Real-time Policy Intelligence

From Wikitia
Jump to navigation Jump to search

Real-time Policy Intelligence or RPI (for health insurance) is a form of data visualization technology that provides insurance technical reporting to policyholders based on online analytics software and the use of artificial intelligence (AI). RPI resolves latency problems for a Real-time Business Intelligence for the health insurance industry. It resolves the first two latency, namely Data latency and Analysis latency, and enables the last latency, namely action latency, as described by analyst Richard Hackathorn.[1]

Traditionally, insurance companies report a policy performance to a policy holder or a broker in the form of a static paper report or excel sheet. Due to the amount of work usually required to compile such a report, insurance companies only deliver it to clients upon request of large clients or once or twice a year. In contrast, RPI has direct links to insurance policy data in real time, making it dynamic and accessible through the internet at any time by the policy holder or broker without requiring preparation work from the insurance company. This eliminates information asymmetry between insurance companies, policy holders and brokers during the life-cycle of a policy. Therefore, it allows for advanced data trends navigation and investigations of day-to-day policy activities flagging an early alarm of any policy troubles.

Historic Evolution

The development of RPI was facilitated through the evolution of two key aspects in the medical insurance claim processing[2]: data capture and data visualization.

Insurance data capture

Information of medical episodes are the basis of health insurance technical report as it contains the expenses on the health insurance policy. Obtaining the complete information in a timely manner is a challenge that required multiple technologies to advance. capturing both of medical approval requests and actual claim costs are essential for a proper understanding of a health insurance policy performance. The documentation of this process has always been paper-based, making it labour intensive, as it consisted of several manual steps including paper collection, organization, shipping, receiving, processing and finally reporting. The operation introduced months of lag-time between claim collection and reporting.

In the 1980s and 1990s, the operation evolved:

  • It became feasible to capture data in customized commercial databases and spreadsheet software.[3]
  • Approval communication between healthcare facilities and insurance companies depended on market adaptation to technologies of the likes of fax machines and email. [4]
  • Few markets adopted the use of Microsoft Excel spreadsheets that were burned on storage devices like disks and CDs to physically ship their digital claims data.

This provided an advancement to data capturing and reporting mechanism and became feasible to produce yearly reports.

During the 2000s, medical approval requests became possible through online portals. Medical providers with internet connection saw its benefits to replace traditional fax communication. Instant capturing of approvals information became possible.

Receiving claim data, however, could not notice an acceleration until a secured and reliable internet file transfer was adopted market wide to link between medical facilities and insurance companies. During 2010s, claims data uploads directly from medical providers to the insurance company system has seen rapid adaptation through insurance specific portals/clouds or by central redistribution storage facilitated by the governments, e.g. Dubai Health Authority.[5] Adopting these technologies allowed instant submission of claims without the need for batching paper claims on a monthly basis and without shipping delays. It has also eliminated the time and resources required for data entry and processing while opened the door for end to end automation minimizing missing data or quality problems. However, much of the processing delays remains today a challenge caused in markets where there is a low adaptation to standard coding for diagnoses and services. Claim processing remains dependent on human elements to adjust those codes in the system causing a queuing lag time.

Data visualization

With the rise of digital technologies that allowed data visualization, insurance companies found useful tools like Microsoft Excel to organize data of policy performance in meaningful charts. However, the complexity of insurance variables and their correlations limited the ability to analyze and interpret the policy performance to actuarial experts. Only high-level simplistic reports became popular to share in what is called a “technical report” upon-request or periodically (quarterly or biannually). Reports are typically provided as hard copies or PDF files to limit insurance clients' exposure to raw data and thereby avoiding misinterpretation of graphs.[6]

As data visualization software advanced and became available in recent years, the ability to produce complex data in simplistic navigable graphs became attractive and shareable. A manager with minimal insurance expertise can intuitively read and manipulate the data.

RPI systems

Real-time Policy Intelligence systems available in the markets are:

  • Saudi Enaya Cooperative Insurance (2020) [7]

Analytics Dimensions

  1. Policy health: Policy health is the indication if a policy is deteriorating (bad health) or within the forecast burning cost[8] (good health).
  2. Policy information: The changes of policy population by addition/deletion of members.
  3. Medical utilization: Medical services approved and costs incurred for most common diagnoses, procedures over a period of time.
  4. Demographic analytics: Members' behavior for different age groups, gender, relation to primary insured, etc.
  5. Geographic analytics: Visits and costs in different geographic regions and per healthcare provider and over a period of time.
  6. Market benchmark: Relative policy performance to general market performance and to same industry performance.
  7. Predictive and forecasts: The use of Artificial Intelligence, AI and other technologies to foresee potential future risks on policy performance. Also includes identification of high risk members and abusive healthcare providers.
  8. Operational KPIs: transparency of the insurance company's service performance. for instant, the time to respond with a decision and the rejection decision ratios.
  9. Consultative: Recommendations and proposed actions the insurance company would prepare for policy holder action.

Impact on Health Insurance Industry

For markets where health insurance is annual renewable policies, uncontrolled policy usage would systematically lead to increased prices year-on-year, increasing burden on the insured and inflating the entire national medical costs.[9] In well priced insurance policies, and in the cases where there is no unforeseen major medical condition, excessive usage is typically driven by abusive policy behavior by insured members and/or healthcare providers.[9] RPI enables active engagement of policy holder to understand area of deterioration in the policy, its severity and its root causes. RPI triggers a continual improvement cycle that leads to effective cost containment as a result. [10]

  1. Identify: RPI enables continuous monitoring of policy expenses and allows early detection of policy deterioration by policy holder and brokers experts. Furthermore, the interactive feature of RPI allows drill down and segmentation of expenses to pinpoint root causes.[11]
  2. Plan: developing corrective measures designed for effective containment of costs and targeting root causes. Plans are drafted solely or jointly with support of health insurance experts from the insurance company or broker.[11][12]
  3. Execute: corrective measure can vary from simple one-time initiative to a lengthy technology supported multi-party #measure. Some corrective measures may be executed by the policyholder alone. For instance, a simple measure can be an #announcement to company employees for behavior adjustment. Other measures are more complex requiring collaboration between #policy holder, insurance company and healthcare provider. A #complex measure could be to implement a PBM system that enforces dispensing generic drugs #instead of expensive brand drugs or replacement of a provider #in network.[11]
  4. Review: RPI provides a visual and intuitive way to monitor the impact of implemented corrective measures enabling a mechanism of trial and error adjustment.[11]

Overall, a strong adaptation of RPI in a market will result in a reduction in national healthcare expenditure. Funds towards medically needed treatments would be efficient allocation and away from excessive utilization (in which a result of behavioral abuse).


Limitation of data:

RPI technology depends on direct connection between the screens used by policyholders and brokers and the data managed by insurance companies. Therefore, the data should have an acceptable level of integrity and cleanliness for external visibility. However, since insurance companies have built their systems and data definitions for internal usage and interpretation, the data relevance to external parties might be confusing or misleading.

Limitation of "Real-time" definition:

The letter R in RPI refers to Real-time. Theoretically, an actual policy performance as of today's utilization would require all medical authorizations and claims data to be reported in the insurance company system. In reality, this is highly subjective to the nature of the market operations and the time required for the claims to be available. In practice, a healthcare provider submits claims in batches once a month to an insurance company. The time between the medical service and the point where the insurance company receives and processes the data is usually a minimum of 6 weeks.

Another aspect of Real-time definition is the rate of data refresh.[13] As insurance core systems are used for data processing, RPI usually extracts its data from a mirrored and synced database in order to maintain an acceptable performance and availability for both operations and RPI. A second-by-second real time refresh is expensive. Insurance companies may opt to sync data on more convenient intervals such as once a week or once a day.

Limitation from regulatory framework:

Regulations for health and financial data availability outside of a country or on public clouds dictate certain limits to the technological options that insurance companies can build advanced RPIs with.


Cyber security: Making availability of medical and financial data available for external access puts pressure on cyber security to protect connections from hacking attempts and misuse of information.


This article "Real-time Policy Intelligence" is from Wikipedia. The list of its authors can be seen in its historical. Articles taken from Draft Namespace on Wikipedia could be accessed on Wikipedia's Draft Namespace.