Data and Analytics
STATISTICAL SAMPLING
Statisticians at Livanta are healthcare data experts with years of experience selecting the optimal sample size and methodological approach to answer questions about the Medicare population, to determine the total number of clinician audits that need to be conducted, and to assess the effectiveness of statewide health policy. Statistical approaches include randomization, stratification, clustering, and convenience sampling. The goal of statistical sampling is to minimize the margin of error in estimation while maintaining reasonable sample size.
Value Proposition: When data collection is costly, such as with auditing or medical record review, identifying an adequate number of cases is essential to promote a high degree of confidence in the findings while efficiently utilizing resources and minimizing disruption in provider practices.
ECONOMETRIC AND STATISTICAL MODELING
The Livanta healthcare analytic team provides its extensive expertise in econometrics, statistics, and clinical healthcare to help government agencies decipher trends in health care improvement, identify key insights into program changes, and delineate healthcare impacts of policy decisions. As modeling experts, the Livanta Team navigates Medicare claims, medical record review outcomes, and other administrative healthcare data with advanced statistical techniques to analyze the effectiveness, efficiency, and equity of national and state healthcare systems and to assist government agencies achieve their goals. Our experience modeling healthcare data applies continuous and discrete dependent variables using quasi-experimental, cross-sectional, longitudinal, clustering, and hierarchical/multi-level approaches.
Value Proposition: Unlocking the incredible power of large clinical and healthcare claims datasets fuels understanding and predicts the impact of policies and programs on populations, such as Medicare beneficiaries, while controlling for socioeconomic and demographic variables.
QUANTITATIVE ANALYSIS
Quantitative analysis utilizes mathematical and statistical approaches to identifying and evaluating measurable data: healthcare data such as direct medical costs, inpatient mortality, 30-day readmissions, and hospital market share and population data such as demographic social determinants of health. Livanta’s quantitative analytics assess clinical, humanistic, actuarial, and health economic outcomes by applying measurement, probability, and economic theory to evaluate and to simulate outcomes of health policy, clinical practice changes, and economic incentives.
Value Proposition: Complex questions require well-specified approaches with measure-able goals to achieve success, and relationships between cause and effect demand a strong conceptual foundation to determine the extent to which the relationships are reliable.
DATA MANAGEMENT AND PROCESSING
Clinical encounter, medical claims, and other large healthcare data sources must be both secure while providing the appropriate access to define relationships between data elements, to identify missing and updated medical billing records, and to query the databases to support analytic reporting and decision making. The Livanta team integrates data from multiple sources through structured processes of extraction, processing and loading for warehoused data or for securely accessed remote data.
Value Proposition: Data is an asset when it provides the ability to perform analytic reporting and support decision making. Data can also be a liability when it is not secure.
VISUAL ANALYTICS
Livanta’s data visualization combines the power of large data sets, quantitative analysis, statistical modeling, and graphical presentation to quickly transform clinical and economic data into information. Relevant correlations between variables are quickly identified when displayed on a map, compared against a benchmark, or trended over time.
Value Proposition: Expertise in visual analytic approaches establishes a reporting environment that gives the end client the ability to ascertain health outcomes information and clinician practice patterns in the data without diverting end use attention to the statistics behind the analysis.