{"id":{"repo_id":"bu","oai_identifier":"oai:open.bu.edu:2144/53629"},"canonical_url":"https://search.ndltd.org/etd/bu/oai:open.bu.edu:2144/53629","repository":{"repo_id":"bu","name":"Boston University","base_url":"https://open.bu.edu/oai/request"},"display":{"title":"Intelligent capacity: AI-driven workflow redesign in healthcare service systems","abstract":"Physician shortages and long waiting times represent a growing crisis in specialist healthcare delivery. The dissertation investigates how AI-assisted telemedicine can be integrated into congested specialty care systems to reduce access delays and geographic disparities, without adding full-time capacity expansion. Chapter 1 analyzes close to 55,000 dermatology appointment records from a large international HMO and document severe access delays and extreme high utilizations across all clinics. Due to excess service pattern variability, we also show that M/M/1 queueing models substantially underestimate actual wait times. Chapter 2 proposes a Dynamic Flow Diversion (DFD) mechanism that activates on-demand AI-assisted auxiliary capacity only when the in-person clinic congestion exceeds a threshold, and develop a tractable two-dimension queueing model to characterize the optimal threshold policy. Closed-form performance measures show that diverting very small percentage of all patients to the auxiliary channel reduces average waiting times dramatically. We extend the model to multi-clinic networks with pooled auxiliary capacity and derive staffing rules using a heavy-traffic approximation. Chapter 3 empirically investigates the patient refusal probability, which is treated as exogenous in the DFD model, using a nationally representative survey of close to 8,000 U.S. adults. We estimate three nested regression models covering willingness to use telehealth, realized utilization, and activation among the willing. A substantial willingness–utilization gap emerges, with provider encouragement the dominant activator. Framing the DFD offer as a provider-endorsed recommendation rather than an automated notification therefore lowers the refusal probability and, through the nonlinearity of queueing at near-unit utilization, amplifies the downstream cost savings. Together, the three chapters provide the theoretical models and empirical analyses for HMOs to integrate AI-assisted care as a system-level solution to specialist access delays.","abstract_html":"Physician shortages and long waiting times represent a growing crisis in specialist healthcare delivery. The dissertation investigates how AI-assisted telemedicine can be integrated into congested specialty care systems to reduce access delays and geographic disparities, without adding full-time capacity expansion. Chapter 1 analyzes close to 55,000 dermatology appointment records from a large international HMO and document severe access delays and extreme high utilizations across all clinics. Due to excess service pattern variability, we also show that M/M/1 queueing models substantially underestimate actual wait times. Chapter 2 proposes a Dynamic Flow Diversion (DFD) mechanism that activates on-demand AI-assisted auxiliary capacity only when the in-person clinic congestion exceeds a threshold, and develop a tractable two-dimension queueing model to characterize the optimal threshold policy. Closed-form performance measures show that diverting very small percentage of all patients to the auxiliary channel reduces average waiting times dramatically. We extend the model to multi-clinic networks with pooled auxiliary capacity and derive staffing rules using a heavy-traffic approximation. Chapter 3 empirically investigates the patient refusal probability, which is treated as exogenous in the DFD model, using a nationally representative survey of close to 8,000 U.S. adults. We estimate three nested regression models covering willingness to use telehealth, realized utilization, and activation among the willing. A substantial willingness–utilization gap emerges, with provider encouragement the dominant activator. Framing the DFD offer as a provider-endorsed recommendation rather than an automated notification therefore lowers the refusal probability and, through the nonlinearity of queueing at near-unit utilization, amplifies the downstream cost savings. Together, the three chapters provide the theoretical models and empirical analyses for HMOs to integrate AI-assisted care as a system-level solution to specialist access delays.","abstract_has_math":false,"creators":["Chen, Fanying"],"institution":"Boston University","degree_name":"Doctor of Philosophy","degree_level":"doctoral","degree_discipline":"Management","degree_department":null,"school":null,"contributors":[],"advisors":["Seidmann, Abraham"],"committee_chairs":[],"committee_members":[],"year":2027,"date_issued":"2027","date_published":"2027","updated_at":"2026-11-19T02:42:12Z","subjects":["Operations research","Health care management","Information science"],"languages":["en_US"],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2144/53629","outbound_label":"Handle","outbound_source":"dc:identifier.uri","record_url":"https://hdl.handle.net/2144/53629","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier.uri","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://open.bu.edu/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Aopen.bu.edu%3A2144%2F53629","prefix":"dim"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Seidmann, Abraham"]},{"key":"dc:creator","label":"Author","values":["Chen, Fanying"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2026-10-08T17:13:11Z"]},{"key":"dc:date.embargo","label":"Dc Date Embargo","values":["2028-10-08T00:00:00Z"]},{"key":"dc:date.issued","label":"Date","values":["2027"]},{"key":"dc:date.updated","label":"Dc Date Updated","values":["2026-10-08T16:15:29Z"]},{"key":"dc:type","label":"Dc Type","values":["Thesis/Dissertation"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Management"]},{"key":"thesis:degree_level","label":"Degree Level","values":["doctoral"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["Boston University"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Operations research","Health care management","Information science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en_US"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/2144/53629"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["2027"]},{"key":"dc:description.abstract","label":"Abstract","values":["Physician shortages and long waiting times represent a growing crisis in specialist healthcare delivery. The dissertation investigates how AI-assisted telemedicine can be integrated into congested specialty care systems to reduce access delays and geographic disparities, without adding full-time capacity expansion. Chapter 1 analyzes close to 55,000 dermatology appointment records from a large international HMO and document severe access delays and extreme high utilizations across all clinics. Due to excess service pattern variability, we also show that M/M/1 queueing models substantially underestimate actual wait times. Chapter 2 proposes a Dynamic Flow Diversion (DFD) mechanism that activates on-demand AI-assisted auxiliary capacity only when the in-person clinic congestion exceeds a threshold, and develop a tractable two-dimension queueing model to characterize the optimal threshold policy. Closed-form performance measures show that diverting very small percentage of all patients to the auxiliary channel reduces average waiting times dramatically. We extend the model to multi-clinic networks with pooled auxiliary capacity and derive staffing rules using a heavy-traffic approximation. Chapter 3 empirically investigates the patient refusal probability, which is treated as exogenous in the DFD model, using a nationally representative survey of close to 8,000 U.S. adults. We estimate three nested regression models covering willingness to use telehealth, realized utilization, and activation among the willing. A substantial willingness–utilization gap emerges, with provider encouragement the dominant activator. Framing the DFD offer as a provider-endorsed recommendation rather than an automated notification therefore lowers the refusal probability and, through the nonlinearity of queueing at near-unit utilization, amplifies the downstream cost savings. Together, the three chapters provide the theoretical models and empirical analyses for HMOs to integrate AI-assisted care as a system-level solution to specialist access delays."]},{"key":"dc:title","label":"Title","values":["Intelligent capacity: AI-driven workflow redesign in healthcare service systems"]}]}],"canonical_facts":{"dc:contributor.advisor":["Seidmann, Abraham"],"dc:creator":["Chen, Fanying"],"dc:date.accessioned":["2026-10-08T17:13:11Z"],"dc:date.embargo":["2028-10-08T00:00:00Z"],"dc:date.issued":["2027"],"dc:date.updated":["2026-10-08T16:15:29Z"],"dc:description":["2027"],"dc:description.abstract":["Physician shortages and long waiting times represent a growing crisis in specialist healthcare delivery. The dissertation investigates how AI-assisted telemedicine can be integrated into congested specialty care systems to reduce access delays and geographic disparities, without adding full-time capacity expansion. Chapter 1 analyzes close to 55,000 dermatology appointment records from a large international HMO and document severe access delays and extreme high utilizations across all clinics. Due to excess service pattern variability, we also show that M/M/1 queueing models substantially underestimate actual wait times. Chapter 2 proposes a Dynamic Flow Diversion (DFD) mechanism that activates on-demand AI-assisted auxiliary capacity only when the in-person clinic congestion exceeds a threshold, and develop a tractable two-dimension queueing model to characterize the optimal threshold policy. Closed-form performance measures show that diverting very small percentage of all patients to the auxiliary channel reduces average waiting times dramatically. We extend the model to multi-clinic networks with pooled auxiliary capacity and derive staffing rules using a heavy-traffic approximation. Chapter 3 empirically investigates the patient refusal probability, which is treated as exogenous in the DFD model, using a nationally representative survey of close to 8,000 U.S. adults. We estimate three nested regression models covering willingness to use telehealth, realized utilization, and activation among the willing. A substantial willingness–utilization gap emerges, with provider encouragement the dominant activator. Framing the DFD offer as a provider-endorsed recommendation rather than an automated notification therefore lowers the refusal probability and, through the nonlinearity of queueing at near-unit utilization, amplifies the downstream cost savings. Together, the three chapters provide the theoretical models and empirical analyses for HMOs to integrate AI-assisted care as a system-level solution to specialist access delays."],"dc:identifier.uri":["https://hdl.handle.net/2144/53629"],"dc:language.iso":["en_US"],"dc:subject":["Operations research","Health care management","Information science"],"dc:title":["Intelligent capacity: AI-driven workflow redesign in healthcare service systems"],"dc:type":["Thesis/Dissertation"],"thesis:degree_discipline":["Management"],"thesis:degree_level":["doctoral"],"thesis:degree_name":["Doctor of Philosophy"],"thesis:institution_name":["Boston University"]},"updated_at":"2026-11-19T02:42:12Z"}