Certified EHRs, patient APIs and Promoting Interoperability: what the federal data show, 2008 to 2024

ONC survey series, CMS program files and our own counts of the 2023 and 2024 Promoting Interoperability data trace how far certified health IT and patient access APIs reach, and where the data stop.

The question and what the data answer

How far have certified health IT and patient-facing APIs actually spread, and how much do patients and clinicians use them? Public federal data answer the first half well and the second half only in part. On adoption, the record is close to complete: more than 99% of non-federal acute care hospitals and 91% of office-based physicians had a certified EHR in 2024 [1]. On patient access, about 87% of hospitals reported API-driven app access to patients' information in 2024 [4], 71% through a standards-based API such as FHIR [5], and 65% of individuals were offered and used an online record or portal [6]. On program performance, clinicians who report Promoting Interoperability now score near the ceiling [12]. What the data do not yet show is how many patients pull their data through payers' Patient Access APIs: payers began reporting those counts to CMS only in 2026, and CMS has not said it will publish them [19] [21].

This report sits beside three Moraph Research outlooks that cover the rules: the payer API and prior authorization outlook, the information blocking enforcement outlook and the TEFCA outlook. Here we stay with what the public data measure.

Data and method

We used five official data sources, each read or downloaded on November 16, 2026.

  • ASTP/ONC Health IT quick-stats on EHR adoption, which draw on the American Hospital Association (AHA) IT Supplement for hospitals and the National Center for Health Statistics' National Electronic Health Record Survey for physicians [1] [2] [3]. We quote AHA figures only as ONC publishes them. "Certified EHR" means an EHR certified under the ONC Health IT Certification Program, as each quick-stat defines it [2] [3].
  • ONC's quick-stat on hospital patient engagement functions and its data brief on hospital API use, both from the AHA IT Supplement; the data brief's percentages are among non-federal acute care hospitals [4] [5].
  • ONC's data brief on individuals' portal and app use, from the Health Information National Trends Survey (HINTS) [6].
  • ONC's linkage of CMS hospital Promoting Interoperability data to the Certified Health IT Product List (CHPL), one row per hospital and certified product, for program years 2023 and 2024 [9] [10].
  • CMS's Quality Payment Program (QPP) Experience public use files for the 2023 and 2024 performance years, one row per TIN/NPI combination [12] [13] [14].

Hospital count method: in the linkage file we kept each CMS Certification Number (CCN) whose "Meets criteria for promoting interoperability of EHRs" field reads Y, then counted distinct CHPL developer names among its rows [9] [10]. Developer names are as the CHPL gives them; we report aggregates and do not rank developers.

Clinician method: in each QPP file we kept rows with a CEHRT ID, the ONC-generated identifier of the certified EHR bundle whose data contributed to the clinician's score [15]. The 2024 count this gives, 268,652, equals the number ONC reports for clinicians who reported for Promoting Interoperability in 2024 [8] [12]. On those rows we took the unweighted Promoting Interoperability category score, which CMS defines as points achieved divided by points available, on a 0 to 100 scale [15], and looked for two measure IDs in the PI measure columns: PI_PEA_1, Provide Patients Electronic Access to Their Health Information [17], and PI_HIE_6, Enabling Exchange Under TEFCA [18]. For PI_PEA_1 we counted rows where the measure earned more than zero points [12] [13]. Measure points cannot be read as performance rates, because CMS's measure score is the score achieved toward the final score [15].

Three limits. The QPP files leave out TIN/NPIs with fewer than 11 patients, so their counts do not match CMS's own totals [14]. A clinician who bills under two TINs appears twice: the 2024 rows with a CEHRT ID cover 252,233 distinct NPIs [12]. And the CHPL itself, the list of every certified product, could not be counted for this report: its download service needs an API key, so all product and developer counts here come from ONC's CHPL-linked program files [9] [11].

Certified EHRs: the adoption curve is finished

Hospitals reached certified EHRs fast. In 2012, 91% of non-federal acute care hospitals reported a certified EHR; by 2014 the figure was 99%, and in 2024 it was above 99% [2].

Physicians moved more slowly and are not finished. 74% of office-based physicians had a certified EHR in 2014, the first year it was measured for them, 84% in 2022 and 91% in 2024; a further 4% used another EHR in 2024 [3].

Moraph's view

For hospitals the adoption question is settled, and the useful questions are now about which certified functions are switched on and used. For physicians, we expect the remaining gap to sit in practices that are hardest to reach with program incentives. We do not expect adoption figures to move much from here; the next national physician figure will tell whether 91% is a plateau.

Patient access in hospitals: near universal, mostly through standards-based APIs

ONC's hospital series shows patient access functions arriving in waves. Patients could view their record at 24% of hospitals in 2012 and at 99% in 2024 [4]. API-driven app access, first measured in 2017 at 39%, reached 86% in 2022 and 87% in 2024 [4]. Viewing clinical notes rose from 57% in 2018 to 96% in 2024 [4]. Two functions lag: transmitting information in a structured format has stayed at 84% to 85% since 2022, and 65% of hospitals let patients submit patient-generated data in 2024 [4].

The 2024 data brief splits API access by type. 71% of hospitals used a standards-based API for patient access and 16% a non-standards-based API [5]. Use of standards-based APIs differs by hospital: 74% for hospitals on one of the three market-leading EHRs against 48% for other EHRs, 75% for medium and large hospitals against 67% for small ones, and 77% for system-affiliated hospitals against 57% for independent ones [5]. ONC notes that since January 1, 2023 all certified EHR users have been required to have standardized FHIR APIs for patient and population services available [5].

Data flowing between EHRs and other software mostly does not use standards. 91% of hospitals integrated third-party data into the EHR for at least one clinical purpose in 2024, but 52% did so through a standards-based API; for administrative purposes, which include prior authorization, the figures were 87% and 35% [5].

Moraph's view

The patient-facing API mandate worked where it was a certification requirement: standards-based access is the norm. The gap by EHR and ownership suggests that independent hospitals on smaller EHRs depend on their developer's roadmap more than on their own choices. The same split is likely to show up when payers' Provider Access and Prior Authorization APIs go live in 2027, because the brief already finds administrative data sharing the least standardized.

Patients: more access, more apps, little aggregation

HINTS shows use catching up with capability. The share of individuals offered and accessing an online medical record or portal rose from 25% in 2014 to 65% in 2024, and more than 3 in 4 were offered access in 2024 [6]. Among people who used a portal, 57% used an app to get to their records in 2024, up from 38% in 2020 [6]. Proxy or caregiver access rose from 24% in 2020 to 51% in 2024 [6].

Records remain scattered. 59% of individuals had more than one online record or portal in 2024, but 7% used an app that combines records from several portals, up from 2% in 2022 [6]. People encouraged by their provider to use their portal accessed it at 87%, against 57% of those not encouraged [6].

Moraph's view

The supply side is close to done; the demand side is where the next gains are. The low use of aggregating apps is the number to watch: it is the clearest public signal of whether API access turns into patients holding their own combined record, which is also the precondition for consumer AI tools that read a whole record rather than one portal.

Promoting Interoperability: scores near the ceiling

In the QPP files, rows reporting a CEHRT ID rose from 259,211 in 2023 to 268,652 in 2024, an increase of 3.6% [12] [13]. The mean unweighted Promoting Interoperability score on those rows rose from 95.86 to 97.03, and the share scoring 100 rose from 65.4% to 70.1%; the median was 100 in both years [12] [13]. CMS's own summary, which covers all MIPS eligible clinicians rather than these rows, describes unweighted Promoting Interoperability scores as consistent between 2023 and 2024 [16].

The patient access measure is almost universally passed. PI_PEA_1 earned more than zero points on 99.3% of rows with a CEHRT ID in both 2023 and 2024 [12] [13] [17]. Enabling Exchange Under TEFCA was reported on 1,614 rows in 2023 and 13,060 in 2024, 4.9% of reporting rows [12] [13] [18]. The HIE bi-directional exchange measure, PI_HIE_5, appeared on 81.2% of rows in 2024 against 78.7% in 2023 [12] [13].

Clinicians reporting for Promoting Interoperability in 2024 named 175 different developers of certified health IT, and 62% reported at least one product from a single developer [8].

Moraph's view

A measure that nearly everyone passes no longer tells CMS or the public much about patient access. The scores say clinicians have the capability and attest to it; they do not say how many patients used an app. The TEFCA measure is the one Promoting Interoperability number still moving fast, and it gives an independent read on TEFCA reach alongside the network's own counts.

Hospitals: certified products and developers

ONC counts 4,058 non-federal acute care hospitals reporting for Medicare Promoting Interoperability in 2024, naming 88 developers of certified health IT, and 48% reported at least one product from a single developer [7]. Our count of the posted linkage file, using the method above, gives 3,931 CCNs marked as meeting the criteria in 2024 against 3,866 in 2023 [10]. Those 3,931 hospitals named 82 distinct developers; the median hospital listed products from 2 developers, and half listed 2 or more [10]. Another 696 CCNs appear in the 2024 file without the Y flag [10]. We could not reproduce ONC's 4,058 from the posted file and report both, each with its definition [7] [10].

Moraph's view

Half of hospitals report certified products from more than one developer, so a hospital's certified technology is often a stack, not a single system. That matters for API work: what a patient or an app can reach depends on every certified product behind the API, and on whether each is kept current.

Payers: the Patient Access API data gap

Medicare Advantage organizations, Medicaid and CHIP programs and plans, and QHP issuers on the federally facilitated exchanges must run a Patient Access API with claims, encounter and clinical data from January 1, 2016 forward [19]. For Medicare Advantage the API has applied since January 1, 2021, and from January 1, 2027 it must also carry prior authorization information [20].

Since 2026 these payers must report two counts to CMS each year, starting with calendar year 2025: unique patients whose data were transferred to a health app they chose, and unique patients whose data were transferred more than once [19] [20]. Medicare Advantage organizations report at contract level by March 31 [20]; Medicaid managed care plans report to their state, which passes the data to CMS in its annual managed care program report [19]. In the final rule CMS said it does not plan to report the metrics at the state, plan or issuer level, but may reference or publish aggregated, de-identified data [21]. As checked on November 16, 2026, CMS's Patient Access API page describes the reporting but not any publication of results, and notes that a 2026 proposed rule, CMS-0062-P, would update the API usage metric requirements [19].

Moraph's view

The payer side is the largest blind spot in public interoperability data. Provider-side adoption is measured every year; payer API use is now measured too, but only CMS sees it. Until CMS publishes a national aggregate, nobody outside can test whether the Patient Access API is used at a scale that justifies the 2027 build-out.

What we left out

We did not count CHPL listings directly, for the reason given above. We did not use AHA survey figures that ONC has not published, or vendor and industry surveys. We did not convert Promoting Interoperability measure points into performance rates, which the public files do not carry.

What this means for you

  • Health system CIOs and interoperability leads: Treat API availability as done and measure use: ONC's data show API-driven patient app access at 87% of hospitals [4] but a 52% standards-based rate for clinical data from third-party tools and 35% for administrative data [5]. Inventory which integrations still run on proprietary interfaces before the 2027 payer APIs arrive.
  • Independent and smaller hospitals: Ask your EHR developer for its standards-based API roadmap in writing: standards-based patient access was 57% at independent hospitals against 77% at system-affiliated ones in 2024 [5].
  • Ambulatory practices reporting MIPS: Expect little score headroom in Promoting Interoperability: 70.1% of reporting rows scored 100 in 2024 [12]. Check whether the TEFCA measure, reported on 13,060 rows in 2024 [12], fits how your EHR already exchanges data.
  • Payers subject to the Patient Access API: Keep the two usage counts you report to CMS each year as internal performance data, not only a filing [20]: they are the only measure of app use you will have, since CMS does not plan plan-level public reporting [21].
  • Digital health and AI app developers: Plan for patients with several portals and few aggregating tools: 59% of individuals had more than one portal in 2024 and 7% used an app to combine them [6].

Moraph predictions

Every prediction is Moraph's view, not a fact. Each has a date, a confidence level (70% means we expect to be wrong about three times in ten) and a test anyone can apply. We score them on this page when the date comes and never remove one. How we make and score predictions.

  1. P1 By Confidence 70% Open

    CMS publishes a QPP Experience public use file for the 2025 performance year in which the mean unweighted Promoting Interoperability category score, across rows with a CEHRT ID, is at least 96.0.

    How we will check: Download the 2025 QPP Experience file from data.cms.gov, keep rows whose CEHRT ID is not blank, and average the 'promoting interoperability (pi) category score' column. Right if the file is published by the date and the mean is 96.0 or more; wrong if it is lower or the file is not published.

    Basis: The same calculation gives 95.86 for 2023 and 97.03 for 2024 [12] [13]. The 2024 file's catalog entry was modified on May 19, 2026 and the 2023 file's on July 21, 2025 [12] [13], so a 2025 file by September 2027 fits the pattern.

  2. P2 By Confidence 75% Open

    The QPP Experience public use file for the 2025 performance year shows more than 13,060 rows with a CEHRT ID that report PI_HIE_6, Enabling Exchange Under TEFCA.

    How we will check: In the 2025 QPP Experience file, count rows with a non-blank CEHRT ID where any 'pi measure id' column equals PI_HIE_6. Right if the count exceeds 13,060 and the file is published by the date.

    Basis: Rows reporting the measure rose from 1,614 in 2023 to 13,060 in 2024 [12] [13]; the measure is an alternative to the referral loop and bi-directional exchange measures [18].

  3. P3 By Confidence 55% Open

    ASTP/ONC publishes a 2025 figure, from the AHA IT Supplement, showing that at least 75% of non-federal acute care hospitals enabled patient access through a standards-based API.

    How we will check: Look for an ASTP/ONC quick-stat or data brief reporting the 2025 share of hospitals using a standards-based API for patient access. Right if published by the date and 75% or more.

    Basis: The 2024 figure was 71%, with ONC describing patient API access as increasingly standards-based [5]; the 2024 brief was first posted in February 2026 [5].

  4. P4 By Confidence 65% Open

    CMS does not publish a national aggregate of the Patient Access API usage counts (unique patients whose data were transferred to an app) for calendar year 2025 or 2026.

    How we will check: Search cms.gov and data.cms.gov for a published total or summary of the Patient Access API usage metrics payers report under 42 CFR 422.119(f) and the parallel rules. Right if none is published by the date.

    Basis: The rule requires reporting to CMS, not public reporting [20]; CMS said it may publish aggregated, de-identified data [21], and its Patient Access API page, last modified September 1, 2026, describes no publication [19].

Facts on this page link to their sources. Text marked "Moraph's view", and every prediction, is our own analysis, not a fact or advice. Methodology.

Spotted an error? Tell us and we will check it and correct the piece where needed.

  • Promoting Interoperability
  • certified EHR technology
  • Patient Access API
  • FHIR
  • CHPL
  • ASTP/ONC
  • CMS
  • QPP