Ingest

Ingest

Integral solution for quality control and statistical analysis of hospital discharges

Ingest is an application developed in collaboration with Innovasalud, an Spanish health consulting company, that addresses health data integration, data quality control and advanced statistical analysis. It is perfectly compatible with databases such as MBDS, registered in Spanish hospitals since the 1990s by ministerial directive.

Key features

Integral solution
Easy to use web user interface
Multi-user fine-grained permissions
Centralized dataset
Standard compilant
Quality control

Description

Collecting and integrating datasets like the MBDS, as well as its subsequent analysis presents a certain level of complexities (different data formats, evolving data records, etc.), specially as the size of the organization gets bigger. Ingest provides an integral solution for the management of this dataset. Ingest was designed with the objective of being able to perform the usual tasks with databases of clinical records like the MBDS in an easy way and with the power of advanced statistical analysis available in its different modules. This are the most important characteristics of Ingest:
  • Automatic data ingesting in the system with customizable data formats
  • Data quality analysis of ingested data
  • Web system accessible from any browser
  • Multi-user with fine grained permissions for data access and analysis capabilities
  • Modular architecture with emphasis on temporal series analysis
  • Analysis with different levels of granularity, with filtering capabilities for all defined variables
  • Support for different versions of the ICD.
  • Integration with 3M™ DRG Grouper, being able to compute the DRG within the application. Ingest also allows to analyse the grouped information with the different versions of the grouper software.
  • Indicators such as PSI (defined by the AHRQ), the ones defined by the Spanish Ministry of Health (in its application iCMBD), Elixhauser's comorbidities or custom indicators defined by the user.
  • Case mix and functionality indicator adjustments
  • Multi-language application
  • Conformance with personal data protection regulations.
  • It has the expected characteristics of an integral data analysis solution: you can store frequently used queries, build custom reports, graphics, export information...

Modular architecture

Ingest has several modules for data analysis. All of them can visualize data with tables and graphs with this common set of features:
  • Create filter with any MBDS variable: dates, categorical variables, numeric variables, texts, etc.
  • Result groping with categorical variables
  • Temporal analysis of results (trends, anomalies, etc.)
  • Define analysis, comparative and reference date intervals
  • Data exportation in different formats (Excel, PDF, etc.), graphic and report generation
  • Aggregated results are connected with micro-scale module that lists individual records
  • Storage of frequently used queries

This are the available Ingest's modules:
  • Micro-scale analysis: consult and edit individual records
  • Record count: reports and basic statistics with grouping and filtering with MBDS variables
  • Data quality metrics: data quality analysis with more than 30 predefined metrics
  • Health care quality indicators: standard indicators defined by the Spanish Ministry of Health and the AHRQ
  • Variable influence analysis on health care quality indicators: identify most influential variables with advanced statistical analysis on indicators
  • Case mix and functionality health care quality indicator adjustments

Data flow

All information provided to Ingest gets homogenized and centralized in a single database. Data dumps have the following characteristics:
  • Different templates can be defined to match different formats of MBDS. Each data dump is managed independently: discharge grouping and indicator calculation jobs can be defined for every dump easily using a wizard.
  • Users can upload a private dump, allowing them to detect possible mistakes before publishing the data to the rest of the users.
  • Data upload can be restricted to a certain time window, facilitating data consolidation.
  • Different data quality analysis metrics are computed at the moment of data dump, being available for later consultation. Valid and invalid records are defined (according to template format), as well as data quality metrics in terms of data coding level.

Work with Predictia! Contact with us to get more information: