Why insurance data integration keeps breaking from the start

The groups use different HR systems, file layouts and definitions of eligibility and dependents.

Instead of getting true straight-through processing, the carrier wrestles with long exception queues.

Brokers, plan administrators and other intermediaries suffer even more than the carriers, because they have to translate both incoming data and outgoing data.

Where enrollment and eligibility break

Sending in enrollment data should be simple, in theory.

In practice, one employer sends weekly full-file replacements.

Another submits files that show only the enrollment changes, using inconsistent change codes.

Keeping the employee eligibility information up to date is even more challenging.

Terminations arrive after claims have been processed.

Information about events that trigger eligibility for COBRA continuation benefits is buried in free-text fields, instead of being highlighted using structured codes.

When enrollment data files come in late or are inaccurate, plans may pay claims for ineligible members or deny valid claims.

The plan design data crisis

Efforts to communicate plan design data get less attention than efforts to communicate enrollment data, but problems in that area can cause enormous disruption.

Someone has to synchronize the rules for a plan's deductibles, copays, network options and formularies across benefits administration systems, claim platforms, ID card vendor systems, and pharmacy benefit manager systems.

There's no universal standard for transmitting the plan design information.

Some carriers send PDFs that require manual entry; others send extracts in a proprietary format.

When the carriers do use standards, their implementation of the standards may vary widely.

Plans often communicate information about mid-year plan design changes by email.

If some systems make the requested updates and some don't, employers may enroll workers in plans that don't exist.

Standards

The industry has invested heavily in developing and promoting standards such as X12, ACORD, and LDEx, and for good reason.

But standards don't eliminate variability.

Companion guides differ.

Elements are used inconsistently.

Many partners can't produce compliant transactions at all.

As a result, benefits organizations often maintain two parallel capabilities: a standards-based path for mature trading partners and a flexible data ingestion path for everyone else.

Most organizations underinvest in the second path, even though that's where the majority of volume originates.

The remediation trap

When benefits organizations get messy data files, remediation becomes manual.

Organizations try to clean things up by using spreadsheets, emails and re-uploads.

Turnaround times are long.

Members wait for ID cards.

Claims stall. Call volumes spike.

Too often, the data integration teams spend more time cleaning up the data files than on efforts to improve and add useful processes.

Partner frustration

Problems with data exchange create relationship problems.

Brokers, TPAs, and employers want to know: Was my file received?

What failed?

What needs fixing?

What actually works

Here are some conclusions about what separates the teams that get ahead from the rest.

First, they design for variability from day one.

Instead of building rigid integrations that assume consistent inputs, they build flexible ingestion frameworks that can accommodate different formats, map them to a common model, and apply consistent validation rules.

When a new employer submits data in a format not previously encountered, onboarding should take hours, not months.

Second, the successful teams give business users control over the rules.

Eligibility logic, validation criteria, and transformation mappings shouldn't require a developer whenever anything changes.

Those who understand the business should be able to manage the rules governing their data.

Third, the successful teams close the loop on errors.

When a record fails, the workflow is clear: triage, correct, reprocess and track the issue to prevent it from happening again.

Finally, the successful teams extend visibility to their partners.

Self-service portals and real-time dashboards help partners track submission status, correct errors, and resubmit updated files without delay.

Accountability shifts to the source of the data, trust improves and organizations can provide good support for more partners without adding employees.

Tim Bond is the chief product officer at Adeptia.