MSMaritza Marie Suarez

Maritza Marie Suarez

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Maritza Marie Suarez has a total of 48 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Safety Risk Prioritization for Patient-Facing Agentic Voice AI in Clinical Care: A Two-Round Modified Delphi Study (Preprint)

    by Arooshi Kumar, Anil Saldanha, Erin Palm, Melanie Fernando, Chethan Sarabu, Vishnu Ravi, Maritza Suarez, Peter McCaffrey, Chiraag Kulkarni, Bryon Frost, et al.

    BACKGROUND Patient-facing agentic voice artificial intelligence (AI) is moving from pilot to production across US health systems, with autonomous agents managing chronic disease check-ins, post-acute follow-up, and quality outreach. Unlike diagnostic models or ambient documentation tools, these agents initiate patient contact, interpret verbal responses in real time, and act with limited synchronous clinician oversight. Existing AI governance frameworks were not designed for autonomous systems of this kind. We previously developed a qualitative risk framework as a multi-stakeholder voice AI safety taskforce. OBJECTIVE This study aimed to establish multidisciplinary expert consensus on the relative priority of safety risks for patient-facing agentic voice AI, stratified by clinical use case, to support risk-proportionate oversight by health systems, payers, and vendors. METHODS A 14-member expert panel spanning 9 constituencies, including clinical practice, health system leadership, clinical informatics, industry, payer operations, patient safety research, and patient advocacy, rated that framework through a two-round modified Delphi conducted across 2025 and 2026. The instrument comprised 21 risk categories across 4 themes (6 agent-level, 5 data-level, 5 patient-level, 5 clinician-level), each rated against 5 clinical use cases, yielding 105 risk-by-use-case cells. Panelists scored likelihood and impact on 1-to-5 scales anchored in ISO 31000 descriptors and case-specific vignettes. Dual consensus was defined a priori as at least 70% of raters within 1 point of the median on both dimensions, per the RAND/UCLA Appropriateness Method. Round 2 non-responder ratings were carried forward, with prespecified sensitivity analyses. RESULTS All 14 panelists completed Round 1 (2940 ratings); 11 of 14 completed Round 2. Dual consensus was reached on 95 of 105 cells (90.5%) after Round 1 and all 105 (100%) after Round 2. Of the 10 Round 1 non-consensus cells, 8 reflected impact disagreement alone, 1 likelihood alone, and 1 both. By theme, clinician-level risks ranked highest (median likelihood by impact 10.5, IQR 9.0-12.0), followed by agent-level (9.5, IQR 9.0–12.0), patient-level (9.0, IQR 8.0–12.0), and data-level (9.0, IQR 6.0–10.5). Hypertension monitoring and post-stroke follow-up ranked highest among use cases; HEDIS gap closure ranked lowest. Identity mis-verification accounted for 3 of the 10 Round 1 non-consensus cells. Theme rank order was preserved in sensitivity analyses excluding framework developers and excluding non-responders. CONCLUSIONS A multidisciplinary panel reached complete consensus on a use-case-stratified prioritization of safety risks for patient-facing agentic voice AI. Clinician-level risks ranked highest, directing attention toward the clinician interface rather than agent-internal failure modes, and risk profiles varied substantially across use cases, indicating uniform governance is unlikely to be efficient. These priorities may inform risk-proportionate oversight, pre-deployment testing, and vendor contracting, although prospective validation against observed incidents is required before they can be used as predictive estimates of harm.

  • Effect of Intensive Blood Pressure Control on Kidney Outcomes

    by Paul E. Drawz, Kristin M. Lenoir, Nayanjot Kaur Rai, Anjay Rastogi, Chi D. Chu, Frederic F. Rahbari-Oskoui, Paul K. Whelton, George Thomas, Andrew McWilliams, Anil K. Agarwal, et al.

    Background: Intensive BP lowering in the Systolic Blood Pressure Intervention Trial (SPRINT) produced acute decreases in kidney function and higher risk for AKI. We evaluated the effect of intensive BP lowering on long-term changes in kidney function using trial and outpatient electronic health record (EHR) creatinine values. Methods: SPRINT data were linked with EHR data from 49 (of 102) study sites. The primary outcome was the total slope of decline in eGFR for the intervention phase and the post-trial slope of decline during the observation phase using trial and outpatient EHR values. Secondary outcomes included a ≥30% decline in eGFR to 2 and a ≥50% decline in eGFR or kidney failure among participants with baseline eGFR ≥60 and 2, respectively. Results: EHR creatinine values were available for a median of 8.3 years for 3041 participants. The total slope of decline in eGFR during the intervention phase was −0.67 ml/min per 1.73 m2 per year (95% confidence interval [CI], −0.79 to −0.56) in the standard treatment group and −0.96 ml/min per 1.73 m2 per year (95% CI, −1.08 to −0.85) in the intensive treatment group (P < 0.001). The slopes were not significantly different during the observation phase: −1.02 ml/min per 1.73 m2 per year (95% CI, −1.24 to −0.81) in the standard group and −0.85 ml/min per 1.73 m2 per year (95% CI, −1.07 to −0.64) in the intensive group. Among participants without CKD at baseline, intensive treatment was associated with higher risk of a ≥30% decline in eGFR during the intervention (hazard ratio, 3.27; 95% CI, 2.43 to 4.40), but not during the postintervention observation phase. In those with CKD at baseline, intensive treatment was associated with a higher hazard of eGFR decline only during the intervention phase (hazard ratio, 1.95; 95% CI, 1.03 to 3.70). Conclusions: Intensive BP lowering was associated with a steeper total slope of decline in eGFR and higher risk for kidney events during the intervention phase of the trial, but not during the postintervention observation phase.


  • Effect of Intensive versus Standard BP Control on AKI and Subsequent Cardiovascular Outcomes and Mortality: Findings from the SPRINT EHR Study

    by Paul E. Drawz, Nayanjot Kaur Rai, Kristin Macfarlane Lenoir, Maritza Suarez, James R. Powell, Dominic S. Raj, Srinivasan Beddhu, Anil K. Agarwal, Sandeep Soman, Paul K. Whelton, et al.

    Key Points: Identifying ways to prevent AKI may reduce mortality further in the setting of intensive BP control. Creatinine-based ascertainment of AKI, enabled by electronic health record data, may be more sensitive and less biased than traditional serious adverse event adjudication. Background: Adjudication of inpatient AKI in the Systolic Blood Pressure Intervention Trial (SPRINT) was based on billing codes and admission and discharge notes. The purpose of this study was to evaluate the effect of intensive versus standard BP control on creatinine-based inpatient and outpatient AKI, and whether AKI was associated with cardiovascular disease (CVD) and mortality. Methods: We linked electronic health record (EHR) data from 47 clinic sites with trial data to enable creatinine-based adjudication of AKI. Cox regression was used to evaluate the effect of intensive BP control on the incidence of AKI, and the relationship between incident AKI and CVD and all-cause mortality. Results: A total of 3644 participants had linked EHR data. A greater number of inpatient AKI events were identified using EHR data (187 on intensive versus 155 on standard treatment) as compared with serious adverse event (SAE) adjudication in the trial (95 on intensive versus 61 on standard treatment). Intensive treatment increased risk for SPRINT-adjudicated inpatient AKI (HR, 1.51; 95% CI, 1.09 to 2.08) and for creatinine-based outpatient AKI (HR, 1.40; 95% CI, 1.15 to 1.70), but not for creatinine-based inpatient AKI (HR, 1.20; 95% CI, 0.97 to 1.48). Irrespective of the definition (SAE or creatinine based), AKI was associated with increased risk for all-cause mortality, but only creatinine-based inpatient AKI was associated with increased risk for CVD. Conclusions: Creatinine-based ascertainment of AKI, enabled by EHR data, may be more sensitive and less biased than traditional SAE adjudication. Identifying ways to prevent AKI may reduce mortality further in the setting of intensive BP control.


  • Open Access

    Factors to Effective Telemedicine Visits During the COVID-19 Pandemic: Cohort Study

    by Kristin Nicole Gmunder, Jose W Ruiz, Dido Franceschi, Maritza M Suarez

    Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).


  • Open Access

    Factors to Effective Telemedicine Visits During the COVID-19 Pandemic: Cohort Study (Preprint)

    by Kristin Nicole Gmunder, Jose W Ruiz, Dido Franceschi, Maritza M Suarez

    BACKGROUND With COVID-19 there was a rapid and abrupt rise in telemedicine implementation often without sufficient time for providers or patients to adapt. As telemedicine visits are likely to continue to play an important role in health care, it is crucial to strive for a better understanding of how to ensure completed telemedicine visits in our health system. Awareness of these barriers to effective telemedicine visits is necessary for a proactive approach to addressing issues. OBJECTIVE The objective of this study was to identify variables that may affect telemedicine visit completion in order to determine actions that can be enacted across the entire health system to benefit all patients. METHODS Data were collected from scheduled telemedicine visits (n=362,764) at the University of Miami Health System (UHealth) between March 1, 2020 and October 31, 2020. Descriptive statistics, mixed effects logistic regression, and random forest modeling were used to identify the most important patient-agnostic predictors of telemedicine completion. RESULTS Using descriptive statistics, struggling telemedicine specialties, providers, and clinic locations were identified. Through mixed effects logistic regression (adjusting for clustering at the clinic site level), the most important predictors of completion included previsit phone call/SMS text message reminder status (confirmed vs not answered) (odds ratio [OR] 6.599, 95% CI 6.483-6.717), MyUHealthChart patient portal status (not activated vs activated) (OR 0.315, 95% CI 0.305-0.325), provider’s specialty (primary care vs medical specialty) (OR 1.514, 95% CI 1.472-1.558), new to the UHealth system (yes vs no) (OR 1.285, 95% CI 1.201-1.374), and new to provider (yes vs no) (OR 0.875, 95% CI 0.859-0.891). Random forest modeling results mirrored those from logistic regression. CONCLUSIONS The highest association with a completed telemedicine visit was the previsit appointment confirmation by the patient via phone call/SMS text message. An active patient portal account was the second strongest variable associated with completion, which underscored the importance of patients having set up their portal account before the telemedicine visit. Provider’s specialty was the third strongest patient-agnostic characteristic associated with telemedicine completion rate. Telemedicine will likely continue to have an integral role in health care, and these results should be used as an important guide to improvement efforts. As a first step toward increasing completion rates, health care systems should focus on improvement of patient portal usage and use of previsit reminders. Optimization and intervention are necessary for those that are struggling with implementing telemedicine. We advise setting up a standardized workflow for staff.