Abstract
Background: Side effect symptoms to blood pressure (BP) medications may be associated with medication nonadherence, treatment dissatisfaction, and worse BP control. This article describes the frequency and characteristics of BP medication side effect symptoms in a primary care population with poorly controlled hypertension and their relationships to treatment satisfaction and adherence.
Methods: Patients in a pragmatic trial were surveyed at baseline to identify and characterize 6 potential BP medication side effect symptoms (tiredness, dizziness, foot swelling, cough, frequent urination, sexual symptoms). Reported symptoms were rated on severity (not a problem, somewhat/moderate problem, big/very big problem) and perceived relatedness to medications (yes/no). Logistic regression models used symptom severity and perceived relatedness to medications to predict BP treatment satisfaction (very satisfied to very dissatisfied) and medication adherence (changing/stopping medications).
Results: Among survey responders (n = 1,719/3,071, 56%), 90% of respondents taking BP medications reported a symptom that was at least somewhat of a problem. Overall, 39% had at least one symptom that was a big or very big problem and 34% had at least one symptom that they perceived as related to their medication. For most symptoms, both higher problem severity and perceived relatedness to medication were significantly associated with lower BP treatment satisfaction and decreased adherence.
Conclusions: BP medication side effect symptoms were very common and often big problems for patients. Identifying and managing them could potentially improve BP outcomes. The brief symptom assessment developed for this study could help identify opportunities to address side effect symptoms and improve patient satisfaction and adherence.Clinical trial registration: www.clinicaltrials.gov NCT02996565
- Blood Pressure
- Drug Side Effects
- Hypertension
- Patient Adherence
- Patient Satisfaction
- Pharmacology
- Surveys and Questionnaires
- Symptom Assessment
Introduction
Blood pressure (BP) control is declining in the US, and only 48% of patients with hypertension had BP controlled to <140/90 mmHg in 2017 to 2020 in a nationally representative longitudinal study.1 In the same analysis, only 31% had BP controlled to <130/80 mmHg. Improving BP control for the population is a major national goal, including strategies to educate the population and medical profession about the importance of controlling hypertension and methods to improve its treatment.2⇓⇓⇓–6 Achieving better BP control requires a multifactorial approach aimed at barriers at the level of policies, clinicians, and patients.
One of the major patient factors contributing to uncontrolled hypertension is nonadherence to BP medications.7⇓–9 Among patients with hypertension, treatment dissatisfaction and nonadherence are highly correlated.10,11 When elicited by surveys among patients taking antihypertensive treatment, patient-reported side effect symptoms are frequent, especially in those newly started on treatment, and are associated with nonadherence and worse health ratings.12,13 Thus, gaining a greater understanding of patient perceptions about side effects to antihypertensive treatment has the potential to improve treatment satisfaction, adherence, and BP control. However, most information about side effects to BP medications comes from randomized controlled trials and from adverse drug event reporting systems in the inpatient setting; less is known about the frequency, severity, and patient perceptions of side effect symptoms in clinical practice and outpatient settings.14
Patient attribution of side effects to BP medications is important as these perceptions may prompt them to stop treatment or seek guidance from their clinician about stopping, which may in turn affect BP control and longer-term cardiovascular risk. However, many side effect symptoms that could be due to BP medications are nonspecific and could also be attributable to other causes. Thus, any effect on adherence might depend on symptom severity and whether the patient believes the symptoms are due to their BP medications. The objectives of this manuscript are 1) to describe the frequency and characteristics of patient-reported BP medication side effect symptoms in primary care practice, 2) to assess patient perceptions of whether side effect symptoms were related to BP medications, and 3) to assess the relationship of symptom severity and perceived relatedness to treatment satisfaction and medication adherence.
Methods
Setting
This cross-sectional analysis of baseline data were part of a pragmatic clinical trial called Hyperlink 3 (NCT02996565) comparing telehealth management of blood pressure with home BP monitoring and pharmacist management to best-practice clinic-based care in 21 clinics within an integrated care system in Minnesota and Western Wisconsin.15
Population
Each patient was screened for study eligibility at all primary care encounters over an 18-month accrual period using an automated algorithm. The inclusion criteria included the following: age 18 to 85, diagnosed hypertension, having a visit at the same clinic as the patient’s assigned primary care clinician, and 2 consecutive visits with systolic BP >150 mmHg or diastolic BP >95 mmHg. Patients were considered enrolled if a study-generated order for hypertension follow-up was signed by the clinician. The study was reviewed by the HealthPartners Institutional Review Board, which granted a waiver of written informed consent.
Measurements
To assess patient-reported baseline data for the study, all enrolled study participants were mailed a survey within one week of enrollment, with telephone outreach to initial nonresponders by trained interviewers (Appendix 2). A cover letter was included with the mailed survey that indicated that returning the survey implied consent to use their survey data for analysis. The total survey time was estimated at 10 to 15 minutes if completed by mail and 15 to 20 minutes if completed by interview in English (complete survey shown in supplement). Written surveys were also available in Spanish, and interviews were also available in a wide variety of languages via bilingual interviewers or a language line. About 10% of interviews were in languages other than English, and the most common languages in which interpreted interviews were conducted were Somali, Vietnamese and Hmong. The baseline survey mailing included a $2 noncontingent cash incentive, and respondents also received a $10 gift card for a completed survey or interview.16⇓–18
Although we found longer instruments that assessed some related domains,12,19,20 we were unable to identify established or validated instruments to briefly assess the presence of potential BP medication side effect symptoms, severity, or perceived relatedness to BP medications. We therefore developed our own brief series of questions to assess 6 common side effect symptoms (tiredness, dizziness, leg swelling, cough, frequent urination, and sexual symptoms) often attributed to BP medications and incorporated them into the larger survey. The questions were developed using best practice and reviewed for face validity by content experts.21 To ensure that the questions were likely to be understood, valued, and useful to the patients and family members, 2 patient coinvestigators played a large role in developing the survey questions (Figure 1).22 The severity rating for each symptom was categorized into 3 levels: not a problem; somewhat or moderate; big or very big. Perceived relatedness to BP medications was dichotomized (yes; no or not sure) for symptoms that were rated as at least somewhat of a problem. The maximum rating across all symptoms was used as an overall measure of symptom severity.
Questions to assess blood pressure medication side effect symptoms.
BP treatment satisfaction was measured by responses on a 5-point Likert scale (very satisfied, somewhat satisfied, neither satisfied nor dissatisfied, somewhat dissatisfied, very dissatisfied) to the question “Taking all things into account, how satisfied or dissatisfied are you with the blood pressure medicine you have taken in the past 6 months?” Satisfaction responses were categorized into 3 levels: very or somewhat dissatisfied; neither satisfied or dissatisfied; somewhat or very satisfied. An adherence question focused on side effect symptoms with a yes/no response was worded as follows: “In the past 6 months, have you changed or stopped your blood pressure medicine, even just for a few days, because of symptoms you think were related to your blood pressure medicine?”
Analysis
The characteristics of survey respondents and nonrespondents were compared using Chi-square statistics or general linear models. For each symptom, respondents who reported taking any blood pressure medications in the prior 6 months were categorized into whether they reported that the symptom was not a problem, somewhat or moderately a problem, or a big or very big problem. Data from those for whom a symptom was at least somewhat of a problem were included in analyses assessing relationships between perceptions of symptom severity and medication relatedness; and whether perceptions of symptom severity and medication relatedness were associated with treatment dissatisfaction and lack of adherence. Unadjusted logistic regression models compared the likelihood that patients reported that each symptom was related to their medication from its rated severity. Unadjusted multinomial logistic regression models predicted higher treatment dissatisfaction from symptom severity ratings, and separately from ratings of medication relatedness, among those who reported on their treatment satisfaction. Unadjusted logistic regression models predicted the likelihood that respondents reported that they had changed or stopped blood pressure medications due to side effects from symptom severity ratings, and separately from medication relatedness ratings, among those who reported on medication adherence.
Results
A total of 3,071 patients met the inclusion criteria, were enrolled in the study, and were sent a baseline survey. A total of 1,719 participants completed the surveys for a response rate of 56%. The mean age of respondents was 62 and mean BP was 158/91 mmHg, 54% were female, 73% were non-Hispanic White, 25% had diabetes, 19% had cardiovascular disease, and 56% were on 2 or more BP medications (Table 1). Compared with nonrespondents, respondents were more likely to be older, White, have cardiovascular disease, take hypertension medications, and were less likely to be Asian.
Participant Characteristics by Baseline Survey Response Status (n = 3,071)
The proportion of symptoms in 1,601 respondents who reported taking any BP medication in the prior 6 months and reported on side effect symptoms is shown in Table 2. Only 10% of participants were completely without any symptom that they considered at least somewhat of a problem, 52% reported at least one moderate problem, and 39% reported at least one symptom that was a big or very big problem. Tiredness (70%) and urinary frequency (56%) were the most frequent symptoms reported, followed by dizziness (43%), leg swelling (37%), cough (34%), and sexual symptoms (24%). Tiredness (18%) and urinary frequency (12%) were also most likely to be rated as a big or very big problem, while cough (7%) and dizziness (7%) were the least likely. See also Appendix 1 Table 1 for severity ratings of selected symptoms stratified by whether participants were taking the drug most commonly associated with the side effect. There were modest increases in patients noting moderate cough with ACE-inhibitors, both moderate and severe leg swelling with dihydropyridine CCBs, and moderate urinary frequency with thiazide diuretics, corresponding with the known side effect profile of these drug classes.
Severity Rating of Each Hypertension Medication Side Effect Symptom Among Patients Reporting Taking Any Blood Pressure Medications in the Prior 6 Months (n = 1,601)
Among the subset of 1,446 participants who had reported at least one symptom of any severity, Table 3 shows the association between severity of each side effect symptom and whether the respondent answered “Yes” to the question about whether they thought that the symptom was related to their BP medication. For example, from Table 2, tiredness was rated as a somewhat/moderate problem by 828 (51.7%), and a big/very big problem by 293 (38.5%). In Table 3, among the 1,121 who reported that tiredness was at least somewhat of a problem, 82% felt that it was unrelated to their BP medication and 18% thought it was related (first column, “Total with symptom”). Those who rated tiredness as a big/very big problem were more likely than those who rated it as a somewhat/moderate problem to answer that they thought the tiredness was related to their BP medication (26.3% vs 15.1%, P < .001). Conversely, those who rated tiredness as only a moderate problem were more likely than those who rated it as a big/very big problem to answer that they thought tiredness was unrelated to their BP medication (84.9% vs 73.7%, P < .001.) Similarly, the group of participants who rated dizziness, leg swelling, and cough to be a big or very big problem were significantly more likely than those who rated it as less severe to believe that the symptoms were related to their BP medications. Tiredness that was a big or very big problem was least likely to be attributed to medications (26%), while the proportion attributing big problems with dizziness, leg swelling and cough to medications was higher (39% to 49%). The relationship between symptom severity and perceived relatedness was not statistically significant for urinary frequency and sexual symptoms.
Associations Between Severity of Symptoms and Perceived Relatedness to Blood Pressure Medications, Among Those Reporting That at Least One Symptom Was at Least Somewhat of a Problem (n = 1,446)
Of the 1,365 participants who had at least one of the symptoms that was a problem and answered the question about BP medication treatment satisfaction, both higher symptom severity (Figure 2A) and perceived relatedness to BP medication (Figure 2B) were associated with a higher likelihood of treatment dissatisfaction (see also Appendix 1 Table 2 for complete data). For example, 173/586 (29.5%) with any symptom that was a big or very big problem reported being somewhat or very dissatisfied with treatment, while 151/779 (19.4%) with a somewhat or moderately severe symptom reported being dissatisfied (P = .001). Similarly, 187/628 (29.8%) who perceived that any symptom was related to medications reported being dissatisfied, while 137/737 (18.6%) who did not think that their symptom was related to medications were dissatisfied (P = .001). These relationships were statistically significant for individual symptoms, except for symptom severity with leg swelling, cough, and urinary frequency.
(A) Association between side effect symptom severity (somewhat/moderate problem vs. big/very big problem) and blood pressure treatment dissatisfaction. (B) Association between perceived relatedness to medication (related vs. unrelated) and blood pressure treatment dissatisfaction.
Both higher symptom severity (Figure 3A) and perceptions of relatedness to BP medications (Figure 3B) were also associated with a higher likelihood of responding “Yes” to the question about changing or stopping medications due to side effects (see also Appendix 1 Table 3 for complete data). The differences were large and statistically significant for symptom severity of tiredness and dizziness, and for all symptoms felt to be related to medications except urinary frequency.
(A) Association between side effect symptom severity (somewhat/moderate problem vs. big/very big problem) and adherence to blood pressure treatment (changing or stopping medications due to side effects). (B) Association between perceived relatedness to medication (related vs. unrelated) and adherence to blood pressure treatment (changing or stopping medications due to side effects).
Discussion
In this primary care population with poorly controlled hypertension 90% of participants had at least one potential side effect symptom, and 39% of the participants considered at least one symptom to be a big or very big problem. Tiredness, urinary frequency and leg swelling were the most commonly reported problems. However, only between a quarter and a half of participants with big or very big problems perceived a relationship between the symptom and their medication, least often for tiredness (26%) and most often for dizziness (49%). Both a higher symptom severity and a perceived relationship of symptoms to their BP medication were independently associated with lower treatment satisfaction and self-reported nonadherence.
The frequency of side effect symptoms we observed was higher than reported in some previous studies that framed the question specifically about medication side effects rather than about symptoms in general. For example, in a study from a hypertension clinic in Hong Kong, when asked directly about adverse drug effects, 33% of patients reported side effects from antihypertensive medication.23 The most common were dizziness (9%), ankle swelling (7%), fatigue (4%), and cough (3%). Among inhabitants of Uppsala County in Sweden, 20% of those using antihypertensive drugs reported perceived side effects, including most of the ones we asked about (dizziness, cough, tiredness, leg swelling, and impotence in men).13 At the low end of the range, only 8.7% of patients in a large cohort of older people living independently in the community in Melbourne, Australia, 43% of whom had hypertension, reported having medication side effects.24
While selection of patients with moderately severe uncontrolled hypertension may have contributed to the high rates of reported symptoms, our study was also unusual in having screening methods designed to recruit typical patients who would be unlikely to volunteer for studies requiring burdensome enrollment procedures. Furthermore, we asked participants about general symptoms they experienced, not just those they believed were medication side effects. A similar approach was used in the Treatment of Mild Hypertension Study, a blinded randomized placebo-controlled trial comparing 5 classes of antihypertensive drugs with placebo.25,26 It used a 49-item symptom checklist that included related items on our survey to compare changes in symptoms over one year of treatment. An overall symptom score was not different from placebo over one year. Among the most common worsening symptoms in the placebo group were fatigue (34%), increased urination (30%), cough (23%) and dizziness (21%). A high frequency of side effect symptoms similar to our study was observed in a study of patients being newly started on antihypertensive medications.12 In response to a survey that included a checklist of 24 side effect symptoms, 85% endorsed one or more symptoms and the median number of symptoms was 4.
We previously reported that the participants with poorly controlled hypertension in our study (mean BP of 158/92 at enrollment) were also substantially less likely to report high satisfaction with hypertension care (only 29% chose 9 to 10 on a scale from 0 to 10) compared with the contemporary rating of physician care among the general primary care population from the same care system (82%).27 The low level of satisfaction with hypertension care is mirrored in the data reported here about low rates of medication treatment satisfaction, particularly among patients who experienced more severe side effect symptoms or felt the symptoms were related to their antihypertensive treatment. Side effect symptom severity and beliefs about relatedness to medications were also correlated with lower treatment adherence in our study. Others have found that self-reported antihypertensive medication nonadherence is in turn associated with higher BP28 and worse cardiovascular outcomes.29 Thus, the results of this study suggest that there probably are large numbers of uncontrolled hypertensive patients in clinical practice experiencing treatment dissatisfaction due to side effect symptoms, and that better management of these symptoms could improve medication adherence, BP control and long-term cardiovascular health.
Time pressures and competing priorities in primary care can be a barrier to the routine assessment of side effect symptoms, but several short questions such those in this survey could help to identify people with perceived side effects who need those problems addressed before intensifying treatment to improve BP control. This could occur during the review of systems or medication history in a clinical encounter. In addition, in many care systems nurses or medical assistants are conducting previsit and/or standardized visit questions based on pre-existing conditions such as diabetes or hypertension. It may be feasible to use these opportunities to ask about common side effect symptoms in patients with uncontrolled hypertension or suspected medication nonadherence. When asking patients about side effects, a quick assessment of severity and perceived relatedness to medications (exemplified in Figure 1), at least for those with poorly controlled hypertension, could provide additional helpful information to the health care team to recognize those most likely to also have adherence issues and treatment dissatisfaction. The instrument developed in this study could also be useful for future research to better assess the impact of quality improvement interventions on medication side effects, adherence, and control.
The results in this analysis are subject to several important limitations. First, this was a cross-sectional analysis of baseline data from a randomized trial and we cannot attribute directionality or causality to the associations. We chose to look at baseline data since the randomized trial did not show an effect of the pharmacist-led telehealth intervention versus clinic-based care on side effect symptoms over 6 months of follow-up, and symptoms were generally comparable at the 6-month follow-up time point to baseline in both groups.30 Second, we used a single question to inquire about changing or missed doses due to side effects to better understand that specific issue, so the responses likely underestimated adherence problems due to other common reasons such as cost and trouble remembering. Nevertheless, about a quarter of the respondents reported side effects that precipitated changed or missed doses. Some patients may also have been instructed by a clinician to stop medication due to side effect symptoms, so some medication stopping may not have been nonadherence. In addition, due to constraints on survey length we did not ask which specific medicine may have been changed or stopped. We might have seen different relationships if we had asked about specific classes of medication. Third, we conducted this analysis in a large patient population with poorly controlled hypertension and do not know if the findings would translate to a population with better BP control. Fourth, the response rate to the study baseline survey was relatively low in this pragmatic trial that relied principally on electronic health record data to measure outcomes and safety. Although this could have led to volunteer and nonresponse bias, the respondents were generally representative of the overall study participants and are likely similar to patients with poorly controlled hypertension in primary care practice. Finally, the side effect symptoms we asked about are commonly observed in controlled trials in patients randomized to placebo and were asked in the context of a survey that focused mainly on blood pressure management.
A new and brief approach to evaluating side effect symptoms to antihypertensive medications was used in this study. In this patient population with poorly controlled hypertension, it revealed very high rates of side effect symptoms that were often seen by patients as big or very big problems and were significantly related to treatment dissatisfaction and medication nonadherence. More research is needed to explore whether better and more direct ways to identify and manage side effects in primary care practice settings could improve hypertension care and outcomes.
Appendix 1.
Associations Between Severity Ratings of Select Hypertension Medication Side Effect Symptoms and Specific BP Medications Current at the Index Visit, Among Patients Reporting Taking Any Blood Pressure Medications in the Prior 6 Months (n = 1,601)
Associations Between Symptom Severity and Perceived Relatedness to Medication to Blood Pressure Treatment Dissatisfaction (n = 1,365)
Associations Between Symptom Severity and Perceived Relatedness to Medication Adherence to Blood Pressure Treatment Adherence (Changing or Stopping Medications Due to Side Effects) and (n = 1,423)
Appendix 2.Blood Pressure Care Survey
Notes
This article was externally peer reviewed.
Funding: This study was supported by the Patient-Centered Outcomes Research Institute (HS-1507-31146).
Conflict of interest: None.
To see this article online, please go to: http://jabfm.org/content/38/2/312.full.
- Received for publication August 2, 2024.
- Revision received October 2, 2024.
- Accepted for publication October 7, 2024.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.