Testing Whether Home Visits by Community Health Workers Help African-American and Hispanic Patients with Low Incomes Better Manage Asthma

Testing Whether Home Visits by Community Health Workers Help African-American and Hispanic Patients with Low Incomes Better Manage Asthma - NCBI Bookshelf An official website of the United States government Here's how you know The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site. The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely. Log in Show account info Close Account Logged in as: username Dashboard Publications Account settings Log out Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Bookshelf Search database Books All Databases Assembly Biocollections BioProject BioSample Books ClinVar Conserved Domains dbVar Gene Genome GEO DataSets GEO Profiles GTR Identical Protein Groups MedGen MeSH NLM Catalog Nucleotide OMIM PMC Protein Protein Clusters Protein Family Models PubChem BioAssay PubChem Compound PubChem Substance PubMed SNP SRA Structure Taxonomy ToolKit ToolKitAll ToolKitBookgh Search term Search Browse Titles Advanced Help Disclaimer NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health. Testing Whether Home Visits by Community Health Workers Help African-American and Hispanic Patients with Low Incomes Better Manage Asthma Andrea J. Apter , MD, MSc, Tyra Bryant-Stephens , MD, Knashawn H. Morales , ScD, Xiaoyan Han , MS, Luzmercy Perez , BA, Alyssa N. Mullen , MHA, John T. Howell, III , MD, and A. Russell Localio , PhD. Author Information and Affiliations Authors Andrea J. Apter , MD, MSc, 1,2,3,4 Tyra Bryant-Stephens , MD, 3,5 Knashawn H. Morales , ScD, 3,6 Xiaoyan Han , MS, 3,6 Luzmercy Perez , BA, 1,2 Alyssa N. Mullen , MHA, 7 John T. Howell, III , MD, 4 and A. Russell Localio , PhD 3,6 . Affiliations 1 Division of Pulmonary, Allergy, and Critical Care Medicine, University of Pennsylvania, Philadelphia, PA 19104, United States 2 Department of Medicine, 3400 Spruce St, University of Pennsylvania, Philadelphia, PA 19104, United States 3 Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, United States 4 University of Pennsylvania Health System, Philadelphia, PA 19104, United States 5 Children's Hospital of Philadelphia, Philadelphia, PA 19104, United States 6 Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, United States 7 Temple Physicians, Inc., Temple University Health System, Philadelphia, PA 19104, United States Washington (DC): Patient-Centered Outcomes Research Institute (PCORI) ; 2020 Jun . Copyright and Permissions Copyright © 2020. University of Pennsylvania. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Structured Abstract Background: Asthma, a chronic treatable disease, disproportionately affects low-income and minority adults, particularly African Americans and Puerto Ricans. Improving access to care and patient-provider communication, believed essential for better outcomes, increasingly rely on information technology including features of the Electronic Health Record (EHR), such as a general –purpose patient portal that offers web-based communication with providers and practices. The portal allows patients to access parts of their medical record, review test results, make appointments, and request refills, and it provides a secure platform for electronic messaging with providers. How patients with limited resources and educational opportunities can benefit from portals is unclear. In contrast, home visits by community health workers (CHWs) specially trained for asthma have improved access to care for children with asthma and promoted caretaker-clinician communication. Objectives: Compare the effects of patient portal training (PT) to patient portal training plus home visits (PT+HV) on adults with moderate to severe asthma. Study outcomes were: 1) asthma outcomes (asthma control [primary outcome], asthma-related quality of life, ED visits, hospitalizations) within each study arm over time, 2) asthma and general health outcomes (ED visits, hospitalizations) between study arms over time. Other study goals were to test whether communication (use of the portal), use of inhaled corticosteroid, and access (appointments made/kept) mediate the interventions' effects on asthma outcomes and to evaluate the effect of the interventions on outcomes in relevant subgroups using latent class models. Methods: Three hundred one predominantly African American and Hispanic/Latino adults with uncontrolled asthma, recruited from low income urban neighborhoods, were directed to the most convenient internet access and taught to use the portal, with and without home visits from a CHW. In home visits, CHWs 1) trained patients to competency in portal use, 2) enhanced care coordination, 3) communicated the complex social circumstances of patients' lives to providers, and 4) compensated for differences in patients' health literacy skills. In a two-arm randomized controlled comparative effectiveness trial, patients across 7 sites were assigned to the intervention (home visits by CHWs plus training in the use of the patient portal) versus an active control (patient portal training). Using an intention-to-treat analysis, longitudinal models were employed to assess changes over time from randomization to 12 months in the primary outcome, asthma control, and the secondary outcomes: asthma-related quality of life, emergency room visits, hospitalizations, and prednisone bursts. All outcomes were self-reported, but where possible validated through the EHR. Results: About half of the patients had access to a computer at home or work. Fewer than 60% of patients in each group used the EHR patient portal during the study follow up. Both groups (PT + HV versus PT) used it at the same rates over time. Both groups improved, with fewer asthma symptoms, better quality of life, less need for oral steroids, fewer asthma-related emergency visits and hospitalizations per year, regardless of whether they received home visits or not. In all measures of effect, the group receiving home visits showed more improvement, but in only one of these outcomes, hospitalization rate per year, did the difference between groups achieve conventional levels of statistical significance (−0.53, 95%CI: −1.08, −0.24). Conclusion: For low-income adults with uncontrolled asthma, particularly African American and Hispanic/Latino patients, use of the EHR patient portal was frequent and regular by only a few patients. In Philadelphia, impoverished adults experience inadequate access to home internet that reduced their ability to use the portal and communicate with CHWs and their medical providers. Even with the addition of home visits, including care coordination and portal training, improvements were limited. Limitations: Larger benefits of the home visits might have been elusive owing to 1) a small number of home visits and over a limited number of weeks; 2) lack of a Spanish-language patient portal; 3) inadequate home internet access that reduced the ability of the clinical practice to remain in frequent contact with the patients; and 4) a portal that is not asthma-specific and therefore does not support patients with day-to-day asthma management, their other health problems, and socioeconomic barriers to health. Background Asthma, a chronic treatable disease, affects 18.7 million US adults.( 1 ) Despite efficacious medications and national guidelines for management,( 2 ) asthma has a disproportionate impact on low-income and minority adults, particularly African Americans and Puerto Ricans.( 3 , 4 ) Compared with children, adults, 65% of whom are women, are more likely to die from asthma.( 1 , 5 ) There is much less research on adults than children with asthma. Comorbidities including hypertension, diabetes, and obesity, not infrequent in adults with asthma, increase the likelihood of adverse asthma-related outcomes and make access to care and patient-provider communication more complex. For example, if a patient's asthma is uncontrolled and requires prednisone which raises blood sugar, options for treatment must consider management of both diabetes as well as asthma. According to the Institute of Medicine, improvements in access and patient-provider communication are needed to reduce health outcome inequities.( 6 ) The patient portal of the Electronic Health Record (EHR), increasingly available to enhance patients' web-based communication with providers and practices, has the potential to improve communication and access. The 2009 Health Information Technology for Economic and Clinical Health Act authorized the Centers for Medicare & Medicaid Services to provide reimbursement incentives for eligible professionals and hospitals who become “meaningful users” of certified EHR technology.( 7 , 8 ) Among the core objectives of Meaningful Use is the promotion of patient portal use by patients and providers. The portal allows patients to access parts of their medical record, review test results, make appointments, and request refills, and it provides a secure platform for electronic messaging with providers. However, the portal is not widely used and is less available to low-income and minority patients.( 9 ) Recently, there have been numerous studies and several systematic reviews of patient portals and patient engagement.( 10-13 ) Several of the portal studies are focused on diabetes but most are not devoted to any one disease. In design most studies are observational and not interventional. Studies consistently found that use of the portal was less likely in older patients; ethnic and racial minority patients and individuals with limited English proficiency, and poorer health status.( 14 ) One study found that older patients with limited health literacy required more time to complete portal tasks and were more like to have navigation barriers to reading, writing typing.( 15 , 16 ) No study clearly showed a benefit to health with use of a portal.( 13 ) The authors recommended research to better align portal features to patients' needs. A recent systematic review of barriers, facilitators, and solutions to optimal portal use argued that qualitative studies identified many barriers to optimal portal use and recommended more qualitative research.( 12 ) Community Health Workers (CHWs) have helped patients to gain access to and communication with health practices and to overcome barriers to care coordination.( 17 , 18 ) Some CHW interventions include home visits.( 19-22 ) CHWs potentially can give clinicians insight into patients' lives and the social and economic barriers to access and communication, including barriers to using the portal. Patients widely accept CHWs as advocates who share cultural, economic and linguistic characteristics with them, who know the community, and who are able to build trust.( 23 , 24 ) We are not aware of any studies using CHWs to facilitate use of patient portals. We are also unaware of any interventions to measure the impact of increasing the use of the patient portal on health outcomes. Limited access to computers among disadvantaged populations may actually exacerbate disparities in health care if medical practices increasingly rely on information technology rather than telephones.( 25 , 26 ) This study examined the benefits to low-income, urban, predominantly African American and Hispanic/Latino adults with uncontrolled asthma, of using the portal with and without home visits by CHWs who encouraged and facilitated portal use, understood patients' social context, and enhanced communication with the medical team.( 27 ) The primary hypothesis was that asthma would improve in adult patients with the use of the portal and that the addition of home visits by CHWs would offer even more improvement, especially for those with low literacy or language barriers. We compared portal training and home visits by a CHW with an active control of portal training (without home visits). Our specific aims were to estimate whether—one year after randomization of adults with uncontrolled asthma to either Patient Portal Training (PT) or Patient Portal Training plus Home Visits from a CHW (PT+HV)—these interventions result in 1) better asthma outcomes (asthma control, quality of life, ED visits, and hospitalizations) within each study arm over time, 2) better asthma and general health outcomes (asthma control, quality of life, ED visits and hospitalizations) in the PT+HV vs. PT arm over time, 3) more communication (use of the portal) and access (appointments made/kept) mediating the interventions' effects on asthma outcomes, and 4) identification of relevant subgroups using latent class models of the effect of the interventions on outcomes. Participation of Patients and Other Stakeholders Patients and their clinicians are important stakeholders and were consulted at all stages of the project. At the start of design of the project, we developed interventions from focus groups of patients and clinicians and then in a pilot project.( 28-30 ) In one set of focus groups, the patient navigator was described to asthmatic adults and these adults were asked what such a navigator could do to help asthmatic patients.( 30 ) In separate focus groups providers were asked what such a navigator could do to help patients.( 30 ) Another set of focus groups described meaningful use and the patient portal and asked patients of their interest and thoughts.( 29 ) We engaged the expertise and insights of patients and their families, their healthcare providers, and health system administrators who must enroll patients in the portal and implement Meaningful Use (https://www.cdc.gov/ehrmeaningfuluse/introduction.html). From these stakeholder groups we recruited members not otherwise currently engaged in related research to our research team. Patient stakeholders were recruited from community organizations (e.g., Congreso de Latinos Unidos) and from participants in the Community Asthma Prevention Program (CAPP), a community asthma education program founded 20 years ago in Philadelphia by Dr. Bryant-Stephens which utilizes CHWs. Patients who had participated in the pilot studies for this project were also invited. Patient stakeholders were consulted in the design, conduct, and monitoring of the study. From them we formed a Community Advisory Board (CAB) of patients and family members who met quarterly to comment on the project's progress and offered suggestions. Of the 10 members invited, some came only once, but eventually 4-5 came consistently. A community member with asthma attended research meetings. As an example of his input, he suggested ways of making items more specific about the portal on the post-study participant feedback forms (i.e. End of Study Questionnaire and Home Visit Satisfaction Survey). In addition, the Data Safety Monitoring Board included a community member. As an example of how engagement of stakeholder partners changed our research, in a CAB meeting we presented the difficulty in contacting patients to arrange home visits and data collection. The CAB members proposed “drive-byes,” driving to participants' home and knocking on their door if unable to contact by phone. With this technique, our team was able to contact difficult-to-reach participants. From treating clinicians, another important stakeholder group, we recruited a member from each practice as a research team member to attend monthly meetings. Our research team regularly visited all practices and communicated with clinician representatives of these practices. Discussions with clinicians focused on ways to prevent ED visits and hospitalizations, and courses of prednisone. We regularly consulted an administrator and representative of information services of each health system (University of Pennsylvania and Temple University). Consultants from each health system provided: ( 1 ) lists of patients' names and contact information for recruitment, ( 2 ) assistance in linking the patient portal use to the main database by transforming raw portal data into discrete episodes of communication, and ( 3 ) a report of appointments made and kept by patients throughout the duration of their participation in the study. Methods Study Overview Prior work by the principal investigators and key staff on asthma patients in both clinical and research settings informed the design and protocol specifics of the current study.( 15 , 16 , 20 , 28 , 31-39 ) The overall study description has been previously published.( 27 ) We describe methods in full here, and in more detail in the attached Appendix. Adults with uncontrolled asthma, generally unfamiliar with the patient portal, (“generally unfamiliar” meant not having logged in more than 3 times.) were recruited from outpatient primary care and specialty clinics of two health systems that serve predominantly but not exclusively low-income urban neighborhoods. Participants were trained in the use of the portal. Participants were then randomized 1:1 to home visits from a CHW (described below), to encourage use of the portal and to facilitate care coordination, or to no home visits. The project was approved by the University of Pennsylvania Institutional Review Board and was registered with ClinicalTrials.gov ( NCT02086565 ). Study Design We conducted a randomized controlled trial utilizing a two-arm comparative effectiveness design with an active control arm, to ensure that all participating patients would have incentives for consenting to participation and for remaining in the study. The longitudinal design uses baseline measures to provide within-subject controls, and the multicenter recruitment also enhances the generalizability to a diversity of social circumstances. The choices for the randomized trial, portal training and home visits, were inspired by the RE-AIM conceptualization framework( 40 ) and the need for implementation research.( 41 ) Implementation of portal training and home visits were hypothesized to improve access and communication, resulting in improved patient-centered outcomes including asthma control and asthma-related quality of life. Although CHWs had no prior contact with the patients randomized to PT+HV, CHWs did work with the same patients with whom they had conducted each patient's portal training sessions. At each home visit, the CHW ensured care coordination, including obtaining an asthma action plan that linked patients' home and community with the clinic. The CHWs reviewed medications including the use of inhalers, provided information on smoking cessation, promoted communication with providers, encouraged appointment keeping, and facilitated familiarity with health information technology. Study Setting A patient portal is a secure website linked to a patient's medical record which gives patients access to their personal health information from anywhere with an Internet connection. Using a secure username and password, patients can view health information such as: appointments, lab results, prescribed medications, and immunizations. It can be used to request an appointment, obtain refills of medications, and to message a clinician can be accessed by a computer or a smart phone. Study sites were chosen to recruit patients who satisfied enrollment criteria and whose clinical experience was diverse (primary care versus specialist care, different health systems – University of Pennsylvania and Temple University). Participating sites included two family medicine, two general internal medicine, two pulmonary, and one allergy-immunology outpatient practice from the University of Pennsylvania Health System (UPHS). Recruiting from these practices in the past had consistently yielded a population that was more than 65% African American and more than 65% female. Candidates for the study were identified using the demographic, diagnostic, and prescription information in the Epic electronic health record systems of the two participating health care organizations. Because University of Pennsylvania clinics are not in Hispanic/Latino areas of Philadelphia, we recruited and enrolled patients from Hunting Park Adult Medicine, of Temple University Health System, a primary care practice serving mainly Spanish-speaking patients, more than 85% Hispanic/Latino, and with more than 700 patients with asthma on their problem lists. Participants We planned to identify 300 adults living in low-income neighborhoods with moderate to severe asthma requiring urgent care and little previous exposure to the portal. Inclusion criteria From the Epic electronic health record we identified adult patients with the following characteristics and recruited only one from a household: 18 years or older living in a Philadelphia neighborhood in which at least 20% of households have incomes below the federal poverty level having a doctor's diagnosis of asthma prescribed an inhaled corticosteroid-containing medication required prednisone or had an ED visit or hospitalization for asthma in the past year were patients in a participating clinic. To restrict our study to patients who were naïve or relatively inexperienced with the patient portal, we further limited our sample to patients who reported having never used the portal or having signed in three times or fewer. This restriction parallels a “new user” design and allows for a separate estimation of the potential impact of the patient portal alone. Exclusion criteria Severe psychiatric or cognitive problems (e.g., obvious mania, schizophrenia, significant mental retardation) that would make it impossible to understand and carryout this protocol. Recruitment Using guidance and policies of the supervising institutional review board and the participating health care systems, we recruited patients as follows. After explaining the protocol in staff meetings of participating sites, we received lists of potentially eligible patients (age > 18 years, asthma in the problem list of the electronic health record, prescribed an inhaled corticosteroid, having an address in a low-income neighborhood) for whom further screening was needed. First, we sent “opt-out” letters to clinicians asking for permission to contact their patients. If providers either did not respond to two letters or gave permission, we sent letters to potential participant patients asking to contact them for screening. If patients gave permission, or did not respond, we called the potential participant or approached them at a clinic visit to explain the study and ask for permission to screen for eligibility. According to their preference, we screened potential participants at their clinic or in their home. Data collectors completed a screening form for each patient contacted for recruitment which included a section to document reasons for declining. The following were reasons for declining: inability to commit to study timeline, disinterest in research participation, and patient illness. Participants all signed informed consent to enroll in the study and to undergo Portal Training and data collection. Those randomized to home visits signed a second consent form informing them they were also randomized to home visits and giving permission for these visits. Randomization and allocation concealment Randomization was stratified by clinical site to guard against chance imbalance of treatment assignment by site and to avoid possible confounding. To maintain allocation concealment, we implemented randomly permuted blocks with varying block sizes (2 to 4). For each clinic we prepared opaque envelopes and instructions on how to preserve allocation concealment until the patient gave consent and the envelope was unsealed. Randomization programs and envelopes were prepared under the supervision of one of the project statisticians who then was blinded to the identities of patients until after recruitment and randomization. Once envelopes were opened, neither patients nor project investigators nor staff were blinded because of the nature of the intervention and the need for constant communication between project staff and patients render precluded continues blinding. Interventions and Comparators: Portal Training and Home Visit with Community Health Worker Community Health Worker (CHW) CHWs live in the community and are familiar with the environment; patients like to work with them. CHWs are able to connect patients to health education, services, and the health care setting.( 21 , 24 , 42 ) In our Community Asthma Prevention Program (CAPP), we identified and trained CHWs as lay health educators to provide home asthma education and environmental intervention. Over 20 years, CAPP CHWs have visited 3000+ families with children( 20 ) in both African American (West Philadelphia) and Hispanic/Latino (North Philadelphia) communities, the same communities for the current project. These CHWs have a high school education or greater with at least three years' work experience and personal knowledge of asthma. They bring their knowledge as a community resident to suggest local resources to meet daily needs. CHWs are able to establish relationships with participants and their families to promote better asthma self-management, connection with community resources, communication with providers, and mitigation of asthma triggers. CAPP's CHWs have functioned not only as home visitors, but also as asthma navigators who have been integrated into the health care team, providing a direct link between families and clinicians. Our experience has allowed us to develop an excellent support system including supervision, training, and systems for scheduling visits and data entry.( 20 , 43 ) The retention rate of participating families in CAPP has been above 90%.( 20 ) Home visits for children with asthma often include tailored asthma education and environmental remediation of allergen and irritant environmental exposures. In the current intervention, each home visit included 1) reinforcement of the use of the portal and 2) care coordination ( Table 1 ) from one of 4 CHWs. Our activities were unique in utilizing CHWs for home visits for promotion of portal use with the intention of linking the patient to the medical practice, thereby improving patient-physician communication. The home visits focused on communication and access, goals potentially enhanced by portal use. Table 1 Activities and data collection (DC) for PT+HV and PT groups. The first home visit (HV1) took 30 to 60 minutes. The subsequent home visits HV2 and HV3 took 20-30 minutes. Components of the Interventions Portal Training and Home Visits represent different approaches to improve patient-provider communication and care coordination as well as access.( 29 ) Although vastly different interventions, both in theory provide potential synergistic benefit particularly for vulnerable patients exposed to poverty and with prevalent comorbidities.( 29 ) Although portals have been widely employed and undergo continuous refinement,( 9 , 25 , 26 ) their benefits in adult inner city asthma patients require investigation. In addition, prior research on portal use has focused on the uptake of portal, the frequency of use, and patient and clinician impressions or attitudes rather than the influence on health outcomes. We hypothesized that home visits may be especially effective for low-income inner city patients facilitating two-way patient-clinician communication; informing clinicians of important environmental, social, and medical barriers to asthma self-management; and reinforcing medical information from clinicians to patients.( 29 ) We also hypothesized that the home visit may be especially effective for those who speak English as a second language, in our sample mostly Spanish-speaking patients, or those with low numeracy and health literacy. We previously had piloted the home visits in families with children( 20 ) and the activities to be performed in home visits with adults.( 28 ) All participants received Patient Portal Training (PT), and those randomized to Home Visits additionally were scheduled for four Home Visits by a CHW to take place over 6 months: at 2-4 weeks, 4-6 weeks, 6-11 weeks, and 23-27 weeks following randomization. All participants were to be followed for at least a year ( Figure 1 ). For that reason, the study team followed no time windows for patient contacts, visits or data collection. Rather they attempted to adhere to the initial, pre-specified time schedules and then noted the actual dates of contacts and visits but did not exceed the pre-specified total number of visits or data collections. Figure 1 Planned time course for participation and data collection. Patient Portal Training (PT) In pilot studies, prior to the current project, we ascertained portal familiarity and interest through separate focus groups of patients and providers exploring their knowledge of and use of the portal.( 29 ) Neither the previous pilot studies nor the current study designed or modified the existing Epic patient portals. Rather, we based training on the design in place in the clinical sites. In the earlier studies, in order to develop and pilot portal training, we had asked 10 focus-group patients who did not have portal access to register for it with our help.( 29 ) All were able to use the point and click mechanism within the portal to accomplish 7 tasks: 1) locate a laboratory test result, 2) look up an upcoming doctor's appointment, 3) learn how to schedule an appointment with their provider (the opportunity to actually make the appointment was offered), 4) locate their medication list, 5) locate their immunization record, 6) determine how to request a refill, and 7) send a secure message to their care team. All were positive that they could accomplish the tasks and found them useful. This protocol became the basis for the two portal training sessions of this project, PT1 and PT2 ( Table 1 ). At PT1 a CHW inquired whether internet access by computer, tablet, or smart phone was available at home, work, community hot spots, or participating clinics, and together the participant and CHW confirmed which was closest and most convenient. As part of inclusion criteria, patients were asked about prior portal use. For this project, in PT1 the CHW described the portal and its functions, and assisted participants in requesting an activation code. PT2, originally planned for a second visit, often followed PT1 on the same day as more convenient for participants. At PT2, working with the same CHW, participants activated and accessed the portal and performed exercises in using it. Internet access To ensure internet access at home, we originally contemplated providing a tablet to all participants. Stakeholders concluded that even with vendor discounts, tablets would require costly internet access not affordable to many and would be subject to theft. Our information technology specialists pointed to the challenges of installation in older homes. Although the Philadelphia Housing Authority, a provider of subsidized housing for low-income city residents, provides internet access, not all participants live in these homes. Philadelphia also has high speed internet discounts to low income families with children, but not all participants lived with school-age children. From patient focus groups and our pilot of portal training, we estimated that half of potential participants had computer access at home or work. Of those with access, only half of these checked email regularly. We ensured internet access for participants at primary care and asthma specialty clinics and in the NIH-funded Clinical and Translational Research Center. The medical facilities of the participating practices are “hot spots.” At enrollment we recommended ways to access the internet: nearest library, nearest hot spot (recognizing low income communities have fewer), and if desired for those with smart phones, using smartphone apps. Home Visit (HV) Protocol Those randomized to PT+HV worked with the same CHW who conducted both portal training sessions. At each home visit, the CHW ensured care coordination, including obtaining an asthma care/action plan that linked patients' home and community with the clinic. The CHW discussed the medications in the asthma care plan, the difference between controller and rescue medications, and how to properly use the devices. They promoted communication with providers, encouraged appointment keeping, and facilitated familiarity with health information technology. Home visits were intended to link the adult patient to the medical practice through the portal, to provide tailored care coordination and coaching on portal use, and to communicate and hopefully improve asthma control.( 43 ) These activities were selected after conducting focus groups of patients( 30 ) and piloting an introduction to the portal to patients unacquainted with it.( 28 ) The CHW was to empower patients to communicate with their health care providers via the portal, to follow asthma guidelines, reconcile medications, facilitate appointment scheduling, and also control environmental exposures (e.g. tobacco).( 29 , 43 ) Each home visit had two parts: 1) care coordination using the portal and 2) improving familiarity with health information technology ( Table 1 ). For care coordination the CHW and participant drafted an asthma action plan for review, revision, and approval by the participant's asthma care provider, and the participant was asked to schedule through the portal an appointment with the provider in which this plan would be discussed. The American Lung Association Asthma Action Plan format was used and included contact information of the providers, and preventative and emergency medications.( 44 ) The CHW helped patients identify specific care coordination goals that could improve asthma control and quality of life and that with patients' permission could be shared with providers. In previous home visits for children living in the same urban Philadelphia neighborhoods, we found that half of families had at least one smoker in the home.( 45 ) Residential exposures to allergens and pollutants contributed to asthma morbidity. CHWs facilitated patients' communication through the portal with their asthma clinician about exposures to tobacco smoke, pollutants, potentially relevant allergens, and about comorbidities. CHWs were knowledgeable in community resources and had access to a resource database which includes smoking cessation programs and housing opportunities. At the first home visit, HV1, designed to occur as early as within 2 weeks of the second portal training session, PT2, the CHW became further acquainted with the participant, if possible met the family, collected information, and conducted a needs assessment. The CHW gathered information on medications in the home to coordinate with the participant's clinician, facilitated appointment scheduling, and if applicable encouraged smoking cessation and reduction of secondhand smoke exposure. From the needs assessment the CHW and the patient created care coordination goals. Examples of these goals chosen by patients were to eliminate asthma triggers from the home, to stop smoking or to exercise more. The action plan was then drafted by the participant with the assistance of the CHW for review, amendment, and approval by the participant's asthma care provider. At the second home visit, HV2, intended to take place within 2-3 weeks of HV1, the CHW reviewed the individualized action plan, asthma control, the goals from the previous visit, and relevant asthma education. Use of the portal was reviewed and encouraged. The third visit, HV3, was to occur approximately 2-4 weeks later. At the final visit, HV4, intended to take place approximately 6 months after enrollment, the home visitor and patient reviewed care coordination and all portal and internet skills ( Table 1 ). To improve familiarity with health information technology, CHWs reviewed use of the portal as the second part of each visit. In addition, at each visit they taught a relevant activity such as using the internet to access health information, how to obtain educational materials through the portal, how to “Google,” how to email, and how to get a weather report. Follow-up At the end of each visit the participant and CHW completed a report to the clinician of topics discussed (e.g. action plan, medications, other topics). Although use of the portal was encouraged, CHWs and patients found it more feasible to use paper, which the CHW gave to clinic personnel for the clinician's mailbox. Goals for the participant to pursue with the asthma doctor to achieve asthma control were also reported. Quarterly phone calls for the remaining 6 months were planned with information on portal training to be reviewed along with information from the prior home visits. The use of the portal was encouraged. Study Outcomes The outcomes selected to be meaningful to patients, were among the outcomes recommended by the recent Asthma Outcomes Workshop ( Table 2 ).( 46 ) The primary outcome, asthma control, reflecting symptoms over the past week, was measured by the 7-item Juniper Asthma Control Questionnaire (ACQ).( 47-49 ) The score is the mean of all responses (0=total control, 6=extremely uncontrolled). The minimal clinically important difference or change is 0.5. A score >1.5 is considered uncontrolled.( 50 ) Asthma-related quality of life (AQOL) was measured with the Mini Asthma Quality of Life Questionnaire (AQLQ).( 51-53 ) This 15-item questionnaire has a 7-point response scale that provides a mean summary score. A 0.5-unit change is considered clinically meaningful.( 53 ) The AQLQ has been shown to be a useful indicator of asthma-related quality of life in low-income adults.( 54 ) Hospitalizations including ICU admissions, ED visits, urgent medical visits (scheduled < 24 hours in advance), prednisone bursts (a new prescription of prednisone or an increase in an already-prescribed dose for an asthma exacerbation), and other medical visits were obtained by self-report because not all such events occur within our health system. The medical records were examined for documentation as feasible. Spirometry was obtained using American Thoracic Society procedures for FEV1 and FVC.( 55 ) Table 2 Summary of data collection. We planned to collect data quarterly at months 3,6,9, and 12 (final visit) but could not always contact the patient to collect this in a timely manner. Statistical Power We estimated as follows that300 patients, 150 per group would be needed to detect a difference in outcome in patients randomized to PT + HV over PT (without home visits). Estimates of power for longitudinal data must account for the added power of following individuals over time. To estimate power with our somewhat complex design, we used statistical simulations, implemented in Stata v 13 (StataCorp, College Station, TX), in which we varied the correlations of outcome within individual over time. Using the asthma control measure as an example, we estimated that power of a balanced design with 300 evaluable subjects was adequate (from 0.82 to 0.87) to detect a difference between groups of 1/3 of a standard deviation of asthma control. According to Juniper's work on asthma control, this detectable difference is less than the minimal clinically important difference.( 49 , 50 ) Additional simulations, for effect modification, suggested good power (0.9) to demonstrate a greater improvement among low literacy patients with home visitors. For all results, the width of reported confidence intervals reflects the actual power of the study as conducted to detect clinically important effects of the interventions. Time Frame for the Study All participants were to be followed for data collection every 3 months for at least 1 year ( Figure 1 ). We chose a 3-month interval for data collection, as the Expert Panel Report on diagnosis and management of asthma uses 3 months before making a change such as “stepping down” (reducing medications). ( 2 ) We chose the times for home visits, based on Dr. Bryant-Stephens' experience with home visits in children. She found it was important to have home visits early in participation.( 38 ) We wanted the last 6 months to be for observation to assess sustainability. Nevertheless, when timing of visits and data collection lagged from the initial timetable, project staff collected data as soon as possible and did not restrict themselves to a preset window of opportunity. Data Collection and Sources The IRB-approved consent and all communications were read/spoken in English or Spanish as preferred by the potential participant. Data collection was accomplished by a CHW, called a Data Collector, who did not act as the home visitor if the patient was so assigned. This choice of Data Collector reduced any pressure on the patient to answer in a way to please the researcher CHW. Baseline questionnaires assessed socio-demographics, asthma severity, health literacy, and comorbidities ( Table 2 ). The first and last data collection visits were in person so that spirometry could be measured. For the other data collection times, we offered a phone or email or in-person visit to minimize the burden of and effect on participation in a study. The Data Collector entered responses either directly into a secure, encrypted, web-based database REDCap (Research Electronic Data Capture) or on paper for transfer to REDCap later.( 61 ) Data Collectors used tablets (iPads) equipped with Wi-Fi but internet access was difficult at times, and CHWs noted many patients were not comfortable with having their information entered onto an iPad. In addition to collecting pre-coded responses, data collectors also documented impressions and observations in free-format text for description and formal discussion at weekly conferences. For each patient we collected their contact information and that of three persons. We also were able to read the electronic record to determine when patients had appointments and to meet them there. Additionally, we wrote letters, visited homes, and left postcards. Data Management REDCap ( http://www.project-redcap.org/ ) allows data attribution and audit capabilities, integrity checks, real-time validation, data storage and backup, and export functions.( 62 ) Whenever any response of the patient required elaboration, the interviewer entered comments into this database. At weekly meetings, the community health workers, project manager, the staff who entered data into REDCap, the project statisticians, and the principal investigators discussed data collection problems. The principal investigator then reviewed all reported adverse events. These meetings then led to the development of project rules for interpreting any of the questions that presented difficulties for patients to answer. Project investigators also queried CHW staff about challenges in contacting patients and about patients' physician condition, living and social environment, and problems with care. A set of 30 tables stored in the REDCap database contain data from interviews and phone assessments of patients over time ( Table 2 and Appendix A1 ). Analytical and Statistical Approaches Analysis summary Three hundred one predominantly African American and Hispanic/Latino adults with uncontrolled asthma, recruited from low income urban neighborhoods, were directed to the most convenient internet access and taught to use the portal, with and without home visits from a CHW. In home visits, community health workers 1) trained patients to competency in portal use, 2) enhanced care coordination, 3) communicated the complex social circumstances of patients' lives to providers, and 4) compensated for differences in patients' health literacy skills. Using an intention-to-treat analysis, longitudinal models were used to estimate treatment effects in the primary outcome: asthma control, and the secondary outcomes: asthma-related quality of life, emergency room visits, hospitalizations, and prednisone bursts (please see Table 1 and Appendix A1 ). Components of our analysis required merging data on portal usage by participants from two health systems: Penn Medicine and Temple Health. Details on our data management approach to the Portal system appear in Appendix A2 . Aim 1 tests if PT alone and PT + HV are individually associated with improved asthma outcomes. Aim 2 assesses whether adding home visits to PT improves asthma outcomes. The merged data set also allowed us to complete the mediation analysis (Aim 3). Candidate mediators were rate of portal use, more appointments kept, and better adherence to inhalers (using self-report from a validated questionnaire). Aim 4 evaluated whether certain subgroups benefitted from the interventions. Statistical Analysis plan Aims 1 and 2: intention to treat approach – primary analysis The primary analysis was “as randomized” (intent to treat), with the assumption that any dropout visits were missing completely at random (MCAR). The “as randomized” approach estimated the effectiveness of the intervention as it was actually conducted in a typical clinical setting rather than the efficacy of the intervention under ideal or optimal settings. But because dropout can defeat the balance achieved by randomization, we conducted sensitivity analyses to the potential effect of dropout. The estimand of interest was the difference between the PT+HV versus PT groups in the change in outcomes over time. We attempted to use mixed effects models, which make less strong missing at random (MAR) assumptions, with varying success. (Details appear below and in the analysis-specific appendix). Estimating within-group effects over the time of the study allowed us to assess the effect of each intervention. In brief, the within-group change in the PT group allowed measurement of PT alone, the active control. Subtracting any change over time in the PT+HV group from the PT group allowed estimates of the additional effect of home visits. These individual estimates within treatment arm over time did not benefit from a third arm of placebo controls. By design we avoided a placebo group because it would adversely affect patient recruitment and retention. Modeling approach – flexible modeling of time Irregular data collection times (see results) required models that used outcome time as a continuous measure. For that reason, analysis models implemented spline-based marginal (GEE) (MCAR assumption) and mixed effects (MAR assumption) longitudinal models, both identity/Gaussian, log/Poisson, and log/gamma with marginal splines to estimate expected values of outcomes at 0 months and 12 months and outcome changes between 0 and 12 months. We used all data collection times and modeled using the actual times. We did not artificially categorize data collection times to equate to the preplanned 0, 12, 24, 36, and 48 weeks from randomization. Nor were data excluded by reason of the lag between preplanned and actual times. The contrast (estimator) of interest was the difference between the treatment groups in the changes in expected values from 0 months to 12 months. In this context the expected value is simply the model-based mean or average effect. Examples in the statistical literature support this approach.( 63 ) All confidence bounds were estimated using 999 bootstrap resamples (percentile-based). By using 999 samples, 95% confidence bounds are easily estimated from the 25 th and 975 th order statistics without interpolation. Details appear in Appendix A3 . For any data collected at home visits we used the same approach as the timing problem was the same. Model form We used both marginal and mixed effects models, with the marginal model being the primary approach owing to convergence problems with mixed effect models (noted below).( 64 ) By including splines for time and group-by-time interaction terms, these models could produce the expected (or mean) values at 0 and 12 months for each patient, even if the patient did not have an observed value at precisely 12 months. This approach avoided the ad hoc practice of assuming that a patient observed at other than 12 months was in fact seen at that exact time. For the mixed effects models we could also estimate the individual patient's “predicted” values for the outcome of interest. For the primary outcome of asthma control, the right skewed distribution of outcomes necessitated that we use log gamma models. The treating clinic was a stratifying factor in all models to account for the stratified randomization. Alternative approaches for estimating asthma control Differences between intervention groups in expected or average values across patients at given follow up times, although statistically appropriate for comparing groups, lacks an immediate connection to the manner in which clinicians might want to interpret results. In research settings a finding that on average a patient might improve asthma control by 1.0 units over the asthma control scale might be important, but the expected value (mean) reflect only the shift in a distribution of patient asthma control scores. This type of report does not translate into a more clinically meaningful interpretation such as the number of individual patients who improved in asthma control to an important degree. For that reason, we pre-planned two methods for translating findings into clinically useful metrics. Fraction of patients who have achieved asthma control (a level below 1.5 on the scale): We avoided dichotomizing data, an ad hoc approach, but instead used mixed effects linear model and predictions to report the change in the fraction of patients who achieved adequate control. This prediction accounts for patient variation about the expected value and in theory represents a better estimate of the patient's true outcome (asthma control) than does the patient's observed value. Once we estimated each patient's predictions at baseline and at 12 months, we dichotomized these model-based predicted values to estimate the fraction of patients who achieved asthma control at baseline and at 12 months in each group. This is not the primary analysis, however, but represents an alternative demonstration of intervention effectiveness. Details appear in Appendix A4 . Minimally important improvement: For estimates of the improvements at the patient level on the asthma control (Juniper) scale, the minimally important difference is 0.5.( 50 , 65 ) Again, the model-based predicted values, in this case of the outcome of asthma control, are the best estimates at the individual level of the level of asthma control at a given time and account for measurement error as well as variation across individual patients. We compared changes in individual predictions of asthma control at 0 months and 12 months and counted the number of predictions that improved by at least 0.5. Methodological details appear in Appendix A4 . Mediation – Aim 3 Mediators explain how the interventions influence asthma outcomes. For this study we designed three candidate mediators to measure communication, access to the health care system, and improved inhaler usage: ( 1 ) the rate of portal use over time, ( 2 ) appointments kept within 6 months of a data collection time, and ( 3 ) corticosteroid inhaler use according to best practices.( 2 , 57 ) To determine whether mediation might be present, we first estimated the association between the randomized assignment (portal only versus portal + home visits) and the hypothesized mediator. In the absence of an association, there could be no indirect effect and no mediation. Details appear in Appendices A5 and A6 . Portal usage as a potential mediator. Patient activity within Patient Portal episodes. After combining the two EPIC Portal databases from the two participating institutions, there were 61 activities that appeared for the patients in our sample ( Appendix A2, Tables 1 and 2 ). An “activity” is the EPIC term for a patient action or request on the Portal, for example, to view lab results. These activities we then combined into a set that corresponded to the basic Portal usage that participating patients were trained to use ( Appendix A2, Table 3 ). Finally, we calculated the frequency of Portal usage episodes during the duration of the patient's study involvement from randomization to the last data collection (subtracting off any Portal usage that coincided with patient connecting with the community health workers or study personnel). We also calculated the frequency of all portal usage by the rate of usage over time. Because patient follow-up varied depending on the time from randomization to last data collection interview, portal use was based as a rate or intensity over time. We hypothesized that the home visitor would lead to greater use of the portal, and that with greater portal use, the patient would have better outcomes. Appointments kept. For health care usage, we hypothesized that access to medical care would improve in that patients assigned to home visitors would be more likely to make and keep regular appointments. According to prevailing guidelines,( 2 ) a patient with moderate or severe asthma should have regular appointments, and at least every 6 months. To effect that standard using data on appointments, we merged the data on appointments with the dates of the data collections and then determined for each collection interview whether the patient had kept an appointment in the 6 prior months. Using appointments kept as a time-varying exposure, we estimated the difference in expected value of asthma control at 0 months (date of randomization) and at 12 months assuming alternatively that patients had kept regular appointments versus had never kept regular appointments. Details of these analysis (and results) appear in Appendix A6 . Use of inhaler. The inhaler adherence scale( 57 ) is scored from 0 through 6. We fit a longitudinal model similar to the one for our outcomes of asthma control and quality of life to estimate the association of treatment assignment and inhaler adherence scale over time. Details appear in Appendix A7 . Heterogeneity of treatment effect (effect modification)—latent class modeling to identify subgroups (Aim 4) Effect modifiers are baseline variables or groups of factors, gleaned from the literature, hypothesized to affect the intervention-outcomes relationships.( 66 , 67 ) Recent literature on heterogeneity of treatment effect suggests that one-by-one testing or estimation of candidate effect modifiers leads to underpowered contrasts, excessive reliance on p-value testing, inadequate pre-specification of candidates and their rationale, and no attention to multiple comparisons.( 68 ) To address these criticisms, we grouped candidate measures into several clearly pre-specified themes from which profiles or latent classes are derived. We implemented a latent class analysis for each group of candidate measures to distinguish profiles of common responses to the set of variables.( 68 ),( 69 ) Then the degree of effect modification of the association of the intervention (home visitors) and the key outcome (asthma control) was estimated for each candidate modifier group by testing interactions between the candidate moderator variable and the randomization indicator. For those candidate effect modifiers that approached conventional levels of statistical significance, we estimated the intervention effect and 95% confidence intervals for each level of the effect modifier. The nine pre-specified candidate effect modifier groupings appear in Table 3 . Our working hypothesis was that patients with comorbidities and with limited computer literacy might (a) not benefit from portal education at all, and (b) might have additional benefit with the home visitor intervention. Further details appear in Appendix A8 . Table 3 Data elements grouped into themed baseline subgroups. Sensitivity analyses We planned the following sensitivity analyses, and pre-specified the endpoint for these analyses to be the between- group differences of the change over time in asthma control from baseline to 12 months. Sensitivity analysis #1. Selection bias from loss to follow-up – covariate adjustment In keeping with recommendations,( 73 , 74 ) we included in our longitudinal analysis model (described previously), in addition to the randomized treatment assignment and the pre-planned stratification variable (clinical site), the covariates that might be related to dropout and the primary outcome. We pre-specified these potential factors early on in the study design; this analysis was not data driven. Details are presented in Appendix A9 . Sensitivity analysis #2. Nonignorable timing of data collection - global sensitivity analysis to irregular visit times and dropout The sensitivity analyses we have described (#1) make assumptions that the dropout and visit times are not related to the values of outcomes, i.e., that dropout is at random and that visit times are ignorable. This is the missing at random (MAR) assumption. Our dropout was limited, but irregular interview times were common. Global sensitivity analyses to the potential from dropout not at random (or missingness not at random MNAR) are now available but only when visit times are regular.( 75 ) Methods for irregular visits are awaiting methodological development (Daniel Scharfstein, personal communications 04/2017). Details appear in Appendix A3 . Sensitivity analysis #3. Non-adherence to home visit protocol As randomized analyses do not answer the question of the effect of an intervention if the patients adhere to the protocol, we explored the potential effect of nonadherence on our results. In this regard, however, simplistic approaches to per-protocol analysis can be biased. Although weighting methods can generate unbiased estimates of the effect of a treatment under full adherence,( 76-78 ) we could not adopt these methods because they assume regular observation times. Use of weighting approach for irregular outcome times awaits methodological development. For that reason, we implemented an alternative based on instrumental variables. Instrumental variable methods adjusted for differences in the characteristics of patients who received all home visits and those who did not.( 79 ) In our case the instrument was randomization at baseline. These methods require several assumptions.( 80 ) Details along with the specific method we used (two- stage residual inclusion) appear in Appendix A10 . CHW Experiences and Observations We asked CHW's to write notes of their experiences during home visits and asked for a description of their experiences during home visits at weekly team meetings. Missing Data Missing baseline data Missing covariate data were infrequent, but some patients were missing an element of a questionnaire, or responded “did not know”. We created latent classes of covariates for both confounder control and for estimating effect modification (please see below). Missing data for subgroup analyses Latent class modeling was used both to reduce the number of covariates used to adjust for baseline factors and to examine possible moderators of the exposure outcome relationship ( Table 3 ). If we were treating each covariate as a potential confounder or as an individual effect modifier, we would need to perform formal imputation of missing data elements of an item that contributes to a latent class or drop patients who had one or more missing items.( 81 , 82 ) By contrast, latent class modeling does not require complete data on every factor that contributes to a class Mixture models generated latent classes and probabilities of class membership for all patients which were used to assign patients to these classes even if some data elements used in the mixture model were missing. This approach assumed missingness completely at random (MCAR), or at least missingness at random given the values of other covariates. In addition, the latent class method allows for the high degree of collinearity among the candidate factors. In short, latent class modeling both accomplished data reduction (too many covariates for adjustment) and facilitated identification and testing of patient subgroup differences in treatment effect. Programs and software Analyses were performed in SAS v9.4 (SAS Institute Cary NC), Stata v 15.0 and 15.1 (College Station Texas), and M Plus v7.0 (Los Angeles CA). Trial monitoring An external Data Safety Monitoring Board reviewed progress and data every six months and monitored adverse events and serious adverse events (unexpected ED visits and hospitalizations). Because patients had moderate or severe asthma, it was expected that ED visits and hospitalizations for these and patients' co-morbid conditions would occur but perhaps change as a result of the interventions. Changes to the Original Study Protocol There were no substantive changes to the original study protocol. Small changes made it easier for patients to adhere to the protocol. For example, if it were easier for the patient, we went to their home for enrollment, also for data collection. For some patients the first portal training session and the second were given on the same day for those for whom a single training session was more convenient. When we could not schedule a patient after several tries for data collection or home visit, or if the neighborhood was dangerous, the data collector went with the home visitor and completed data collection when the home visit was complete. These changes minimized missing data. Results Participant Flow We electronically screened the clinical databases of the University of Pennsylvania Health System and Temple University Health System for patients in a participating clinic. We received daily reports of adults with an asthma diagnosis, prescribed an inhaled steroid, and having an upcoming appointment: approximately 5000 patients with the repeated assessments. Of 814 screened further, 742 were potentially eligible (age 18 or more, prescribed ICS, living in a designated zip code). Seventy-two of these were found to be ineligible: did not identify with asthma (10 patients); no longer lived in designated zip code ( 12 ); had not had ER visits, hospitalization, or prednisone burst (addition of prednisone or increase in its dose) in 12 months (23 patients); no longer were prescribed ICS (19 patients); or were no longer a patient at one of the participating recruitment sites ( 8 ) ( Figure 2 ). Further screening as described in Figure 2 (CONSORT diagram) allowed randomization: 151 to PT+HV and 150 to PT. This diagram also describes withdrawals, deaths, and patients lost to follow-up. Nine in the PT + HV group (3%) and seven (2%) in the PT group were lost to follow-up. Figure 2 Participant flow diagram (CONSORT). Patient Baseline Characteristics Baseline sociodemographic, asthma, and other health characteristics of the 301 patients are described in Tables 4 and 5 . Mean age was 49; 90% were female. Two hundred twenty-seven (75%) are self-described as African American, 66 (22%) as Hispanic/Latino. Of these 51 were Spanish-only-speaking and 15 English proficient. One-hundred ninety (63%) were enrolled in Medicaid. More than half had been hospitalized for asthma and 83% had had at least one ED visit in the past year from baseline. 56% were past or present smokers, and more than 25% were current smokers. Comorbidities were prevalent: 58% had hypertension, 32% diabetes and mean BMI was in the obese range: BMI 35.5 ± 9.6. Twenty-one had been without housing the past 6 months, 49 had some days without telephone or electricity. Table 4 Baseline characteristics of 301 adults with moderate to severe asthma, assigned randomly 1:1 to PT + HV or PT, expressed as frequency (percent) for discrete and mean (±SD) for continuous variables. Table 5 Baseline characteristics of patients' homes, exposures to and perceptions of information technology in 301 adults with moderate to severe asthma, assigned randomly 1:1 to PT + HV or PT, expressed as frequency (percent) for discrete and mean (±SD) (more...) Previous (baseline) exposure to information technology is described in Table 5 . About half of participants (53%) owned a computer. In comments to the home visitor, some participants who reported they had a computer considered their cell phones to be computers. 68% had a smart phone, and 68% had an active email account. 45% had heard of a patient portal. Patient characteristics were well balanced between intervention groups at baseline. Home Visits CHWs made 501 home visits. Of the 151 participants assigned to home visits, 140 (93%) had at least one home visit. The home visits occurred at irregular intervals (see Statistical Analysis Plan under the Methods section for details). All four visits were given to 103 (68%) patients ( Table 6 ; Figure 3 ). Table 6 Actual times of home visits among the 140 patients assigned to the treatment intervention and had home visits. Figure 3 Time from randomization to home visits. Timing of Asthma Outcomes Data collection outcomes: The study protocol assumed that data collection would occur at baseline and then at months 3, 6, 9, and 12. The challenges of contacting patients, making appointments, and then interviewing the patients meant that the interviewers needed to phone repeatedly, leave messages, contact relatives, and in some cases drive by the patient's home to locate the patient. This process could take days, weeks, and sometimes months with this difficult-to-reach population. Across the 301 patients, there were 1232 data collection visits: 625 in the PT group and 607 in the PT + HV group. Numbers of data collection visits per patient ranged from 1 to 5, with 72% of patients having 4 or more data collections. The numbers did not differ by treatment assignment (See Table 7 ). Figure 4 reflects the dates of actual data collection for the five data collection visits. Table 7 Frequency of data collections over time by intervention group. Figure 4 Time from randomization to data collection by treatment group. Mean time to the final data collection ranged from 0 months (no data collection after baseline) to 33 months. The median time to the final data collection interview did not differ between treatment groups: 13.2 months for the home visitor (PT+HV) group and 12.9 months for the control (PT) patients. Mean times to the last data collection differed by 0.12 months (p=0.85). Twenty-five percent of patients had total follow-up times in excess of 16 months, and 10% of patients required more than 20 months of follow-up, compared to the predefined goal of having all data collections completed at 12 months. These findings reflect the months of effort needed to find and interview a substantial portion of the patients in our sample. The rate, or intensity of the data collection visits, did not differ between the two groups (incidence rate ratio= 0.95, 95% CI= 0.87 to 1.03). Main Results: Aims 1 and 2 Table 8 describes the change in asthma control, the primary outcome, and the other asthma outcomes (asthma-related quality of life, exacerbations, ED visits, hospitalizations) over the study period. Because data collection did not occur uniformly at 3, 6, 12, 15 months and at 3-month intervals and because we had insufficient data to measure outcomes at 6 months, we used 12 months from the date of randomization at baseline as the time at which effectiveness with respect to outcomes was assessed ( Figure 4 ). Table 8 Asthma outcomes at baseline and changes over time by treatment assignment in 301 adults with uncontrolled asthma. Intention to treat (as randomized) analysis. The outcomes other than hospitalizations, although not statistically significant, were all in the direction of favoring the PT+ HV intervention. At baseline, mean asthma control, the primary outcome, was in the uncontrolled range (mean=2.4 with minimum 0.0, maximum 6.0) ( Table 4 ). Quality of life was low compared to other populations (mean at baseline =3.6 with minimum 1 and maximum 6.9) ( Table 4 ). Patients had experienced considerable exacerbations, ED visits, and hospitalizations in the year before enrollment. These outcomes all improved over baseline in both arms. Hospitalizations decreased in both arms, more in the PT + HV arm. For the outcomes of hospitalization and ED visits, data for the baseline estimate were based on a different question to the patient than were data during follow-up. For that reason, estimates of improvement over time within the intervention group are not strictly comparable, as they could be influenced by errors in patient recall. This potential flaw in recall would not, however, bias the contrasts across intervention groups of the differences in ED visits or hospitalizations over time. Asthma control, by contrast, the primary outcome was measured in the same manner in all patients at all measurement times. The predicted number of patients who achieved asthma control (Asthma Control Questionnaire score equal to or less than 1.5) at 12 months compared to 0 months was greater among the PT+HV patients. At baseline 20 (13.3%) in the PT group and 15 (9.9%) in the PT+HV group had an asthma control level of 1.5 or better. The estimated number to achieve control at 12 months was 35 (23.3%) in the PT group and 42 (27.8%) in the PT+HV group (See Appendix 4 and Table 2 of Appendix 4 ). Our results also predicted that 37 persons assigned to the home visit group had a clinically important degree of improvement (0.5 points in the Control score) ( Appendix A4, Table 1 ), while none in the PT group had such a predicted improvement. Mediation – Aim 3 (1) Portal use as a potential mediator. Patient activity within Patient Portal episodes. Portal use was low across all patients: forty-four percent did not use the portal at all and another 30% used it only once per quarter (Details appear in Table 9 .). We examined whether communication (measured as independent use of the patient portal) mediated a relationship between PT and PT+HV with asthma control but found no evidence. Portal use did not differ between the randomization groups ( Table 9 ) and therefore cannot be a mediator. We were unable to determine whether a use of the portal was specific for asthma, because we could read only the type of activity and not the contents of that activity. For example, we could identify that a patient sent a message at a certain time, but we could not identify the addressee nor evaluate the contents of the message. Table 9 Hypothesized mediators: portal usage during course of study, appointments kept, and self-reported use of inhaled corticosteroid did not differ between groups. We defined an “episode” as independent portal activity by the patient between the log-on and log-off. We excluded episodes that occurred in the presence of a CHW, that is, occurring on the day of a home visit. 170 participants (56%) used the portal independently at least once, 58% in the PT+HV group and 55% in the PT group. Both groups, PT + HV and PT, used the portal with the same frequency ( Table 9 ). The most frequent portal activities by patients were reading messages (85% of portal users), reviewing appointments (83% of portal users), and reviewing lab results (82% of portal users) ( Appendix A11 ). Least used functions were scheduling appointments, reviewing immunizations, requesting refill of medications, and sending messages to health care personnel ( Appendix A11, Tables 1 , 2 , & 3 ). There was little difference between groups in type or frequency of portal use activity ( Appendix A11 ). Although there was no difference in portal use between the two groups, PT and PT+HV, portal use was more likely among younger, more highly educated participants, and those with higher reported household income ( Appendix 12 ). Participants with a computer or internet or smartphone or an active email account for internet access, and those who were previously aware of the patient portal were more likely to use it ( Appendix 13 ). Those concerned about the confidentiality of the patient portal were more likely to use it ( Appendix 13 ). Patients with diabetes, hypertension, hospitalizations or ED visits for asthma in the past year were less likely to use the portal ( Appendix 14 ). There was no association with asthma severity among the 203 patients who had baseline FEV1 measurements ( Appendix 15 ). (2) Appointments kept did not significantly differ between randomization groups ( Appendix A6, Table 1 ). Patients who kept their appointments had worse control ( Appendix A6, Table 2 ), perhaps because those with poor control over time needed more contacts with their physicians. More details appear in Appendix A6 . (3) Use of inhaler We examined whether reported inhaled corticosteroid use was a mediator and likewise found no significant difference between groups although both groups improved over time ( Appendix A7, Table 1 ). We also found no association of treatment assignment and inhaler adherence score between 0 and 12 months of observation ( Appendix A7, Table 1 ). Heterogeneity of treatment effect (effect modification) – latent class modeling to identify subgroups (Aim 4) The initial analysis suggested that, of the nine potential subgroups, Spanish as the primary language and trust in the internet and the patient portal for clinical information were possible effect modifiers, but these subgroups did not prove to have different benefits for the PT+HV intervention. ( Appendix A8 , Tables 5 & 6 ). Sensitivity Analyses There were 12 and 16 withdrawals in the PT and HV + PT groups, respectively. Of these, 4 and 10 participants completed only baseline data collection ( Figure 2 ). Covariate adjustment used to account for selection bias due to loss to follow-up yielded qualitatively similar results to the main analysis ( Appendix A9 , Table 1 ). While data collection times did not occur as planned there was no difference between intervention groups ( Table 7 ). More detail appears in Appendix A3 . Instrumental variable methods suggest a somewhat larger improvement for the home visitor group over time among patients who would comply ( Appendix A9 , Table 1 ). Community Health Worker Experiences and Observations CHWs found some participants could not walk independently, struggled with stairs for the subway and buses and otherwise had difficulty getting to clinic appointments. Home visits required that CHWs drive into blighted neighborhoods with abandoned homes ( Figure 5a , b ). CHWs found some participants living as tenants in homes poorly maintained by the owners, sometimes having poor ventilation, and, on one occasion containing fierce dogs that threatened them. In several of the homes, the gas or electricity was about to be or had been shut off. Some homes visited in the winter were cold and experienced inadequate heating. In some homes, many family members shared crowded space. One woman, a double amputee who lived alone, was wheel-chair bound. One home had 7 cats and 2 dogs and the animals urinated in the home; another housed 5 cats. Many homes had strong odors of tobacco smoke. Rodents and cockroaches were common. Figure 5a Google Street View images of the home and neighborhood of two participants. Figure 5b Google Street View images of the home and neighborhood of two participants. CHWs reported challenges in making home visits owing to dangerous neighborhoods, even during the daylight hours. One woman told the CHW she only sleeps in her home because the neighborhood is dangerous. She spends daytime at the home of a family member. CHWs arranged visits that were cancelled at the last minute or “no shows,” meaning that the participant was not home when the CHW arrived. On one occasion CHWs driving to a home visit found themselves in the middle of an active crime scene. On another occasion while interviewing an older woman, the woman's daughter ran into the room distraught because her son, the participant's grandson, was being attacked by someone wielding a machete just outside. The biggest challenge for CHWs was scheduling appointments. Phone numbers frequently changed or phones were temporarily or permanently disconnected. The former happened more frequently at the end of the month when purchased phone time was exhausted. The CHWs learned to schedule visits as much as possible at the beginning of the month. When demonstrating the use of the portal to patients, CHWs found that several participants could not learn to log on. Some of these, both young and old, were unable to read and could not find the login portion of the screen. Many older participants had never used a computer or the internet. Some participants called in family members to help them with the computer exercise. Many patients (48%) stated that they did not have a computer, or did not use the internet (38%), or did not have an active email account (32%) ( Table 5 ). Qualitative data from the CHWs support these accounts ( Appendix A16 ). When absolutely necessary to communicate via internet, they relied on family members. Some participants strongly preferred calling the medical practice to using the portal, because the staff members recognized the patients and offered warm responses. No participant went to a library to use a computer and participants frequently noted that neighborhood libraries had been closed. Some distrusted public computers. Participants also frequently expressed concern about privacy of their health information in using a computer even at home ( Table 5 ) and “did not want the government knowing my personal business. They are always watching me.” Most participants had no easy or convenient way of accessing the portal. Some Spanish speakers were unable to use the portal, and there was no portal option for Spanish or any language other than English. Nevertheless, these adult participants generally appreciated the home visits and those that used the portal found its features potentially useful. They liked the options of requesting refills and making appointments but without a computer most, particularly the older participants, could not take advantage of it. Discussion Context for Study Results In this longitudinal two-arm randomized controlled trial we compared the effect over time of ( 1 ) introducing patients to, and training them in the use of, an electronic health-record-based patient portal to enhance patient-clinician communication, versus ( 2 ) an intervention that added to this portal training a series of regular home visits by specially training community health workers. All 301 study participants were adults with uncontrolled asthma and who lived in low-income, inner-city neighborhoods. By design we did not add a third arm, a placebo control, because in this patient population to do so would severely impair our ability to recruit consenting patients and follow them over time. Thus, while both groups improved in the asthma outcomes over time, we could not estimate what would have happened in the absence of any intervention. Nevertheless, our two-arm design assumed that the addition of home visits to portal-based intervention would sufficiently benefit patients so that their asthma outcomes would improve significantly. In both groups asthma control, asthma-related quality of life, improved and rate of ED visits, hospitalizations and prednisone bursts decreased. All comparative outcomes had point estimates in favor of the PT + HV group, but only for hospitalizations was the difference between groups statistically significant. Thus, while we found only limited benefit from the addition of home visitors, the changes, especially of hospitalizations, were all consistently in the direction of the home visits. We do not have a clear reason why both groups had improvement in asthma outcomes. It is possible attention to the patients was important in identifying their needs for care but we have no evidence to support this hypothesis. Patient portal – the control intervention Although patient portals have been championed by the Centers for Medicare Services (CMS) as a means to engage patients in their own care and thus to improve their outcomes, we found that usage of the patient portal in both groups was infrequent and variable. Although we found that outcomes in this active control group improved over the course of the study, we also found that there was no association between the frequency of portal usage and outcome. Given the lower overall portal usage, we suspect that the opportunity for the portal alone to influence outcome would be limited. With this level of non-adherence to an intervention, evaluation of the potential effect of the portal if patients used it become problematic. This latter finding suggests that patient contact by our study team and not patient contact with health care providers through the portal was the real reason for any improvement in outcomes. Not only did many patients, more than 40%, never use the patient portal independently during the course of the study, nearly 10% of the enrolled subjects never activated their portals. At the other extreme, very few patients used the portal consistently. In addition, patients who needed communication and access most, those who were older with co-morbidities, were least likely to use the portal. If the patient portal is going to be a helpful adjunct to patient care for the type of patients with asthma in our sample, the design of and access to the portal will need substantial improvements. Access to the portal must reflect the current reality that many patients in low income settings will not have home access to a computer, an iPad, or a smart phone. For them access means finding a school or library with workstations for mail and internet access or buying a cell phone and a usage plan. We cannot expect patients with chronic diseases to leave their homes, too often in poor and unsafe neighborhoods, and venture to a neighborhood library. It is easier for many patients simply to use a telephone. While broadband access might improve for inner-city patients over time, these patients face fundamental barriers to portal use today. We found the patient portal activity with most limited usage was appointment scheduling. That function might not be well-suited to a portal because the confirmation of an appointment time agreeable to both the clinician and the patient often requires give and take in real time. Checking appointment times or reports of lab values were functions that patients used more often. These functions offer immediate information and avoid wait times on the telephone, and perhaps call backs that the telephone often requires. Regardless of the function in the portal, however, the access to the portal information, even for the patient who has a computer at home, and/or a smart phone, and who has internet savvy, still mandates display screens that are easily read and navigated. We noted in our design of the study and our confirmation of portal functioning, that for both the computer and iPhone/iPad interface, the user had to hunt for the desired button or label on which to click. The interface for the computer screen version had small typefaces that might be adequate for younger, healthier, and more computer literate patients but not so for the many, less educated, chronically ill patients in our asthma population. Thus, for the patient portal to realize the legitimate goals of enhancing patient participation in his or her healthcare, portal design must reduce these barriers to those patients who wants to use the portal. Home visits – the added intervention According to our study design, the estimable contrast based on randomization was the effect of the CHW home visits. The longitudinal contrast of the effect of portal training was observational only since both arms included portal training. The addition of home visits in theory could represent the complement of care and communication that could improve patient engagement with healthcare providers and patient outcomes. Home visits by CHWs have been shown to be acceptable and feasible in low-income populations. Nevertheless, at least in this randomized sample of patients assigned to the addition of home visitors to the intervention, asthma outcomes were generally no better than in the active control arm. The lack of large (and statistically significant) incremental improvements might reflect the unusual challenges of conducting these interventions in these patients, as well as the overriding influences of alternative exposures, e.g., cigarette smoke, and environmental factors, e.g., ambient air pollution or allergens in the home and neighborhood. First, and most notably, we found that contacting patients for both home visits and data collection represented a formidable challenge. The irregularity of visits might have dampened the impact of home visits, which ideally would have occurred as originally scheduled over several weeks rather than over months. Likewise, we had to evaluate patients' outcomes over often several more months than the originally designed and more concentrated 12-month period. We suspect that these delays in actual versus preplanned scheduling of visits and data collection are the norm in studies of patients who live in low income and distressed neighborhoods, and we have taken care to document and report these delays. The degree of delay did not differ across intervention group, and therefore we do not suspect that delay introduced bias in the randomized contrast. Nonetheless, the design of interventions for these hard-to-reach patients must anticipate inevitable delays in reaching patients. Second, the time horizon of the study restricted the maximum number of home visits that were possible. Increasing the dose (number and intensity) of home visits might have revealed a critical level of contact that home visits must have with their patients to result in improvements in patient outcomes. Longer studies should be planned, and they will require added resources to maintain ongoing patient contact and to collect data over longer periods. Third, we found during the home visits that patients experienced many exposures that might not only interfere with their ability to communicate with their clinical team but also might undermine their ability to adhere to their disease management. We also found that CHWs' visits with patients in their homes allowed more insight into the barriers that asthmatic adults face.( 38 ) Many times there was little social support for these adults which led to reliance on the CHWs. In this study, the intervention specified in the research protocol did not include asthma education but only care coordination. The addition of education regarding asthma self-management along with care coordination may have led to larger differences in outcomes in the home visit study arm. For example, the provision of advice for addressing asthma triggers like the presence of tobacco smoke could become a part of the home visitor's intervention package. Home environment interventions including pest remediation must be considered also. The incremental effect of the addition of self-management to communication and care coordination should become the focus of future research on the elements of home visits that make a difference. Generalizability of the Findings Our analysis demonstrates that those who are most vulnerable are least likely to have access to the portal, raising concern that the portal may contribute to increased health disparities. As Lyles et al( 84 ) in their editorial state, there are ethical and legal and practical standards to ensure that this information technology does not worsen existing health and health care disparities. Older adults with asthma, unlike their younger counterparts tend to have comorbidities. Any intervention for asthma must take account of these. We have found that the patient portal is least helpful for those who could potentially benefit most: those for whom travel to an office appointment is difficult and those in need of extra resources. We did not consider geographic diversity; our cohort lived entirely in the inner-city. We do not know if home visits can be efficiently achieved in rural environments where prolonged travel might be required. We did not attempt having CHWs provide HVs by telemedicine in our study, but telemedicine may be more feasible or even essential in rural settings. Our country is ethnically very diverse and the resources of our study did not allow us to consider other subpopulations with respect to portal use and home visits: other languages, other cultures, other exposures. The portal has great potential for accommodating non-English speakers and different cultural backgrounds. Thus the next generation portal must not only be specific to patients' settings, it also must be generalizable across the diversity of individuals and their living situations. Approximately 90% of the participants were female. Asthma is more prevalent among females beyond the age of puberty, 1 and women tend to seek medical help more often than males. This percentage is only slightly higher than the percentage we have found in other studies of the same population. 15 , 28 , 31 We could not find an additional reason; we did not have data showing men were less likely to enroll, or more likely to drop out. Subpopulation Consideration – Heterogeneity of Treatment Effects There was a suggestion in our findings that some subgroups of patients are more likely to benefit from the portal plus home visit intervention, but the differences are too small to be clinically meaningful. Owing to the diversity of patients, even among this group of low income, inner city patients, and the limits of the overall power with a sample of 301 patients, we could not identify important subgroups and estimate other than very substantial differences in separate incremental effects of home visits over the active control intervention. Any future work will have to enroll much larger populations to allow for the identification of subgroups with meaningful numbers of subjects. In addition, in planning these studies, investigators should focus on defining subgroups that are based on more than one patient characteristic, to recruit patients deliberately from those specific subgroups of interest, and then to balance randomization within those subgroups. Study Limitations Owing to the constraints of the PCORI funding mechanisms, we could not enroll larger numbers of patients over longer periods and follow them uniformly over more months. In a study of patients who suffer from chronic diseases and comorbidities, short term improvement might be both unrealistic to expect and difficult to maintain. With longer observation, investigators might be able to detect improvements with longer latency periods, not only of patients but of their children. Longer follow up might have supported an analysis of the sustainability of immediate gains. We were unable to consistently determine if some uses of the portal were specific for asthma. In addition, inadequate home internet access reduced the ability of the clinical practice to remain in frequent contact with the patients. That the patient portals we observed were only in English is another limitation. A portal available in Spanish might provide significant more of interest to Hispanic patients. And a portal that could be easily adapted to other languages in today's diverse country would be of substantial potential benefit. The extreme poverty witnessed by CHWs making home visits must also be understood by clinicians and will require long-term effort by society to address. Poverty limited patients' ability to complete the protocol, e.g., to meet for home visit and data collection at the protocol-defined times, and made travel by staff for home visits sometimes dangerous. It often was responsible for the delays and difficulties with data collection. Not only do living conditions affect current health, but recent studies have shown that the childhood socioeconomic status of parents influences the asthma outcomes of their children.( 85 ) That is, not only must we follow patients over years, but we also must continue to follow the influence of poverty on asthma well-being of the next generation. Additional study limitations include our inability to support enough staff in order to separate staff who collected data from those who recruited and visited patients and thus permit blinding of the data collectors to the treatment allocation of the patients. There was no usual care group for comparison. Finally, hospitalizations, ED visits and other health care utilization were self-reported and could only be partially validated as we were unable to access most records outside of the participating health systems. Future Research Future research should combine an intervention targeting patients both at home and in the clinic. In future work we are proposing to study a multi-component intervention to acknowledge home and clinic environments and to include clinicians more directly in communications with the home visitor/patient advocate. It will be important to include more affluent persons, to address the question of whether there is a difference between PT and PT+ HV across a spectrum of patients with varying home resources and environments. Future research also should propose ways to better align the portal and other modes of patient-clinician contact and communication to patients' needs and interests. Finally, investigators will need to anticipate the challenges of scheduling patients for home visits and telephone interviews and the impact of these challenges on study design and analysis. Traditional longitudinal designs that assume patient contact and measurements at specific months or weeks from baseline and the subsequent modeling of time as a categorical factor will need to give way to protocols that allow for irregular visits and that model time flexibly as a continuous variable. Forcing patient contact time to obey an inflexible schedule within pre-determined time windows will only lead to loss of patient contacts and data collections and suboptimal methods for imputing missing outcome data. Conclusions Although results for the incremental effect of home visits on asthma outcomes revealed that only the finding of reduced hospitalizations reached significance, the direction of all the outcomes favored the home visitor group, which may indicate a potential positive effect for the home visitor intervention. While the patient portal may be helpful for more affluent younger groups with greater familiarity with information technology, we learned that in the portal's present form, patients with the highest risk for morbidity are least likely to use it. This finding does not rule out the possibility that adjusting the portal content for older high-risk patient, including minority patients and those with comorbidities, may be beneficial. Our findings on the effects of home visits and the patient portal suggest an extremely potent influence on health of patients with asthma, one that is often ignored in the medical setting because it is not observed: the effect of poverty and all the associated deprivations.( 40 , 86 ) Health care interventions, even those that venture beyond the clinic to the home and community, are unlikely to succeed without additional commitments to improved incomes and housing and safer neighborhoods. Future interventions should attempt to document the effect of these barriers on improved outcomes for adults with asthma. REFERENCES 1. Akinbami LJ, Moorman JE, Bailey C, Zahran HS, King M, Johnson CA, et al. Trends in asthma prevalence, health care use, and mortality in the United States, 2001-2010. NCHS Data Brief. 2012(94):1-8. [ PubMed : 22617340 ] 2. Expert Panel Report 3: Guidelines for the diagnosis and management of asthma Bethesda, MD: National Institutes of Health, National Heart, Lung, and Blood Institute; 2007:NIH Publication 08-5846:1-416. 3. 2012;Pages. Accessed at U.S. Department of Health and Human Services at http://minorityhealth.hhs.gov/templates/content.aspx?lvl=3&lvlID=532&ID=6170 on January 24, 2014 2013. 4. 2016;Pages. Accessed at US Department of Health and Human Sercices, Office of Minority Health at http: ​//minorityhealth ​.hhs.gov/omh/browse ​.aspx?lvl=4&lvlid=60 on 6/3/2016 2016. 5. Schiller JS, Lucas JW, Ward BW, Peregoy JA. Summary health statistics for U.S. adults: national Health Interview Survey, 2010. National Center for Health Statistsics. Vital Health Stat 10(252).; 2012. [ PubMed : 22834228 ] 6. Smedley BD, Stith AY, Nelson AR. Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, D.C.: Institute of Medicine. The National Academies Press; 2003. [ PubMed : 25032386 ] 7. 2009;Pages http://www.recovery.gov/arra/About/Pages/The_Act.aspx on 3/25/2015. 8. Marcotte L, Seidman J, Trudel K, Berwick DM, Blumenthal D, Mostashari F, et al. Achieving Meaningful Use of Health Information Technology: A Guide for Physicians to the EHR Incentive Programs. Arch Intern Med. 2012;172(9):731-6. [ PubMed : 22782203 ] 9. Seidman J, Wolf MS. The role of health literacy in health information technology. Innovations in Health Literacy Research. Washington, DC: The National Academies Press; 2011:23-32. 10. Irizarry T, DeVito Dabbs A, Curran CR. Patient Portals and Patient Engagement: A State of the Science Review. J Med Internet Res. 2015;17(6):e148. [ PMC free article : PMC4526960 ] [ PubMed : 26104044 ] 11. Kontos E, Blake KD, Chou WY, Prestin A. Predictors of eHealth usage: insights on the digital divide from the Health Information National Trends Survey 2012. J Med Internet Res. 2014;16(7):e172. [ PMC free article : PMC4129114 ] [ PubMed : 25048379 ] 12. Zhao JY, Song B, Anand E, Schwartz D, Panesar M, Jackson GP, et al. Barriers, Facilitators, and Solutions to Optimal Patient Portal and Personal Health Record Use: A Systematic Review of the Literature. AMIA Annu Symp Proc. 2017;2017:1913-22. [ PMC free article : PMC5977619 ] [ PubMed : 29854263 ] 13. Goldzweig CL, Orshansky G, Paige NM, Towfigh AA, Haggstrom DA, Miake-Lye I, et al. Electronic patient portals: evidence on health outcomes, satisfaction, efficiency, and attitudes: a systematic review. Ann Intern Med. 2013;159(10):677-87. [ PubMed : 24247673 ] 14. Sarkar U, Karter AJ, Liu JY, Adler NE, Nguyen R, Lopez A, et al. Social disparities in internet patient portal use in diabetes: evidence that the digital divide extends beyond access. J Am Med Inform Assoc. 2011;18(3):318-21. [ PMC free article : PMC3078675 ] [ PubMed : 21262921 ] 15. Tieu L, Sarkar U, Schillinger D, Ralston JD, Ratanawongsa N, Pasick R, et al. Barriers and Facilitators to Online Portal Use Among Patients and Caregivers in a Safety Net Health Care System: A Qualitative Study. J Med Internet Res. 2015;17(12):e275. [ PMC free article : PMC4704882 ] [ PubMed : 26681155 ] 16. Tieu L, Schillinger D, Sarkar U, Hoskote M, Hahn KJ, Ratanawongsa N, et al. Online patient websites for electronic health record access among vulnerable populations: portals to nowhere? J Am Med Inform Assoc. 2017;24(e1):e47-e54. [ PMC free article : PMC6080722 ] [ PubMed : 27402138 ] 17. Kangovi S, Mitra N, Grande D, Huo H, Smith RA, Long JA. Community Health Worker Support for Disadvantaged Patients With Multiple Chronic Diseases: A Randomized Clinical Trial. Am J Public Health. 2017;107(10):1660-7. [ PMC free article : PMC5607679 ] [ PubMed : 28817334 ] 18. Kangovi S, Mitra N, Norton L, Harte R, Zhao X, Carter T, et al. Effect of Community Health Worker Support on Clinical Outcomes of Low-Income Patients Across Primary Care Facilities: A Randomized Clinical Trial. JAMA Intern Med. 2018;178(12):1635-43. [ PMC free article : PMC6469661 ] [ PubMed : 30422224 ] 19. Krieger J, Song L, Philby M. Community health worker home visits for adults with uncontrolled asthma: the HomeBASE Trial randomized clinical trial. JAMA Intern Med. 2015;175(1):109-17. [ PubMed : 25419871 ] 20. Bryant-Stephens T, Kurian C, Guo R, Zhao H. Impact of a household environmental intervention delivered by lay health workers on asthma symptom control in urban, disadvantaged children with asthma. Am J Public Health. 2009;99 Suppl 3:S657-65. [ PMC free article : PMC2774193 ] [ PubMed : 19890172 ] 21. Krieger JW, Takaro TK, Song L, Weaver M. The Seattle-King County Healthy Homes Project: a randomized, controlled trial of a community health worker intervention to decrease exposure to indoor asthma triggers. Am J Public Health. 2005;95(4):652-9. [ PMC free article : PMC1449237 ] [ PubMed : 15798126 ] 22. Carrasquillo O, Lebron C, Alonzo Y, Li H, Chang A, Kenya S. Effect of a Community Health Worker Intervention Among Latinos With Poorly Controlled Type 2 Diabetes: The Miami Healthy Heart Initiative Randomized Clinical Trial. JAMA Intern Med. 2017;177(7):948-54. [ PMC free article : PMC5818804 ] [ PubMed : 28459925 ] 23. Shah M, Kaselitz E, Heisler M. The role of community health workers in diabetes: update on current literature. Curr Diab Rep. 2013;13(2):163-71. [ PMC free article : PMC3929361 ] [ PubMed : 23345198 ] 24. Singh P, Chokshi DA. Community health workers--a local solution to a global problem. N Engl J Med. 2013;369(10):894-6. [ PubMed : 24004115 ] 25. Goel MS, Brown TL, Williams A, Cooper AJ, Hasnain-Wynia R, Baker DW. Patient reported barriers to enrolling in a patient portal. J Am Med Inform Assoc. 2011;18 Suppl 1:i8-12. [ PMC free article : PMC3241181 ] [ PubMed : 22071530 ] 26. Goel MS, Brown TL, Williams A, Hasnain-Wynia R, Thompson JA, Baker DW. Disparities in enrollment and use of an electronic patient portal. J Gen Intern Med. 2011;26(10):1112-6. [ PMC free article : PMC3181306 ] [ PubMed : 21538166 ] 27. Apter AJ, Bryant-Stephens T, Morales KH, Wan F, Hardy S, Reed-Wells S, et al. Using IT to improve access, communication, and asthma in African American and Hispanic/Latino Adults: Rationale, design, and methods of a randomized controlled trial. Contemp Clin Trials. 2015;44:119-28. [ PubMed : 26264737 ] 28. Apter AJ, Wan F, Reisine S, Bogen DK, Rand C, Bender B, et al. Feasibility, acceptability and preliminary effectiveness of patient advocates for improving asthma outcomes in adults. J Asthma. 2013;50(8):850-60. [ PMC free article : PMC4118639 ] [ PubMed : 23800333 ] 29. Black H, Gonzalez R, Priolo C, Schapira MM, Sonnad SS, Hanson CW, 3rd, et al. True “meaningful use”: technology meets both patient and provider needs. Am J Manag Care. 2015;21(5):e329-37. [ PubMed : 26167781 ] 30. Black HL, Priolo C, Akinyemi D, Gonzalez R, Jackson DS, Garcia L, et al. Clearing clinical barriers: enhancing social support using a patient navigator for asthma care. J Asthma. 2010;47(8):913-9. [ PMC free article : PMC2950230 ] [ PubMed : 20846085 ] 31. Apter AJ, Boston RC, George M, Norfleet AL, Tenhave T, Coyne JC, et al. Modifiable barriers to adherence to inhaled steroids among adults with asthma: it's not just black and white. J Allergy Clin Immunol. 2003;111(6):1219-26. [ PubMed : 12789220 ] 32. Apter AJ, Garcia LA, Boyd RC, Wang X, Bogen DK, Ten Have T. Exposure to community violence is associated with asthma hospitalizations and emergency department visits. J Allergy Clin Immunol. 2010;126(3):552-7. [ PMC free article : PMC2936964 ] [ PubMed : 20816190 ] 33. Apter AJ, Reisine ST, Affleck G, Barrows E, ZuWallack RL. Adherence with twice-daily dosing of inhaled steroids. Socioeconomic and health-belief differences. Am J Respir Crit Care Med. 1998;157(6 Pt 1):1810-7. [ PubMed : 9620910 ] 34. Apter AJ, Wan F, Reisine S, Bender B, Rand C, Bogen DK, et al. The association of health literacy with adherence and outcomes in moderate-severe asthma. J Allergy Clin Immunol. 2013;132(2):321-7. [ PMC free article : PMC3745278 ] [ PubMed : 23591273 ] 35. Apter AJ, Wang X, Bogen DK, Rand CS, McElligott S, Polsky D, et al. Problem solving to improve adherence and asthma outcomes in urban adults with moderate or severe asthma: A randomized controlled trial. J Allergy Clin Immunol. 2011;128:516-23. [ PMC free article : PMC3164914 ] [ PubMed : 21704360 ] 36. Sarkar U, Karter AJ, Liu JY, Adler NE, Nguyen R, Lopez A, et al. The literacy divide: health literacy and the use of an internet-based patient portal in an integrated health system-results from the diabetes study of northern California (DISTANCE). J Health Commun. 2010;15 Suppl 2:183-96. [ PMC free article : PMC3014858 ] [ PubMed : 20845203 ] 37. Bryant-Stephens T, Kurian C, Chen Z. Brief report of a low-cost street-corner methodology used to assess inner-city residents' awareness and knowledge about asthma. J Urban Health. 2011;88 Suppl 1:156-63. [ PMC free article : PMC3042069 ] [ PubMed : 21337061 ] 38. Bryant-Stephens T, Reed-Wells S, Canales M, Perez L, Rogers M, Localio AR, et al. Home visits are needed to address asthma health disparities in adults. J Allergy Clin Immunol. 2016;138(6):1526-30. [ PubMed : 27777181 ] 39. Bryant-Stephens T, West C, Dirl C, Banks T, Briggs V, Rosenthal M. Asthma prevalence in Philadelphia: description of two community-based methodologies to assess asthma prevalence in an inner-city population. J Asthma. 2012;49(6):581-5. [ PubMed : 22747076 ] 40. Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322-7. [ PMC free article : PMC1508772 ] [ PubMed : 10474547 ] 41. Glasgow RE, Vinson C, Chambers D, Khoury MJ, Kaplan RM, Hunter C. National Institutes of Health approaches to dissemination and implementation science: current and future directions. Am J Public Health. 2012;102(7):1274-81. [ PMC free article : PMC3478005 ] [ PubMed : 22594758 ] 42. Peretz PJ, Matiz LA, Findley S, Lizardo M, Evans D, McCord M. Community health workers as drivers of a successful community-based disease management initiative. Am J Public Health. 2012;102(8):1443-6. [ PMC free article : PMC3464827 ] [ PubMed : 22515859 ] 43. Krieger JW, Philby ML, Brooks MZ. Better home visits for asthma lessons learned from the Seattle-King County Asthma Program. Am J Prev Med. 2011;41(2 Suppl 1):S48-51. [ PubMed : 21767735 ] 44. 2015;Pages. Accessed at American Lung Association at http://www ​.lung.org/associations ​/states/colorado ​/asthma/asthma-action-plan.html on July 8, 2015. 45. Bryant-Stephens T. Asthma disparities in urban environments. J Allergy Clin Immunol. 2009;123(6):1199-206; quiz 207-8. [ PubMed : 19501229 ] 46. Busse WW, Morgan WJ, Taggart V, Togias A. Asthma outcomes workshop: overview. J Allergy Clin Immunol. 2012;129(3 Suppl):S1-8. [ PMC free article : PMC4259286 ] [ PubMed : 22386504 ] 47. Juniper EF, O'Byrne PM, Roberts JN. Measuring asthma control in group studies: do we need airway caliber and rescue beta2-agonist use? Respir Med. 2001;95(5):319-23. [ PubMed : 11392570 ] 48. Juniper EF, O'Byrne PM, Ferrie PJ, King DR, Roberts JN. Measuring asthma control. Clinic questionnaire or daily diary? Am J Respir Crit Care Med. 2000;162(4 Pt 1):1330-4. [ PubMed : 11029340 ] 49. Juniper EF, O'Byrne PM, Guyatt GH, Ferrie PJ, King DR. Development and validation of a questionnaire to measure asthma control. Eur Respir J. 1999;14(4):902-7. [ PubMed : 10573240 ] 50. Juniper EF, Bousquet J, Abetz L, Bateman ED. Identifying ‘well-controlled' and ‘not well-controlled' asthma using the Asthma Control Questionnaire. Respir Med. 2006;100(4):616-21. [ PubMed : 16226443 ] 51. Juniper EF, Guyatt GH, Cox FM, Ferrie PJ, King DR. Development and validation of the Mini Asthma Quality of Life Questionnaire. Eur Respir J. 1999;14(1):32-8. [ PubMed : 10489826 ] 52. Juniper EF, Buist AS, Cox FM, Ferrie PJ, King DR. Validation of a standardized version of the Asthma Quality of Life Questionnaire. Chest. 1999;115(5):1265-70. [ PubMed : 10334138 ] 53. Juniper EF, Guyatt GH, Willan A, Griffith LE. Determining a minimal important change in a disease-specific Quality of Life Questionnaire. J Clin Epidemiol. 1994;47(1):81-7. [ PubMed : 8283197 ] 54. Leidy KN, Chan KS, Coughlin C. Is the asthma quality of life questionnaire a useful measure for low-income asthmatics? Am J Respir Crit Care Med. 1998;158(4):1082-90. [ PubMed : 9769264 ] 55. Standarization of Spirometry, 1994 Update. Am J Respir Crit Care Med. 1995;152:1107-36. [ PubMed : 7663792 ] 56. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care. 1986;24(1):67-74. [ PubMed : 3945130 ] 57. Dolce JJ, Crisp C, Manzella B, Richards JM, Hardin JM, Bailey WC. Medication adherence patterns in chronic obstructive pulmonary disease. Chest. 1991;99(4):837-41. [ PubMed : 2009784 ] 58. Apter AJ, Cheng J, Small D, Albert C, Fein DL, Bennett IM, et al. Asthma Numeracy Skill and Health Literacy. J Asthma. 2006;43:705-10. [ PubMed : 17092853 ] 59. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J. Development of a brief test to measure functional health literacy. Patient Educ Couns. 1999;38(1):33-42. [ PubMed : 14528569 ] 60. Norman CD, Skinner HA. eHEALS: The eHealth Literacy Scale. J Med Internet Res. 2006;8(4):e27. [ PMC free article : PMC1794004 ] [ PubMed : 17213046 ] 61. 2015;Pages http: ​//project-redcap.org/ on 3/26/2015. 62. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-81. [ PMC free article : PMC2700030 ] [ PubMed : 18929686 ] 63. Howe LD, Tilling K, Matijasevich A, Petherick ES, Santos AC, Fairley L, et al. Linear spline multilevel models for summarising childhood growth trajectories: A guide to their application using examples from five birth cohorts. Stat Methods Med Res. 2016;25(5):1854-74. [ PMC free article : PMC4074455 ] [ PubMed : 24108269 ] 64. Fitzmaurice GM, Laird NM, J.H. W. Applied Longitudinal Analysis, 2nd Edition. Hoboken, NJ: John Wiley & Sons; 2011. 65. Juniper EF, Svensson K, Mork AC, Stahl E. Measurement properties and interpretation of three shortened versions of the asthma control questionnaire. Respir Med. 2005;99(5):553-8. [ PubMed : 15823451 ] 66. Holmbeck GN. Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: examples from the child-clinical and pediatric psychology literatures. J Consulting and Clinical Psychology. 1997;65:599-610. [ PubMed : 9256561 ] 67. VanderWeele TJ. Explanation in Causal Inference: Methods for Mediation and Interaction. Oxford: Oxford University Press; 2015. 68. Alosh M, Huque MF, Bretz F, D'Agostino RB, Sr. Tutorial on statistical considerations on subgroup analysis in confirmatory clinical trials. Stat Med. 2017;36(8):1334-60. [ PubMed : 27891631 ] 69. McCutcheon AL. Latent Class Analysis. Beverly Hill, CA: Sage University Press Series on Quantiative Applications in the Social Sciences; 1987. 70. Knapp C, Madden V, Wang H, Sloyer P, Shenkman E. Internet use and eHealth literacy of low-income parents whose children have special health care needs. J Med Internet Res. 2011;13(3):e75. [ PMC free article : PMC3222184 ] [ PubMed : 21960017 ] 71. Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. PG - 473-83. Med Care. 1992;30(6). [ PubMed : 1593914 ] 72. Sherbourne CD, Stewart AL. The MOS social support survey. Soc Sci Med. 1991;32(6):705-14. [ PubMed : 2035047 ] 73. Council NR. The Prevention and Treatment of Missing Data in Clinical Trials. Panel on Handling Missing Data in Clinical Trials. Committee on National Statistics, Division of Behavioral and Social Sciences and Education. Washington, DC: The National Academies Press.; 2010:106. 74. Council NR. The Prevention and Treatment of Missing Data in Clinical Trials Washington, DC: The National Academies Press.; 2010:ISBN: 978-0-309-15814-5. [ PubMed : 24983040 ] 75. Scharfstein DO, McDermott A, Olson W, Wiegand F. Global Sensitivity Analysis for Repeated Measures Studies With Informative Dropout: A Fully Parametric Approach. Statistics in Biopharmaceutical Research 2014;6(4):338-48 Taylor and Francis 76. Toh S, Hernan MA. Causal inference from longitudinal studies with baseline randomization. Int J Biostat. 2008;4(1):Article 22. [ PMC free article : PMC2835458 ] [ PubMed : 20231914 ] 77. Toh S, Hernandez-Diaz S, Logan R, Robins JM, Hernan MA. Estimating absolute risks in the presence of nonadherence: an application to a follow-up study with baseline randomization. Epidemiology. 2010;21(4):528-39. [ PMC free article : PMC3315056 ] [ PubMed : 20526200 ] 78. Toh S, Hernandez-Diaz S, Logan R, Rossouw JE, Hernan MA. Coronary heart disease in postmenopausal recipients of estrogen plus progestin therapy: does the increased risk ever disappear? A randomized trial. Ann Intern Med. 2010;152(4):211-7. [ PMC free article : PMC2936769 ] [ PubMed : 20157135 ] 79. Nagelkerke N, Fidler V, Bernsen R, Borgdorff M. Estimating treatment effects in randomized clinical trials in the presence of non-compliance. Stat Med. 2000;19(14):1849-64. [ PubMed : 10867675 ] 80. Jo B. Statistical power in randomized intervention studies with noncompliance. Psychol Methods. 2002;7(2):178-93. [ PubMed : 12090409 ] 81. Rubin DB. Inference and missing data. Biometrika. 1976;63:581-92. 82. Skrondal A, Rabe-Hesketh S. Generalized Latent Variable Modeling: Multilevel, longitudinal and structural equation models. New York: Chagman & Hall/CRC; 2004. 83. Radloff LS. The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385-401. 84. Lyles CR, Fruchterman J, Youdelman M, Schillinger D. Legal, Practical, and Ethical Considerations for Making Online Patient Portals Accessible for All. Am J Public Health. 2017:e1-e4. [ PMC free article : PMC5607665 ] [ PubMed : 28817324 ] 85. Chen E, Shalowitz MU, Story RE, Ehrlich KB, Manczak EM, Ham PJ, et al. Parents' childhood socioeconomic circumstances are associated with their children's asthma outcomes. J Allergy Clin Immunol. 2017;140(3):828-35 e2. [ PMC free article : PMC5509526 ] [ PubMed : 28089871 ] 86. Glasgow RE. Implementation Science Models (And Related Metrics) to Help Reduce Health Disparities. The Science of Eliminating Health Disparities. National Harbor, Maryland; 2012. Related Publications 1. Apter AJ, Bryant-Stephens T, Morales KH, et al. Using IT to Improve Access, Communication, and Asthma Outcomes in African American and Puerto Rican Adults. Am J Respir Crit Care Med 2018; 197: A2713. 2. Apter AJ, Bryant-Stephens T, Perez L, et al. Electronic Health Record Patient Portal Access for Low Income Minority Adults with Uncontrolled Asthma. Am J Respir Crit Care Med 2018; 197: A1074. 3. Gleeson, PK, Perez L, Localio AR, et al. Inhaler technique among adults with uncontrolled asthma from low-income Philadelphia neighborhoods. Journal of Allergy and Clinical Immunology 2018; 141(2): AB211. [ PMC free article : PMC6842684 ] [ PubMed : 31173936 ] 4. Apter AJ, Localio AR, Bryant-Stephens T, et al. Are Low-Income Asthmatic Adults Using the Patient Portal for Communicating With Providers? Am J Respir Crit Care Med 2017; 193: A1083. 5. Kramer CB, LeRoy L, Donahue S, et al. Enrolling African-American and Latino patients with asthma in comparative effectiveness research: Lessons learned from 8 patient-centered studies. The Journal of allergy and clinical immunology 2016;138:1600-07. [ PubMed : 27789250 ] 6. Bryant-Stephens T, Reed-Wells S, Canales M, et al. Home visits are needed to address asthma health disparities in adults. The Journal of allergy and clinical immunology 2016;138:1526-30. [ PubMed : 27777181 ] 7. Apter AJ, Bryant-Stephens T, Morales KH, et al. Using IT to improve access, communication, and asthma in African American and Hispanic/Latino Adults: Rationale, design, and methods of a randomized controlled trial. Contemporary clinical trials 2015;44:119-28. [ PubMed : 26264737 ] Acknowledgments This work was supported through a Patient-Centered Outcomes Research Institute (PCORI) Program Award (AS-1307-05218). All statements in this report, including its findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute (PCORI), its Board of Governors or Methodology Committee. Research reported in this report was [partially] funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (#AS-1307-05218) Further information available at: https://www.pcori.org/research-results/2013/testing-whether-home-visits-community-health-workers-help-african-american-and Appendices Appendix A1. PCORI Visit Schedule and List of Case Report Forms Used in REDCap (PDF, 160K) Appendix A2. Patient Portal in EPIC—Data Management Details (PDF, 172K) Appendix A2, Table 1. Portal usage at the keystroke level (n=18462 keystrokes) across 170 patients between randomization and last contact for patient data collection. (PDF, 151K) Appendix A2, Table 2. Portal activity – Writing/sending. (PDF, 149K) Appendix A2, Table 3. Major functions (see Appendix A2, Table 1) used by patients on the Patient Portal. (PDF, 94K) Appendix A3. Methods for Handling Irregular Data Collection Interviews—Unified Modeling Approach (PDF, 153K) Appendix A4. Fitting Mixed Effects Models to Estimate Individual Improvement in Asthma Control and Achievement of Asthma Control (PDF, 143K) Appendix A4, Table 1. Improvement over 12 months of individual-level asthma control by intervention group. Number and percentage of patients who improved by at least 0.5 in asthma control. (PDF, 93K) Appendix A4, Table 2. Control level of 1.5 or better at baseline and 12 months by intervention group. (PDF, 93K) Appendix A5. Mediation—Portal Use and Asthma Outcomes Over Time (PDF, 140K) Appendix A5, Table 1. Asthma control expected values at 0 and 12 months, by patient portal use. (PDF, 93K) Appendix A5, Table 2. Quality-of-life expected values at 0 and 12 months, by patient portal use. (PDF, 93K) Appendix A6. Appendix A6 (PDF, 137K) Appendix A6, Table 1. Patients cross classified by intervention group and number of data collection interviews (maximum of 5) for which patients had an appointment kept with a treating clinician within 6 months of the interview data. (PDF, 93K) Appendix A6, Table 2. Kept appointments, expected asthma control, and changes over time. (PDF, 89K) Appendix A7. Association of Treatment Assignment (Home Visitor Versus Portal Only) and Adherence to Recommended Regimens for Inhaler Adherence (PDF, 130K) Appendix A7, Table 1. Association of treatment assignment (home visitor versus portal training only) and inhaler adherence over time (expected values at months 0 and month 12). (PDF, 92K) Appendix A8. Effect Modification and Identification of Subgroups (PDF, 191K) Appendix A8, Table 1. Skills that would support use of portal and asthma management. (PDF, 98K) Appendix A8, Table 2. Social/Community Barriers. (PDF, 97K) Appendix A8, Table 3. Depression and chronic disease load. (PDF, 97K) Appendix A8, Table 4. Asthma severity. (PDF, 94K) Appendix A8, Table 5. Subgroup analysis by primary language (Spanish). Change over 12 months. (PDF, 94K) Appendix A8, Table 6. Subgroup analysis by concern about using internet for personal health information. (PDF, 97K) Appendix A9. Sensitivity Analysis for Dropout—Using Baseline Covariates (PDF, 100K) Appendix A9, Table 1. Asthma control outcomes at baseline, expected values at 12 months, and changes over time by treatment assignment in 301 adults with uncontrolled asthma. (PDF, 98K) Appendix A10. Estimating Effect of Home Visitors Adjusted for Incomplete Adherence – A Sensitivity Analysis (PDF, 185K) Appendix A10, Table 1. The effect of the home visitor (HV) intervention among the compliers with the home visits. (PDF, 92K) Appendix 11. Detailed Portal Use Activity (PDF, 120K) Appendix 11, Table 1. Portal use activity by patient and intervention group. (PDF, 96K) Appendix 11, Table 2. Portal usage by major activity and by episode (n=2720) during active study involvement among 170 patients who used the portal at least once. (PDF, 95K) Appendix 11, Table 3. Portal usage by major activity and by patient. (PDF, 93K) Appendix A12. The Association of Portal Use With Socio-demographics in 301 African American and Hispanic/Latino Adults With Uncontrolled Asthma (PDF, 112K) Appendix A13. Portal Use and Familiarity With Technology of 301 Participants (PDF, 109K) Appendix A14. For 301 Patients, Association of Portal Use and Diabetes, Hypertension, Hospitalizations, ED Visits for Asthma (PDF, 101K) Appendix A15. Association of Portal Use and Asthma Severity Measured by FEV-1 (n=203) (PDF, 94K) Appendix A16. Patient Comments After Completing Portal Training (PDF, 91K) Institution Receiving PCORI Award: University of Pennsylvania Original Project Title: Using Information Technology to Improve Access, Communication and Asthma in African American and Hispanic/Latino Adults PCORI ID: AS 1307-05218 ClinicalTrials.gov ID: NCT02086565 Suggested citation: Apter AJ, Bryant-Stephens T, Morales KH, et al. (2020). Using IT to Improve Access, Communication, and Asthma in African American and Hispanic/Latino Adults . Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/06.2020.AS.130705218 Disclaimer The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee. Copyright © 2020. University of Pennsylvania. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Bookshelf ID: NBK621035 PMID: 41712744 DOI: 10.25302/06.2020.AS.130705218 Share Views PubReader Print View Cite this Page Apter AJ, Bryant-Stephens T, Morales KH, et al. Testing Whether Home Visits by Community Health Workers Help African-American and Hispanic Patients with Low Incomes Better Manage Asthma [Internet]. Washington (DC): Patient-Centered Outcomes Research Institute (PCORI); 2020 Jun. doi: 10.25302/06.2020.AS.130705218 PDF version of this title (2.2M) In this Page Background Participation of Patients and Other Stakeholders Methods Results Discussion Conclusions REFERENCES Related Publications Acknowledgments Appendices Other titles in this collection PCORI Final Research Reports Related information NLM Catalog Related NLM Catalog Entries PMC PubMed Central citations PubMed Links to PubMed Recent Activity Clear Turn Off Turn On Testing Whether Home Visits by Community Health Workers Help African-American an... Testing Whether Home Visits by Community Health Workers Help African-American and Hispanic Patients with Low Incomes Better Manage Asthma Your browsing activity is empty. Activity recording is turned off. Turn recording back on See more... Follow NCBI Twitter Facebook LinkedIn GitHub NCBI Insights Blog Connect with NLM Twitter Facebook Youtube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov