Health care organizations generate increasing amounts of electronic data, as more and more of the process becomes digitized – for example, through e-prescribing, electronic medical records (EMRs), digital imaging scans, pharmacy data, lab data, admissions systems, billing systems, insurance claims data, and regional health information exchanges. Recent industry initiatives to improve care, lower costs, and track trends have created a need to aggregate these volumes of data and analyze it in ways that can provide insights to maximize performance. Yet, information collected in different formats by systems that are not interoperable is likely to yield few insights. This disparate data is characterized by three major problems:
- The data exist in silos, so the health care organization does not have one complete, integrated repository of all its data. For example, each provider the patient has an encounter with – from hospitals to physician specialists – most likely maintains a separate patient record. New reimbursement models (like bundled payments and penalties for avoidable hospital readmissions) mean providers must work together more closely to coordinate care, yet they aren’t able to work from a central data repository. To further complicate matters, payers now also collect patient data, with the goal of applying clinical analytics to manage population health – creating yet another silo of data that probably cannot be shared or combined with other data sets for a “whole picture” view.
- The data are highly redundant throughout the organization – and we’re not talking about back-up systems. Problematic data redundancy occurs when providers duplicate patient data in unlinked files – for example, a list of medication allergies that exists in both the EMR and the pharmacy system. It not only wastes storage space, but creates inconsistencies when a provider updates information in one place but not another.
- The data are variable in format and content. When information isn’t standardized, the organization cannot combine data sets or compare them, either internally or externally. For example, providers and payers may be collecting information about the same types of events – diabetes management techniques and their outcomes – but cannot pool their data to coordinate improvements to the care process.
Data disparity practically ensures that an organization cannot tap into the true meaning of all the information it has captured – or use it to enhance performance. Once these challenges are overcome, however, health care organizations can apply business intelligence and predictive analytics processes to improve disease management, optimize financial performance, streamline operations, and become more patient-centric.
It’s necessary to transform disparate data (both internal and external to the organization) into a usable data set in order to provide a solution to a particular problem or analyze the data to show trends and predict outcomes. The transformation involves creating a set of processes, governance policies, standards, and tools that permit the disparate data integration. This means there will be burgeoning professional opportunities for health care informatics team members as the health care industry struggles with this big data challenge.
American Sentinel’s online health care informatics degree, the Master of Health Care Informatics (MHCI), offers a degree designed to provide skills in gathering, analyzing, and presenting health care data for clinical use.