Healthcare information systems benefit in many ways using Ubiquiti’s data-driven technologies and text analytics capabilities. From concept-based searching in medical literature to statistical instance-based diagnostic systems – a wide range is enabled by our core competencies. The sectors cover provider hospitals and physicians, the payer insurances, the pharmaceutical drug manufacturers, the government regulatory agencies, and ancillary sectors such as TPAs and software vendors. Healthcare is an ideal sector for data-intensive applications, and we describe example uses.
At hospitals and physician practice datasets, entity extraction of doctor-patient encounter narratives produces structured data. This structured data includes automated encoding into ICD or CPT to optimize payments, and enables outcomes analysis. Use of standardized vocabularies and ontologies such as SNOMED enables searching and analysis to address quality of care, efficacy of treatment, and thereby, provides the evidence for performance-based payments. Better patient care is helped by powerful semantics-based search of medical literature, and by tracking within heretofore simple EMR systems. Ubiquiti also helps with medical research efforts as well.
Payers and insurers monitor and adjudicate medical claim transactions using complex business rules logic to alert for common and uncommon cases of over-payments. Ubiquiti data mining helps identify patterns of unacceptable payments, find the outliers in historical claims reimbursement, and also with the actuarial activity. The hierarchical information navigation in Ubiquiti software allows setting flags for complex situations, and to find cases of abuse of the system. Changes to the rules, alerts and information organization can be made very easily. This is contrast to the cumbersome and inflexible data warehousing and OLAP systems in relatively common use.
Healthcare organizations track data on ailments, medications, therapies, procedures etc. to identify, anticipate and report on a wide-range of issues. For example, clinical trials for new medications are constrained in the number of cases that may be studied. And so, when a larger population is supplied new drugs and therapies, potentially unexpected problematic outliers emerge – which have to be quickly identified and reported. Ubiquiti’s data mining of adverse event reports for such newly introduced patient-care products provide automated tracking alerts for potential problems. Similarly, tracking of patient populations, medications, and events, together provide results important, say, to epidemiology studies.
Advanced statistical instance-based machine-learning techniques supplied by Ubiquiti help in the complex science of diagnosis and recommended therapy guidelines. Our technology, as applied to large repositories of healthcare data, automatically find cases similar to a particular patient condition, and rapidly suggest differential diagnoses and best guidelines to therapy. Our approach effectively enables evidence-based medicine practices, and can be deployed easily. Ubiquiti’s approach, in more technical terms, uses a model-free similarity-search approach to identify the appropriate information, and as such, reflects the reasoning used by experienced medical personnel.
