Several factors in this group are associated with suboptimal adherence [15, 21–23]. However, a composite score developed from a basket of variables associated with adherence may not provide clinically useful discrimination, even for individuals from which the predictive model was developed [23]. Other evidence suggests that access to care and medications is more important than sociodemographic factors in mediating adherence and hypertension control [24, 25]. Rather than attempting to estimate adherence from sociodemographic factors, an alternative strategy is, to begin with, efficient scalable methods that support medication adherence for all patients. The clinician and care team can then use reliable methods to detect suboptimal adherence in individual patients who do not have the expected response to effective pharmacotherapy. Addressing the specific factors impacting adherence for those individuals allows a more focused and efficient use of time and resources.

War and disasters are among socioenvironmental factors impacting medication adherence. For example, the COVID-19 pandemic with mandates for social distancing and sheltering-in-place for extended periods of time had an adverse impact on adherence to antihypertensive medications and BP control [3, 26, 27]. Adherence to medications during the 6 months following guidance to distance and shelter was much more strongly associated with adherence in the 6 months prior to this guidance than age, sex, type of healthcare insurance, several major comorbid conditions, and region of the country [3]. Our findings are consistent with other reports that past adherence is the strongest predictor of current or future adherence [28].

In the U.S., most adults ≥ 65 years old have healthcare insurance as this age group qualifies for Medicare. Lack of healthcare insurance was more common among adults < 65 years old, especially in sociodemographic minorities and those with limited education and income status prior to the Affordable Care Act, with Medicaid expansion in 2014 [29]. From 1988 to 2010, adults < 65 years old without healthcare insurance in the U.S. did not have a significant improvement in hypertension control [24]. Despite sociodemographic characteristics similar to uninsured adults, those with public healthcare insurance, mainly Medicaid, had control rates similar to adults with private healthcare insurance, including a 22-percentage point absolute improvement from 1988 through 2010. Healthcare insurance, whether public or private, served as a proxy for access to healthcare and medications [24, 25].

The relationship between the patient and clinician, the clinician’s communication style, the extent of shared decision making, and the quality of team-based care can all impact patient adherence [15, 21, 22]. Trust is critical. Patients must be confident their clinician is competent and accounts for their interests in management decisions.

Direct questioning such as ‘Did you take your medication(s)?’ or ‘Why don’t you take your medication?’ is generally less effective at understanding the patient’s perspective and promoting adherence than reflective questions such as ‘Are you having any difficulties obtaining and taking your medications such as cost or side effects?’.

Patients who participate in shared decisions on their medication regimens are more adherent than patients who are not engaged in the decision [30]. Racial-ethnic minorities are less often engaged in the decisions regarding their treatment than racial-ethnic majorities, which may be a contributing factor to lower adherence in the former [31].

Team-based care and well-functioning patient-centered medical homes are associated with better adherence and risk factor control than when these factors are not present [19–21, 32]. Clinician burn out can adversely impact patient adherence. Clinicians and staff are happier and more productive and have less burn out in an effective team-based care arrangement [20].

Clinicians and staff may not recognize clues linked to suboptimal adherence, including missed appointments or prescription refills, or a poor therapeutic response to medications or combinations of medications that are usually effective. Identifying and addressing factors underlying ‘red flags’ can improve patient adherence and risk factor control [33].

Including access to and cost of care and medications, impact adherence and clinical outcomes [19, 21, 22, 32]. Out-of-pocket costs account for roughly one-third of the variance in adherence, affecting adherence even in patients with the capacity to pay [34].

Complex regimens with asynchronous dosing are barriers to adherence [10, 12, 14]. Fewer medications, and especially fewer pills, which can be implemented using once daily single-pill combinations (SPCs), are associated with better adherence and hypertension control [35, 36]. Patients who reach therapeutic targets more rapidly, who require fewer adjustments in their medication regimen, and who experience no or limited adverse effects are more likely to adhere than patients with a longer period to control, who often undergo multiple changes to their medication regimens, and experience adverse effects, are less likely to adhere to treatment [15, 21, 37, 38].

Adults with hypertension, especially older adults, often have multiple chronic conditions and polypharmacy [39]. Major depression and other psychoses, which are more prevalent among individuals with multiple chronic conditions, can adversely influence adherence as can drug or alcohol abuse and dementia [40, 41].

Some patients do not accept the diagnosis and related treatment. Other patients may not perceive potentially severe complications of a currently asymptomatic disease. If patients believe that prescription medications are ineffective in controlling hypertension or are associated with major adverse effects, then adherence is likely to be adversely impacted [42]. While education is often perceived to improve adherence, education alone often produces minimal change in adherence [15, 21, 43].

Forgetfulness is an important contributor to suboptimal adherence. Several interventions from pill organizers to electronic applications that provide timely and recurrent prompts can improve adherence [15, 21]. Low self-efficacy, or lacking confidence in one’s ability to self-manage a condition or disease, is another documented barrier to adherence [15, 37, 44].

Patients who endorse alternatives to traditional or Western medicine are less adherent to prescription medications [45, 46]. Individuals who discount the future at higher rates are less likely to engage in preventive health behaviors [47–50]. For example, among 422 adults with hypertension receiving care in safety net clinics, their estimated future discount rate was approximately 44% annually [48]. Their discount rate suggested that events perceived to be two or more years in the future did not merit current attention. While these individuals knew that uncontrolled hypertension led to stroke, heart disease, and kidney disease, each 1% increase in the annual discount rate reduced the probability of getting BP checked by 3.5% and not following their doctor’s treatment plan by 1.6%.

While the patient interview is simple, studies suggest it is no better than tossing a coin for assessing adherence [21, 53]. Potential explanations include: (i) time available for the interview, (ii) communications skills of clinicians vary, (iii) inaccuracies in patient self-report as they tend to overstate their adherence either because they do not recall missing doses or because they want to please their clinicians, (iv) adherence is a variable and dynamic process [21, 54].

Another clinically useful method for detecting non-adherence is to measure the antihypertensive response to prescribed medication in the clinical setting [55]. With direct observation, patients are advised not to take medications before their clinical appointments.

Questionnaires are rarely used in routine clinical practice due to time constraints. The relationship between adherence measured by questionnaires and more reliable methods is relatively weak, and questionnaires, like the patient interview, tend to over-estimate adherence [21, 56].

Pill count is frequently used and provides a relatively good view of the medication taken by the patient over time. Among limitations of this method are included the fact that pills can be removed from their container but not taken, medication may be shared with another individual, and the patient may obtain medication from another source [21, 57].

The percentage of days covered by prescriptions provides a rough estimate of drug adherence and persistence [24, 58, 59]. This approach is particularly useful when an electronic monitoring of drug prescriptions in pharmacies is available and assumes that patients take their drugs when therapy is available. Ideally, the data source for drug dispensation covers all sources of supply, which is not always the case.

A criticism of the Medication Electronic Monitoring System (MEMS) is that opening the pill container does not ensure medication was taken. Studies comparing MEMS data and drug concentrations have shown 97% concordance between the two methods, suggesting that when the pillbox is opened drugs are taken [21, 59, 60]. While MEMS is one the most reliable techniques for detecting poor adherence, availability outside research settings remains limited.

Most BP medications can be detected in urine or blood using high-performance liquid chromatography-tandem mass spectrometry [4, 5, 7, 21]. As noted previously, monitoring of drugs and metabolites is useful for detecting and addressing suboptimal adherence and can be cost effective to cost saving. Absence of medication typically indicates it has not been taken for several half-lives. Testing for antihypertensive drugs other than diuretics is limited in many clinical settings. However, even in these settings, if diuretics are prescribed in SPCs with other classes of antihypertensive medications, then adherence to other drugs in the combination can be indirectly assessed.

Prescribing SPCs represents a simple and effective, yet under-utilized strategy for improving adherence, hypertension control, and clinical outcomes. In a claims-based study, 17,465 adults with hypertension were prescribed olmesartan or valsartan, amlodipine, and hydrochlorothiazide in one-, two-, or three-pill regimens [61]. Adherence assessed by at least 80% of days covered was roughly 50% lower with the two-pill regimen and 75% lower with the three-pill than one-pill regimen. Medication gaps of 60 days or greater were also significantly greater with the two- and three-pill than one-pill regimen.

In another, claims-based study, 28,999 adults with hypertension receiving SPCs and 28,999 adults receiving equivalent free-pill combinations (FPCs) were selected by propensity-score matching [62]. Several adverse clinical outcomes were reduced when patients received SPCs rather than equivalent FPCs, including myocardial infarction ~4–38%, coronary artery disease ~23–68%, stroke ~23–46%, and heart failure ~32–60%. The clinical outcomes advantage for SPCs over equivalent FPCs coincided with higher levels of medication persistence at one year with SPCs, which is consistent with greater adherence documented for SPCs [35]. Importantly, this and other claims-based studies showing improved outcomes with SPCs suggest that reducing pill counts represents a long-term strategy for improving adherence and clinical outcomes.

A greater number of days of supply with each prescription reduces refill frequency and improves adherence [63, 64]. In a claims-based analysis of 379,658 adults with hypertension, non-persistence during the last 90 days of a one-year period following treatment initiation was lower when 90 as compared to 30 days of medication were prescribed [65].

In some healthcare settings, access to SPCs of antihypertensive medications may be limited, costs may be higher than the separate medications, and coverage for 90- as compared to 30-day refills may be limited. In these settings, the preceding recommendations may not result in better adherence given limited access and financial barriers to SPCs and longer days of supply.

Most adults with hypertension require combination therapy, typically prescribed as FPCs. Moreover, patients with hypertension often have multiple chronic conditions requiring additional pharmacotherapy [66]. Pharmacy visits can be further reduced by consolidating refills, which improves persistence with anti-hypertensive and lipid lowering medications [67].

Even when the patient agrees with the treatment plan, misunderstandings are common. A simple and efficient approach to ensuring the treatment regimen is understood, even when patient-centered and engaged, is to have the patient explain the treatment plan in their own words. Teach back has been associated with improvement in the response and outcomes from pharmacotherapy [68, 69].

SPCs, increasing days of supply, refill consolidation, and teach back are standard components in training clinicians and healthcare teams to implement the American Medical Association’s MAP™ (Measure accurately, Act rapidly, Partner with patients) Hypertension program. MAP Hypertension is associated with a 3–5 mmHg greater SBP response to adding an antihypertensive medication during the intervention than the baseline period [70, 71]. The effect of each component in this training package on the SBP response to treatment intensification has not been assessed.

To optimize this component, monthly clinical encounters are recommended until hypertension is controlled [71].

A majority of adults with uncontrolled hypertension are untreated or receive antihypertensive monotherapy when most require combination therapy to attain BP control [72, 73]. Upon verification that the uncontrolled BP values are accurate and represent usual BP, adding a single antihypertensive medication at half-standard, standard, or maximum recommended dose lowers SBP by ~7, 9, and 11 mmHg, respectively [74]. In contrast, doubling the dose of a single antihypertensive medication from standard to maximum recommended dose lowers SBP ~2–3 mmHg. Initiating therapy with half- or quarter-standard doses of two to four antihypertensive medications, then doubling the doses of all components concurrently is more efficient than initial monotherapy with a traditional stepped algorithm [75–77]. The initial SPC approach is generally well tolerated for patients who are not frail and have baseline SBP more than 10 mmHg from goal.

Adding a BP medication class for patients with verified uncontrolled hypertension is appropriate for most patients on less than three antihypertensive medications. When the patient is uncontrolled on three or more antihypertensive medications, then a management approach for apparent treatment resistant hypertension is appropriate [72, 73, 78].

Self-measured BP (SMBP) values relayed to the healthcare team, combined with timely advice from the healthcare team to the patient (advice), can be accomplished without an in-person healthcare visit. SMBP with support is effective for improving BP control [79] and lowering BP for patients with adverse social determinants of health [80]. Team-based care, which trains patients to measure their BP and guides them on actions to take for uncontrolled values, which is supported by efficient electronic transfer of SMBP readings, can provide an efficient, cost effective method for improving hypertension medication adherence and control [81, 82].