An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial

An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial View ORCID Profile Emre Sezgin , Elizabeth Clarkson , Faith Logan , View ORCID Profile Daniel I. Jackson , Syed-Amad Hussain , Joshua Stokes , Alicia Bunger , View ORCID Profile Guy Brock , Eric Fosler-Lussier , View ORCID Profile Alex R. Kemper , Ahna L. Pai doi: https://doi.org/10.1101/2025.11.17.25340433 Emre Sezgin 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH 2 The Ohio State University College of Medicine , Columbus OH PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Emre Sezgin For correspondence: Emre.Sezgin{at}nationwidechildrens.org Elizabeth Clarkson 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH BA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Faith Logan 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH BS, MEd Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel I. Jackson 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH BSc Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Daniel I. Jackson Syed-Amad Hussain 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH BSc Find this author on Google Scholar Find this author on PubMed Search for this author on this site Joshua Stokes 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH BA, MA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alicia Bunger 2 The Ohio State University College of Medicine , Columbus OH PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Guy Brock 2 The Ohio State University College of Medicine , Columbus OH PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guy Brock Eric Fosler-Lussier 3 Nationwide Children’s Hospital , Columbus OH PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Alex R. Kemper 2 The Ohio State University College of Medicine , Columbus OH 3 Nationwide Children’s Hospital , Columbus OH MD, MPH, MS Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Alex R. Kemper Ahna L. Pai 1 The Abigail Wexner Research Institute at Nationwide Children’s Hospital , Columbus, OH 2 The Ohio State University College of Medicine , Columbus OH PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abstract Full Text Info/History Metrics Data/Code Preview PDF Abstract Unmet health-related social needs (HRSNs) are major drivers of poor health outcomes in early childhood. Children with unmet HRSNs are at greater risk for developmental delays, caregiver stress, and increased healthcare utilization, yet current screening approaches in pediatric primary care are resource-intensive and inconsistently implemented. AI-powered chatbots (conversational agents or virtual assistants) may offer a private, secure, scalable, and cost-effective alternative for identifying unmet needs and connecting families to services. This protocol describes a pilot randomized controlled trial designed to evaluate the feasibility, acceptability, and usability of DAPHNE, an AI-driven chatbot developed to facilitate the identification of unmet HRSNs and provide personalized community resource referrals. One hundred caregivers of children under two years of age will be recruited from Nationwide Children’s Hospital pediatric primary care clinics and randomized to either the standard care (control) group or DAPHNE+ Standard care (intervention) group (n=50 each arm). Caregivers will complete surveys at baseline, 1 month, 3 months, and 6 months post-intervention (depending on the measure). For the intervention group, participants will receive weekly chatbot prompts and on-demand access throughout the 6-month study period. Primary outcomes include study feasibility (recruitment, retention, and survey completion across both arms), acceptability (caregiver-reported ratings in both arms and intervention-specific ratings), and usability of the DAPHNE chatbot (System Usability Scale among intervention participants). Secondary outcomes include caregiver-reported outcome measures (caregiver stress, self-efficacy, satisfaction with resource access, quality of life), and electronic health record-derived measures (including documentation of HRSN screening and referrals, adherence to well-child visits, missed appointments, emergency department utilization, and estimated healthcare costs). In addition, ten primary care providers will also participate to assess workflow integration and report on current HRSN practices. Mixed-methods analyses will integrate survey data, chatbot engagement metrics, and qualitative interviews to refine both the intervention and the study protocol. The results of this study will inform the design of a future multi-site trial to evaluate the efficacy and implementation of DAPHNE for addressing HRSNs in pediatric primary care. Trial registration: NCT07168382 . INTRODUCTION Health-related social needs (HRSNs), including food insecurity, unstable housing, transportation challenges, and financial instability, account for more than 50% of modifiable health outcomes.[ 1 , 2 ] Children with unmet HRSNs have substantially elevated risks for cognitive, behavioral, and developmental difficulties, having greater risk than peers from resource secure households.[ 3 – 6 ] They are also 18% more likely to have emergency department (ED) visits and 36% more likely to experience hospitalizations compared with children without social risks (IRR=1.18, 95% CI [1.12- 1.23]; 1.36, 95% CI [1.26–1.47]).[ 7 ] This burden is especially pronounced among infants younger than 2 years, for whom early-life HRSNs, such as food insecurity or benefit instability, are linked to measurable behavioral and regulatory difficulties by six months of age (aOR= 1.64-1.86 across domains; aOR= 2.16 for 2 or more identified HRSNs).[ 8 ] This early-life concentration represents a high-yield prevention target and reinforces primary care as the locus for proactive screening and linkage to services.[ 7 – 9 ] Pediatric primary care clinics (PCCs) are therefore critical access points for early identification and intervention but current screening practices are constrained by staffing shortages, workflow inefficiencies, and high operational costs.[ 10 ] Caregivers are also often reluctant to disclose needs due to stigma, language barriers, or confusion about service navigation, which further exacerbates under-identification.[ 11 ] Referral completion is constrained by a combination of structural, provider-level, and caregiver-level barriers. Structural barriers include complex eligibility requirements, poor coordination, and limited availability of community resources. Provider-level barriers, such as limited familiarity with available services and uncertainty about referral processes, further hinder follow-through. Caregivers also face barriers related to time, transportation, and competing demands that make service access difficult.[ 12 – 14 ] Despite the availability of standardized screening instruments, fewer than 25% of hospitals and 16% of physician practices routinely screen for HRSNs.[ 15 ] When implemented, screening is often limited to episodic encounters such as annual well-child visits, missing acute and urgent social needs at other timepoints.[ 16 , 17 ] In this cohort, particularly the caregivers of infants (ages 0–2) are disproportionately affected by the lack of screening.[ 3 , 8 ] For instance,food insecurity in this age group is strongly associated with adverse developmental outcomes,[ 18 ] higher hospitalization rates,[ 19 ] and increased risk of chronic conditions including obesity and diabetes.[ 20 ] The barriers are particularly pronounced for Medicaid-insured children, who experience greater challenges in securing basic necessities such as infant formula, diapers, and transportation.[ 21 , 22 ] Rationale for AI-Driven Solutions AI-based chatbots (also known as conversational agents and virtual assistants) offer an opportunity to overcome some of these structural and provider-related barriers in HRSN identification and referral. Unlike traditional clinician-initiated screening, chatbots can provide continuous, user-initiated engagement across multiple contexts and time points, enabling real-time identification of urgent needs.[ 23 – 25 ] By facilitating private conversations, chatbots play a positive role to reduce stigma and support honest disclosure for some groups.[ 11 , 26 – 28 ] They can also automate referrals using up-to-date community resource databases, and reduce provider burden while offering tailored, context-specific recommendations.[ 25 , 29 , 30 ] These advantages are further supported by our preliminary engagement sessions with social work teams, caregivers, and patient advocates, who indicated high acceptance of chatbot-based social needs assessment, provided that privacy and usability are prioritized.[ 23 , 31 ] Conceptual Framework The methods for this study were informed by the Obesity-Related Behavioral Intervention Trials (ORBIT) model[ 32 , 33 ], which provides a systematic framework for developing behavioral interventions by progressing through iterative phases (from early conceptualization and definition of intervention components, to feasibility and pilot testing, and ultimately to efficacy and effectiveness trials). Prior applications of conversational AI in healthcare, including tools for mental health support, symptom tracking, and caregiver-facing applications, demonstrate early evidence and implications for digital engagement strategies.[ 34 – 39 ] Building on this foundation, the present pilot randomized controlled trial (pRCT) corresponds to Phase IIb of ORBIT, which emphasizes feasibility testing, establishing acceptability, and refinement of study protocols before progression to full-scale efficacy trials. Study Objectives The long-term goal of this research is to develop an empirically supported, scalable, AI-driven solution to facilitate HRSN identification and referral in pediatric primary care. This pilot study has three specific objectives: (1) To assess the feasibility, acceptability, and usability of the DAPHNE chatbot with caregivers and providers of pediatric primary care patients compared with standard care; (2) To evaluate the feasibility and acceptability of the study protocol, including recruitment, retention, and measurement strategies, for informing design of a future large-scale trial; and (3) To characterize the current standard of care for addressing unmet HRSNs, thereby informing selection of appropriate comparator conditions and outcomes in future trials. Preliminary Work We have conducted several studies to inform DAPHNE’s development. A scoping review of patient-facing chatbots and voice agents revealed high interest but a lack of rigorous trials on chatbot-based interventions and indicating the need for systematic evaluation.[ 40 ] Building upon this knowledge, we developed a voice-interactive diary for caregivers of children with special health care needs; over 80% reported ease of use and approximately half increased health tracking frequency, demonstrating caregiver receptivity to conversational tools.[ 39 ] To enable integration of unstructured caregiver data, we created a natural language processing pipeline that accurately structured caregiver-generated notes and transcripts (F1 >0.7), validating real-time data capture for clinical use.[ 41 , 42 ] Finally, in a feasibility study of a semi-functional DAPHNE prototype (n=13), community health and social workers rated usability at 72/100 on the System Usability Scale and emphasized its value for triaging referrals.[ 23 ] Together, these studies establish feasibility, provider and caregiver acceptance, and the need for a pilot randomized controlled trial. METHODS Study Design and Setting This study is designed as a pilot randomized controlled trial (pRCT) guided by the ORBIT model (Phase IIb).[ 32 , 33 ] One hundred caregivers of children ≤2 years of age will be recruited from Nationwide Children’s Hospital Primary Care Network (NCH PCN). Following baseline assessment, participants will be randomized in a 1:1 ratio to either the standard care group or DAPHNE + standard of care group (n=50 per arm). In addition, 10 providers will be recruited to complete surveys on workflow integration and HRSN practices. The trial will be conducted in NCH PCCs, which provided nearly 250,000 visits in 2024 for mostly Medicaid-enrolled patients. The PCN includes 14 clinics across the Columbus metropolitan area. For the purpose of this study, we identified 2 clinics with the capacity to support the research project and a relatively higher number of patients with unmet HRSNs. Study Status and Timeline This study is currently in the Initiation and Planning phase, with the protocol finalized and Institutional Review Board (IRB) approval obtained (NCH IRB #00004369). We initiated the process of participant recruitment in Q1, Year 1, and are actively enrolling participants. The time schedule is provided in Figure 1 . We project that participant recruitment will be completed in Q2 of Year 2, primary data collection (6-month follow-up) will be completed in Q1 of Year 3, and final results are expected for manuscript submission in Q4 of Year 3. Download figure Open in new tab Figure 1. Time schedule of enrolment, interventions, and assessments on participant outcome (informed by the SPIRIT 2025 reporting guideline)[ 43 ] Participants Caregivers. Recruitment will occur through the NCH PCN, with research staff approaching caregivers during routine visits. We estimate to recruit n=50 participants per arm. We facilitate recruitment using flyers and social worker referrals. Furthermore, we will communicate with potential participants who are interested via virtual methods (phone call, teleconference). Inclusion criteria are: (1) primary caregiver of a child ≤2 years old receiving care at the selected clinics in the PCN, (2)self-identify with at least one unmet HRSN (3) ability to provide informed consent and complete study surveys in English, (4) ownership of a mobile device for chatbot use, and (5) willingness to participate in follow-up assessments. Exclusion criteria: none beyond failure to meet the above criteria. Caregivers will be compensated for their participation. Providers. To evaluate the feasibility of integrating the DAPHNE chatbot into clinical workflows at primary care clinics, we will recruit approximately 10 providers across the NCH PCN, ensuring representation from at least 50% of the 14 PCC sites. Eligible providers include physicians, nurses, social workers, and care coordinators who are directly involved in patient care. Providers on extended leave or those not engaged in patient care will be excluded. Recruitment will be facilitated through the NCH Primary Care Research Network. DAPHNE Chatbot Development and structure DAPHNE was developed using secure cloud infrastructure (via AWS). At the backend, we integrated state-of-the-art large language models (e.g. Claude, GPT, Mistral, Gemini, LLaMA) adapted for healthcare settings and enabling intent recognition, contextual follow-up, and dynamic retrieval of resources from the resources API. Unlike rule-based decision trees, the AI allows free-text input from caregivers, interpreting varied language patterns (e.g., “I can’t get to the clinic” means transportation needs) and providing tailored responses. It demonstrated >99% intent accuracy over 1,500 turns of conversations.[ 31 ] DAPHNE is provided as a web-based conversational agent accessible on iOS, Android, or web browsers ( Figure 2 ). It is deployed using HIPAA-compliant servers and validated for accuracy and security by the internal Information Services governance committee. The chatbot will not require computational power on the user’s side, therefore a simple smartphone could be used to engage with the app. Our conversational model is also optimized for scaling at the cloud side with minimal cost and computational resource needs. Download figure Open in new tab Figure 2. Overview of how the DAPHNE chatbot connects users to community resources To ensure safety, DAPHNE automatically flags out-of-scope or potentially concerning interactions, and all conversations are subject to monitoring and auditing protocols. The research team reviews flagged transcripts to prevent harmful content and safeguard patient well-being. In addition, the research team periodically reviews the chatbot logs. For urgent or high-risk needs, an escalation protocol is used directing caregivers to professional resources, while the research team coordinates with the social work team as appropriate. Intervention: DAPHNE Chatbot Caregivers randomized to the intervention arm will receive onboarding consisting of instructions on how to download the app and a brief tutorial on the app’s features. The chatbot engages caregivers in natural conversations to understand HRSNs, provide tailored resource recommendations (e.g., food pantries, housing, transportation), and follow-up with automated reminders. Features include periodic personalized notifications (e.g., informational quick tips, updates), information about local services, and guidelines to sustain engagement. Control: Standard Care Caregivers in the control arm will receive the existing clinic-based HRSN screening and referral procedures as currently implemented, without access to the chatbot. These caregivers will complete the same study assessments as the intervention group except those related to chatbot usability and feasibility of the intervention. Procedures All caregivers will complete caregiver-reported outcome measures (cPROMs; see “secondary outcomes” section below) at baseline, 3 months, and 6 months post-intervention. Demographic information will be collected from all caregivers at baseline. Intervention group participants will have on-demand access to the chatbot throughout the 6-month study period, and will receive weekly notifications from the chatbot app. 10 caregivers from both groups will also complete a qualitative interview at 3 months and 6 months to evaluate their experience using the chatbot app versus receiving the standard of care. Provider Component After receiving a brief demonstration of the DAPHNE chatbot, providers will share their feedback and complete a structured survey on current HRSN screening and referral practices. Providers will complete the Workflow Integration Expectancy (WIE) tool and the Feasibility of Intervention Measure (FIM) at baseline and 6 months; providers will also complete the SUS at baseline. Additionally, providers will complete a survey on the feasibility of implementing the chatbot into their clinical workflows at baseline and 6 months. Semi-structured interviews will be conducted to elicit feedback on integration challenges, potential facilitators, and perceived impact on team workflow. Data from provider surveys will be analyzed descriptively, and interview transcripts will undergo thematic analysis to identify patterns in feasibility, barriers, and opportunities for scaling. Outcomes and Measures Primary outcomes (Quantitative) The primary outcomes will evaluate study feasibility (recruitment, retention, and survey completion across both arms) and intervention usability/acceptability. Feasibility will be assessed through enrollment rates (target ≥ 70%), retention rates (target ≥ 70%), and implementation feasibility using the Feasibility of Intervention Measure (FIM; target ≥ 80%).[ 44 , 45 ] Acceptability will be measured with the Website Evaluation Questionnaire (WEQ), with success defined as ≥ 80% endorsement of top rating categories.[ 46 ] Usability will be assessed using the System Usability Scale (SUS), with a mean score ≥ 68 considered acceptable.[ 47 ] Technical performance will also be evaluated, including chatbot response accuracy (target F1 ≥ 0.7 [ 48 ]), response latency (<3 seconds), and differences across platforms (iOS vs. Android).In addition, caregiver comprehension will be assessed using the Patient Comprehension Questionnaire (PCQ), with success defined as a mean score ≥4.The PCQ will be administered to caregivers in the intervention arm at baseline and 3 months. Primary outcomes (Qualitative) To further assess the DAPHNE chatbot, qualitative analysis will be completed for caregiver and provider interviews. This qualitative data will overlap with quantitative measures to gain a stronger understanding of provider and caregiver perspectives on feasibility, acceptability, and usability of the DAPHNE chatbot. For caregivers in the intervention group, qualitative interviews will be coded to evaluate factors that drive caregiver engagement with the DAPHNE chatbot and encourage use of the chatbot as well as barriers that may impact caregiver ability or want to use the chatbot. Interviews for caregivers in the control group will provide information about the current standard of care for caregivers with social needs and allow a comparison between resources or suggested resources caregivers receive in the current standard of care versus resources or suggested resources caregivers received while using the chatbot. Caregiver interviews across both groups will also be evaluated for caregiver perspectives toward study burden to determine the acceptability and feasibility of completing a larger scale trial of the DAPHNE chatbot. Clinician interviews will give first-hand insight into how well the DAPHNE chatbot could be integrated into clinic workflow by focusing on how useful the chatbot is perceived to be in a clinic setting, the challenges that may arise from using the chatbot in a clinic setting, and the ease with which the chatbot would be able to be integrated and possibly enhance the current standard of care. Secondary outcomes Secondary outcomes address caregiver- and provider-level effects. Caregiver-reported outcomes (cPROMs) include stress,[ 49 ] self-efficacy,[ 50 ] satisfaction with access to community resources,[ 51 ] and quality of life for caregiver and child as proxy.[ 52 ] These cPROMs will be completed by caregivers in both arms at baseline, 3, and 6 months. In addition to cPROMs, caregivers in both arms will complete the standardized HRSN screener. This survey will be administered at 3 and 6 months to capture changes in reported needs over time. Caregiver engagement with the chatbot is measured through system logs (e.g., minutes of use, number of logins, session durations, screenings completed, and resources accessed). Caregiver burden associated with study procedures will be measured in both arms using the Participant Burden Assessment (PeRBA; target ≥4) at 1, 3, and 6 months.[ 53 ] Provider perspectives on workflow integration will be assessed using the Workflow Integration Expectancy tool, with a target rating ≥ 4.[ 54 ] These secondary outcomes will be analyzed to see how they change over the course of the trial and to eventually address whether there are any associations with outcome changes and chatbot use. Exploratory outcomes Exploratory analyses will focus on healthcare utilization and economic outcomes to inform future large-scale trials. Electronic health record (EHR) data will be extracted to capture documentation of HRSN screening and referrals, adherence to well-child visits, missed appointments, emergency department utilization, and estimated healthcare costs. EHR data will be abstracted for all participants at 3- and 6-month follow-up period to capture pediatric primary care visits, social work referrals, and emergency department utilization. Billing and visit data will be used to estimate healthcare costs, including cost avoidance from reduced no-shows or ED visits. While the study is not powered to detect cost effects, descriptive comparisons between the DAPHNE and standard care arms will be conducted to identify patterns in service utilization and associated costs. Findings will inform the selection of economic outcomes and power calculations for a future R01 trial. Additional exploratory outcomes will include continuous tracking of recruitment rates and dropout rates throughout the study period. Outcome measures, descriptions and planned timeline are also available at our registered protocol.[ 55 ] Sample Size Justification The target enrollment of 100 caregivers (50 per arm) is appropriate for a pilot trial. The study is not powered for efficacy testing; rather, the focus is feasibility, acceptability, and usability outcomes to inform the design of a future fully powered R01 trial. The sample size provides a precision (half-width of 95% confidence interval) of 0.4 standard deviations for estimating mean differences between groups. Data Analysis Quantitative data (survey scores, engagement metrics, retention rates) will be summarized using descriptive statistics (e.g., mean, median, SD and IQR). Qualitative data will be transcribed and analyzed using thematic content analysis (hybrid inductive-deductive approach) in NVivo by two independent coders with ≥85% inter-rater reliability. Integration of quantitative and qualitative findings will follow a convergent mixed-methods design,[ 56 , 57 ] in which both data streams are analyzed separately and then merged during interpretation. Integration will occur through joint display and narrative weaving, identifying points of convergence, divergence, and complementarity to refine the intervention and inform future trials. Exploratory analyses will compare intervention and control arms on secondary outcomes (caregiver-reported stress, self-efficacy, satisfaction; EHR-derived service use) using t-tests or Wilcoxon rank-sum tests and descriptive comparisons. Effect sizes will be estimated to inform future power calculations. Randomization and Analysis Population Participants will be randomized 1:1 to DAPHNE or standard care using a pre-generated allocation sequence from the NCI Clinical Trial Randomization Tool (maximal procedure; maximum tolerated imbalance = 2). Allocation concealment will be maintained by an unblinded coordinator who releases the assignments via secure research tool (RedCAP) to be shared with the study team during recruitment. The primary analytic approach will follow the intention-to-treat (ITT) principle, whereby all randomized participants are analyzed in their originally assigned group regardless of adherence or withdrawal. This ensures comparability between arms and reflects real-world implementation. Missing Data For feasibility endpoints (e.g., retention, completion), denominators will include all randomized participants; nonresponders will be classified as not completing that timepoint. For continuous outcomes (e.g., SUS, WEQ, PCQ), available-case summaries will serve as the primary analysis, with prorating applied when allowed by instrument scoring manuals. If missingness exceeds ∼10–20% for a given outcome, multiple imputation using chained equations will be conducted as a sensitivity analysis, incorporating arm, site, baseline measures, and predictors of missingness. Engagement metrics for the intervention arm will be analyzed under ITT, with non-users coded as zero activity to avoid survivorship bias. Sensitivity analyses will compare complete-case, imputed, and worst-case assumptions to evaluate robustness. Data Monitoring and Risk Management All study data will be collected through HIPAA-compliant electronic systems and stored on encrypted Nationwide Children’s Hospital servers. Data integrity will be ensured through automated validation checks, routine audits, and adherence to standardized abstraction protocols. Technical support will be available via a dedicated helpdesk during working hours, with all issues logged for iterative improvement. Potential risks include participant frustration due to usability issues, privacy concerns, or inaccurate chatbot responses. To address these risks, the study incorporates caregiver training during onboarding, visual and video guides, and reassurance that DAPHNE complements rather than replaces clinical care. The chatbot’s performance will be continuously monitored for response accuracy and timeliness, with misinformation or errors flagged and corrected during weekly team reviews. Provider oversight will be integrated through periodic reporting of chatbot-identified needs and referrals to ensure appropriate follow-up. These strategies are designed to maintain caregiver trust, ensure accurate outputs, and safeguard participants throughout the study. An independent Data Safety and Monitoring Board, consisting of 4 experts in primary care practices, AI/ML research, and biostatistics will oversee the trial to ensure participant safety, data integrity, and overall trial conduct. The board, which is independent of the study sponsor and funders, will meet semi-annually to review cumulative safety data, enrollment metrics, and protocol adherence. They will provide recommendations directly to the Principal Investigator and the managing institution regarding the continuation, modification, or termination of the study. Discussion This protocol describes a pilot randomized controlled trial of DAPHNE, an AI-driven conversational agent designed to improve identification of health-related social needs and facilitate caregiver access to essential community resources. The study has potential to address the access barriers in pediatric primary care, where unmet HRSNs remain a major driver of adverse child health outcomes, yet routine screening and referral processes are underutilized.[ 15 – 17 ] By focusing on feasibility, acceptability, and usability, this study will generate critical data to guide refinement of both the intervention and the study protocol. Specifically, it will assess caregiver and provider engagement, evaluate technical performance across devices, and characterize barriers to study participation. Complementary analyses of provider surveys and EHR data will contextualize the current standard of care, providing essential information for selecting comparator conditions and outcomes in future clinical trials. The results will directly inform a larger, multi-site R01 trial evaluating the efficacy and implementation of AI-driven HRSN screening. Consistent with the ORBIT model, [ 32 , 33 ] this study represents an intermediate phase (IIb) in behavioral intervention development, building the foundation for robust testing of clinical and operational outcomes. Longer-term directions include expansion to non-English-speaking populations, integration of voice interactivity, and cost-effectiveness analyses to assess healthcare savings from reduced no-shows and emergency utilization. If successful, DAPHNE could advance personalized, equitable pediatric care by enabling real-time HRSN screening, reducing stigma in caregiver disclosures, and streamlining referral pathways. Beyond pediatrics, this approach has the potential to provide a generalizable framework for integrating AI-driven conversational tools into routine healthcare delivery and public health systems. Strengths and Limitations This study has several important strengths. First, DAPHNE introduces an innovative, AI-driven approach to health-related social need (HRSN) screening, using natural language, private, and context-aware interactions that move beyond static questionnaires.[ 23 – 25 , 29 , 30 ] Second, the intervention was co-designed with caregivers, clinicians, and advocates, ensuring that usability, engagement, and contextual relevance are embedded in its design.[ 23 ] Third, the study emphasizes integration into pediatric primary care workflows, which is essential to reduce burden on care teams while maintaining alignment with provider responsibilities.[ 58 , 59 ] Fourth, the trial is guided by the ORBIT model,[ 32 , 33 ] which provides a systematic framework for developing and iteratively evaluating the DAPHNE behavioral intervention. Finally, the modular design of backend resource integration allows integration with diverse community resource databases, making the intervention scalable and adaptable to local or health-domain specific populations and clinical contexts.[ 60 – 65 ] Nonetheless, there are key limitations. This is a single-site pilot study with a relatively small sample size (n=100) from specific region/ state, which restricts the generalizability of findings and precludes formal efficacy testing. Results will instead focus on feasibility, acceptability, and usability outcomes to inform future large-scale trials. The study is also limited to English-speaking, smartphone and data plan owner caregivers, which constrains equity in early testing. The 6-month follow-up period may not capture longer-term outcomes such as sustained engagement, resource utilization, or downstream effects on clinical care and health outcomes. Finally, technology-specific challenges, including variable device access, inconsistent internet connectivity, and risks inherent to conversational AI (such as inaccurate responses or hallucinations), may influence feasibility and participant trust. [ 66 – 69 ] Even though we hypothesize the benefits of chatbot, such as the automation of referrals, increased utilization of resources and the potential reduction of provider burden by efficiently identifying HRSNs, we invite readers and practitioners to be cautious. There are implicit factors including potential AI biases, AI related errors and inaccuracies and human overreliance on AI for resource support, which may potentially limit implementation of chatbot at a scale. Technology in our project focuses on supplementing human support rather than replacing, therefore not diminishing the value of personalized assistance and human interaction in practice.[ 70 ] Conclusions This pilot randomized controlled trial will provide essential feasibility, acceptability, and usability data on the DAPHNE chatbot for addressing health-related social needs in pediatric primary care. Findings will inform refinement of both the intervention and the study protocol, directly supporting the design of a future multi-site R01 trial. Guided by the ORBIT framework, the study emphasizes early-phase testing as a foundation for scalable, rigorous evaluation. In future research, the trial will expand to larger and more diverse populations, incorporate non-English language options and voice interactivity, and integrate cost-effectiveness analyses to assess the impact on healthcare utilization. Ultimately, this work will determine whether AI-driven conversational agents can serve as a sustainable and equitable model for integrating social needs identification and referral into routine pediatric care. Data Availability No datasets were generated or analysed during the current study. All relevant data from this study will be made available upon study completion. Funding This publication was supported, in part, by The Ohio State University Clinical and Translational Science Institute (CTSI) and the National Center for Advancing Translational Sciences of the National Institutes of Health under Grant Number UM1TR004548. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Ethics The study has Institutional Review Board (IRB) approval at Nationwide Children’s Hospital (#00004369). Informed consent will be obtained from all participants. Data Availability Statement Not available Competing Interests None declared Acknowledgements We gratefully acknowledge the support from the AWS Health Equity Initiative (via AWS Social Responsibility and Impact credits), as well as the FindHelp.org Fellowship to maintain community resource navigation. Our appreciation is also extended to the NCH IT-RI team for their expertise in platform development and to the NCH Office of Technology Commercialization for their guidance in scaling and dissemination. References 1. ↵ Hood CM , Gennuso KP , Swain GR , Catlin BB . County Health Rankings: Relationships between determinant factors and health outcomes . Am J Prev Med . 2016 ; 50 : 129 – 135 . doi: 10.1016/j.amepre.2015.08.024 OpenUrl CrossRef PubMed 2. ↵ Magnan S . Social determinants of health 201 for health care: Plan, do, study, act . NAM Perspect . 2021 ; 2021 . doi: 10.31478/202106c OpenUrl CrossRef 3. ↵ COUNCIL ON COMMUNITY PEDIATRICS. Poverty and child health in the United States. Pediatrics. 2016 ; 137 : e20160339 . doi: 10.1542/peds.2016-0339 OpenUrl CrossRef PubMed 4. Cree RA , Bitsko RH , Robinson LR , Holbrook JR , Danielson ML , Smith C , et al. Health care, family, and community factors associated with mental, behavioral, and developmental disorders and poverty among children aged 2-8 years - United States, 2016 . MMWR Morb Mortal Wkly Rep. 2018 ; 67 : 1377 – 1383 . doi: 10.15585/mmwr.mm6750a1 OpenUrl CrossRef PubMed 5. National Academies of Sciences, Engineering, and Medicine, Health and Medicine Division, Board on Health Care Services , Committee on Integrating Social Needs Care into the Delivery of Health Care to Improve the Nation’s Health . Integrating social care into the delivery of health care: Moving upstream to improve the nation’s health . Washington, D.C., DC : National Academies Press ; 2020 . doi: 10.17226/25467 OpenUrl CrossRef 6. ↵ Boch S , Keedy H , Chavez L , Dolce M , Chisolm D . An integrative review of social determinants of health screenings used in primary care settings . J Health Care Poor Underserved . 2020 ; 31 : 603 – 622 . doi: 10.1353/hpu.2020.0048 OpenUrl CrossRef PubMed 7. ↵ Rigdon J , Montez K , Palakshappa D , Brown C , Downs SM , Albertini LW , et al. Social risk factors influence pediatric emergency department utilization and hospitalizations . J Pediatr . 2022 ; 249 : 35 – 42 .e4. doi: 10.1016/j.jpeds.2022.06.004 OpenUrl CrossRef PubMed 8. ↵ Anyigbo C , Liu C , Ehrlich S , Reyner A , Ammerman RT , Kahn RS . Household health-related social needs in newborns and infant behavioral functioning at 6 months . JAMA Pediatr . 2024 ; 178 : 160 – 167 . doi: 10.1001/jamapediatrics.2023.5721 OpenUrl CrossRef PubMed 9. ↵ Chang L , Stewart AM , Monuteaux MC , Fleegler EW . Neighborhood conditions and recurrent emergency department utilization by children in the United States . J Pediatr . 2021 ; 234 : 115 – 122 .e1. doi: 10.1016/j.jpeds.2020.12.071 OpenUrl CrossRef PubMed 10. ↵ Nerlinger AL , Kopsombut G . Social determinants of health screening in pediatric healthcare settings . Curr Opin Pediatr . 2023 ; 35 : 14 – 21 . doi: 10.1097/MOP.0000000000001191 OpenUrl CrossRef PubMed 11. ↵ Hsu C , Cruz S , Placzek H , Chapdelaine M , Levin S , Gutierrez F , et al. Patient Perspectives on Addressing Social Needs in Primary Care Using a Screening and Resource Referral Intervention . J Gen Intern Med . 2020 ; 35 : 481 – 489 . doi: 10.1007/s11606-019-05397-6 OpenUrl CrossRef PubMed 12. ↵ Savin K , Morales A , Levi R , Alvarez D , Seligman H . “now I feel a little bit more secure”: The impact of SNAP enrollment on older adult SSI recipients . Nutrients . 2021 ; 13 : 4362 . doi: 10.3390/nu13124362 OpenUrl CrossRef PubMed 13. Algert SJ , Reibel M , Renvall MJ . Barriers to participation in the food stamp program among food pantry clients in Los Angeles . Am J Public Health . 2006 ; 96 : 807 – 809 . doi: 10.2105/AJPH.2005.066977 OpenUrl CrossRef PubMed Web of Science 14. ↵ Kreuter M , Garg R , Thompson T , McQueen A , Javed I , Golla B , et al. Assessing the capacity of local social services agencies to respond to referrals from health care providers . Health Aff (Millwood ). 2020 ; 39 : 679 – 688 . doi: 10.1377/hlthaff.2019.01256 OpenUrl CrossRef PubMed 15. ↵ Fraze TK , Brewster AL , Lewis VA , Beidler LB , Murray GF , Colla CH . Prevalence of screening for food insecurity, housing instability, utility needs, transportation needs, and interpersonal violence by US physician practices and hospitals . JAMA Netw Open . 2019 ; 2 : e1911514 . doi: 10.1001/jamanetworkopen.2019.11514 OpenUrl CrossRef PubMed 16. ↵ Dolce M , Keedy H , Chavez L , Boch S , Zaim H , Dias B , et al. Implementing an EMR-based health-related social needs screen in a pediatric hospital system . Pediatr Qual Saf . 2022 ; 7 : e512 . doi: 10.1097/pq9.0000000000000512 OpenUrl CrossRef 17. ↵ Victorino CC , Gauthier AH . The social determinants of child health: variations across health outcomes - a population-based cross-sectional analysis . BMC Pediatr . 2009 ; 9 : 53 . doi: 10.1186/1471-2431-9-53 OpenUrl CrossRef PubMed 18. ↵ Bitsko RH , Visser SN , Schieve LA , Ross DS , Thurman DJ , Perou R . Unmet health care needs among CSHCN with neurologic conditions . Pediatrics . 2009 ; 124 Suppl 4 : S343 – 51 . doi: 10.1542/peds.2009-1255D OpenUrl CrossRef PubMed 19. ↵ Pantell MS , Kaiser SV , Torres JM , Gottlieb LM , Adler NE . Associations between social factor documentation and hospital length of stay and readmission among children . Hosp Pediatr . 2020 ; 10 : 12 – 19 . doi: 10.1542/hpeds.2019-0123 OpenUrl CrossRef PubMed 20. ↵ Ryan JL , Franklin SM , Canterberry M , Long CL , Bowe A , Roy BD , et al. Association of health-related social needs with quality and utilization outcomes in a Medicare Advantage population with diabetes . JAMA Netw Open . 2023 ; 6 : e239316 . doi: 10.1001/jamanetworkopen.2023.9316 OpenUrl CrossRef PubMed 21. ↵ Cook JT , Frank DA , Meyers AF , Berkowitz C , Black MM , Casey PH , et al. Food insecurity is associated with adverse health outcomes among human infants and toddlers . J Nutr . 2004 ; 134 : 1432 – 1438 . doi: 10.1093/jn/134.6.1432 OpenUrl Abstract / FREE Full Text 22. ↵ Gray TW , Podewils LJ , Rasulo RM , Weiss RP , Tomcho MM . Examining the implementation of health-related social need (HRSN) screenings at a pediatric community health center . J Prim Care Community Health . 2023 ; 14 : 21501319231171519 . doi: 10.1177/21501319231171519 OpenUrl CrossRef 23. ↵ Sezgin E , Kocaballi AB , Dolce M , Skeens M , Militello L , Huang Y , et al. Chatbot for social need screening and resource sharing with vulnerable families: Iterative design and evaluation study . JMIR Hum Factors . 2024 ; 11 : e57114 . doi: 10.2196/57114 OpenUrl CrossRef 24. Abbasian M , Khatibi E , Azimi I , Oniani D , Shakeri Hossein Abad Z , Thieme A , et al. Foundation metrics for evaluating effectiveness of healthcare conversations powered by generative AI . NPJ Digit Med. 2024 ; 7 : 82 . doi: 10.1038/s41746-024-01074-z OpenUrl CrossRef 25. ↵ Xue J , Zhang B , Zhao Y , Zhang Q , Zheng C , Jiang J , et al. Evaluation of the current state of chatbots for digital health: Scoping review . J Med Internet Res . 2023 ; 25 : e47217 . doi: 10.2196/47217 OpenUrl CrossRef PubMed 26. ↵ Johnson CB , Luther B , Wallace AS , Kulesa MG . Social determinants of health: What are they and how do we screen . Orthop Nurs . 2022 ; 41 : 88 – 100 . doi: 10.1097/NOR.0000000000000829 OpenUrl CrossRef PubMed 27. Wallace AS , Luther BL , Sisler SM , Wong B , Guo J-W . Integrating social determinants of health screening and referral during routine emergency department care: evaluation of reach and implementation challenges . Implement Sci Commun . 2021 ; 2 : 114 . doi: 10.1186/s43058-021-00212-y OpenUrl CrossRef PubMed 28. ↵ Tian W , Lu Y , Yu J , Fan J , Tang P , Zhang L. A privacy-preserving framework for mental health chatbots based on confidential computing. 2022 IEEE Smartworld, Ubiquitous Intelligence & Computing, Scalable Computing & Communications, Digital Twin, Privacy Computing, Metaverse, Autonomous & Trusted Vehicles (SmartWorld/UIC/ScalCom/DigitalTwin/PriComp/Meta) . IEEE ; 2022 . pp. 1119 – 1124 . doi: 10.1109/smartworld-uic-atc-scalcom-digitaltwin-pricomp-metaverse56740.2022.00160 OpenUrl CrossRef 29. ↵ Grassini E , Buzzi M , Leporini B , Vozna A . A systematic review of chatbots in inclusive healthcare: insights from the last 5 years . Univers Access Inf Soc . 2024 . doi: 10.1007/s10209-024-01118-x OpenUrl CrossRef 30. ↵ Ilne-Ives M . The effectiveness of artificial intelligence conversational agents in health care: Systematic review . J Med Internet Res . 2020 ; 22 . 31. ↵ Sezgin E , Jackson DI , Hussain S-A , Kocaballi AB , Richardson C , Skeens M , et al. Evaluating User Experiences with an AI Chatbot for Health-Related Social Needs: A Cross-Sectional Mixed Methods Study . Health Informatics. medRxiv; 2025 . Available: https://www.medrxiv.org/content/10.1101/2025.06.16.25329054v1.full.pdf 32. ↵ Czajkowski SM , Powell LH , Adler N , Naar-King S , Reynolds KD , Hunter CM , et al. From ideas to efficacy: The ORBIT model for developing behavioral treatments for chronic diseases . Health Psychol . 2015 ; 34 : 971 – 982 . doi: 10.1037/hea0000161 OpenUrl CrossRef PubMed 33. ↵ Powell LH , Freedland KE , Kaufmann PG . Behavioral clinical trials for chronic diseases . 1st ed. Cham, Switzerland : Springer Nature ; 2021. doi: 10.1007/978-3-030-39330-4 OpenUrl CrossRef 34. ↵ Boucher EM , Harake NR , Ward HE , Stoeckl SE , Vargas J , Minkel J , et al. Artificially intelligent chatbots in digital mental health interventions: a review . Expert Rev Med Devices . 2021 ; 18 : 37 – 49 . doi: 10.1080/17434440.2021.2013200 OpenUrl CrossRef PubMed 35. Fitzpatrick KK , Darcy A , Vierhile M . Delivering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (Woebot): A randomized controlled trial . JMIR Ment Health . 2017 ; 4 : e19 . doi: 10.2196/mental.7785 OpenUrl CrossRef PubMed 36. Winn AN , Somai M , Fergestrom N , Crotty BH . Association of use of online symptom checkers with patients’ plans for seeking care . JAMA Netw Open . 2019 ; 2 : e1918561 . doi: 10.1001/jamanetworkopen.2019.18561 OpenUrl CrossRef 37. Abd-Alrazaq AA , Rababeh A , Alajlani M , Bewick BM , Househ M . Effectiveness and safety of using chatbots to improve mental health: Systematic review and meta-analysis . J Med Internet Res . 2020 ; 22 : e16021 . doi: 10.2196/16021 OpenUrl CrossRef PubMed 38. You Y , Gui X. Self-diagnosis through AI-enabled chatbot-based symptom checkers: User experiences and design considerations . arXiv [cs.HC]. 2021. Available: http://arxiv.org/abs/2101.04796 39. ↵ Sezgin E , Oiler B , Abbott B , Noritz G , Huang Y . “hey Siri, help me take care of my child”: A feasibility study with caregivers of children with special healthcare needs using voice interaction and automatic speech recognition in remote care management . Front Public Health . 2022 ; 10 : 849322 . doi: 10.3389/fpubh.2022.849322 OpenUrl CrossRef 40. ↵ Sezgin E , Militello LK , Huang Y , Lin S . A scoping review of patient-facing, behavioral health interventions with voice assistant technology targeting self-management and healthy lifestyle behaviors . Transl Behav Med . 2020 ; 10 : 606 – 628 . doi: 10.1093/tbm/ibz141 OpenUrl CrossRef PubMed 41. ↵ Sezgin E , Hussain S-A , Rust S , Huang Y . Extracting medical information from free-text and unstructured patient-generated health data using natural language processing methods: Feasibility study with real-world data . JMIR Form Res . 2023 ; 7 : e43014 . doi: 10.2196/43014 OpenUrl CrossRef 42. ↵ Hussain S-A , Sezgin E , Krivchenia K , Luna J , Rust S , Huang Y . A natural language processing pipeline to synthesize patient-generated notes toward improving remote care and chronic disease management: a cystic fibrosis case study . JAMIA Open . 2021 ; 4 : ooab084. doi: 10.1093/jamiaopen/ooab084 OpenUrl CrossRef 43. ↵ Chan A-W , Boutron I , Hopewell S , Moher D , Schulz KF , Collins GS , et al. SPIRIT 2025 statement: Updated guideline for protocols of randomised trials . PLoS Med . 2025 ; 22 : e1004589 . doi: 10.1371/journal.pmed.1004589 OpenUrl CrossRef PubMed 44. ↵ Anastasi JK , Capili B , Norton M , McMahon DJ , Marder K . Recruitment and retention of clinical trial participants: understanding motivations of patients with chronic pain and other populations . Front Pain Res (Lausanne ). 2023 ; 4 : 1330937 . doi: 10.3389/fpain.2023.1330937 OpenUrl CrossRef 45. ↵ Weiner BJ , Lewis CC , Stanick C , Powell BJ , Dorsey CN , Clary AS , et al. Psychometric assessment of three newly developed implementation outcome measures . Implement Sci . 2017 ; 12 : 108 . doi: 10.1186/s13012-017-0635-3 OpenUrl CrossRef PubMed 46. ↵ Elling S , Lentz L , de Jong M. Website evaluation questionnaire: Development of a research-based tool for evaluating informational websites. Lecture Notes in Computer Science. Berlin , Heidelberg: Springer Berlin Heidelberg ; 2007 . pp. 293 – 304 . doi: 10.1007/978-3-540-74444-3_25 OpenUrl CrossRef 47. ↵ Lewis JR , Sauro J . The factor structure of the system usability scale. Human Centered Design. Berlin , Heidelberg: Springer Berlin Heidelberg ; 2009 . pp. 94 – 103 . doi: 10.1007/978-3-642-02806-9_12 OpenUrl CrossRef 48. ↵ Yuwono SK , Wu B , D’Haro LF . Automated scoring of chatbot responses in conversational dialogue . Lecture Notes in Electrical Engineering . Singapore : Springer Singapore ; 2019 . pp. 357 – 369 . doi: 10.1007/978-981-13-9443-0_31 OpenUrl CrossRef 49. ↵ Barroso NE , Hungerford GM , Garcia D , Graziano PA , Bagner DM . Psychometric properties of the Parenting Stress Index-Short Form (PSI-SF) in a high-risk sample of mothers and their infants . Psychol Assess . 2016 ; 28 : 1331 – 1335 . doi: 10.1037/pas0000257 OpenUrl CrossRef PubMed 50. ↵ Bandura A . Self efficacy . New York, NY : W.H. Freeman ; 1997 . 51. ↵ Owen R . Net promoter score and its successful application . Management for Professionals. Singapore: Springer Singapore ; 2019 . pp. 17 – 29 . doi: 10.1007/978-981-10-7724-1_2 OpenUrl CrossRef 52. ↵ Flanagan JC . Measurement of quality of life: current state of the art . Arch Phys Med Rehabil . 1982 ; 63 : 56 – 59 . Available: https://www.ncbi.nlm.nih.gov/pubmed/6460487 OpenUrl PubMed Web of Science 53. ↵ Lingler JH , Schmidt KL , Gentry AL , Hu L , Terhorst LA . A new measure of research participant burden: Brief report . J Empir Res Hum Res Ethics . 2014 ; 9 : 46 – 49 . doi: 10.1177/1556264614545037 OpenUrl CrossRef PubMed 54. ↵ Flanagan M , Arbuckle N , Saleem JJ , Militello LG , Haggstrom DA , Doebbeling BN . Development of a workflow integration survey (WIS) for implementing computerized clinical decision support . AMIA Annu Symp Proc . 2011 ; 2011 : 427 – 434 . Available: https://www.ncbi.nlm.nih.gov/pubmed/22195096 OpenUrl PubMed 55. ↵ AI-Based Personalized Health and Self-Care . In: clinicaltrials.gov [Internet]. [cited 14 Nov 2025]. Available: https://clinicaltrials.gov/study/NCT07168382 56. ↵ Fetters MD , Curry LA , Creswell JW . Achieving integration in mixed methods designs-principles and practices . Health Serv Res . 2013 ; 48 : 2134 – 2156 . doi: 10.1111/1475-6773.12117 OpenUrl CrossRef PubMed Web of Science 57. ↵ Creswell JW , David Creswell J. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches . SAGE Publications ; 2017 . Available: https://play.google.com/store/books/details?id=KGNADwAAQBAJ 58. ↵ Xu L , Sanders L , Li K , Chow JCL . Chatbot for Health Care and Oncology Applications Using Artificial Intelligence and Machine Learning: Systematic Review . JMIR Cancer . 2021 ; 7 : e27850 . doi: 10.2196/27850 OpenUrl CrossRef 59. ↵ Sezgin E , Jackson DI , Kaufman K , Skeens MA , Gerhardt CA , Moscato E . Perceptions about the use of virtual assistants for seeking health information among caregivers of young childhood cancer survivors . Digit Health . 2025 ; 11 : 20552076251326160 . doi: 10.1177/20552076251326160 OpenUrl CrossRef 60. ↵ Pilato G , Augello A , Gaglio S. A Modular Architecture for Adaptive ChatBots . 2011 IEEE Fifth International Conference on Semantic Computing . IEEE ; 2011 . doi: 10.1109/icsc.2011.68 OpenUrl CrossRef 61. Nadarzynski T , Knights N , Husbands D , Graham CA , Llewellyn CD , Buchanan T , et al. Achieving health equity through conversational AI: A roadmap for design and implementation of inclusive chatbots in healthcare . PLOS Digit Health . 2024 ; 3 : e0000492 . doi: 10.1371/journal.pdig.0000492 OpenUrl CrossRef 62. Thirunavukarasu AJ . Large language models will not replace healthcare professionals: curbing popular fears and hype . J R Soc Med . 2023 ; 116 : 181 – 182 . doi: 10.1177/01410768231173123 OpenUrl CrossRef PubMed 63. Solanki RK , Rajawat AS , Gadekar AR , Patil ME . Building a conversational chatbot using machine learning. Advances in Medical Education , Research, and Ethics. IGI Global ; 2023 . pp. 285 – 309 . doi: 10.4018/978-1-6684-7164-7.ch013 OpenUrl CrossRef 64. Singh B , Olds T , Brinsley J , Dumuid D , Virgara R , Matricciani L , et al. Systematic review and meta-analysis of the effectiveness of chatbots on lifestyle behaviours . NPJ Digit Med . 2023 ; 6 : 118 . doi: 10.1038/s41746-023-00856-1 OpenUrl CrossRef 65. ↵ Denecke K , Abd-Alrazaq A , Househ M , Warren J . Evaluation metrics for health chatbots: A Delphi study . Methods Inf Med . 2021 ; 60 : 171 – 179 . doi: 10.1055/s-0041-1736664 OpenUrl CrossRef 66. ↵ Colasacco CJ , Born HL . A case of artificial intelligence chatbot hallucination . JAMA Otolaryngol Head Neck Surg . 2024 ; 150 : 457 – 458 . doi: 10.1001/jamaoto.2024.0428 OpenUrl CrossRef PubMed 67. Woo B , Huynh T , Tang A , Bui N , Nguyen G , Tam W . Transforming nursing with large language models: from concept to practice . Eur J Cardiovasc Nurs . 2024 ; 23 : 549 – 552 . doi: 10.1093/eurjcn/zvad120 OpenUrl CrossRef PubMed 68. Oca MC , Meller L , Wilson K , Parikh AO , McCoy A , Chang J , et al. Bias and inaccuracy in AI chatbot ophthalmologist recommendations . Cureus . 2023 ; 15 : e45911 . doi: 10.7759/cureus.45911 OpenUrl CrossRef 69. ↵ Heo S , Son S , Park H . HaluCheck: Explainable and verifiable automation for detecting hallucinations in LLM responses . Expert Syst Appl . 2025 ; 272 : 126712 . doi: 10.1016/j.eswa.2025.126712 OpenUrl CrossRef 70. ↵ Sezgin E . Artificial intelligence in healthcare: Complementing, not replacing, doctors and healthcare providers . Digit Health . 2023 ; 9 : 20552076231186520 . doi: 10.1177/20552076231186520 OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted November 19, 2025. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial Emre Sezgin , Elizabeth Clarkson , Faith Logan , Daniel I. Jackson , Syed-Amad Hussain , Joshua Stokes , Alicia Bunger , Guy Brock , Eric Fosler-Lussier , Alex R. Kemper , Ahna L. Pai medRxiv 2025.11.17.25340433; doi: https://doi.org/10.1101/2025.11.17.25340433 Share This Article: Copy Citation Tools An AI-Based Chatbot to Support Health-Related Social Needs among Pediatric Primary Care Population: Protocol for a Pilot Randomized Controlled Trial Emre Sezgin , Elizabeth Clarkson , Faith Logan , Daniel I. Jackson , Syed-Amad Hussain , Joshua Stokes , Alicia Bunger , Guy Brock , Eric Fosler-Lussier , Alex R. Kemper , Ahna L. Pai medRxiv 2025.11.17.25340433; doi: https://doi.org/10.1101/2025.11.17.25340433 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Health Informatics Subject Areas All Articles Addiction Medicine (568) Allergy and Immunology (863) Anesthesia (300) Cardiovascular Medicine (4434) Dentistry and Oral Medicine (444) Dermatology (382) Emergency Medicine (608) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1508) Epidemiology (15227) Forensic Medicine (30) Gastroenterology (1124) Genetic and Genomic Medicine (6595) Geriatric Medicine (668) Health Economics (997) Health Informatics (4534) Health Policy (1368) Health Systems and Quality Improvement (1613) Hematology (540) HIV/AIDS (1264) Infectious Diseases (except HIV/AIDS) (15915) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (145) Nephrology (667) Neurology (6597) Nursing (346) Nutrition (998) Obstetrics and Gynecology (1144) Occupational and Environmental Health (957) Oncology (3332) Ophthalmology (973) Orthopedics (369) Otolaryngology (420) Pain Medicine (436) Palliative Medicine (130) Pathology (663) Pediatrics (1692) Pharmacology and Therapeutics (691) Primary Care Research (711) Psychiatry and Clinical Psychology (5447) Public and Global Health (9230) Radiology and Imaging (2198) Rehabilitation Medicine and Physical Therapy (1370) Respiratory Medicine (1196) Rheumatology (593) Sexual and Reproductive Health (712) Sports Medicine (530) Surgery (712) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a000275cdb3a52ad',t:'MTc3OTQ5ODM0MQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();