About

Fernanda Bravo is an Assistant Professor of Decisions, Operations and Technology Management at UCLA Anderson. She designs analytical models and data-driven frameworks that aim to improve strategic and operational decision-making within health care organizations. Her research includes studying how to better share and mitigate risk in service-based B2B supply chains through innovative pricing contract designs, as well as developing optimization-driven approaches to understand costs, guide resource allocation, and support network capacity building decisions at the system level.

Research topics

• Actuarial science
• Artificial Intelligence
• Computer Science
• Medicine
• Machine Learning
• Business
• Operations research
• Risk analysis (engineering)
• Epistemology
• Management science
• Data science
• Mathematics
• Engineering
• Nursing
• Economics
• Microeconomics
• Environmental health
• Finance

Selected publications

• Closer to Home: A Structural Estimate-Then-Optimize Approach to Improve Access to Healthcare Services

  Management Science · 2025-07-23 · 2 citations

  articleOpen access1st authorCorresponding

  Geographic inequalities in healthcare access extend beyond rural–urban divides to include socioeconomic, racial, and other disparities. Proximity to hospitals, clinics, healthcare providers, and pharmacies varies widely, posing a challenge in deciding where to strategically locate such facilities. Demand for each service depends on local population health, individual preferences, provider capacity, and other factors. This study introduces a novel structural estimate-then-optimize (SETO) framework, combining structural demand estimation using a modified Berry–Levinsohn–Pakes approach that accounts for provider capacity with a choice-based optimal facility location model to maximize health service utilization. Our methodology is illustrated with a case study on the Federal Retail Pharmacy Program in California, a public–private partnership that administered millions of COVID-19 vaccinations. Demand estimates indicate that residents of socioeconomically vulnerable communities are more sensitive to travel distances to pharmacy-based vaccination sites. Strategically adding 500 retail stores serving lower-income communities increases predicted vaccinations by 2.9% overall (770,000 additional vaccinations statewide) and by 5.3% in the least healthy neighborhoods. Our integrative SETO approach outperforms heuristics that allocate resources based on current vaccination rates, existing service gaps, population density, or predicted demand. The case study demonstrates the importance of (1) accounting for heterogeneity in estimating demand and (2) selecting partnerships to complement existing networks with spatially heterogeneous supply and efficiently fill service gaps. Our study provides a systematic approach to optimize healthcare delivery networks, using publicly available aggregate data while accounting for individuals’ preferences, highlighting the value of combining a structural demand model with prescriptive analytics. This paper was accepted by Stefan Sholtes, healthcare management. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.06274 .

  PublisherOA PDFDOI

• Enhancing Safety Signaling: Integrating Clinical Trials and Post-Marketing Adverse Event Reports

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access1st authorCorresponding
  PublisherDOI

• Incentivizing Drug Rediscovery: Exclusivity Extension and Priority Review for Repurposed Therapeutics

  SSRN Electronic Journal · 2025-01-01

  articleOpen accessSenior author
  PublisherDOI

• Algorithm Adoption and Explanations: An Experimental Study on Self and Other Perspectives

  SSRN Electronic Journal · 2024-01-01

  preprintOpen access1st authorCorresponding
  PublisherDOI

• Supervivencia en pacientes con cáncer de mama estadio IV inicial con manejo sistémico y quirúrgico

  Oncología (Ecuador) · 2024-04-30

  articleOpen access1st authorCorresponding

  Introducción: El manejo sistémico es el pilar del tratamiento en las pacientes con cáncer de mama metastásico al debut. Sin embargo, la terapia conjunta (sistémica con cirugía local/locorregional) es objeto de investigación para determinar si ofrece un beneficio adicional en los resultados oncológicos. Los ensayos clínicos aleatorizados tienen reportes contradictorios en cuanto a supervivencia global, mientras que los estudios retrospectivos muestran un impacto favorable. Esta investigación tuvo como objetivo describir la supervivencia global y la supervivencia libre de progresión de pacientes con carcinoma de mama metastásico al debut, tratadas con terapia sistémica exclusiva o terapia conjunta. Materiales y método: Estudio retrospectivo de una cohorte de pacientes con carcinoma de mama metastásico al debut de una clínica de referencia oncológica. Se evaluaron dos grupos de manejo: con terapia sistémica exclusiva vs. terapia conjunta. Los resultados principales evaluados fueron la supervivencia libre de progresión y la supervivencia global, calculados mediante las funciones de supervivencia de Kaplan-Meier y ajustados a las variables confusoras con modelos de Cox. Resultados: Recibieron terapia sistémica exclusiva 174 pacientes y 88 pacientes, terapia conjunta. La mediana de seguimiento fue de 58,38 meses; la supervivencia libre de progresión fue de 38,56 meses en el grupo de terapia sistémica exclusiva vs. 72,25 meses para el grupo de terapia conjunta. La supervivencia global fue de 42,4 meses (IC 95 % 33,23-51,56) en terapia sistémica exclusiva vs. 82,33 (IC 95 % 62,1-102,55) en terapia conjunta, ambos resultados estadísticamente significativos para el grupo quirúrgico. Conclusión: En pacientes con carcinoma de mama metastásico al debut, la supervivencia global y la supervivencia libre de progresión fueron mejores en los tratados con terapia conjunta que en los manejados con terapia sistémica exclusiva.

  PublisherOA PDFDOI

• Supervivencia en pacientes con cáncer de mama triple negativo con enfermedad residual post-neoadyuvancia tratadas con capecitabine adyuvante

  Oncología (Ecuador) · 2024-08-30

  articleOpen access

  Antecedentes: La escasez de terapias eficaces ha contribuido a que el cáncer de mama triple negativo tenga resultados desfavorables. Objetivo: Evaluar supervivencia global y libre de progresión en pacientes con cáncer de mama triple negativo con enfermedad residual postneoadyuvancia, tratadas con capecitabine. Métodos: Estudio de cohorte retrospectiva. Se calcularon funciones de supervivencia de Kaplan-Meier. Adicionalmente se desarrollaron modelos de regresión de Cox para análisis de asociación. Resultados: Se incluyeron 41 pacientes, de las cuales 25 (61%) eran postmenopáusicas, 23 (56,1%) tenían tumores iniciales ?5.1cm. La mediana de SLP fue de 25.03 meses (IC 95%, 13.37 – 36.68). El 26,8% de las pacientes presentaron progresión a los 36 meses de seguimiento, de ellas 54,5% que presentaron progresión eran premenopáusicas, En las mujeres con estado postmenopáusico se observó mayor SLP (HR 0,32, IC95% 0,09 -0,98, p 0,045). La mediana de SG fue de 55.60 meses (IC 95%, 46.5-58.5). No se observaron diferencias significativas entre el score RCB (Residual Cancer Burden) y la SLP y SG. Conclusión: En pacientes con enfermedad residual postneoadyuvancia tratadas con capecitabine adyuvante se observaron resultados favorables particularmente, en aquellas pacientes postmenopáusicas y con menor tamaño tumoral previo.

  PublisherOA PDFDOI

• Interpretable Prediction Rules for Congestion Risk in Intensive Care Units

  Stochastic Systems · 2023-07-17 · 5 citations

  articleOpen access1st authorCorresponding

  We study the problem of predicting congestion risk in intensive care units (ICUs). Congestion is associated with poor service experience, high costs, and poor health outcomes. By predicting future congestion, decision makers can initiate preventive measures, such as rescheduling activities or increasing short-term capacity, to mitigate the effects of congestion. To this end, we consider well-established queueing models of ICUs and define “high-risk states” as system states that are likely to lead to congestion in the near future. We strive to formulate rules for determining whether a given system state is high risk. We design the rules to be interpretable (informally, easy to understand) for their practical appeal to stakeholders. We show that for simple Markovian queueing systems, such as the [Formula: see text] queue with multiple patient classes, our rules take the form of linear and quadratic functions on the state space. For more general queueing systems, we employ methods from queueing theory, simulation, and machine learning (ML) to devise interpretable prediction rules, and we demonstrate their effectiveness through an extensive computational study, which includes a large-scale ICU model validated using data. Our study shows that congestion risk can be effectively and transparently predicted using linear ML models and interpretable features engineered from the queueing model representation of the system. History: This paper has been accepted for the Service Science/Stochastic Systems Joint Special Issue. Supplemental Material: The online appendix is available at https://doi.org/10.1287/stsy.2022.0018 .

  PublisherDOI

• Primary Care First Initiative: Impact on Care Delivery and Outcomes

  Manufacturing & Service Operations Management · 2023-04-03 · 11 citations

  articleOpen accessSenior author

  Problem definition: The Centers for Medicare & Medicaid Services launched the Primary Care First (PCF) initiative in January 2021. The initiative builds upon prior innovative payment models and aims at incentivizing a redesign of primary care delivery, including new modes of delivery, such as remote care. To achieve this goal, the initiative blends capitation and fee-for-service (FFS) payments and includes performance-based adjustments linked to service quality and health outcomes. We analyze a model motivated by this new payment system, and its impact on the different stakeholders, and derive insights on how to design it to reach the best possible outcome. Methodology/results: We propose an analytical model that captures patient heterogeneity in terms of health complexity, provider choice of care-delivery mode (referral to a specialist, in-person visit, or remote care), and quality of service (health outcomes and wait time). We analyze the provider decision on the mode of care delivery under both FFS and PCF and study whether PCF can be designed to yield a socially optimal outcome. We characterize analytically when patients, payer, and providers are better off under PCF and show that, in many cases, PCF can be designed to yield a socially optimal outcome. We numerically calibrate our model for 14 states in the United States. We observe that the average health status in a state is a source of heterogeneity that crucially drives the performance of PCF. We find that the model motivated by the current PCF implementation results in too much adoption of referral care and too little adoption of remote care. In addition, states with poor average health status may use more in-person care than socially optimal under a baseline (low) level of capitation. Moreover, relying on high levels of capitation leads to low adoption of in-person care. Managerial implications: Our results have health policy implications by shedding light on how PCF might impact patients, payer, and providers. Under the current performance-based adjustments, low levels of capitation should be preferred. PCF has the potential to be designed to achieve socially optimal outcomes. However, the fee per visit may need to be tailored to the local population’s health status. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2023.1207 .

  PublisherOA PDFDOI

• Care coordination for healthcare referrals under a shared‐savings program

  Production and Operations Management · 2022 · 19 citations

  1st authorCorresponding
  • Business
  • Actuarial science
  • Finance

  Accountable care organizations (ACOs) are responsible for the quality and cost of care of specified patient populations, including the cost of referrals. Motivated by this environment, we study care coordination for healthcare referrals. We consider an ACO that refers an uncertain number of patients from its attributed population to a preferred external provider for specialized health services. ACOs are typically paid under the Medicare Shared Savings Program (MSSP). Under the MSSP, the payer sets a spending benchmark for the beneficiary population during a fixed time period and shares any gains (losses) relative to it with the ACO. During the billing period, all services delivered to the attributed population by the ACO and external providers continue to be reimbursed under fee‐for‐service. Gains (losses) are determined at the end of the period by comparing the actual spending, which includes all care expenses (regular visits, referrals, and failed treatments) incurred by the payer in the period to the predefined benchmark. In this environment, the ACO and external providers—the latter not compensated under the MSSP—lack incentives to invest enough in care coordination initiatives. We study financial incentive mechanisms between the ACO and its preferred external provider to achieve integrated care coordination in referral markets under the MSSP. We show that traditional fee‐for‐service and capitation agreements do not provide sufficient incentives for care coordination in referral markets. However, a risk‐ and cost‐sharing mechanism can induce integrated care coordination efforts while satisfying the ACO and provider's participation constraints. We characterize a family of such mechanisms and numerically study the variability of the ACO and the external provider's profit. We demonstrate that this type of agreement can be used not only to induce integrated care coordination but can also result in a Pareto improvement in profit variability. We also illustrate the impact of the different MSSP risk tracks parameters on the performance of this care coordination mechanism, including their effect on the quality of care and the payer's mean spending.

  PublisherOA PDFDOI

• Optimal COVID-19 Vaccination Facility Location

  SSRN Electronic Journal · 2022-01-01 · 11 citations

  articleOpen access1st authorCorresponding
  PublisherDOI

Frequent coauthors

• José Carlos Oliveira

  16 shared

• Elisa F Long

  The University of Texas Rio Grande Valley

  9 shared

• Graça Porto

  Hospital de Santo António

  9 shared

• António Cabrita

  Hospital de Santo António

  8 shared

• Maria de Sousa

  7 shared

• Yaron Shaposhnik

  7 shared

• Carla Cardoso

  7 shared

• Rosa Lacerda

  6 shared