4 min read

How to choose a medical records retrieval vendor

There are four ways to retrieve patient medical records at scale. Here is how they differ on coverage, speed, and cost, and the questions to ask any vendor before you sign.

By Prabha Dublish, CEO

If your product depends on patient medical records, the retrieval vendor you pick decides how complete those records are, how long your customers wait, and what each case costs. The short answer: match the vendor's retrieval method to where your patients' records actually live. If they live in large connected health systems, an electronic network API may be enough. If they are scattered across small offices, imaging centers, and specialists, you need a vendor that can reach providers outside the networks too.

This guide compares the four approaches and ends with the questions worth asking any vendor before you sign.

The four ways to retrieve medical records

1. Network APIs

Network APIs query health information exchanges (HIEs) and national networks on your behalf and return whatever connected systems share electronically, usually as structured data (FHIR resources or C-CDA documents).

They are fast and developer-friendly. The limits are coverage and purpose. Small independent practices are underrepresented in exchange feeds, so the records that decide a case are often the ones missing. And many network queries are only permitted for treatment, which rules out legal, insurance, and most research use cases unless the vendor offers a patient-directed path such as TEFCA Individual Access Services.

2. Release-of-information services

Release-of-information (ROI) services take a signed HIPAA authorization and request the chart from each provider, by portal, fax, phone, or mail. Because a person is asking for the actual chart, they can return scanned documents, handwritten notes, and records no API exposes.

The trade-off is time and cost. Turnaround is usually measured in weeks, follow-up is manual, and pricing often includes per-page fees passed through from the provider's own records vendor.

3. Patient-access apps

Patient-access apps have the patient sign in to their patient portals or payer accounts and pull the data available there, typically through standardized FHIR APIs.

This works well for engaged, tech-comfortable patients with portal accounts. It breaks down for everyone else. Patients forget logins, abandon the flow, or never had a portal account at all, and each provider they skip is a gap in the record.

4. Hybrid retrieval

Hybrid retrieval starts electronic and keeps going. It pulls what connected systems and payers can return quickly, then works the remaining providers through their existing release-of-information channels, using software agents, staff, or both.

The goal is a more complete record without waiting weeks on the parts that could have come back instantly. The questions to ask are how much of the long tail the vendor actually reaches and how they handle providers that never answer.

How the approaches compare

Network APIsROI servicesPatient-access appsHybrid retrieval
Typical speedSeconds to minutesWeeksMinutes, if the patient finishesMinutes for connected systems, days for the rest
Small, unconnected officesOften missingReachedOnly if the patient has a loginReached
What comes backStructured data and summariesFull chart, often as PDFsPortal and payer dataStructured data plus full documents
Works for legal and insuranceOnly with a patient-directed pathYes, with a signed authorizationDepends on downstream useYes, with patient authorization
Patient effortNone to lightSign an authorizationLog in to every portalAuthorize once
Common pricingPer query or subscriptionPer request plus page feesSubscriptionPer retrieval or annual contract

These are typical patterns, not guarantees. Vendors within each category vary, so verify coverage on your own cases.

Try it yourself in a demo.
We will walk you through patient-mediated access live, on test data.

Questions to ask any vendor

1. What legal basis does each retrieval run on? A signed authorization, the HIPAA right of access, a treatment relationship, and TEFCA Individual Access Services each unlock different sources. Make sure the basis fits your use case, especially for legal, insurance, and research work where treatment-only access does not apply.

2. How do you reach providers outside electronic networks? Ask for the share of providers reached electronically versus manually, and what happens when an office never responds.

3. What exactly comes back? A medication list and a problem list are not the same as progress notes, operative reports, imaging reports, and bills. If your decisions depend on the full chart, ask for a sample.

4. What format, and how is it delivered? Raw PDFs, structured FHIR, or both, and whether results arrive by API, webhook, or a portal someone has to log in to.

5. What does turnaround look like, by source? Ask separately for connected systems and for the long tail. A single average hides the difference.

6. How is pricing structured? Look for per-page pass-through fees, minimums, and charges for retries or requests that return nothing.

7. How is the patient's identity verified? Patient-directed access should verify identity to a recognized standard (NIST IAL2 is the bar under TEFCA Individual Access Services) and let the patient revoke consent.

8. How are specially protected records handled? Substance use treatment records (42 CFR Part 2), psychotherapy notes, and similar categories need their own authorization. A vendor should never sweep them into a general request.

9. Do you sign a BAA, and what is in the audit trail? You should be able to see who requested what, when, from which source, and under which consent.

Where Hubble fits

Hubble is a hybrid retrieval platform built on patient-mediated access. The patient verifies their identity once, authorizes the request, and Hubble retrieves their records across EHRs (reached directly and through HIE networks) and payers, then uses voice and browser agents to reach the providers those miss. Records come back as raw PDFs or structured FHIR through one API.

You can see how this plays out by segment on our industries pages, or read the legal records case study for one team's results.

Try it yourself in a demo.
We will walk you through patient-mediated access live, on test data.

Frequently asked questions

What are the main types of medical records retrieval vendors?

There are four. Network APIs query health information exchanges and return what connected systems share electronically. Release-of-information services request the chart from each provider with a signed authorization. Patient-access apps have the patient log in to portals or payer APIs. Hybrid retrieval combines electronic access with agents or staff who reach the providers networks miss.

Why do network APIs miss records?

Health information exchanges only return what participating systems share, and small independent practices are underrepresented. Many network queries also require a treatment purpose of use, which legal, insurance, and research use cases do not have.

What should I ask a medical records retrieval vendor?

Ask what legal basis each retrieval runs on, how they reach providers outside electronic networks, what comes back (full chart or a summary), the format and turnaround, whether there are per-page fees, how they verify patient identity, and whether you get an API with status updates and an audit trail.

How long does medical records retrieval take?

Electronic retrieval from connected systems can return records in seconds to minutes. Manual retrieval from unconnected providers usually takes weeks, because it depends on each office answering calls and faxes. Hybrid approaches use electronic access first and work the remaining providers in parallel.

Is patient-authorized retrieval HIPAA compliant?

Yes. It runs on the HIPAA individual right of access (45 CFR 164.524), which lets a patient obtain their own records and direct a copy to a recipient they choose. The patient's identity should be verified, and consent should be revocable.