Using spreadsheets and non-digital, manual mechanisms to roll up accurate and timely reporting from hospitals, urgent care providers, nursing homes, medical suppliers, and others is proving to be daunting. As a result, delayed or inaccurate data is feeding into the prediction models, artificial intelligence, advanced analytics tools, and dashboards that leaders use to make policy. Executive decisions are frequently based on old data, which itself is increasingly deficient in quality and comprehensiveness.
Adding to this challenging environment is the federal government’s reporting and monitoring requirements for the recently allocated emergency funding flowing down to states and municipalities through a variety of federal social programs via the two-trillion-dollar CARES Act. Several oversight and watchdog organizations are standing up and establishing operations to protect these essential funds from fraud, waste, and abuse, which is just as important as their dispersal. The organizations are meant to ensure that the funds are meeting their intended purpose to help citizens, businesses, and health providers engaged in the COVID-19 fight.
Crisis-Induced Data Challenges
When routine data collection and flows rapidly became overwhelmed, state and municipal governments realized their legacy systems could not adequately respond. Of immediate concern was:- Mass Data Ingestion: Government agencies are inundated with massive amounts of structured and unstructured pandemic-related data from existing and new sources. Ensuring the completeness of that data is critical—essentially addressing the gap from what is supposed to be collected and what is collected. Of concern as well is inefficient use of limited bandwidth, data storage, and application consumption by collecting the same data sets multiple times rather than just the new information.
- Data Quality: Agencies need to have confidence in the quality and accuracy of collected data used for intelligent data-driven decisions. Well-meaning but inadequate decisions based on inaccurate or incomplete data can lead to public-government trust issues, such as life-saving medicine and protective and testing equipment being improperly prioritized and distributed.
- Payment Recipient Data and Citizen Privacy: Agencies have to guard against improper payments and fraudulent activity related to the massive stimulus package. Funds must be distributed to the citizens, health organizations, and businesses that need them the most to either fight the pandemic or maintain solvency for our future economic recovery. It’s also important to continue to ensure data privacy and protection in accordance with the law, which also helps maintain public trust.
Innovation on the Fly
Enterprise-level data management programs aren’t established or fixed overnight. If you’re starting from scratch or with legacy systems based on siloed application-database environments, you need to address the fundamentals first. If you’re starting from some level of enterprise data management maturity but have neglected core functions, the challenges such as those faced by state and municipal governments now are just as painful.While most are doing some level of data integration—or extract, transform, and load (ETL) operations—a neglect of data quality, data governance, and metadata management or data cataloging can create significant issues when a crisis erupts. Given the availability of data tools to automate these core functions, it’s hard to watch as organizations struggle through the current crisis and seek immediate corrections now.
That said, there are some quick fixes that some of the hardest hit state and municipal governments are implementing now. I encourage all government leaders to take similar steps as soon as possible. These include establishing the following capabilities:
- Data Integration and Data Quality: Massively ingest and integrate data from/to cloud, multi-cloud, hybrid, or on-premises sources, in any format in batch or real time, to critical applications in minutes. Easily support new and complex integration patterns. Achieve high performance, reliability, and near-universal connectivity for mission-critical business processes. Build trust in data by enabling automated delivery of the right data quality capabilities at the right time and the right scale to the right users.
- Data as a Service (DaaS): Ensure stimulus payment accuracy and reduce fraud with citizen, employee, business address, phone number, and email address verification to include autocorrelation and corrections, all through a web-based managed service. Include geo-coding capabilities that allow agencies to understand trends, hot spots, and other issues.
- Data Privacy via Data Masking: De-identify, de-sensitize, and anonymize sensitive data from inadvertent release or unauthorized access for application users, business/mission intelligence, application testing, and outsourcing.
- Citizen/Business 360-Degree Insights: In existing data lakes or Hadoop environments, bring together transaction, interaction, and other citizen data types and manage billions of records across all data sources. Infer citizen or patient attributes, prognoses, lifecycle events, and more from unstructured data using natural language processing (NLP). Link citizen, patient, or claimant data and create relationships using machine learning, right out of the box. Visualize relationships, contact points, families or households, social networks, and organizational hierarchies at scale using a graph data store.
