% reflect that a single visit may fit multiple communication categories.
DISCUSSION
We found that documented communication about hypertension or blood pressure rarely occurred in visits with RA patients who had uncontrolled hypertension despite their increased CVD risk [6, 7, 29]. Discussions about high BP were documented in fewer than one in three RA visit notes even when severely elevated (≥160/100 mmHg). At this blood pressure range, using estimates from the general population with one CVD risk factor, experts predict that providers need to treat seven patients to prevent one cardiovascular event [30]. This highlights an opportunity for improving outcomes by managing or referring RA patients with high blood pressures back to primary care for hypertension management. Blood pressure was routinely measured, yet high BPs most often went unmentioned.
In this study, we also reported that despite compounded CVD risk, tobacco users received less BP-communication, and only those with the highest BMI and prevalent CVD were more likely to receive BP-communication. In a prior study, we also noted lower rates of incident hypertension diagnosis in RA patients than peers, with particularly low rates in RA patients who currently use tobacco [31]. In another study, we also reported that only 10% of encounters with RA patients who smoked documented cessation counselling[32] suggesting that competing priorities alone did not explain gaps in CVD risk factor counselling for either hypertension or smoking cessation. Likewise, noting more BP-communication after CVD events highlights a missed opportunity to reduce CVD by addressing modifiable risk factors like hypertension before events occur.
Moreover, when documented BP-communication did occur in this study, rheumatologists rarely took action. Recommended follow-up for elevated blood pressures was documented in only 10% of eligible RA visits. Our findings are similar to a previous study in which only 31% of rheumatologists said they would routinely treat elevated BP [16]. As we have previously reported from interviews [15], rheumatologists might hesitate to act on BP elevations, even though PCPs requested that they simply “send patients back” to address hypertension and other modifiable risks. This lack of action highlights the opportunity for systematic approaches to help RA clinics connect patients to the blood pressure care and CVD preventive care they need.
The Centers for Disease Control and CMS have called for the use of staff hypertension protocols as a way to “save more life years than any other intervention” [33]. Staff hypertension protocols are proven to improve BP control from 50% nationally to >80% in primary care settings [34]. Such protocols have not been extensively studied in specialty settings like rheumatology clinics, but they hold potential. Specialty clinic protocols could support timely follow-up of high BPs as defined by quality metrics using simple new thresholds (BPs ≥140/90 mmHg for all adults) [35]. BP protocols do not require risk calculations to determine who should be treated, making BP follow up referrals easier to execute than the “annual CVD risk evaluations” called for by the European League Against Rheumatism and others [12, 36]. Moreover, staff-driven protocols can address group practice quality metrics for BP follow up without burdening busy rheumatology clinicians. Future studies should test staff BP protocols in rheumatology clinics to address gaps observed in this study.
A strength of our study is that it captured a large number of clinical encounters with RA patients who received both primary and specialty care in the same health network. Presumably, this setting would be an ideal environment for communication by rheumatologists to connect patients back to primary care for BP follow-up, yet this rarely occurred. As with any study, one must also consider limitations. First, given that this was a medical record abstraction study, we recognize that undocumented BP-communication discussions would not be captured. However, if an action was not documented, it is less likely to have occurred or to be followed up by the patient or primary provider. Second, as a method, manual abstraction could introduce error although quality control measures aimed to decrease such errors [24]. Third, our results reflect rheumatology clinics from a single health network. These clinics might not reflect the care practices of other rheumatologists, and BP communication in other specialty clinics was not compared. Likewise, beyond BP and BMI, these clinics did not measure waist circumference representing abdominal fat mass for additional CVD risk estimation, although this would not be customary in most US non-cardiology specialty clinics. In addition, we also acknowledge that high BP at a single visit might not indicate hypertension. In this study, we did not have access to ambulatory blood pressure data or blood pressures outside of clinic. Still some authors recently confirmed that variability in BP itself heightens CVD risk in RA [37] so even intermittent BP elevations likely confer risk and merit discussion with patients. Moreover, in our study, multiple elevated BPs over time were required to meet clinical criteria for hypertension, and given known CVD risk in the RA population, recurrent elevated BPs are a risk factor that should be identified and addressed.
Overall, regardless of BP category, most eligible RA visits lacked documented communication about high blood pressures despite heightened cardiovascular risk in RA patients. Future work should examine systematic approaches to improve identification and referral of high BP and other modifiable risk factors to reduce CVD risk and promote longevity in patients with RA and other rheumatologic conditions.
Key Points.
In this study, documented communication about elevated blood pressures was discussed in only 22% of visits, and only 31% of visits when blood pressure was severely elevated.
Less than 10% of eligible RA visits with uncontrolled hypertension resulted in action including a recommendation to follow-up elevated blood pressures; <1% resulted in antihypertensive medication initiation or titration.
Our findings highlight the potential benefits of systematic strategies to facilitate followup or management of hypertension or other CVD risk factors in at-risk RA populations.
Acknowledgments
Authors would like to thank Joanna Wong, MD for data abstraction support, and Courtney Maxcy and Amanda Perez for supporting manuscript preparation.
Source of Funding:
Research reported in this publication was primarily supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases, part of the National Institutes of Health (NIH), under Award Number K23AR062381 (Bartels). Additional support from NIH-NCATS 9U54TR000021 (the Health Innovation Program/Community-Academic Partnerships core of UW ICTR-CTSA), NIH-NHLBI 1 K23 HL112907 (Johnson), and the University of Wisconsin Shapiro Research Program (Voelker). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Bartels currently receives institutional grant funding from Independent Grants for Learning and Change (Pfizer) unrelated to this manuscript.
Footnotes
These research results were presented at the American College of Rheumatology Annual Meeting, San Diego, CA USA, October 2013, which was highlighted by Medscape News http://www.medscape.com/viewarticle/813294?src=rss but have not been published previously.
Conflicts of Interest: All other authors declare no relationships or conflicts of interest.
References
- 1.Avina-Zubieta JA, Thomas J, Sadatsafavi M, et al. Risk of incident cardiovascular events in patients with rheumatoid arthritis: a meta-analysis of observational studies. Ann Rheum Dis. 2012;71(9):1524–9. doi: 10.1136/annrheumdis-2011-200726. [DOI] [PubMed] [Google Scholar]
- 2.Roger VL, S GA, Lloyd-Jones DM, et al. Circulation. American Heart Association; 2012. Heart disease and stroke statistics–2012 update: a report from the American Heart Association. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Osiri M, Sattayasomboon Y. Prevalence and out-patient medical costs of comorbid conditions in patients with rheumatoid arthritis. Joint Bone Spine. 2013;80(6):608–12. doi: 10.1016/j.jbspin.2013.01.013. [DOI] [PubMed] [Google Scholar]
- 4.Tiippana-Kinnunen T, Kautiainen HPL, Leirisalo-Repo M. Co-morbidities in Finnish patients with rheumatoid arthritis: 15-year follow-up. Scand J Rheumatol. 2013;42(6):451–6. doi: 10.3109/03009742.2013.790073. [DOI] [PubMed] [Google Scholar]
- 5.Panoulas VF, Toms TE, Metsios GS, et al. Target organ damage in patients with rheumatoid arthritis: the role of blood pressure and heart rate. Atherosclerosis. 2010;209(1):255–60. doi: 10.1016/j.atherosclerosis.2009.08.047. [DOI] [PubMed] [Google Scholar]
- 6.Innala L, Moller B, Ljung L, et al. Cardiovascular events in early RA are a result of inflammatory burden and traditional risk factors: a five year prospective study. Arthritis Res Ther. 2011;13(4):R131. doi: 10.1186/ar3442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Serelis J, Panagiotakos DB, Mavrommati M, et al. Cardiovascular disease is related to hypertension in patients with rheumatoid arthritis: a greek cohort study. J Rheumatol. 2011;38(2):236–41. doi: 10.3899/jrheum.100564. [DOI] [PubMed] [Google Scholar]
- 8.Baghdadi LR, Woodman RJ, Shanahan EM, et al. The impact of traditional cardiovascular risk factors on cardiovascular outcomes in patients with rheumatoid arthritis: a systematic review and meta-analysis. PLoS One. 2015;10(2):e0117952. doi: 10.1371/journal.pone.0117952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bartels C, Johnson H, Jacobs E, et al. American College of Rheumatology. San Diego, CA: Arthritis and Rheumatism; 2013. Frequency and Predictors Of Hypertension Communication In Rheumatoid Arthritis Visits; pp. S83–S4. [Google Scholar]
- 10.Chung CP, Giles JT, Petri M, et al. Prevalence of traditional modifiable cardiovascular risk factors in patients with rheumatoid arthritis: comparison with control subjects from the multi-ethnic study of atherosclerosis. Semin Arthritis Rheum. 2012;41(4):535–44. doi: 10.1016/j.semarthrit.2011.07.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Panoulas VF, Douglas KM, Milionis HJ, et al. Prevalence and associations of hypertension and its control in patients with rheumatoid arthritis. Rheumatology (Oxford) 2007;46(9):1477–82. doi: 10.1093/rheumatology/kem169. [DOI] [PubMed] [Google Scholar]
- 12.Semb AG, Rollefstad S, van Riel P, et al. Cardiovascular disease assessment in rheumatoid arthritis: a guide to translating knowledge of cardiovascular risk into clinical practice. Ann Rheum Dis. 2014 doi: 10.1136/annrheumdis-2013-204792. [DOI] [PubMed] [Google Scholar]
- 13.Schappert SM, Rechtsteiner EA. Ambulatory medical care utilization estimates for 2007. Vital Health Stat 13. 2011;(169):1–38. [PubMed] [Google Scholar]
- 14.Bartels CM, Kind AJ, Everett C, et al. Low frequency of primary lipid screening among medicare patients with rheumatoid arthritis. Arthritis Rheum. 2011;63(5):1221–30. doi: 10.1002/art.30239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bartels CM, Roberts TJ, Hansen KE, et al. Rheumatologist and Primary Care Management of Cardiovascular Disease Risk in Rheumatoid Arthritis: Patient and Provider Perspectives. Arthritis Care Res (Hoboken) 2015 doi: 10.1002/acr.22689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nguyen-Oghalai TU, Hunnicutt SE, Smith ST, et al. Factors that impact decision making among rheumatologists in the initiation of treatment for hypertension in rheumatoid arthritis. J Clin Rheumatol. 2007;13(6):307–12. doi: 10.1097/RHU.0b013e318156bcc2. [DOI] [PubMed] [Google Scholar]
- 17.Center for Medicare Services 2013 Physician Quality Reporting System (PQRS) Group Practice Reporting Option (GPRO) Participation for the PQRS Incentive Payment Made Simple. 2013:4. [Google Scholar]
- 18.Barber CE, Marshall DA, Alvarez N, et al. Development of Cardiovascular Quality Indicators for Rheumatoid Arthritis: Results from an International Expert Panel Using a Novel Online Process. J Rheumatol. 2015;42(9):1548–55. doi: 10.3899/jrheum.141603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.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(19):2560–72. doi: 10.1001/jama.289.19.2560. [DOI] [PubMed] [Google Scholar]
- 20.Katz JN, Barrett J, Liang MH, et al. Sensitivity and positive predictive value of Medicare Part B physician claims for rheumatologic diagnoses and procedures. Arthritis Rheum. 1997;40(9):1594–600. doi: 10.1002/art.1780400908. [DOI] [PubMed] [Google Scholar]
- 21.MacLean CH, Louie R, Leake B, et al. Quality of care for patients with rheumatoid arthritis. JAMA. 2000;284(8):984–92. doi: 10.1001/jama.284.8.984. [DOI] [PubMed] [Google Scholar]
- 22.Thorpe CT, Flood GE, Kraft SA, et al. Effect of patient selection method on provider group performance estimates. Med Care. 2011;49(8):780–5. doi: 10.1097/MLR.0b013e31821b3604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tu K, Campbell NR, Chen ZL, et al. Accuracy of administrative databases in identifying patients with hypertension. Open Med. 2007;1(1):e18–26. [PMC free article] [PubMed] [Google Scholar]
- 24.Reisch LM, Fosse JS, Beverly K, et al. Training, quality assurance, and assessment of medical record abstraction in a multisite study. Am J Epidemiol. 2003;157(6):546–51. doi: 10.1093/aje/kwg016. [DOI] [PubMed] [Google Scholar]
- 25.Borzecki AM, Wong AT, Hickey EC, et al. Identifying hypertension-related comorbidities from administrative data: what’s the optimal approach? Am J Med Qual. 2004;19(5):201–6. doi: 10.1177/106286060401900504. [DOI] [PubMed] [Google Scholar]
- 26.Hebert PL, Geiss LS, Tierney EF, et al. Identifying persons with diabetes using Medicare claims data. Am J Med Qual. 1999;14(6):270–7. doi: 10.1177/106286069901400607. [DOI] [PubMed] [Google Scholar]
- 27.Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27. doi: 10.1097/00005650-199801000-00004. [DOI] [PubMed] [Google Scholar]
- 28.Reid RJ, Roos NP, MacWilliam L, et al. Assessing population health care need using a claims-based ACG morbidity measure: a validation analysis in the Province of Manitoba. Health Serv Res. 2002;37(5):1345–64. doi: 10.1111/1475-6773.01029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Dougados M, Soubrier M, Antunez A, et al. Prevalence of comorbidities in rheumatoid arthritis and evaluation of their monitoring: results of an international, cross-sectional study (COMORA) Ann Rheum Dis. 2014;73(1):62–8. doi: 10.1136/annrheumdis-2013-204223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ogden LG, He J, Lydick E, et al. Long-term absolute benefit of lowering blood pressure in hypertensive patients according to the JNC VI risk stratification. Hypertension. 2000;35(2):539–43. doi: 10.1161/01.hyp.35.2.539. [DOI] [PubMed] [Google Scholar]
- 31.Bartels CM, Johnson H, Voelker K, et al. Impact of rheumatoid arthritis on receiving a diagnosis of hypertension among patients with regular primary care. Arthritis Care Res (Hoboken) 2014;66(9):1281–8. doi: 10.1002/acr.22302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Vreede AP, Johnson HM, Piper M, et al. Rheumatologists Modestly More Likely to Counsel Smokers in Visits without Rheumatoid Arthritis Control: An observational study. J Clin Rheumatol. 2017 doi: 10.1097/RHU.0000000000000559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Frieden TR, King S, Wright JS. Protocol-based treatment of hypertension: A critical step on the pathway to progress. JAMA. 2014;311(1):21–2. doi: 10.1001/jama.2013.282615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Jaffe MG, Lee GA, Young JD, et al. Improved blood pressure control associated with a large-scale hypertension program. JAMA. 2013;310(7):699–705. doi: 10.1001/jama.2013.108769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Weber MA, Schiffrin EL, White WB, et al. Clinical practice guidelines for the management of hypertension in the community a statement by the American Society of Hypertension and the International Society of Hypertension. J Hypertens. 2014;32(1):3–15. doi: 10.1097/HJH.0000000000000065. [DOI] [PubMed] [Google Scholar]
- 36.Peters MJ, Symmons DP, McCarey D, et al. EULAR evidence-based recommendations for cardiovascular risk management in patients with rheumatoid arthritis and other forms of inflammatory arthritis. Ann Rheum Dis. 2010;69(2):325–31. doi: 10.1136/ard.2009.113696. [DOI] [PubMed] [Google Scholar]
- 37.Myasoedova E, Crowson CS, Green AB, et al. Longterm blood pressure variability in patients with rheumatoid arthritis and its effect on cardiovascular events and all-cause mortality in RA: a population-based comparative cohort study. J Rheumatol. 2014;41(8):1638–44. doi: 10.3899/jrheum.131170. [DOI] [PMC free article] [PubMed] [Google Scholar]