|Home | About | Journals | Submit | Contact Us | Français|
Efforts to increase blood pressure (BP) control rates in African Americans, a traditionally underserved, high risk population must address both provider practice and patient adherence issues. The Home-Based BP Intervention for African Americans study is a three-arm randomized controlled trial designed to test two strategies to improve HTN management and outcomes in a decentralized service setting serving a vulnerable and complex home care population. The primary study outcomes are systolic BP, diastolic BP, and BP control; secondary outcomes are nurse adherence to HTN management recommendations, and patient adherence to medication, healthy diet and other self-management strategies.
Nurses (N=312) in a nonprofit Medicare-certified home health agency are randomized along with their eligible hypertensive patients (N=845). The two interventions being tested are: (i) a “basic” intervention delivering key evidence-based reminders to home care nurses and patients while the patient is receiving traditional post-acute home health care; and (ii) an “augmented” intervention that includes that same as the basic intervention, plus transition to an ongoing HTN Home Support Program that extends support for 12 months. Outcomes are measured at 3 and 12 months post baseline interview. The interventions will be assessed relative to usual care and to each other.
Systems change to improve BP management and outcomes in home health will not easily occur without new intervention models and rigorous evaluation of their impact. Results from this trial will provide important information on potential strategies to improve BP control in a low income, chronically ill patient population.
Hypertension (HTN) is a chronic condition for which adherence to prescribed medication, dietary and behavioral regimens yields a high “payoff” in terms of improved blood pressure (BP) control (1). Adherence interventions found to be effective over the long term generally have been multifaceted. However, no single strategy or programmatic focus has shown a clear advantage over others (2). This paper describes a randomized controlled trial (RCT) designed to assess the effectiveness and cost-effectiveness of two organizational interventions intended to increase adherence to prescribed HTN regimens and improve BP control among African American patients admitted to home health care. Home care is a relatively unstudied part of the health care system that serves a chronically ill population.
Home health nursing personnel are uniquely positioned to help vulnerable individuals achieve adequate BP control. Keeping abreast of evidence-based practices, teaching patients about their conditions, promoting patient self-management, helping patients adhere to medical, dietary and exercise regimens established by their physicians, and linking them to medical and community resources are central to the home health nurse’s role (3). However, HTN management has not been a high priority for guideline adoption or quality improvement activities in the home care setting (4). By comparing two approaches – one “basic,” shorter-term and less complex, and one longer-term, “augmented” and more comprehensive – and by examining their impact at two points in time (3 months and 12 months post-study enrollment), this study will allow us to derive new and important inferences about the relationship between the complexity and duration of interventions and the scope and duration of their impacts. The data provided by this study potentially will provide significant policy-relevant information in support of revamping or extending the current Medicare home health care benefit.
Home health care is a non-institutional, diverse health care setting that provides skilled nursing and therapy services to a high-risk population characterized by multiple chronic conditions and significant needs for both medical and self-care management. In 2005, Medicare-Medicaid certified home health agencies discharged over 3.3 million patients, most of whom entered care from the hospital (5). Approximately 85% of home care episodes are for people 65 years of age or older; HTN and diabetes are the most common chronic conditions in this population, followed by arthritis and musculoskeletal diseases, heart failure, chronic pulmonary disease and ischemic heart disease (5).
Despite the high prevalence of HTN in the home care population, home health nurses often see elevated BP readings as “a set of numbers rather than a condition to be managed.” A variety of factors may contribute to this view. Clinician “inertia” has been cited as a common phenomenon leading to suboptimal HTN treatment (6, 7). Perceived lower benefits or heightened risk of adverse events associated with treatment may lead clinicians to overlook high BP among older patients, despite the fact that their baseline BP rates are higher and that HTN therapy has been shown to reduce short term risks of stroke and heart failure even among people over 80 (8, 9). Further, very few patients come into home care with a primary diagnosis of HTN. Multiple comorbidities, sometimes compounded by challenging socio-economic or familial circumstances (e.g., low income, poor housing or weak family supports), may lead to competing treatment priorities and divert the nurse’s attention away from HTN within the typical 24 day median length of home care stay (10).
This study focuses on influencing home health care nurse and patient behavior – and indirectly on influencing the patient’s physician – to improve implementation of Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) hypertension management recommendations. The study’s multifaceted interventions were informed by Andersen’s Behavioral Model (11, 12) and Green and Kreuter’s Precede-Proceed Model (13) (see Figure 1). Andersen’s model focuses on the role of predisposing and enabling factors, in conjunction with perceived need, as predictors or drivers of personal health practices and use of health services. The Precede-Proceed model focuses on the role of these factors, in conjunction with reinforcement mechanisms, in promoting provider and patient behavior change.
Andersen’s Behavioral Model has evolved over time as the dynamics of the health care system have changed, and has been adapted for vulnerable populations (11, 12). The core of the model postulates that certain population characteristics can either facilitate or impede patient health practices and service use. “Predisposing” characteristics include demographics, health beliefs/knowledge and social structure. “Enabling” characteristics include personal/family resources and community resources. “Need” includes perceived health and evaluated health. Some of these characteristics are relatively fixed, while others are potentially mutable via outside intervention. This study focuses on mutable patient characteristics to influence BP control behavior and service use. The basic or less intensive intervention study arm seeks to influence patients through strategies aimed at their predisposing knowledge, awareness and perceived need for treatment. Accordingly, patients receive educational materials, home BP monitors, log books and recommended BP targets to help them better understand their condition and gauge their own HTN-related risk or need. The more intensive augmented arm supplements these tools with a 12-month HTN support program that extends beyond the typical home health stay and seeks to influence a broader range of patient-level predisposing and enabling characteristics, including “activation” level or self-management preparedness, link to a regular primary care physician, and appropriate health service use.
We used the Precede-Proceed model (13) to conceptualize factors influencing nurse practice change. The model emphasizes the influence of predisposing, enabling, and reinforcing factors on practitioner behavior. Predisposing factors include individual practitioner characteristics – e.g., training, knowledge and beliefs – that affect motivation to change. Enabling factors include organizational and structural factors – e.g., reminders, checklists or information systems – that facilitate change. Finally, reinforcing factors include incentives that reward selected behaviors. Both intervention arms of the study focus primarily on enabling factors to influence the behavior of home care nurses randomized to intervention status. Specifically, we use email reminders to promote evidence-based practices. The “dual task theory” of performance (14) posits that organizing and presenting cogent information at critical times to reduce clinician overload, along with directing clinicians to tasks needing their attention, will improve the performance of “secondary” preventive tasks in conjunction with hands-on medical care. Chueh and Barnett (15) describe “just-in-time” reminders to identify patients who are candidates for a particular guideline and to send that guideline to the clinician “at the right time and place.” Numerous studies support reminders to improve clinical practice (16, 17). A recent Cochrane review of HTN RCTs found that such reminders effectively directed physician attention to recommended HTN practices and improved patient appointment-keeping (18). In addition, our own heart-failure (HF) research found that just-in-time email reminders significantly improved both nurses’ adherence to HF guidelines (19) and patients’ HF-specific outcomes (20). This study builds on that experience by examining the impact of evidence-based reminders and information feedback for another condition (HTN) in home care, a distinct subpopulation (African Americans), and both separately and in conjunction with a longer lasting and more comprehensive intervention (the HTN Home Support Program).
The project employs a randomized design to assess the effectiveness and cost-effectiveness of the basic and augmented interventions relative to usual care. Each intervention is intended to improve BP control among the target population by increasing evidence-based, JNC 7-consistent practice among nurses, achieving effective pharmacologic management for patients, and strengthening patient adherence to appropriate medication and behavior regimens. The basic intervention delivers key evidence-based information to nurses and patients while the patient is receiving traditional post-acute home health care; the augmented intervention is a more ambitious organizational change that transitions patients to a HTN Home Support Program and extends the information, monitoring and feedback available to patients and their primary care physicians for a 12-month period beyond an index home care admission. Randomization occurs at the nurse level, and a nurse’s initial random assignment to a specific group (usual care, basic treatment or augmented treatment) determines the status for all new patients allocated to that particular nurse’s care for the duration of the study.
The study is being conducted in the acute care division of a large urban, non-profit, Medicare-certified home health care organization. The majority of patients are admitted to home care immediately after a hospital discharge; others come through community referrals (primarily physicians). Usual care involves a nurse “coordinator of care” responsible for developing and implementing a plan of care based on the physician’s orders and the nurse’s patient evaluation at admission. Other skilled professionals (therapists; social workers) provide visits as needed, and home health aides may be assigned to assist with personal care. The agency’s nursing workforce, like that employed by most home health agencies, is mainly generalist, widely dispersed, and receives infrequent on-site clinical supervision.
The study population consists of home health nurses and the eligible HTN patients they serve. Potentially eligible patients are those identified through agency electronic records as 21 through 80 years of age, black, English speaking and admitted to the agency with a primary, secondary, or tertiary diagnosis of HTN (ICD9-CM401, 402, 403, or 404). Only those whose BP at the time of the recruitment interview is >=140mmHg systolic and/or >=90 mmHg diastolic (>=130 mmHg SBP and/or >=80 mmHg DBP for patients with diabetes or chronic kidney disease) are included. The eligibility BP is taken by trained field interviewers blinded to intervention status using a Microlife Model 3AA1-2 device programmed to take three readings and provide an average. This monitor uses an oscillometric algorithm that has been validated using the British Hypertension Society (BHS) criteria. Exclusion criteria include a co-diagnosis of dementia, other organic brain disorder, or other cognitive impairment, severe heart failure, or end stage renal disease. Also excluded are patients on dialysis, those with a kidney transplant, participants in another disease management program, or those with a predicted life expectancy of less than 6 months. A study baseline accrual of 845 patients was targeted to yield an ultimate sample size of 600 at 12-month follow-up, accounting for approximately 30% attrition due to study subject death, nursing home admission, moves out of catchment area, survey refusals and lack of availability within specified 12-month interview window.
The basic intervention relies on care provided by home health nurses during the routine home health stay. The augmented intervention adds an HTN study nurse (hereafter called the “HTN nurse”) and a health educator, who are responsible for assuring a patient’s smooth transition to the HTN Home Support Program and for delivering the main components of that intervention. See table 1. Both interventions are multifaceted in that they focus on at least two of three key individuals – the patient, the home care nurse and the primary care physician – and target two or more aspects of care, including information and feedback, decision support, community linkages, patient self-management tools, and/or delivery system design. They are also multifaceted in that they address not only appropriate pharmacologic management of HTN but also evidence-based behavior changes including modification of sodium intake, physical inactivity and other risk behaviors. All of these intervention components have been tested individually or in various combinations in published RCTs and other research studies, many HTN specific. However, the literature does not “advantage” any individual component or combination, even though it supports multifaceted interventions overall (1, 18, 21).
Patients whose nurses are randomized to both the basic and augmented intervention groups receive educational material. The JNC 7 patient guide “Lowering your Blood Pressure” is mailed to basic group members with the recommendation that they review the guide and ask their home care nurse or primary care physician if they have any questions. Patients in the augmented group receive a home visit by the HTN nurse and a detailed workbook with 10 chapters to be used over the course of the intervention. Each chapter provides information on a specific topic followed by a practical activity or interactive tool to reinforce the learning. For example, the chapter on how to properly take your own blood pressure is followed by a BP log that the patient is asked to complete each week; the chapter on diet is followed by a variety of recipes and a menu planning guide. Whenever possible culturally specific information is integrated into the content, and information is presented at a 5th grade reading level. When the content is more challenging, the study protocol calls for the HTN nurse to review with the patient until the material is assimilated.
Patients in both intervention groups receive their own Microlife Model 3AA1-2 device. Patients receive written and verbal instructions on the proper procedures to follow in taking BP and how to use the device, as well as a recommendation to record their readings regularly to share with their home care nurses and their primary care physicians. The purpose of the device is to increase a patient’s awareness of his/her BP and, ideally, increase participation in behaviors that will help maintain BP control. In the augmented arm the HTN nurse regularly reviews the BP logs with the patient in order to determine if the patient’s physician should be contacted to consider medication adjustments.
Nurses in both intervention groups receive an e-mail targeted to a specific HTN patient shortly after a patient on their caseload has been interviewed and enrolled as a study participant. The message relays information on the key nursing-specific practices required to support the JNC 7 recommendations relayed in a study team created “4M” approach – Measure, Monitor, Modify, Motivate – and a weblink to the guideline. The top box of table 2 displays the first screen of the email message. Additional information follows on the 4 Ms, including instructions on proper BP measurement, medication information, and suggestions on how to work with the patient on promoting healthy behavior. The email suggests a lower target BP for in-home readings as per recommendations in the JNC 7. A second email message goes out 7 days later with information on the “5A” model (see second box on table 2) to encourage nurses to provide their patients with self-management support (22). Agency policy specifies that home care nurses review their email at least once each work day to access new messages and announcements.
The Home Support Program, which is the key additional component of the augmented intervention, is intended to complement the traditional home care service and then extend patient support beyond discharge from the acute home care episode. In view of the Behavioral Model, the HTN Home Support Program provides enabling support to patients to strengthen their self-management readiness and skills, help them adhere to recommended medication and behavioral regimens and communicate more effectively with their primary care physician. Trained intervention staff employ patient-centered motivational interviewing (MI) techniques, found to enhance readiness to change in varied contexts (23, 24), to help patients overcome ambivalence about making health behavior changes and move toward a commitment to change. Each structured counseling session includes: 1) setting a small behavioral change goal in collaboration with the patient; 2) using a 0-10 numeric scale to assess the importance of the set goal to the patient and confidence that the goal can be achieved before the next contact; and 3) collaborative, individualized problem solving to overcome identified barriers to achieving the set goal. Intervention staff participate in a total of 12 hours training on MI. Initially they have two 4-hour “in classroom” educational sessions with an outside MI specialist. These sessions include: 1) the history, purpose, and evidence for MI; 2) the role of MI in the augmented intervention; and 3) a video of MI clinician-patient sessions; and 4) role playing. Two 2 hour “booster” sessions follow over the course of the intervention period.
The HTN Support Program is divided into two phases conducted over a 12-month period (see table 3). In the 3-month Phase 1, the HTN nurse conducts 6 structured phone calls to follow the home visit. At the end of Phase 1, patients and their physicians receive a BP chart. The frequency of contact and level of HTN nurse and health educator involvement in Phase 2 depend on the patient’s BP at the end of Phase 1 and are adjusted throughout Phase 2 depending on the patient’s BP. The health educator alone conducts quarterly calls with patients achieving pre-HTN or normal BP levels and monthly calls with patients at Stage 1 HTN. Patients with Stage 2 BP receive weekly calls (1/month from the nurse and 3/month from the educator) until they achieve better BP control. The duration of each call depends on the need to review medication changes and health updates and the time required to establish goals in the MI component. In general, the HTN nurse’s role is to encourage and support the patient to be the prime contact with the patient’s physician; however, patients who “linger” in Stage 2 sometimes trigger direct and frequent nurse-physician communication to find an effective medication regimen. Involvement in the home support program is also a health behavior intermediate outcome in itself, inasmuch as involvement is hypothesized to be the key vehicle for achieving BP control.
A structured database generates monthly summaries, which are reviewed by a study investigator to monitor fidelity to the intervention protocol. Length of contact time, patient reported BP, goals set, patient ratings of the goals’ importance and patients’ reported confidence in ability to achieve them are entered by intervention staff after each patient contact. In addition, the database allows entry of free text comments, which study staff are directed to include whenever a patient reports difficulty in adhering to an HTN management recommendation. Examples reported to date include new diagnosis of cancer, out of control diabetes, eviction, death of spouse. The quantitative data captured in the database will allow tabulation of frequency/duration of contacts to be used in analyzing intervention dose/response relationships. The qualitative data will be used to help interpret study results.
This study has three major aims: 1) to examine the effect of the basic and the augmented interventions on HTN practice among home health care nurses; 2) to examine the effect of the basic and augmented interventions on patients’ BP outcomes, medications, and HTN self-management; 3) to estimate the costs associated with the respective intervention strategies and subsequent care, and to compare these costs across interventions. To address these aims, we are making use of data from five main sources: 1) mainframe data and patient-level clinical and functional assessment data derived from OASIS, the nationally mandated home care Outcomes Assessment and Information Set; 2) patient clinical records, which will be reviewed for nurse documentation of HTN assessment, management and teaching of self-care; 3) patient interviews conducted at baseline (enrollment in the study) and at 3-, and 12- months post study enrollment; 4) administrative data routinely collected by the agency’s billing and human resources departments; and 5) intervention cost data collected especially for this study. Information to be obtained is summarized in Table 4.
The analysis will estimate the impact of the basic and augmented interventions on nursing practices and processes of care, on patient outcomes and HTN self-management, and on costs. Regression models will be used to control for differences among nurses in the three groups with respect to characteristics that might influence practices or processes of care. Regression models also will be used to control for differences across patient groups that might influence behavior, service use, BP or other health outcomes. Estimates of treatment-control group differences (for both treatments) generated by these models will be tested for statistical significance to determine the extent to which observed differences at the nurse and patient levels are attributable to the alternative intervention strategies.
While this research team has conducted several intervention studies in the urban home care environment, each time we are in the field we are reacquainted with the complexity of the lives of so many of the patients in our low income population. Reflecting upon the adapted Behavioral Model for vulnerable populations developed by Gelberg, Andersen and Leake (13), we have identified several predisposing factors that present significant challenges to patients in all three arms of the study, that are influencing the extent of patient engagement in the home care support program (e.g., BP self-monitoring, participating in scheduled phone calls). Among the most vexing predisposing barriers to full participation are: housing issues (foreclosures, evictions), substance abuse and literacy issues (inability to read or comprehend educational material even at the elementary school level). An additional challenge for staff has been advocating HTN management when many patients have significant co-morbidities (e.g., diabetes, cancer) that cause greater apparent discomfort and fully occupy the immediate attention of patients, families and formal caregivers. HTN support staff acknowledge these issues and to the extent possible provide referrals to additional support services outside the scope of what the project provides (e.g., support groups, housing authority, department of the aging). However, given the limitations of the underlying social safety net, housing stock and educational and health care systems, the staff simply do not have access to a rich enough panoply of resources to fully mitigate or resolve the array of obstacles that constitute the “worry burden” of some patients.
In the category of “enabling” factors, the intervention team also has found challenges in connecting patients to necessary primary care resources. The team is serving an inner-city patient population who commonly use clinic-based delivery services in academic health centers. These clinics, in turn, primarily use rotating interns and medical residents, so a patient may be assigned to a new physician every few months, making the identification and ability to reach a patient’s primary care provider difficult. This challenge was anticipated, but the amount of time spent trying to make a connection and implement an improved HTN management plan before a rotation change has been more time consuming than expected. Another physician issue relates to resistance to making medication changes or focusing on the target ranges recommended by the JNC 7 guideline. In the words of one physician, “As long as her blood pressure is 180 or under that is fine with me.” In instances like this the HTN nurse provides the physician with the patient’s BP log along with JNC 7 information. If the patient continues to have such an elevated BP level and the physician does not make any care plan changes, the home support team assists the patient with scheduling a visit with an HTN specialist.
In all cases, the HTN nurse works with the patient on how to keep a BP log, provides concrete suggestions to support treatment adherence, and queries about possible medication side effects. In addition, the intervention staff work with all patients to develop more effective ways of communicating with their doctor. For example, all patients are encouraged to routinely bring a list of all their medications to each physician visit, prepare a list of questions and concerns in advance, and ask about treatment choices. The strategy is focused on providing patients with improved activation and self-management skills that can be used over the life course for HTN management, as well as for any other chronic condition that needs to be addressed.
Uncontrolled BP contributes significantly to morbidity, mortality, and patient care costs. The underlying premise of the project is that systems change to improve BP management and outcomes will not easily occur in home health care without new intervention models, rigorous evaluation of their impact, or efforts by opinion leaders in the clinical and research communities to disseminate information on strategies that are efficacious and cost-effective. The evidence for strategies that improve HTN is available. Getting patients and providers to put this evidence into practice is what we are striving toward.
Funding Source: This work was supported by a grant from the National Heart, Lung and Blood Institute: Grant #: R01 HL078585 –Home-Based BP Interventions for African Americans (PI: Penny H. Feldman)
Conflict of Interest Disclosures: None