About

Pedro Luis Gozalo is a Professor of Health Services, Policy and Practice at Brown University. His research focuses on health services, particularly in the context of long-term care, nursing homes, hospice care, and post-acute care. He has contributed to the development of methodologies to identify cohorts of Medicare beneficiaries residing in assisted living facilities using administrative data and has studied various aspects of end-of-life care, including hospice discharges, site of death, and quality of care for patients with advanced illnesses. His work often involves analyzing Medicare and other administrative data to evaluate healthcare utilization, costs, and outcomes in vulnerable populations, with an emphasis on improving quality and transparency in end-of-life and long-term care settings.

Research topics

• Nursing
• Medicine
• Medical emergency
• Family medicine

Selected publications

• Association of Medicare Advantage Enrollment With Post‐Acute Care Use and Associated Patient Outcomes

  Journal of the American Geriatrics Society · 2026-04-06

  articleOpen accessSenior author

  IMPORTANCE: Enrollees in Medicare Advantage (MA) receive less intensive post-acute care (PAC) than those in traditional Medicare, but the implications of this lower intensity, particularly for patients with complex needs, remain poorly understood. OBJECTIVES: To estimate the association of MA enrollment with PAC use and patient outcomes for hospitalized beneficiaries with hip fracture or stroke. DESIGN, SETTING, AND PARTICIPANTS: A quasi-experimental difference-in-differences analysis leveraging the geographic expansion of MA from 2012 to 2017. The study included 148,396 stroke and 126,046 hip fracture hospitalizations, representing quasi-exogenous hospitalization events in high MA-growth counties. MAIN OUTCOME MEASURES: Initial PAC setting, 30-day all-cause hospital readmission, and 30- and 90-day all-cause mortality. RESULTS: MA enrollment was associated with fewer discharges to inpatient rehabilitation facilities (stroke: -8.9 pp; 95% CI, -9.88 to -7.92; hip fracture: -14.4 pp; 95% CI: -15.38 to -13.42). While 30-day readmissions were modestly lower for MA enrollees in both cohorts, MA enrollees experienced a 7.1% relative increase in 30-day mortality for stroke (0.6 pp; 95% CI: 0.01 to 1.19) and an 11.9% relative increase in 90-day mortality for hip fracture (1.3 pp; 95% CI: 0.52 to 2.08). This adverse mortality effect was concentrated in markets with high baseline IRF use (> = 33.3% of discharges, top tercile), where MA enrollment was associated with an 18.0% relative increase in 90-day mortality for stroke (2.0 pp; 95% CI: 0.82 to 3.18) and a 22.3% relative increase in 90-day mortality for hip fracture (2.3 pp; 95% CI: 0.93 to 3.67). CONCLUSIONS: MA enrollment was associated with lower IRF use, modestly lower readmissions, and a higher mortality risk for hip fracture and stroke. These findings suggest that MA's strategy of shifting patients to lower-cost settings may carry unintended adverse consequences for clinically complex patients.

  PublisherOA PDFDOI

• Symptom Burden Across Disease Groups in Palliative Care Consultations

  American Journal of Hospice and Palliative Medicine® · 2026-02-24

  articleSenior author

  BackgroundThe goal of palliative care (PC) is to reduce suffering and improve quality of life for patients with life-limiting illnesses and their families. Prior studies consistently demonstrate high symptom burden among PC patients; however, most evidence comes from cancer populations outside the US, leaving a gap in large-scale US data across disease groups.MethodsWe conducted a retrospective cohort study using quality metrics from a large, nonprofit hospice agency in the US Northeast. The study included 5871 patients who received palliative care services and had at least one Edmonton Symptom Assessment Scale (ESAS) assessment between July 2022 and December 2023. Five symptoms are described by severity on a 0-10 scale in a sample with malignant and non-malignant diagnoses.ResultsWe observed high prevalence of pain (34.3%), anxiety (32.3%), and dyspnea (28.0%) at first consultation. Many patients also reported moderate (4-6) or severe (7-10) intensity for pain (17.6% and 16.7%, respectively). Symptom burden also varied across disease groups: patients with solid tumors (63.3%) and liver disease (57.5%) exhibited particularly high rates of pain, while heart (46.9%) and lung disease (66.8%) exhibited high dyspnea. Among the 2852 patients with repeated consultations, there was substantial symptom improvement for pain (33.1% of patients), anxiety (22.1%), and dyspnea (19.3%).ConclusionPalliative care patients experience high symptom burden, varying by diagnosis. Planning access to tailored PC services to meet varying physical and emotional needs at a population level remains critical.

  PublisherDOI

• Comparative Effectiveness of Inpatient Rehabilitation Versus Skilled Nursing Facilities for Stroke and Hip Fracture Patients

  Journal of the American Geriatrics Society · 2025-10-23 · 2 citations

  articleOpen accessSenior author

  BACKGROUND: Prior research suggests discharge to inpatient rehabilitation facilities (IRF) leads to improved outcomes for stroke and hip fracture patients relative to skilled nursing facilities (SNF), while incurring greater costs. However, these estimates are likely biased by non-random patient selection. METHODS: We used a quasi-experimental design to compare post-acute care outcomes among Medicare beneficiaries hospitalized for stroke or hip fracture in 55 US hospitals that closed their IRF units between 2009 and 2017. Primary and secondary outcomes were 30-, 90-, and 180-day readmission and mortality, and successful community discharge. RESULTS: Among 10,761 stroke and 13,963 hip fracture hospitalizations, IRF discharge declined sharply, offset by increases to SNF and home health. Relative to IRF, SNF discharge was associated with no significant differences in readmissions but an increase in 90-day mortality for stroke (+6.5%, 95% CI 1.5%-11.4%) and hip fracture (+5.8%, 95% CI 2.5%-9.0%). Successful community discharge did not differ for patients redirected to SNF, but stroke patients redirected to home health had significantly higher rates of successful discharge (DID estimate: +6.8%; 95% CI 0.1%-13.5%). The protective effect of IRF was concentrated within 20 days post-discharge. CONCLUSIONS: Following hospitalization for stroke and hip fracture, discharge to an IRF was associated with lower mortality relative to SNF. However, given the potential for unmeasured confounding, this association should be interpreted with caution. Careful post-acute care referral protocols are critical to ensure good patient outcomes.

  PublisherOA PDFDOI

• Association of Shift Patterns with Overwork, Turnover, and Care Quality in U.S. Skilled Nursing Facilities

  Innovation in Aging · 2025-12-01

  articleOpen access

  Abstract Workforce challenges, including high turnover and staffing shortages, persist in skilled nursing facilities (SNFs). Understanding how shift structures influence workforce stability and care quality is critical for policy and operational decisions. This study identified the dominant shift for SNFs, defined as the most frequent daily work hours for the majority of nurses (Registered Nurses, Licensed Practical Nurses, and Certified Nurse Assistants), and classified SNFs into three shift categories: predominantly 8-hour shifts, 12-hour shifts, or a mix of both. Using facility-level data from 2021, integrating Payroll-Based Journal data, Nursing Home Care Compare quality measures, and Long-Term Care Focus facility characteristics, we fit linear regression models with state fixed effects to assess the association between dominant shift type and key workforce and care quality outcomes, adjusting for facility characteristics such as ownership, size, and case mix. Of the total 13,071 facilities studied, 10,821 (83%) had a dominant 8-hour shift, 1,918 (15%) had a dominant 12-hour shift, and 332 (2%) used mixed shifts. Facilities with 12-hour or mixed-shifts had significantly higher overwork prevalence (2.1 and 1.3 percentage points higher than 8-hour facilities, respectively) and turnover rates (7.6 percentage points higher for 12-hour shifts). Quality of care was also worse in 12-hour shift facilities across 16 of 22 measures, including higher rates of pressure ulcers, falls, and depressive symptoms. Policies should address the implications of different types of shifts, and future research should explore interventions to balance staffing efficiency with staff well-being and patient care.

  PublisherOA PDFDOI

• Do Higher-Cost Settings Deliver Better Outcomes? Post-Acute Care Efficiency in Medicare Skilled Nursing Facilities and Home Health Agencies

  SSRN Electronic Journal · 2025-01-01 · 1 citations

  preprintOpen access
  PublisherDOI

• Does the Veterans Health Administration Purchase High-Quality Home Health Care?

  Journal of the American Medical Directors Association · 2025-06-05

  articleOpen access

  OBJECTIVES: Veterans are frequently referred for skilled home health care (HHC) after hospital discharge. We examined access to high-quality VA-contracted home health agencies (HHAs) for Veterans following a VA medical center (VAMC) hospitalization and by rural or urban residence. DESIGN: Retrospective observational study. SETTING AND PARTICIPANTS: We used national VA data from April 2015 to September 2019 and included all Veterans discharged from a VAMC hospitalization who received skilled HHC paid for by VA. METHODS: We assigned every eligible discharge a choice set of VA-contracted HHAs serving the Veteran's county and a choice set of all HHAs serving the Veteran's county. We calculated a quality ratio (QR): the percentage of high-quality VA-contracted (4- or 5-star) HHAs relative to all high-quality HHAs serving the county reported in Home Health Compare. We then aggregated the QR by VAMC. High QRs (average QR and 95% CI >1) indicate VA contracts with higher-quality HHAs, relative to all HHAs serving the county. Neutral QRs (positive or negative average QR and 95% CI includes 1) indicate no difference in quality. Low QRs (average QR and 95% CI <1) indicate a lower-quality network. We examined bootstrapped QRs by VAMCs, including the subset serving mostly rural Veterans. RESULTS: We identified 60,406 VA-paid HHC episodes for 42,010 Veterans discharged from 113 VAMCs. Although 61.1% of VAMCs had high QRs, only 27.3% of the 33 VAMCs serving rural Veterans had high QRs. Rural-residing Veterans had lower proportions of high QRs than urban Veterans (46.3% vs 64.0%) and nearly double the proportion of neutral QRs (31.5% vs 16.2%), but similar proportions of low QRs (22.2% vs 19.1%). CONCLUSIONS AND IMPLICATIONS: Overall, VA-contracted HHAs were of higher quality compared to available HHAs serving Veterans' counties. Although VAMCs serving mostly rural Veterans provided less access to higher-quality HHAs, this effect was driven by more access to neutral quality HHAs in rural counties, not by differentially contracting with lower-quality HHAs.

  PublisherOA PDFDOI

• A Multilevel Analysis of Determinants of Dying in Place among Assisted Living Residents

  Innovation in Aging · 2025-12-01

  articleOpen accessSenior author

  Abstract Many older adults residing in assisted living (AL) consider this setting their home; yet, there is limited evidence about determinants of their ability to die in place, with state regulations explaining only some variation in this outcome. We examined the likelihood of residents dying in place as a function of state regulations and AL-community factors. To do this, we combined an inventory of state regulations with a nationally-representative survey of over 2,000 AL administrators from 48 states about care processes, and claims data for 9,803 deceased fee-for-service Medicare beneficiaries in 2021-2022. Variables included state regulations related to third-party services (e.g. license explicitly allows hospice), survey variables about organizational characteristics (size, neighboring facilities, for-profit status, accepting Medicaid), care processes (staffing, quality of collaboration with hospice), resident characteristics (demographics, chronic conditions) and a geographic (HSA) indicator of intensity of end-of-life care. Multilevel models were estimated with sampling weights to obtain nationally-representative estimates. State regulations were not significantly associated with dying in place in fully adjusted models. Residents were more likely to die in place when residing in AL communities with for-profit status and those that did not accept Medicaid. Residents at communities near a nursing home were significantly less likely to die in place. In terms of care processes, dying in place was associated with administrators arranging for hospice care (rather than referring out), and reporting higher quality of collaboration with hospice providers. AL characteristics and care processes are more important determinants of dying in place than state regulations.

  PublisherOA PDFDOI

• Causal inference with cross-temporal design

  Biometrics · 2025-01-07 · 2 citations

  articleOpen access

  When many participants in a randomized trial do not comply with their assigned intervention, the randomized encouragement design is a possible solution. In this design, the causal effects of the intervention can be estimated among participants who would have experienced the intervention if encouraged. For many policy interventions, encouragements cannot be randomized and investigators need to rely on observational data. To address this, we propose a cross-temporal design, which uses time to mimic a randomized encouragement experiment. However, time may be confounded with temporal trends that influence the outcomes. To disentangle these trends from the intervention effects, we replace the commonly used exclusion restrictions with temporal assumptions. We develop Bayesian procedures to estimate the causal effects and compare it to instrumental variables and matching approaches in simulations. The Bayesian approach outperforms the other 2 approaches in terms of estimation accuracy, and it is relatively robust to various violations of the common trends assumption. Taking advantage of the expansion of the Medicare Advantage (MA) program between 2011 and 2017, we implement the proposed method to estimate the effects of MA enrollment on the risk of skilled nursing facility residents being re-hospitalized within 30 days after discharge from the hospital.

  PublisherOA PDFDOI

• Using Latent Class Analysis to Identify Subgroups of Post-Operative Older Adults

  medRxiv · 2025-01-05

  preprintOpen accessSenior authorCorresponding

  Abstract Objective Identify subgroups of postoperative older adults using electronic health record data. Summary of Background Data Postoperative older adults represent a vulnerable population who may benefit from tailored postoperative care pathways. Identifying clinical subgroups can inform the development of these pathways. Methods Retrospective cohort study of postoperative adults &gt; 65 years (N=2,036) from a single healthcare system. Latent class analysis was used to identify patient subgroups based on measures of frailty, mobility, activities of daily living, and general health status. Hospital outcomes were described among each subgroup, including extended lengths of stay (LOS) (&gt;0.5 SD beyond mean LOS by surgical category), discharge disposition (i.e., home versus non-home discharge), and utilization (weekly visit frequency) of physical therapy (PT) and occupational therapy (OT). Results We identified 3 subgroups that we labeled Low Frailty-High Mobility (LF-HM), High Frailty-Low Mobility (HF-LM), and Low Frailty-Low Mobility (LF-LM), representing 15.3%, 27.6%, and 57.1% of the cohort, respectively. Discharge to home was highest among the LF-HM group (99%), followed by LF-LM (96%), and HF-LM (77%). Extended LOS was most common among the HF-LM group (27%), followed by LF-LM (18%), and LF-HM (6%). PT and OT visit frequencies were highest in the HF-LM group followed by the LF-LM and LF-HM groups. Conclusions This study identified 3 subgroups of postoperative older adults using routinely collected patient data. These groups may help to identify patients with increased odds of non-home discharge, extended LOS, and higher utilization of PT and OT and may inform the development of tailored postoperative care pathways for older adults.

  PublisherOA PDFDOI

• Opportunities to Improve End-of-Life Care in Assisted Living: Results From a National Survey of Administrators

  Journal of the American Medical Directors Association · 2025-11-24

  articleOpen access
  PublisherOA PDFDOI

Recent grants

• The Impact of Accountable Care Organizations on Post Acute Care

  NIH · $2.2M · 2016–2023

• NIH Grant R03HS011457

  NIH · $100k · 2003

• Data Management and Methods Core

  NIH · $54.2M · 2007–2029

• Data Management and Methods Core

  NIH · $5.0M · 2007–2024

Frequent coauthors

• Vincent Mor

  Providence College

  297 shared

• Joan M. Teno

  Providence College

  231 shared

• Kali S. Thomas

  Johns Hopkins University

  142 shared

• David Dosa

  Brown University

  99 shared

• Susan L. Mitchell

  Harvard University

  88 shared

• Susan C. Miller

  Brown University

  72 shared

• Emmanuelle Bélanger

  Brown University

  70 shared

• Jessica Ogarek

  Brown University

  65 shared

Labs

• Pedro Gozalo's LabPI

  Not provided

Education

• Ph.D., Health Services, Policy and Practice

  Brown University

  2000

• M.S., Health Services, Policy and Practice

  Brown University

  1996

• B.A., Economics

  Brown University

  1993