Maria Marecka

Maria Marecka

Auto-generated profile, based on publicly available information

Maria Marecka

Infermedica, New York, NY

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 August 2024 in Medinformatics
Verified
Journal Article
Open access
PDF file
GGGeorge Gellert
George Gellert
MPMateusz Palczewski
Mateusz Palczewski
MMMaria Marecka
Maria Marecka
KPKatarzyna Paczkowska
Katarzyna Paczkowska
ASAleksandra Suwinska
Aleksandra Suwinska
SGStanislaw Gorski
Stanislaw Gorski
Reports indicate that only 19% of patients with prodromal symptoms of sudden cardiac arrest seek the appropriate level of care acuity, which may be due partly to individuals being unable to distinguish which acute symptoms require urgent care. This study evaluated the degree to which patient-users of artificial intelligence-based virtual triage presenting with prodromal symptoms of a possibly life-threatening cardiac condition intended to seek the needed emergency care. A dataset of 746,282 virtual triage patient-users was examined to assess the characteristics of those reporting symptoms of potentially life-threatening cardiac conditions with no intention to seek emergency care. Of 3,968 patient-users over age 35 with severe cardiac symptoms, 89.4% did not plan to seek emergency care. A majority of patient-users did not appear to understand the level of care acuity to seek, or planned to engage self-care or to visit an outpatient physician, instead of seeking emergency care. These findings reinforce the need for much-improved patient education among patients at risk for potentially life-threatening cardiac symptoms which require urgent medical care. Virtual triage (or symptom checkers) with care referral may have value in patient engagement and education to achieve this objective.

20 December 2025 in Mayo Clinic Proceedings: Digital Health
Verified
Journal Article
Open access
GGGeorge Gellert
George Gellert
BMBettina McMahon
Bettina McMahon
TPTim Price
Tim Price
ZLZhixin Liu
Zhixin Liu
AKAleksandra Kabat-Karabon
Aleksandra Kabat-Karabon
NLNirvana Luckraj
Nirvana Luckraj
To evaluate whether an artificial intelligence–based national virtual triage and care referral (VTCR) service in Australia improved care acuity level alignment, increased patient engagement of telemedicine services, and reduced emergency department demand by offering lower acuity, less costly options for urgent, virtual, or in-person care services. Cross-sectional analyses examined changes in patient care intent following VTCR to determine whether it facilitated patient adoption of new emergency and nonurgent telemedicine and virtual care services. Virtual triage and care referral more than doubled the number of patients selecting appropriate, lower acuity nonurgent care from 330,279 (21.3%) to 820,800 (52.9%), an increase of 31.6 percentage points [PPs] (P<.01), and effectively eliminated uncertainty in patient care seeking from 670,502 to 2557 patients, a decrease of 99.6%. Intent for in-person emergency care fell significantly from 119,414 (36.7%) to 105,349 patients (24.6%) (–12.1 PP; P<.01), replaced by substantial growth in patient intent to use virtual emergency care (from 612 to 11,840 patients or +10.1 PP) and nonurgent virtual care use (from 20,467 to 26,289 patients or +2.9 PP) (P<.01). Victoria, a state within Australia, recorded the highest uptake. Extrapolated nationally, these shifts could prevent an estimated 2409 unnecessary in-person nonurgent visits and 19,286 unnecessary emergency department visits annually in Australia. Aboriginal and Indigenous patients showed similar benefits and engaged VTCR at higher rates than other patients. Artificial intelligence–based VTCR improved alignment between patient perceived needs and recommended care pathways, not only driving greater use of appropriate, lower acuity, and telemedicine services but also reducing unnecessary in-person emergency visits. By eliminating uncertainty in care seeking and advancing adoption of new virtual emergency and nonurgent care options, VTCR offers a scalable, evidence-based solution for optimizing emergent and urgent care delivery and easing pressure on emergency departments across Australia.

19 September 2025 in JMIR Preprints
Verified
Preprint
Open access
MMMaria Marecka
Maria Marecka
ANAnna Nowicka
Anna Nowicka
ASAleksandra Suwińska
Aleksandra Suwińska
POPiotr Orzechowski
Piotr Orzechowski
BACKGROUND Virtual clinical triage is central to digital front door strategies, yet how interaction mode—structured, closed-ended AI questionnaires vs large language model (LLM)–enabled conversational dialogue—affects information capture, engagement, and alignment with care recommendations is unclear. OBJECTIVE This study compared demographics, clinical content, engagement, and alignment with recommended care between a traditional, structured questionnaire interface (traditional triage [TT]) and an LLM-enabled conversational interface (conversational triage [CT]) sharing the same validated Bayesian reasoning engine. METHODS This comparative, retrospective observational study analyzed 116,890 encounters over 28 weeks (January 2025 to August 2025) completed on a virtual triage website. Users self-selected TT or CT. Primary end points included demographic and clinical characteristics, engagement patterns, and self-reported intended adherence to triage recommendations. Analyses used poststratification weighting by age and sex. Group differences were assessed with the chi-square test (with Rao-Scott correction for weighted data) and survey-weighted Wilcoxon rank-sum test; P values for comparisons of proportions were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate procedure. For CT, opening and closing sentiment (positive, neutral, or negative) was labeled using a prompt in Gemini 2.5 Flash. RESULTS Of 116,890 respondents, 100,533 (86%) used TT and 16,357 (14%) used CT. Female users were the majority in both groups but less predominant in CT (10,402/16,357, 64% vs 71,039/100,533, 71%) than in TT. TT users skewed younger (aged 18-29 years >50%), whereas CT was more evenly distributed with higher shares at ages 12 to 17, 30 to 44, and ≥45 years. Median session duration was longer with CT than with TT (8 minutes 21 seconds, IQR 6 minutes 10 seconds to 11 minutes 51 seconds vs 4 minutes 25 seconds, IQR 3 minutes 14 seconds to 6 minutes 22 seconds, respectively). CT elicited more clinical findings (median 36, IQR 31-43 vs 32, IQR 27-38; P<.001) and surfaced more mental health evidence (eg, depressive symptoms, 8.5% vs 6.2%; P<.001). Posttriage intent survey completion was higher with CT (5104/16,357, 31.2% vs 5968/100,533, 5.9%). Self-reported intended adherence to the recommended level of care was higher with CT (1749/5104, 34.3% vs 1740/5968, 29.2%; P<.001), especially at extremes of acuity: self-care (534/625, 85.4% vs 325/525, 61.9%), emergency room (171/723, 23.7% vs 100/949, 10.5%), and ambulance (39/339, 11.5% vs 27/554, 4.9%; all P<.001). Among CT encounters, positive sentiment increased markedly, while negative sentiment rose slightly. CONCLUSIONS The LLM-enabled CT was used by a more demographically diverse user base and was associated with richer clinical context, deeper engagement, and higher self-reported intended adherence among survey respondents, particularly for low-acuity self-care and high-acuity emergency scenarios, where reassurance or urgent escalation is critical. This highlights the potential of hybrid LLM tools that integrate validated clinical logic to support engagement, information gathering, and care navigation in digital front door settings.

22 September 2026 in JMIR AI
Verified
Journal Article
Open access
MMMaria Marecka
Maria Marecka
ANAnna Nowicka
Anna Nowicka
ASAleksandra Suwińska
Aleksandra Suwińska
POPiotr Orzechowski
Piotr Orzechowski
Background: Virtual clinical triage is central to digital front door strategies, yet how interaction mode—structured, closed-ended AI questionnaires vs large language model (LLM)–enabled conversational dialogue—affects information capture, engagement, and alignment with care recommendations is unclear. Objective: This study compared demographics, clinical content, engagement, and alignment with recommended care between a traditional, structured questionnaire interface (traditional triage [TT]) and an LLM-enabled conversational interface (conversational triage [CT]) sharing the same validated Bayesian reasoning engine. Methods: This comparative, retrospective observational study analyzed 116,890 encounters over 28 weeks (January 2025 to August 2025) completed on a virtual triage website. Users self-selected TT or CT. Primary end points included demographic and clinical characteristics, engagement patterns, and self-reported intended adherence to triage recommendations. Analyses used poststratification weighting by age and sex. Group differences were assessed with the chi-square test (with Rao-Scott correction for weighted data) and survey-weighted Wilcoxon rank-sum test; P values for comparisons of proportions were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate procedure. For CT, opening and closing sentiment (positive, neutral, or negative) was labeled using a prompt in Gemini 2.5 Flash. Results: Of 116,890 respondents, 100,533 (86%) used TT and 16,357 (14%) used CT. Female users were the majority in both groups but less predominant in CT (10,402/16,357, 64% vs 71,039/100,533, 71%) than in TT. TT users skewed younger (aged 18-29 years >50%), whereas CT was more evenly distributed with higher shares at ages 12 to 17, 30 to 44, and ≥45 years. Median session duration was longer with CT than with TT (8 minutes 21 seconds, IQR 6 minutes 10 seconds to 11 minutes 51 seconds vs 4 minutes 25 seconds, IQR 3 minutes 14 seconds to 6 minutes 22 seconds, respectively). CT elicited more clinical findings (median 36, IQR 31-43 vs 32, IQR 27-38; P<.001) and surfaced more mental health evidence (eg, depressive symptoms, 8.5% vs 6.2%; P<.001). Posttriage intent survey completion was higher with CT (5104/16,357, 31.2% vs 5968/100,533, 5.9%). Self-reported intended adherence to the recommended level of care was higher with CT (1749/5104, 34.3% vs 1740/5968, 29.2%; P<.001), especially at extremes of acuity: self-care (534/625, 85.4% vs 325/525, 61.9%), emergency room (171/723, 23.7% vs 100/949, 10.5%), and ambulance (39/339, 11.5% vs 27/554, 4.9%; all P<.001). Among CT encounters, positive sentiment increased markedly, while negative sentiment rose slightly. Conclusions: The LLM-enabled CT was used by a more demographically diverse user base and was associated with richer clinical context, deeper engagement, and higher self-reported intended adherence among survey respondents, particularly for low-acuity self-care and high-acuity emergency scenarios, where reassurance or urgent escalation is critical. This highlights the potential of hybrid LLM tools that integrate validated clinical logic to support engagement, information gathering, and care navigation in digital front door settings.