August 27, 2026: Carequality retrieval now adapts to each responding organization

We've rebuilt how Zus retrieves records over Carequality. Document list and document download requests are now dispatched and tracked separately for every responding organization on the network, instead of flowing through one shared pipeline that treated every responder the same.

What's changed

Pacing tuned to each responder. Every responding organization now has its own dedicated path through our retrieval system. Zus adjusts to what each responder can actually handle, and dials back automatically when a responder starts showing strain. A small practice whose endpoints struggle under load is queried more gently than a large health system that can absorb heavy parallel load.

Improved isolation of struggling responders. Traffic separation between responders eliminates any impact that a struggling responder has on you receiving data from other responders.

Improved error handling. Zus now assesses every error a responder returns and retries errors that may be recoverable, calibrated to that specific organization's behavior. This closes real data gaps by retrying responders that have returned positive matches for a patient but are struggling to keep up with document request load.

What you'll notice: jobs may stay in progress longer

Because Zus now backs off and retries against slower or struggling responders, individual jobs can remain in an in-progress state noticeably longer than they did before. Retries against a strained responder may extend over hours, and for document lists they can span more than a day before Zus stops trying.

This is deliberate. That additional time is spent recovering documents we would previously have abandoned after a single API error. A job that completes quickly but silently skips three organizations is a worse outcome than a job that takes longer and comes back complete.

If your workflows key off job completion, expect a wider distribution of completion times — especially for patients with records at many organizations, or at organizations that are having a rough day. The tradeoff is a more complete record at the end of it.

What's coming up

We're continuing to tune retry behavior and pacing per organization. We are also working to expose more granular organization-level response status so users have a clearer understanding of why their jobs are still in progress.

Questions?

Reach out to your Zus contact if you'd like to talk through how this affects your workflows.