Alex Milinovich

Alex Milinovich

Auto-generated profile, based on publicly available information

Alex Milinovich

Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

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

Representative Publications

View All Publications
24 September 2020 in Medical Decision Making
Verified
Journal Article
Open access
PDF file
GTGlen B. Taksler
Glen B. Taksler
JDJarrod E. Dalton
Jarrod E. Dalton
APAdam T. Perzynski
Adam T. Perzynski
MRMichael B. Rothberg
Michael B. Rothberg
AMAlex Milinovich
Alex Milinovich
DEDouglas Einstadter
Douglas Einstadter
Electronic health records (EHRs) offer the potential to study large numbers of patients but are designed for clinical practice, not research. Despite the increasing availability of EHR data, their use in research comes with its own set of challenges. In this article, we describe some important considerations and potential solutions for commonly encountered problems when working with large-scale, EHR-derived data for health services and community-relevant health research. Specifically, using EHR data requires the researcher to define the relevant patient subpopulation, reliably identify the primary care provider, recognize the EHR as containing episodic (i.e., unstructured longitudinal) data, account for changes in health system composition and treatment options over time, understand that the EHR is not always well-organized and accurate, design methods to identify the same patient across multiple health systems, account for the enormous size of the EHR, and consider barriers to data access. Associations found in the EHR may be nonrepresentative of associations in the general population, but a clear understanding of the EHR-based associations can be enormously valuable to the process of improving outcomes for patients in learning health care systems. In the context of building 2 large-scale EHR-derived data sets for health services research, we describe the potential pitfalls of EHR data and propose some solutions for those planning to use EHR data in their research. As ever greater amounts of clinical data are amassed in the EHR, use of these data for research will become increasingly common and important. Attention to the intricacies of EHR data will allow for more informed analysis and interpretation of results from EHR-based data sets.

27 February 2024 in Neurology
Verified
Journal Article
Closed access
AMAnnalisa Morgan
Annalisa Morgan
YLYadi Li
Yadi Li
NTNicolas R. Thompson
Nicolas R. Thompson
AMAlex Milinovich
Alex Milinovich
JAJustin Abbatemarco
Justin Abbatemarco
AKAmy Kunchok
Amy Kunchok
Longitudinal outcomes in anti-NMDA receptor encephalitis (anti-NMDARe) are still not fully understood and may not be adequately captured with the modified Rankin Scale (mRS), often the sole reported outcome. We aim to characterize longitudinal outcomes in anti-NMDARe using multiple outcome measures. This single-center, retrospective, observational study examined outcome measures (mRS and Clinical Assessment Scale in Autoimmune Encephalitis [CASE]) in adults with NMDA receptor-IgG in CSF at short- and long-term follow-ups using linear and logistic regression modeling. Patients with evaluations for cognitive impairment (Montreal Cognitive Assessment/Mini-Mental State Examination), depression (Patient Health Questionnaire-9), and anxiety (General Anxiety Disorder-7) >6 months from symptom onset were correlated with final CASE scores. Thirty-eight patients (76% female, median disease onset age = 28 years, range = 1-75 years) were included. The majority received first-line immunosuppressants (97%) at a median of 3.9 weeks (interquartile range [IQR] = 2.1-9.7) from symptom onset and 68% received second-line therapies. At baseline, median/mean mRS and CASE were 4 (IQR = 3-5) and 12.9 (SD = 7.2), respectively. At short-term follow-up (median = 10 weeks, IQR = 6-17), factors associated with higher CASE and mRS included dysautonomia, coma/lethargy, seizures/status epilepticus, and intensive care unit admission (p < 0.05). At long-term follow-up (median = 70 weeks, IQR = 51-174), median/mean mRS and CASE were 2 (IQR = 1-3) and 4.4 (SD = 4.2), respectively. Only weakness at symptom onset predicted higher mRS scores (odds ratio = 5.6, 95% confidence interval 1.02-30.9, p = 0.047). Despite both mRS and CASE improving from baseline (p < 0.001), only 9 patients (31%) returned to their premorbid function. Among patients with cognitive and mood evaluations >6 months from onset, moderate-severe cognitive impairment (42%), depression (28%), and anxiety (30%) were frequent. Cognitive and depression measures were associated with final CASE subscores (including memory, language, weakness, and psychiatric). Multiple clinical factors influenced short-term outcomes, but only onset weakness influenced long-term mRS, highlighting that mRS is predominantly affected by global motor function. Although mRS and CASE improved over time for most patients, these outcome measures did not capture the full extent of long-term functional impairment in terms of mood, cognition, and the ability to return to premorbid function. This emphasizes the need for increased utilization of more nuanced cognitive and mood outcome measures.

20 April 2021 in Journal of the American Heart Association
Verified
Journal Article
Open access
PDF file
JRJay Ramchand
Jay Ramchand
PPPooja Podugu
Pooja Podugu
NONancy Obuchowski
Nancy Obuchowski
SHSerge C. Harb
Serge C. Harb
MCMichael Chetrit
Michael Chetrit
SFScott D. Flamm
Scott D. Flamm
Background: Left ventricular non‐compaction remains a poorly described entity, which has led to challenges of overdiagnosis. We aimed to evaluate if the presence of a thin compacted myocardial layer portends poorer outcomes in individuals meeting cardiac magnetic resonance criteria for left ventricular non‐compaction . Methods and Results: This was an observational, retrospective cohort study involving individuals selected from the Cleveland Clinic Foundation cardiac magnetic resonance database (N=26 531). Between 2000 and 2018, 328 individuals ≥12 years, with left ventricular non‐compaction or excessive trabeculations based on the cardiac magnetic resonance Petersen criteria were included. The cohort comprised 42% women, mean age 43 years. We assessed the predictive ability of myocardial thinning for the primary composite end point of major adverse cardiac events (composite of all‐cause mortality, heart failure hospitalization, left ventricular assist device implantation/heart transplant, ventricular tachycardia, or ischemic stroke). At mean follow‐up of 3.1 years, major adverse cardiac events occurred in 102 (31%) patients. After adjusting for comorbidities, the risk of major adverse cardiac events was nearly doubled in the presence of significant compacted myocardial thinning (hazard ratio [HR], 1.88 [95% CI, 1.18‒3.00]; P =0.016), tripled in the presence of elevated plasma B‐type natriuretic peptide (HR, 3.29 [95% CI, 1.52‒7.11]; P =0.006), and increased by 5% for every 10‐unit increase in left ventricular end‐systolic volume (HR, 1.01 [95% CI, 1.00‒1.01]; P =0.041). Conclusions: The risk of adverse clinical events is increased in the presence of significant compacted myocardial thinning, an elevated B‐type natriuretic peptide or increased left ventricular dimensions. The combination of these markers may enhance risk assessment to minimize left ventricular non‐compaction overdiagnosis whilst facilitating appropriate diagnoses in those with true disease.

1 January 2016 in Clinical Medicine Insights: Endocrinology and Diabetes
Verified
Journal Article
Open access
KPKevin Pantalone
Kevin Pantalone
THTodd M. Hobbs
Todd M. Hobbs
BWBrian J. Wells
Brian J. Wells
SKSheldon X. Kong
Sheldon X. Kong
MKMichael Kattan
Michael Kattan
RZRobert S. Zimmerman
Robert S. Zimmerman
To assess changes in the clinical characteristics and treatment patterns of patients with newly diagnosed type 2 diabetes (T2D), the electronic health record system at Cleveland Clinic was used to create cross-sectional summaries of all patients with new-onset T2D in 2008 and 2013. Differences between the 2008 and 2013 data sets were assessed after adjusting for age, gender, race, and income. Approximately one-third of patients with newly diagnosed T2D in 2008 and 2013 had an A1C ≤8%, suggesting the continued presence of a delayed recognition of the disease. Patients with newly diagnosed T2D in 2008 were older than those in 2013. Hypertension, cardiovascular disease, and neuropathy were highly prevalent among patients diagnosed with T2D. The prevalence of neuropathy, cerebrovascular disease, and peripheral vascular disease increased from 2008 to 2013. Metformin was the most commonly prescribed antidiabetic medication. Sulfonylurea usage remained unchanged, while use of thiazolidinediones decreased considerably.

21 November 2023 in Journal of the American Heart Association
Verified
Journal Article
Open access
CHCatherine M. Heinzinger
Catherine M. Heinzinger
NTNicolas R. Thompson
Nicolas R. Thompson
AMAlex Milinovich
Alex Milinovich
MAMatheus Lima Diniz Araujo
Matheus Lima Diniz Araujo
COCinthya Pena Orbea
Cinthya Pena Orbea
RMReena Mehra
Reena Mehra
Background: We leverage a large clinical cohort to elucidate sleep‐disordered breathing and sleep‐related hypoxia in incident atrial fibrillation (AF) development given the yet unclear contributions of sleep‐related hypoxia and pulmonary physiology in sleep‐disordered breathing and AF. Methods and Results: Patients who underwent sleep studies at Cleveland Clinic January 2, 2000, to December 30, 2015, comprised this retrospective cohort. Cox proportional hazards models were used to examine apnea hypopnea index, percentage time oxygen saturation <90%, minimum and mean oxygen saturation, and maximum end‐tidal carbon dioxide on incident AF adjusted for age, sex, race, body mass index, cardiopulmonary disease and risk factors, antiarrhythmic medications, and positive airway pressure. Those with spirometry were additionally adjusted for forced expiratory volume in 1 second, forced vital capacity, and forced expiratory volume in 1 second/forced vital capacity. This cohort (n=42 057) was 50.7±14.1 years, 51.3% men, 74.1% White individuals, had median body mass index 33.2 kg/m 2 , and 1947 (4.6%) developed AF over 5 years. A 10‐unit apnea hypopnea index increase was associated with 2% higher AF risk (hazard ratio [HR], 1.02 [95% CI, 1.00–1.03]). A 10‐unit increase in percentage time oxygen saturation <90% and 10‐unit decreases in mean and minimum oxygen saturation were associated with 6% (HR, 1.06 [95% CI, 1.04–1.08]), 30% (HR, 1.30 [95% CI, 1.18–1.42]), and 9% (HR, 1.09 [95% CI, 1.03–1.15]) higher AF risk, respectively. After adjustment for spirometry (n=9683 with available data), only hypoxia remained significantly associated with incident AF, although all coefficients were stable. Conclusions: Sleep‐related hypoxia was associated with incident AF in this clinical cohort, consistent across 3 measures of hypoxia, persistent after adjustment for pulmonary physiologic impairment. Findings identify a strong role for sleep‐related hypoxia in AF development without pulmonary physiologic interdependence.

1 August 2025 in Circulation: Cardiovascular Quality and Outcomes
Verified
Journal Article
Closed access
MRMichael B. Rothberg
Michael B. Rothberg
AHAaron C. Hamilton
Aaron C. Hamilton
BHBo Hu
Bo Hu
MSMegan Sheehan
Megan Sheehan
JFJacqueline Fox
Jacqueline Fox
MPMatthew A. Pappas
Matthew A. Pappas
Background: There are multiple risk assessment models (RAMs) for venous thromboembolism prophylaxis, but it is unknown whether they increase appropriate prophylaxis. Methods: To determine the impact of a RAM embedded in the electronic health record, we conducted a stepped-wedge hospital-level cluster-randomized trial conducted from October 1, 2017 to February 28, 2019 at 10 Cleveland Clinic hospitals. We included consecutive general medical patients aged 18 years or older. Patients were excluded if they had a contraindication to prophylaxis, including anticoagulation for another condition, acute bleeding, or comfort-only care. A RAM was embedded in the general admission order set and physicians were encouraged to use it. The decisions to use the RAM and act on the results were reserved to the treating physician. The primary outcome was the percentage of patients receiving appropriate prophylaxis (high-risk patients with pharmacological thromboprophylaxis plus low-risk patients without prophylaxis) within 48 hours of hospitalization. Secondary outcomes included total patients receiving prophylaxis, venous thromboembolism among high-risk patients at 14 and 45 days, major bleeding, heparin-induced thrombocytopenia, and length of stay. Mixed-effects models were used to analyze the study outcomes. Results: A total of 26 506 patients (mean age, 61; 52% female; 73% White) were analyzed, including 11 134 before and 15 406 after implementation of the RAM. After implementation, the RAM was used for 24% of patients, and the percentage of patients receiving appropriate prophylaxis increased from 43.1% to 48.8% (adjusted odds ratio, 1.11 [1.00–1.23]), while overall prophylaxis use decreased from 73.5% to 65.2% (adjusted odds ratio, 0.87 [0.78–0.97]). Rates of venous thromboembolism among high-risk patients (adjusted odds ratio, 0.72 [0.38–1.36]), rates of bleeding and heparin-induced thrombocytopenia (adjusted odds ratio, 0.19 [0.02–1.47]), and length of stay were unchanged. Conclusions: Implementation of a RAM for venous thromboembolism increased appropriate prophylaxis use, but the RAM was used for a minority of patients. REGISTRATION: URL: https://www.clinicaltrials.gov/study/NCT03243708?term=nct03243708&rank=1 ; Unique identifier: NCT03243708.