Interactive Voice Response Calls Research Articles

Implementing a Hospital Call Center Service for Mental Health in Uganda: User-Centered Design Approach.

Mental health conditions are a significant public health problem globally, responsible for >8 million deaths per year. In addition, they lead to lost productivity, exacerbate physical illness, and are associated with stigma and human rights violations. Uganda, like many low- and middle-income countries, faces a massive treatment gap for mental health conditions, and numerous sociocultural challenges exacerbate the burden of mental health conditions. This study aims to describe the development and formative evaluation of a digital health intervention for improving access to mental health care in Uganda. This qualitative study used user-centered design and design science research principles. Stakeholders, including patients, caregivers, mental health care providers, and implementation experts (N=65), participated in focus group discussions in which we explored participants' experience of mental illness and mental health care, experience with digital interventions, and opinions about a proposed digital mental health service. Data were analyzed using the Consolidated Framework for Implementation Research to derive requirements for the digital solution, which was iteratively cocreated with users and piloted. Several challenges were identified, including a severe shortage of mental health facilities, unmet mental health information needs, heavy burden of caregiving, financial challenges, stigma, and negative beliefs related to mental health. Participants' enthusiasm about digital solutions as a feasible, acceptable, and convenient method for accessing mental health services was also revealed, along with recommendations to make the service user-friendly, affordable, and available 24×7 and to ensure anonymity. A hospital call center service was developed to provide mental health information and advice in 2 languages through interactive voice response and live calls with health care professionals and peer support workers (recovering patients). In the 4 months after launch, 456 calls, from 236 unique numbers, were made to the system, of which 99 (21.7%) calls went to voicemails (out-of-office hours). Of the remaining 357 calls, 80 (22.4%) calls stopped at the interactive voice response, 231 (64.7%) calls were answered by call agents, and 22 (6.2%) calls were not answered. User feedback was positive, with callers appreciating the inclusion of peer support workers who share their recovery journeys. However, some participant recommendations (eg, adding video call options) or individualized needs (eg, prescriptions) could not be accommodated due to resource limitations or technical feasibility. This study demonstrates a systematic and theory-driven approach to developing contextually appropriate digital solutions for improving mental health care in Uganda and similar contexts. The positive reception of the implemented service underscores its potential impact. Future research should address the identified limitations and evaluate clinical outcomes of long-term adoption.

Technology-Assisted Buprenorphine Treatment in Rural and Nonrural Settings

Expansion of opioid use disorder treatment is needed, particularly in rural communities. To evaluate technology-assisted buprenorphine (TAB) efficacy (1) over a longer period than previously examined, (2) with the addition of overdose education, and (3) among individuals residing in rural communities. Two parallel, 24-week randomized clinical trials were conducted at the University of Vermont between February 1, 2018, and June 30, 2022. Participants were adults with untreated opioid use disorder from nonrural (trial 1) or rural (trial 2) communities. These trials are part of a programmatic effort to develop TAB protocols to improve treatment availability in underserved areas. Within each trial, 50 participants were randomized to TAB or control conditions. Participants in the TAB group completed bimonthly visits to ingest medication and receive take-home doses via a computerized device. They received nightly calls via an interactive voice response (IVR) system, IVR-generated random call-backs, and iPad-delivered HIV, hepatitis C virus (HCV), and overdose education. Control participants received community resource guides and assistance with contacting resources. All participants received harm reduction supplies and completed monthly assessments. The primary outcome was biochemically verified illicit opioid abstinence across monthly assessments. Secondary outcomes included self-reported opioid use in both groups and abstinence at bimonthly and random call-back visits, treatment adherence, satisfaction, and changes in HIV, HCV, and overdose knowledge among TAB participants. Fifty individuals (mean [SD] age, 40.6 [13.1] years; 28 [56.0%] male) participated in trial 1, and 50 (mean [SD] age, 40.3 [10.8] years; 30 [60.0%] male) participated in trial 2. Participants in the TAB group achieved significantly greater illicit opioid abstinence vs controls at all time points in both trial 1 (85.3% [128 of 150]; 95% CI, 70.7%-93.3%; vs 24.0% [36 of 150]; 95% CI, 13.6%-38.8%) and trial 2 (88.0% [132 of 150]; 95% CI, 72.1%-95.4%; vs 21.3% [32 of 150]; 95% CI, 11.4%-36.5%). High abstinence rates were also observed at TAB participants' bimonthly dosing visits (83.0% [95% CI, 67.0%-92.0%] for trial 1 and 88.0% [95% CI, 71.0%-95.0%] for trial 2). Treatment adherence was favorable and similar between trials (with rates of approximately 99% for buprenorphine administration, 93% for daily IVR calls, and 92% for random call-backs), and 183 of 187 urine samples (97.9%) tested negative for illicit opioids at random call-backs. iPad-delivered education was associated with significant and sustained increases in HIV, HCV, and overdose knowledge. In these randomized clinical trials of TAB treatment, demonstration of efficacy was extended to a longer duration than previously examined and to patients residing in rural communities. ClinicalTrials.gov Identifier: NCT03420313.

Home Hospice Family Caregivers' Use of Audio Diaries and Reported Prevalence of Patient and Caregiver Symptoms.

Family caregivers are essential to home hospice care for patients with advanced cancer, including reporting patient symptoms to hospice providers for follow-up. Hospice caregiving can also impact personal well-being. 1) Assess home hospice caregivers' use of prospective, longitudinal audio diaries tracking patient and caregiver wellbeing; 2) Explore how patient-focused vs. caregiver-focused diary prompts perform; 3) Examine the prevalence of interactive voice response (IVR)-tracked symptoms and whether diaries revealed additional symptoms. Caregivers (N=102) were asked to report patient and caregiver symptoms via daily IVR calls and could record optional diaries responding to patient-focused or caregiver-focused prompts. Diaries were transcribed, classified by presence/absence of new information, and compared by prompt type. Content coding for IVR-tracked symptoms and inductive coding for additional symptoms were summarized by frequency counts and exemplary quotes. Sixty-nine percent of participants (n=70) recorded diaries, and of these 72.86% (n=51) recorded ≥ one new-information diary. The median recording length was 53.00 seconds (SD=53.36). Participants responding to the caregiver-focused prompt (n=33) recorded more diaries than those in the patient-focused group (n=37; U=437.500, P=0.04. Most prevalent IVR-tracked symptoms were patient fatigue/weakness (26.54% of symptoms mentioned) and pain (23.08%), and caregiver anxiety/nervousness (47.51%) and fatigue (22.10%). The most prevalent additional symptoms were patient increasing sleepiness/sleeping (26.32%) and breathing difficulties (24.32%), and negative caregiver emotions (e.g., guilt, resentment, anger; 29.17%). Prospective audio diaries offer a viable avenue for communicating symptoms and support needs. Future research will focus on leveraging longitudinal data for developing focused and tailored caregiver support interventions.

Performance and Resource Requirements of In-Person, Voice Call, and Automated Telephone-Based Socioeconomic Data Collection Modalities for Community-Based Health Programs

Gathering data on socioeconomic status (SES) is a prerequisite for health programs that aim to improve equity. There is a lack of evidence on which approaches offer the best combination of reliability, cost, and acceptability. To compare the performance of different approaches to gathering data on SES in community health programs. A search of the Cochrane Library, MEDLINE, Embase, Global Health, ClinicalTrials.gov, the World Health Organization International Clinical Trials Registry Platform, and OpenGrey from 1999 to June 29, 2021, was conducted, with no language limits. Google Scholar was also searched and the reference lists of included articles were checked to identify further studies. The search was performed on June 29, 2021. Any empirical study design was eligible if it compared 2 or more modalities to elicit SES data from the following 3 categories: in-person, voice call, or automated telephone-based systems. Two reviewers independently screened titles, abstracts, and full-text articles and extracted data. They also assessed the risk of bias using Cochrane tools and assessed the certainty of the evidence using the Grading of Recommendations, Assessment, Development and Evaluation approach. Findings were synthesized thematically without meta-analysis. Response rate, equivalence, time, costs, and acceptability to patients and health care professionals. The searches returned 3943 records. The 11 included studies reported data on 14 036 individuals from 7 countries, collecting data on 11 socioeconomic domains using 2 or more of the following modes: in-person surveys, computer-assisted telephone interviews (CATIs), and 2 types of automated data collection: interactive voice response calls (IVRs) and web surveys. Response rates were greater than 80% for all modes except IVRs. Equivalence was high across all modes (Cohen κ > 0.5). There were insufficient data to make robust time and cost comparisons. Patients reported high levels of acceptability providing data via IVRs, web surveys, and CATIs. Selecting an appropriate and cost-effective modality to elicit SES data is an important first step toward advancing equitable effective service coverage. This systematic review did not identify evidence that remote and automated data collection modes differed from human-led and in-person approaches in terms of reliability, cost, or acceptability.

Patient-Centered Pain Care Using Artificial Intelligence and Mobile Health Tools

Cognitive behavioral therapy for chronic pain (CBT-CP) is a safe and effective alternative to opioid analgesics. Because CBT-CP requires multiple sessions and therapists are scarce, many patients have limited access or fail to complete treatment. To determine if a CBT-CP program that personalizes patient treatment using reinforcement learning, a field of artificial intelligence (AI), and interactive voice response (IVR) calls is noninferior to standard telephone CBT-CP and saves therapist time. This was a randomized noninferiority, comparative effectiveness trial including 278 patients with chronic back pain from the Department of Veterans Affairs health system (recruitment and data collection from July 11, 2017-April 9, 2020). More patients were randomized to the AI-CBT-CP group than to the control (1.4:1) to maximize the system's ability to learn from patient interactions. All patients received 10 weeks of CBT-CP. For the AI-CBT-CP group, patient feedback via daily IVR calls was used by the AI engine to make weekly recommendations for either a 45-minute or 15-minute therapist-delivered telephone session or an individualized IVR-delivered therapist message. Patients in the comparison group were offered 10 therapist-delivered telephone CBT-CP sessions (45 minutes/session). The primary outcome was the Roland Morris Disability Questionnaire (RMDQ; range 0-24), measured at 3 months (primary end point) and 6 months. Secondary outcomes included pain intensity and pain interference. Consensus guidelines were used to identify clinically meaningful improvements for responder analyses (eg, a 30% improvement in RMDQ scores and pain intensity). Data analyses were performed from April 2021 to May 2022. The study population included 278 patients (mean [SD] age, 63.9 [12.2] years; 248 [89.2%] men; 225 [81.8%] White individuals). The 3-month mean RMDQ score difference between AI-CBT-CP and standard CBT-CP was -0.72 points (95% CI, -2.06 to 0.62) and the 6-month difference was -1.24 (95% CI, -2.48 to 0); noninferiority criterion were met at both the 3- and 6-month end points (P < .001 for both). A greater proportion of patients receiving AI-CBT-CP had clinically meaningful improvements at 6 months as indicated by RMDQ (37% vs 19%; P = .01) and pain intensity scores (29% vs 17%; P = .03). There were no significant differences in secondary outcomes. Pain therapy using AI-CBT-CP required less than half of the therapist time as standard CBT-CP. The findings of this randomized comparative effectiveness trial indicated that AI-CBT-CP was noninferior to therapist-delivered telephone CBT-CP and required substantially less therapist time. Interventions like AI-CBT-CP could allow many more patients to be served effectively by CBT-CP programs using the same number of therapists. ClinicalTrials.gov Identifier: NCT02464449.

Utility of interactive voice response self-monitoring in stabilizing initial change during natural recovery attempts among persons with alcohol use disorder

BackgroundProviding extensive low threshold, low intensity interventions during self-initiated recovery attempts may help to stabilize initial change during early recovery when relapse risk is high. This study provided untreated persons with alcohol use disorder (AUD) daily access to an interactive voice response (IVR) self-monitoring system during the early months of a natural recovery attempt to report drinking-related variables. The study prospectively examined event-level associations among IVR call frequency, urges to drink, and drinking practices to evaluate whether more frequent IVR utilization helped participants to manage urges and attenuated the established relationship between urges and relapse. MethodsRecently resolved untreated problem drinkers (N = 128) recruited from the community received daily IVR access via telephone for 4–5 months to report their urges to drink and drinking practices during the preceding 24 h. Among IVR callers, analyses using SAS Proc Glimmix evaluated associations among IVR utilization, urges to drink, and drinking practices. ResultsConsistent with previous research, greater urges were associated with a higher likelihood of heavy drinking. Moreover, heavier drinking was associated overall with higher IVR utilization, and the more frequently participants called the IVR system over time, the likelihood of heavy drinking episodes was less strongly related to urges. ConclusionsRecently resolved persons with AUD who experienced higher urges to drink and were consuming alcohol above heavy drinking thresholds received some limited benefits in slowing their return to a heavy drinking pattern if they called the IVR system frequently, even though they were not successfully maintaining their initial abstinent or moderation drinking recoveries. These results, using a prospective design, indicated complex event-level associations among IVR calls, urges, and drinking during early natural recovery that merit further investigation.

Adherence to Daily Interactive Voice Response Calls for a Chronic Pain Intervention

Adherence to Daily Interactive Voice Response Calls for a Chronic Pain Intervention

IVR System Use by Patients with Heart Failure: Compliance and Services Utilization Patterns.

Heart failure (HF) is the leading cause of cardiovascular morbidity and health care utilization globally. Much of the cost for HF is related to hospitalization, strategies to decrease cost need to focus on avoiding unnecessary hospital readmission. Interactive voice response (IVR) is an automated telephony system that leverages existing telephone lines to monitor patients post-discharge, for early intervention. This study explores the pattern of IVR use by HF patients in the IVR program at the University of Ottawa Heart Institute (UOHI) and assesses IVR use by patients in relation to symptoms, compliance behavior, lifestyle, and hospital readmission. A total of 902 HF patients were considered; the mean age was 70 years, and 59.4% were male. Over a 12-week period of IVR use, there was an overall increase in medication adherence and a decrease in symptoms occurrence, weight gain and readmission rate. The highest and lowest compliance rates were associated with medication adherence and exercise, respectively. Overall, older, female patients from rural/community hospitals were more likely to complete the IVR calls, have less symptoms occurrence, comply with medications, weight, and lifestyle recommendations. The findings suggest that IVR system use can have a positive impact on HF patients’ management. The increased use of IVR in remote patient monitoring will allow for a cheaper and more accessible form of home monitoring. Leveraging IVR technology to support other conditions, especially during a pandemic, may be beneficial for patients to avoid unnecessary visits to the hospital and complications due to delay in seeking care.

Patient and supporter factors affecting engagement with diabetes telehealth.

To assess what patient, family supporter, andcall characteristics predicted whether patients completed automated and coach-provided calls in a telehealth diabetes intervention. A total of 123 adults with type 2 diabetes and high glycated hemoglobin A1c (HbA1c) or blood pressure, enrolled with a family supporter, received automated interactive voice response (IVR) and coach-provided visit preparation calls over 12 months. Data from baseline surveys and diabetes-related clinical information from patient medical records were entered into multilevel, multivariate regression models of associations between participant and call characteristics with call completion. A total of 76.3% of 2784 IVR calls and 75.8% of 367 visit preparation calls were completed. For IVR calls, patients with recent call-triggered provider alerts had higher odds of call completion (adjusted odds ratio [AOR], 3.5; 95% CI, 2.2-5.5); those with depressive symptoms (AOR, 0.4; 95% CI, 0.2-0.9), higher HbA1c (AOR, 0.8; 95% CI, 0.6-0.99), and more months in the study (AOR, 0.9; 95% CI, 0.87-0.94 per month) had lower odds. For visit preparation calls, higher patient activation scores predicted higher call completion (AOR, 1.4; 95% CI, 1.1-1.9); patient college education predicted less call completion (AOR, 0.3; 95% CI, 0.2-0.6). Supporter help taking medications predicted less completion of both call types. Patient age did not predict call completion. Patients of all ages completed telehealth calls at a high rate. Automated IVR calls were completed more often when urgent issues were identified to patients' providers, but less often if patients had high HbA1c ordepression. Visit preparation call content should be tailored to patient education level. Family help with medications may identify patients needing additional support to engage with telehealth.

Sustained Care Smoking Cessation Intervention for Individuals Hospitalized for Psychiatric Disorders

Smoking among individuals with serious mental illness (SMI) represents a major public health problem. Intervening during a psychiatric hospital stay may provide an opportunity to aid engagement in smoking cessation treatment and facilitate success in quitting. To examine the effectiveness of a multicomponent, sustained care (SusC) smoking cessation intervention in adults with SMI receiving inpatient psychiatric care. The Helping HAND 3 randomized clinical trial compared SusC with usual care (UC) among individuals with SMI who smoked daily and were receiving inpatient psychiatric care in Austin, Texas, in a single hospital. The study was conducted from July 2015 through August 2019. The UC intervention involved brief smoking cessation information, self-help materials and advice from the admitting nurse, and an offer to provide nicotine replacement therapy during hospitalization. The SusC intervention included 4 main components designed to facilitate patient engagement with postdischarge smoking cessation resources: (1) inpatient motivational counseling; (2) free transdermal nicotine patches on discharge; (3) an offer of free postdischarge telephone quitline, text-based, and/or web-based smoking cessation counseling, and (4) postdischarge automated interactive voice response calls or text messages. The primary outcome was biochemically verified 7-day point-prevalence abstinence at 6-month follow-up. A secondary outcome was self-reported smoking cessation treatment use at 1, 3, and 6 months after discharge. A total of 353 participants were randomized, of whom 342 were included in analyses (mean [SD] age, 35.8 [12.3] years; 268 White individuals [78.4%]; 280 non-Hispanic individuals [81.9%]; 169 women [49.4%]). They reported smoking a mean (SD) of 16.9 (10.4) cigarettes per day. Participants in the SusC group evidenced significantly higher 6-month follow-up point-prevalence abstinence rates than those in the UC group (8.9% vs 3.5%; adjusted odds ratio, 2.95 [95% CI, 1.24-6.99]; P = .01). The number needed to treat was 18.5 (95% CI, 9.6-306.4). A series of sensitivity analyses confirmed effectiveness. Finally, participants in the SusC group were significantly more likely to report using smoking cessation treatment over the 6 months postdischarge compared with participants in the UC group (74.6% vs 40.5%; relative risk, 1.8 [95% CI, 1.51-2.25]; P < .001). The findings of this randomized clinical trial provide evidence for the effectiveness of a scalable, multicomponent intervention in promoting smoking cessation treatment use and smoking abstinence in individuals with SMI following hospital discharge. ClinicalTrials.gov Identifier: NCT02204956.

Combining Wearable Devices and Mobile Surveys to Study Child and Youth Development in Malawi: Implementation Study of a Multimodal Approach.

BackgroundMultimodal approaches have been shown to be a promising way to collect data on child development at high frequency, combining different data inputs (from phone surveys to signals from noninvasive biomarkers) to understand children’s health and development outcomes more integrally from multiple perspectives.ObjectiveThe aim of this work was to describe an implementation study using a multimodal approach combining noninvasive biomarkers, social contact patterns, mobile surveying, and face-to-face interviews in order to validate technologies that help us better understand child development in poor countries at a high frequency.MethodsWe carried out a mixed study based on a transversal descriptive analysis and a longitudinal prospective analysis in Malawi. In each village, children were sampled to participate in weekly sessions in which data signals were collected through wearable devices (electrocardiography [ECG] hand pads and electroencephalography [EEG] headbands). Additionally, wearable proximity sensors to elicit the social network were deployed among children and their caregivers. Mobile surveys using interactive voice response calls were also used as an additional layer of data collection. An end-line face-to-face survey was conducted at the end of the study.ResultsDuring the implementation, 82 EEG/ECG data entry points were collected across four villages. The sampled children for EEG/ECG were 0 to 5 years old. EEG/ECG data were collected once a week. In every session, children wore the EEG headband for 5 minutes and the ECG hand pad for 3 minutes. In total, 3531 calls were sent over 5 weeks, with 2291 participants picking up the calls and 984 of those answering the consent question. In total, 585 people completed the surveys over the course of 5 weeks.ConclusionsThis study achieved its objective of demonstrating the feasibility of generating data through the unprecedented use of a multimodal approach for tracking child development in Malawi, which is one of the poorest countries in the world. Above and beyond its multiple dimensions, the dynamics of child development are complex. It is the case not only that no data stream in isolation can accurately characterize it, but also that even if combined, infrequent data might miss critical inflection points and interactions between different conditions and behaviors. In turn, combining different modes at a sufficiently high frequency allows researchers to make progress by considering contact patterns, reported symptoms and behaviors, and critical biomarkers all at once. This application showcases that even in developing countries facing multiple constraints, complementary technologies can leverage and accelerate the digitalization of health, bringing benefits to populations that lack new tools for understanding child well-being and development.

Assessment of Psychological Distress in Adults With Type 2 Diabetes Mellitus Through Technologies: Literature Review.

BackgroundThe use of technological devices can support the self-management of individuals with type 2 diabetes mellitus (T2DM), particularly in addressing psychological distress. However, there is poor consistency in the literature regarding the use of psychological instruments for the web-based screening of patients’ psychological distress and subsequent monitoring of their psychological condition during digital interventions.ObjectiveThis study aims to review previous literature on the types of psychological instruments delivered in digital interventions for assessing depression, anxiety, and stress in patients with T2DM.MethodsThe literature review was conducted using the PsycINFO, CINAHL and PubMed databases, in which the following terms were considered: diabetes mellitus, measure, assessment, self-care, self-management, depression, anxiety, stress, technology, eHealth, mobile health, mobile phone, device, and smartphone.ResultsIn most studies, psychological assessments were administered on paper. A few studies deployed self-reporting techniques employing automated telephonic assessment, a call system for screening and monitoring patients’ conditions and preferences, or through telephone interviews via interactive voice response calls, a self-management support program leveraging tailored messages and structured emails. Other studies used simple telephone interviews and included the use of apps for tablets and smartphones to assess the psychological well-being of patients. Finally, some studies deployed mood rating scales delivered through tailored text message–based support systems.ConclusionsThe deployment of appropriate psychological tools in digital interventions allows researchers and clinicians to make the screening of anxiety, stress, and depression symptoms faster and easier in patients with T2DM. Data from this literature review suggest that mobile health solutions may be preferred tools to use in such digital interventions.

User Perceptions and Experiences of an Interactive Voice Response Mobile Phone Survey Pilot in Uganda: Qualitative Study

BackgroundWith the growing burden of noncommunicable diseases in low- and middle- income countries, the World Health Organization recommended a stepwise approach of surveillance for noncommunicable diseases. This is expensive to conduct on a frequent basis and using interactive voice response mobile phone surveys has been put forth as an alternative. However, there is limited evidence on how to design and deliver interactive voice response calls that are robust and acceptable to respondents.ObjectiveThis study aimed to explore user perceptions and experiences of receiving and responding to an interactive voice response call in Uganda in order to adapt and refine the instrument prior to national deployment.MethodsA qualitative study design was used and comprised a locally translated audiorecorded interactive voice response survey delivered in 4 languages to 59 purposively selected participants' mobile phones in 5 survey rounds guided by data saturation. The interactive voice response survey had modules on sociodemographic characteristics, physical activity, fruit and vegetable consumption, diabetes, and hypertension. After the interactive voice response survey, study staff called participants back and used a semistructured interview to collect information on the participant’s perceptions of interactive voice response call audibility, instruction clarity, interview pace, language courtesy and appropriateness, the validity of questions, and the lottery incentive. Descriptive statistics were used for the interactive voice response survey, while a framework analysis was used to analyze qualitative data.ResultsKey findings that favored interactive voice response survey participation or completion included preference for brief surveys of 10 minutes or shorter, preference for evening calls between 6 PM and 10 PM, preference for courteous language, and favorable perceptions of the lottery-type incentive. While key findings curtailing participation were suspicion about the caller’s identity, unclear voice, confusing skip patterns, difficulty with the phone interface such as for selecting inappropriate digits for both ordinary and smartphones, and poor network connectivity for remote and rural participants.ConclusionsInteractive voice response surveys should be as brief as possible and considerate of local preferences to increase completion rates. Caller credibility needs to be enhanced through either masking the caller or prior community mobilization. There is need to evaluate the preferred timing of interactive voice response calls, as the finding of evening call preference is inconclusive and might be contextual.

The Effectiveness of Artificial Intelligence Conversational Agents in Health Care: Systematic Review.

BackgroundThe high demand for health care services and the growing capability of artificial intelligence have led to the development of conversational agents designed to support a variety of health-related activities, including behavior change, treatment support, health monitoring, training, triage, and screening support. Automation of these tasks could free clinicians to focus on more complex work and increase the accessibility to health care services for the public. An overarching assessment of the acceptability, usability, and effectiveness of these agents in health care is needed to collate the evidence so that future development can target areas for improvement and potential for sustainable adoption.ObjectiveThis systematic review aims to assess the effectiveness and usability of conversational agents in health care and identify the elements that users like and dislike to inform future research and development of these agents.MethodsPubMed, Medline (Ovid), EMBASE (Excerpta Medica dataBASE), CINAHL (Cumulative Index to Nursing and Allied Health Literature), Web of Science, and the Association for Computing Machinery Digital Library were systematically searched for articles published since 2008 that evaluated unconstrained natural language processing conversational agents used in health care. EndNote (version X9, Clarivate Analytics) reference management software was used for initial screening, and full-text screening was conducted by 1 reviewer. Data were extracted, and the risk of bias was assessed by one reviewer and validated by another.ResultsA total of 31 studies were selected and included a variety of conversational agents, including 14 chatbots (2 of which were voice chatbots), 6 embodied conversational agents (3 of which were interactive voice response calls, virtual patients, and speech recognition screening systems), 1 contextual question-answering agent, and 1 voice recognition triage system. Overall, the evidence reported was mostly positive or mixed. Usability and satisfaction performed well (27/30 and 26/31), and positive or mixed effectiveness was found in three-quarters of the studies (23/30). However, there were several limitations of the agents highlighted in specific qualitative feedback.ConclusionsThe studies generally reported positive or mixed evidence for the effectiveness, usability, and satisfactoriness of the conversational agents investigated, but qualitative user perceptions were more mixed. The quality of many of the studies was limited, and improved study design and reporting are necessary to more accurately evaluate the usefulness of the agents in health care and identify key areas for improvement. Further research should also analyze the cost-effectiveness, privacy, and security of the agents.International Registered Report Identifier (IRRID)RR2-10.2196/16934

A pilot randomized controlled trial (RCT) of daily versus weekly interactive voice response calls to support adherence among antiretroviral treatment patients in India.

There are more than two million people living with HIV (PLH) in India, with more than 30% on antiretroviral treatment (ART) estimated to be non-adherent. This study aimed to (I) document adherence rates and related factors among ART patients in a large ART clinic in India, and (II) pilot test daily and weekly interactive voice response (IVR) calls to improve ART adherence and related outcomes. ART patients reporting missing at least one dose in prior 6 months (N=362) were enrolled and assessed via self-report and medical record review. Participants were randomized to one of two conditions: (I) twice-daily IVR call reminders with self-management support messaging, plus a weekly IVR adherence assessment; or (II) an attention control, with only weekly IVR adherence assessment. Participants completed study assessments at baseline, 2-, 4-, and 6-months with high retention (88% to 96%). Intention-to-treat analyses found limited support for intervention effects for improving or maintaining ART adherence or CD4 counts between the two study arms over 6-months follow-up. Adherence increased significantly in the six months prior to baseline from about 65% to >95% with perfect adherence based on pill counts from medical records and consistent with patient self-report measures, which presented ceiling effects for detecting improvements in ART adherence in response to IVR intervention exposure. There was also limited support for intervention effects on secondary, self-management outcomes. High levels of adherence were sustained throughout the 6-month RCT. IVR regulation changes in India delayed study launch for 6 months, which likely allowed mobilization of improved adherence at the clinic, provider and patient levels in anticipation of the study launch. Therefore, ceiling effects limited inferences on intervention effects to improve adherence. Results suggest that clinic-level adherence monitoring may be sufficient to mobilize adherence improvements by providers and patients. ClinicalTrials.gov registration #NCT02118454.

Electronic Health Record Portal Messages and Interactive Voice Response Calls to Improve Rates of Early Season Influenza Vaccination: Randomized Controlled Trial.

BackgroundPatient reminders for influenza vaccination, delivered via an electronic health record patient portal and interactive voice response calls, offer an innovative approach to engaging patients and improving patient care.ObjectiveThe goal of this study was to test the effectiveness of portal and interactive voice response outreach in improving rates of influenza vaccination by targeting patients in early September, shortly after vaccinations became available.MethodsUsing electronic health record portal messages and interactive voice response calls promoting influenza vaccination, outreach was conducted in September 2015. Participants included adult patients within a large multispecialty group practice in central Massachusetts. Our main outcome was electronic health record–documented early influenza vaccination during the 2015-2016 influenza season, measured in November 2015. We randomly assigned all active portal users to 1 of 2 groups: (1) receiving a portal message promoting influenza vaccinations, listing upcoming clinics, and offering online scheduling of vaccination appointments (n=19,506) or (2) receiving usual care (n=19,505). We randomly assigned all portal nonusers to 1 of 2 groups: (1) receiving interactive voice response call (n=15,000) or (2) receiving usual care (n=43,596). The intervention also solicited patient self-reports on influenza vaccinations completed outside the clinic. Self-reported influenza vaccination data were uploaded into the electronic health records to increase the accuracy of existing provider-directed electronic health record clinical decision support (vaccination alerts) but were excluded from main analyses.ResultsAmong portal users, 28.4% (5549/19,506) of those randomized to receive messages and 27.1% (5294/19,505) of the usual care group had influenza vaccinations documented by November 2015 (P=.004). In multivariate analysis of portal users, message recipients were slightly more likely to have documented vaccinations when compared to the usual care group (OR 1.07, 95% CI 1.02-1.12). Among portal nonusers, 8.4% (1262/15,000) of those randomized to receive calls and 8.2% (3586/43,596) of usual care had documented vaccinations (P=.47), and multivariate analysis showed nonsignificant differences. Over half of portal messages sent were opened (10,112/19,479; 51.9%), and over half of interactive voice response calls placed (7599/14,984; 50.7%) reached their intended target, thus we attained similar levels of exposure to the messaging for both interventions. Among portal message recipients, 25.4% of message openers (2570/10,112) responded to a subsequent question on receipt of influenza vaccination; among interactive voice response recipients, 72.5% of those reached (5513/7599) responded to a similar question.ConclusionsPortal message outreach to a general primary care population achieved a small but statistically significant improvement in rates of influenza vaccination (OR 1.07, 95% CI 1.02-1.12). Interactive voice response calls did not significantly improve vaccination rates among portal nonusers (OR 1.03, 95% CI 0.96-1.10). Rates of patient engagement with both modalities were favorable.Trial RegistrationClinicalTrials.gov NCT02266277; https://clinicaltrials.gov/ct2/show/NCT02266277

Digital Health for Assessment and Intervention Targeting Tobacco and Cannabis Co-Use

This article aims to summarize current research on digital health for assessment and intervention targeting tobacco and cannabis co-use and to answer the following questions: Which digital tools have been used? Which populations have been targeted? And what are implications for future research? Ecological Momentary Assessment (EMA) via text messages or interactive voice response calls has been used to capture co-use patterns within a time window or co-administration of both substances via blunts among young adults. Feasibility of multicomponent interventions targeting dual cessation of both substances among adult co-users with cannabis use disorder, delivered via smartphone apps, online, and computer modules has been demonstrated. Digital tools, particularly those using EMAs and mobile sensors, should be expanded to assess co-use of emerging tobacco and cannabis products. Digital cessation interventions should be tailored to different groups of co-users and address specific mechanisms underlying different co-use patterns.

Rapid Assessment Procedures Formative Research Approach Used to Design a Mobile-technology Enhanced Social and Behavior Change Communication Nutrition Strategy in Nigeria (P16-053-19)

Objectives1To design of a tailored social and behavior change communication (SBCC) strategy to improve program coverage and key nutrition behaviors of internally displaced persons (IDP) in northeast Nigeria,2To determine preferred and appropriate ways of integrating one- and two-way mobile communication tools into SBCC activities,3To demonstrate application of the Rapid Assessment Procedures (RAP) formative research approach in a humanitarian context with severe insecurity. MethodsA RAP approach and was conducted over 4 weeks in August, 2018. It included multiple qualitative methods and various participant types for triangulation. In-depth interviews (IDI), open-ended interviews and focus group discussions (FGD) were conducted among mothers and fathers of children under 5 years, local community leaders and health workers. Semi-structured interviews were conducted among professionals (nutrition and technology experts) with contextual programming knowledge. Interviews and FGD were digitally recorded in Hausa or Kanuri before translation and transcription into English. Textual data were managed in NVivo software, where salient themes pertinent to the research questions were identified. ResultsA total of 29 IDI, 12 FGD and 10 semi-structured interviews were conducted among diverse stakeholders. An SBCC strategy was designed with preferred communication channels, defined audience segments and tailored messages using emic terms in local languages. Primary caregivers, typically female in this context, were identified as the primary audience. Community health workers, local leaders (e.g., imams) and husbands were identified as secondary audiences who influence the primary caregivers’ nutrition behaviors. Interpersonal (care groups), media (radio) and mobile technologies (interactive voice response calls; SMS; chatbots in low-cost websites) emerged as preferred and viable communication channels. Salient terms and phrases were applied to standard biomedical nutrition messages about breastfeeding and complementary feeding. ConclusionsA RAP approach offers a viable formative research model in humanitarian contexts where security challenges might not allow for desired lengthy, in-depth formative work typical of high-quality SBCC strategy design. Funding SourcesWorld Food Program, Korea International Cooperation Agency (KOICA). Supporting Tables, Images and/or Graphs▪

Provider Counseling and Weight Loss Outcomes in a Primary Care-Based Digital Obesity Treatment.

Primary care-based digital health weight loss interventions offer promise for addressing obesity in underserved populations. To determine if primary care providers' weight counseling is associated with weight change during a weight loss intervention. This is a secondary analysis of a randomized clinical trial testing a 12-month primary care-based digital health weight loss intervention. Participants were community health center patients with body mass indexes of 30-44.9kg/m2. The weight loss intervention included tailored behavioral goal setting; weekly goal monitoring via text messaging or interactive voice response calls; counseling calls; skills training material; and participant-tailored recommendations for provider counseling. At 6 and 12months, participants' weight was measured and they reported if their provider delivered weight counseling (general or intervention-specific) at their most recent visit and their perception of providers' empathy. Providers' documentation of weight counseling was extracted from health records. Participants (n = 134-141) were predominantly female (70%) and African American (55%) with a mean age of 51years and BMI of 36kg/m2. Participant-reported provider weight counseling was not associated with weight change. However, participants whose providers documented intervention-specific counseling at any point during the intervention (n = 35) lost 3.1kg (95% CI 0.4 to 5.7kg) more than those whose providers documented only general weight counseling (n = 82) and 4.0kg (95% CI 0.1 to 7.9kg) more than those whose providers did not document weight counseling (n = 17). Perceptions of provider empathy were associated with greater weight loss from 6 to 12months (0.8kg per measure unit, 95% CI 0.07 to 1.5kg, p = .03). Provider counseling that focuses specifically on engagement in a weight loss intervention may enhance weight loss outcomes relative to more general weight loss advice. Counseling that enhances patients' perceptions of empathy may be most beneficial for patients' weight loss. NCT01827800.

Implementation and barriers to uptake of interactive voice response technology aimed to improve blood pressure control at a large academic medical center

ObjectivesBlood pressure control among patients with hypertension is a widely recognized quality metric, but many large health systems fail to reach targets set by the Healthcare Effectiveness Data and Information Set. We developed an interactive voice response (IVR) system called the “Mobile You Blood Pressure Program” at a large academic medical center and linked it to the health system’s electronic health record (EHR). The goal of the program was to capture home blood pressure readings in the EHR and to alert ambulatory care clinical pharmacists automatically of readings below or above clinical thresholds through direct messaging in the EHR. The goal of this report is to describe implementation of IVR, initial patient participation rates, and pharmacist-identified barriers to patient enrollment. SettingAmbulatory care clinical pharmacist specialists’ practice in 14 clinics in family medicine and internal medicine at Michigan Medicine, an academic health system serving more than 24,000 patients with a diagnosis of hypertension. Practice DescriptionThis study describes implementation and initial patient enrollment in IVR linked to the EHR for home blood pressure monitoring. EvaluationWe tracked the number of hypertensive patients enrolled and IVR call completion rates between September 2017 and February 2018. We also assessed pharmacist-identified barriers to patient enrollment during 2 separate 2-week intervals in January and February 2018. ResultsBetween September 1, 2017, and February 28, 2018, a total of 71 patients were enrolled from 14 clinics. Patients were scheduled for 1-3 IVR calls per week focusing on medication adherence and blood pressure control. A total of 936 IVR phone calls were made, with 488 (52%) calls completed. Access to a validated home blood pressure monitor was the largest pharmacist-identified barrier to patient enrollment. ConclusionsThe IVR Mobile You Blood Pressure Program represents a new application of digital technology within our health system. Pharmacist-identified barriers to patient participation included access to a validated home blood pressure monitor.