Health App Statistics 2026
By Axis Intelligence Research
Co-author: Jennifer Miller (Digital Health) | Last updated: September 8, 2026 | License: CC BY 4.0
England runs the largest health app dataset any government publishes, and it contains a number nobody quotes. Axis Intelligence Research finds that 32.79% of all NHS App accounts ever created — 13,826,254 of them — recorded no login at all in the twelve months to July 2026. Registration is not adoption. The gap between the two is the whole category.
Quick Answer
42,170,269 NHS App accounts had been created in England by 31 July 2026, but only 28,344,015 distinct users logged in during the trailing twelve months — meaning Axis Intelligence Research calculates a health app dormancy rate of 32.79%. In the month of July 2026 alone, 15,409,367 distinct users generated 79,987,913 login sessions and 6,131,504 of them ordered a repeat prescription, per NHS England. Axis Intelligence Research scores the whole funnel at a Health App Activation Index (HAAI™) baseline reading of 35.7 out of 100. On the reimbursed side, Germany has issued more than 1 million prescriptions for digital health applications since September 2020, of which roughly 81% were ever activated by the patient.
Key Findings
- Axis Intelligence Research finds a health app dormancy rate of 32.79% in England as of 31 July 2026 — 13,826,254 of 42,170,269 registered NHS App accounts recorded no login during the preceding twelve months, calculated from NHS England’s July 2026 management information release.
- Axis Intelligence Research sets the Health App Activation Index (HAAI™) baseline at 35.7 for July 2026, scoring four stages — account activation, session intensity, message attention and clinical transaction — on the only national health app funnel published in full anywhere in the world.
- NHS App login sessions grew 42% year on year to 79,987,913 in July 2026 while new account registrations fell 6% to 281,792, per NHS England — the first published national evidence that health app growth has shifted from acquisition to intensity.
- 39.79% of NHS App monthly active users ordered a repeat prescription in July 2026, an Axis Intelligence Research calculation on NHS England data, making prescription reordering the single largest clinical transaction in consumer health software.
- Roughly 19% of Germany’s more than 1 million prescribed digital health applications were never activated by the patient, per the GKV-Spitzenverband figures reported in npj Digital Medicine — the sharpest published measure of leakage between a doctor prescribing an app and a patient opening it.
The Health App Activation Index (HAAI™) — July 2026 Baseline Reading
An original Axis Intelligence Research metric. This figure is not published in this form by any primary source.
Every published health app statistic measures one of two things: how many people downloaded something, or how many people say they use something. Neither survives contact with a product manager. Downloads count intent. Surveys count self-report. What decides whether a health app changes anything is a chain: does the account stay alive, does the person come back often enough for the app to see a pattern, does the app’s clinical communication get read, and does any of it end in an action a health system records.
The Health App Activation Index (HAAI™) scores that chain end to end. It is built on England’s NHS App because NHS England publishes every link of it — registrations, monthly uniques, sessions, message read rates and transactions — in a single monthly release. No commercial app store dataset and no consumer survey contains all four.
Formula:
HAAI = (0.30 x Activation) + (0.15 x Intensity) + (0.20 x Attention) + (0.35 x Transaction)
Weights, and why. Transaction carries 0.35 because a health app that produces no recorded clinical action produces no clinical outcome, however beautiful the retention curve. Activation carries 0.30 because a dormant account is a marketing number wearing a product costume. Attention carries 0.20 because a delivered message that nobody opens is a delivery statistic, not a communication. Intensity carries the lightest 0.15 deliberately: in this category frequency is a weak proxy for value — a patient who opens the app twice a month to reorder medication is better served than one who opens it daily and does nothing.
HAAI™ components, inputs and scores
| Stage | Definition | Input values | Score | Weight | Source |
|---|---|---|---|---|---|
| Activation | Monthly unique users as % of all accounts ever registered | 15,409,367 / 42,170,269 | 36.54 | 0.30 | NHS England, July 2026 |
| Intensity | Sessions per monthly user, scored against daily use (30/month = 100) | 79,987,913 / 15,409,367 = 5.19 | 17.30 | 0.15 | NHS England, July 2026 |
| Attention | % of delivered in-app messages read within 24 hours | 41 (published) | 41.00 | 0.20 | NHS England, July 2026 |
| Transaction | Users completing a clinical transaction as % of monthly users | 6,131,504 / 15,409,367 | 39.79 | 0.35 | NHS England, July 2026 |
| HAAI™ | Composite, 0–100 | 35.7 | 1.00 | Axis Intelligence Research |
Worked arithmetic: (0.30 × 36.540832) + (0.15 × 17.302877) + (0.20 × 41) + (0.35 × 39.790758) = 10.9622 + 2.5954 + 8.2000 + 13.9268 = 35.6844, rounded to 35.7
This is a baseline reading. It is the first computation of the index and describes no history; subsequent readings will be published against it as NHS England releases each month.
What the baseline says. A health app used by 15.4 million people in a single month, delivering 36.8 million messages and processing 7.6 million prescription orders, still scores 35.7 out of 100 — because two thirds of the accounts it has ever created are asleep and 59% of the messages it sends go unread within a day. Scale and activation are different products.
What would move it. Transaction is the lever, and the arithmetic is unforgiving: with Activation, Intensity and Attention held where they are, the Transaction score would have to reach 80.69 to lift the composite to 50. That is roughly four in five monthly users completing a recorded clinical action every month — plausible only if the app’s transaction set expands well beyond repeat prescriptions. Every 10-point gain in Transaction adds 3.5 to the composite; the same gain in Activation adds 3.0.
What the index does not capture. HAAI™ scores a single national health service app, so it measures a system where the transaction exists by design. It does not score consumer wellness apps, where no equivalent clinical transaction is available to complete, and it does not measure clinical outcomes — a completed prescription order is an action, not a result. Cross-country readings will follow only where a health service publishes the same four measures.
HAAI™ is an original metric of Axis Intelligence Research, published under CC BY 4.0. Cite: Axis Intelligence Research (2026). Health App Activation Index. https://axis-intelligence.com/health-app-statistics/
How Many People Actually Use Health Apps in 2026?
The honest answer is that the download figures everybody quotes are the least informative number in the category, and the most reliable national counts come from health services rather than app stores.
England publishes the deepest series. NHS England’s monthly NHS App management information records 42,170,269 accounts created since recording began, 28,344,015 distinct users logging in across the twelve months to 31 July 2026, and 15,409,367 in July alone. Those three numbers describe three different populations, and conflating them is how the category’s inflated user counts get built.
For consumer health apps outside a national health service, the best population estimate available from peer-reviewed work is European. A cross-sectional survey of 6,581 respondents across eight EU countries, published in the European Journal of Public Health, found 21.87% currently using health apps, with a further 42.71% interested but not using. Axis Intelligence Research calculates that as 1.95 interested non-users for every current user — the category’s addressable market is roughly triple its installed base, which is either the most encouraging or the most damning number here depending on how long it has looked that way.
Health app users: the counts that actually mean different things
| Measure | Value | Population | Period | Source |
|---|---|---|---|---|
| Accounts ever registered | 42,170,269 | NHS App, England | to 31 Jul 2026 | NHS England |
| Distinct users, trailing 12 months | 28,344,015 | NHS App, England | Aug 2025 – Jul 2026 | NHS England |
| Distinct users, single month | 15,409,367 | NHS App, England | Jul 2026 | NHS England |
| Dormant accounts | 13,826,254 | NHS App, England | to 31 Jul 2026 | Axis Intelligence Research |
| Currently using a health app | 21.87% | Adults, 8 EU countries | Apr 2023 | Causio et al., EJPH 2025 |
| Interested in future use | 42.71% | Adults, 8 EU countries | Apr 2023 | Causio et al., EJPH 2025 |
| Highest national use rate | 29.51% | Spain | Apr 2023 | Causio et al., EJPH 2025 |
| Lowest national use rate | 14.29% | France | Apr 2023 | Causio et al., EJPH 2025 |
Axis Intelligence Research does not blend these two series. NHS England counts logins to one app in one country; the EJPH survey counts self-reported use of any health app across eight. A merged figure would describe nothing that exists.
Jennifer Miller, Digital Health: The Spain-to-France spread is more than a two-fold difference in adoption inside a single regulatory bloc, with comparable smartphone penetration and comparable healthcare digitisation budgets. That rules out infrastructure as the explanation and points at what the apps are actually offering. Where a health app carries a transaction the patient needs — a prescription, an appointment, a result — use follows. Where it offers a dashboard, it competes with everything else on the phone, and loses.
Why Do So Many Health App Accounts Go Dormant?
Because registration is usually a one-time administrative act, and the category keeps reporting it as if it were a relationship.
The NHS App data lets this be measured precisely for the first time. Against 42,170,269 accounts ever created, 28,344,015 distinct users logged in over the trailing twelve months. Axis Intelligence Research calculates the difference — 13,826,254 accounts, or 32.79% of everything ever registered — as the health app dormancy rate.
Two further ratios sharpen the picture. Monthly uniques run at 54.37% of twelve-month uniques, so roughly half the people who use the app in a year use it in any given month. And sessions per monthly user sit at 5.19, which is closer to a weekly errand than a daily habit.
The registration trend is the leading indicator worth watching. New account registrations in July 2026 came in at 281,792, 6% below July 2025, while login sessions over the same comparison rose 42% and monthly unique users rose 26%. The acquisition phase is ending; the usage phase is accelerating. Both things are true at once, and reporting only the growth figure — as the category’s promotional material tends to — misses that the pool of new people to acquire is visibly draining.
The NHS App funnel, July 2026
| Measure | Value | Year-on-year | Source |
|---|---|---|---|
| New account registrations in month | 281,792 | −6% | NHS England |
| Login sessions in month | 79,987,913 | +42% | NHS England |
| Distinct users in month | 15,409,367 | +26% | NHS England |
| Distinct users, trailing 12 months | 28,344,015 | +19% | NHS England |
| Sessions per monthly user | 5.19 | — | Axis Intelligence Research |
| Monthly users as share of annual users | 54.37% | — | Axis Intelligence Research |
| Dormancy rate | 32.79% | — | Axis Intelligence Research |
What Do People Actually Do Inside a Health App?
They reorder medication. Overwhelmingly, and by a margin that should reorganise how the category thinks about feature roadmaps.
In July 2026, 6,131,504 distinct NHS App users placed 7,636,604 repeat prescription orders — 1.25 orders per transacting user, and 39.79% of everyone who opened the app that month. Both figures grew 27% year on year. No other function in the app comes close.
Compare that with the newest feature in the same release. Prescription tracking had reached 3,210 pharmacy sites by the end of July 2026, covering 31% of pharmacies, and 495,651 distinct users viewed a tracking update that month. That is up 134% year on year — genuinely fast growth — and still only 3.22% of monthly active users. A feature can triple and remain a rounding error when its supply side is only a third built out. The 2,211 sites offering “Ready to Collect” notifications, 21% of pharmacies, are the constraint, not user appetite.
Messaging is the third leg and the most instructive. NHS England delivered 36,814,662 in-app messages in July 2026, up 60% year on year — 2.39 messages per monthly active user — of which 41% were read within 24 hours. Notification opt-in stands at 65% of all registered users but 88% of users who were active that month, which is the same dormancy story arriving from a different direction: the people who left also stopped listening.
What NHS App users did in July 2026
| Function | Distinct users | Volume | Reach of monthly users | Source |
|---|---|---|---|---|
| Repeat prescription ordering | 6,131,504 | 7,636,604 orders | 39.79% | NHS England / Axis calculation |
| Prescription tracking updates | 495,651 | — | 3.22% | NHS England / Axis calculation |
| In-app messages received | — | 36,814,662 delivered | 2.39 per user | NHS England / Axis calculation |
| Messages read within 24 hours | — | 41% of delivered | — | NHS England |
| Notification opt-in (registered) | — | 65% | — | NHS England |
| Notification opt-in (active users) | — | 88% | — | NHS England |
Jennifer Miller, Digital Health: Thirty-nine per cent of monthly users completing a prescription order and three per cent viewing a tracking update is not a story about user preference. It is a story about which feature has its supply side finished. Repeat prescribing works because every GP practice is connected; tracking does not because two thirds of pharmacies are not. Health app engagement is a function of institutional integration far more than of interface design, and every roadmap that inverts that ordering produces a dashboard nobody opens.
How Do App-Based Health Records Compare in the US?
The American series measures a different thing — how people reach their medical records rather than what they do in a single national app — and it shows apps winning decisively.
ASTP/ONC Data Brief 77, drawn from the 2024 Health Information National Trends Survey (7,278 complete responses, 27.31% response rate), reports that 57% of individuals who accessed their records used an app in 2024, up from 51% in 2022 and 38% in 2020. Web-only access fell from 60% to 42% across the same period. That is a 19 percentage point swing to apps in four years.
The complication is that the American health app experience is not one app. In 2024, 68% of individuals held a portal with their primary care provider, 40% with another provider, 30% with an insurer, 29% with a clinical laboratory, 24% with a pharmacy and 22% with a hospital. 59% held multiple portals, up from 50% in 2022, while only 7% used a portal-organising app to consolidate them. Axis Intelligence Research calculates that as 8.43 people living with fragmented records for every one who has consolidated them — and fragmentation is growing faster than consolidation.
Where Americans hold an online medical record, 2024
| Organisation | Share of individuals | 2022 | Source |
|---|---|---|---|
| Primary care provider | 68% | 63% | ASTP/ONC Data Brief 77 |
| Multiple portals | 59% | 50% | ASTP/ONC Data Brief 77 |
| Another health care provider | 40% | 32% | ASTP/ONC Data Brief 77 |
| Insurer | 30% | 29% | ASTP/ONC Data Brief 77 |
| Clinical laboratory | 29% | 26% | ASTP/ONC Data Brief 77 |
| Pharmacy | 24% | 23% | ASTP/ONC Data Brief 77 |
| Hospital | 22% | not asked | ASTP/ONC Data Brief 77 |
| No portal at all | 16% | 22% | ASTP/ONC Data Brief 77 |
| Used a portal-organising app | 7% | 2% | ASTP/ONC Data Brief 77 |
The structural contrast with England is the finding. One country has a single app with a dormancy problem; the other has six apps per person with a consolidation problem. Broader adoption and clinical data-sharing patterns across devices and monitoring programmes are covered in our connected health statistics analysis.
What Happens When a Doctor Prescribes an App?
Germany is the only country that has run this experiment at national scale for long enough to publish results, and the results are more sobering than the programme’s reputation suggests.
Under the 2019 Digital Healthcare Act, approved digital health applications — DiGA — can be prescribed by a physician and reimbursed in full by statutory health insurance. Figures from the GKV-Spitzenverband’s statutory report, as documented in a systematic review in npj Digital Medicine, show more than 1 million prescriptions issued by the end of 2024, cumulative spending of €234 million between September 2020 and December 2024, and an average 2024 price of €541 per prescription against a range from €119 to €2,077.
The number that matters is activation. Approximately 81% of prescriptions issued were activated by patients — meaning Axis Intelligence Research puts the non-activation rate at 19%. Roughly one in five app treatments a German doctor prescribed was never opened. No consumer app category publishes an equivalent figure, because no consumer app category can observe the moment of prescription separately from the moment of use.
The evidence layer underneath is thinner than the programme’s international reputation implies, and the review says so plainly. Of 23 approval studies publicly available as of March 2024, all were judged at high overall risk of bias; 21 of 23 — 91.30% — carried a high risk of bias specifically in outcome measurement, because outcomes were almost entirely patient-reported in trials where participants knew which arm they were in. Mean sample size was 321 participants and mean intervention duration 3.8 months. Dropout averaged 21.7% in intervention arms against 11.8% in controls, an Axis-calculated 9.9 percentage point gap in favour of leaving the app.
Attrition also shows up in the directory itself. As of 16 June 2025, 44 DiGA held permanent approval and 14 provisional, for 58 listed; 11 had been delisted, six of them — 54.55% — for failing to demonstrate a positive healthcare effect.
Germany’s prescribed health apps, by the numbers
| Metric | Value | Period | Source |
|---|---|---|---|
| Prescriptions issued | >1,000,000 | Sep 2020 – Dec 2024 | GKV-Spitzenverband |
| Activated by patients | ~81% | Sep 2020 – Dec 2024 | GKV-Spitzenverband |
| Never activated | ~19% | Sep 2020 – Dec 2024 | Axis Intelligence Research |
| Cumulative statutory insurance spend | €234 million | Sep 2020 – Dec 2024 | GKV-Spitzenverband |
| Average price per prescription | €541 | 2024 | GKV-Spitzenverband |
| Price range per use period | €119 – €2,077 | 2024 | GKV-Spitzenverband |
| DiGA listed (44 permanent + 14 provisional) | 58 | 16 Jun 2025 | BfArM directory |
| Delisted, of which for absent evidence | 11, of which 6 | to 16 Jun 2025 | BfArM directory |
| Approval studies at high risk of bias in outcome measurement | 21 of 23 | to 15 Mar 2024 | Sippli et al., npj Digital Medicine |
| Mean trial dropout, intervention vs control | 21.7% vs 11.8% | 2012–2022 studies | Sippli et al., npj Digital Medicine |
Jennifer Miller, Digital Health: A nineteen per cent non-activation rate and a ten-point excess dropout in the trials that won approval are the same signal measured twice. Adherence, not efficacy, is the binding constraint on prescribed software, and the approval framework barely measures it — mean intervention duration of 3.8 months in conditions that are mostly chronic tells you the evidence stops long before the therapy would need to. Germany deserves credit for publishing numbers that make its own programme look harder than the brochure. Nobody else is publishing them at all.
Are Health Apps Safe to Share Data With?
The regulatory answer in the United States is that a health app outside HIPAA is not outside the law, and the FTC has said so with money attached.
In its first enforcement action under the Health Breach Notification Rule, announced 1 February 2023, the Commission secured a $1.5 million civil penalty against GoodRx for failing to notify consumers of unauthorised disclosures of identifiable health information to advertising platforms. The FTC noted that more than 55 million consumers had visited or used the company’s website or mobile apps since January 2017. The order permanently barred the sharing of user health data with third parties for advertising. A second action, against the publisher of the ovulation-tracking app Premom, followed in 2023.
Users are not unaware of this. The EJPH survey found 72.34% of European respondents perceive data misuse as a risk of health app data sharing and 63.68% perceive hacking as a risk, yet 52.82% would still share app data with their provider or health portal — and only 25.48% would share with both public and private research institutions. The demand is specific: 52.59% want to know how their data will be used and who can access it before installing, and 44.02% want to know their right to delete it.
Health app privacy: enforcement and perception
| Measure | Value | Geography | Source |
|---|---|---|---|
| First HBNR civil penalty | $1.5 million | United States | FTC, Feb 2023 |
| Consumers who used the platform since Jan 2017 | >55 million | United States | FTC, Feb 2023 |
| Perceive data misuse as a risk | 72.34% | 8 EU countries | Causio et al., EJPH 2025 |
| Perceive data hacking as a risk | 63.68% | 8 EU countries | Causio et al., EJPH 2025 |
| Willing to share app data with provider | 52.82% | 8 EU countries | Causio et al., EJPH 2025 |
| Willing to share with public and private research | 25.48% | 8 EU countries | Causio et al., EJPH 2025 |
| Want data usage and access disclosed pre-install | 52.59% | 8 EU countries | Causio et al., EJPH 2025 |
| Want right-to-delete disclosed pre-install | 44.02% | 8 EU countries | Causio et al., EJPH 2025 |
Who Is Not Using Health Apps?
Older adults and people without tertiary education, and the effect sizes are large enough to be a design problem rather than a marketing one.
The EJPH regression puts health app use among Baby Boomers at an odds ratio of 0.54 against Generation Z, and the Silent Generation at 0.47 — roughly half the odds. Tertiary education runs the other way at 1.41. Geography compounds it: Eastern European respondents were more likely to use health apps than Central European ones (OR 1.47), as were Southern Europeans (OR 1.40), which cuts against the assumption that adoption tracks GDP.
The same older cohorts that use health apps least are the most demanding of disclosure before they will start — Baby Boomers were more than twice as likely as Generation Z to request detailed information before using an app (OR 2.11) and more than twice as likely to hold a distrustful attitude (OR 2.45). That is not technophobia. It is an informed refusal, and it responds to disclosure rather than to onboarding tutorials.
| Segment | Odds of current health app use vs reference | Reference | Source |
|---|---|---|---|
| Generation X | 0.74 | Generation Z | Causio et al., EJPH 2025 |
| Baby Boomers | 0.54 | Generation Z | Causio et al., EJPH 2025 |
| Silent Generation | 0.47 | Generation Z | Causio et al., EJPH 2025 |
| Tertiary education | 1.41 | Below tertiary | Causio et al., EJPH 2025 |
| Eastern Europe | 1.47 | Central Europe | Causio et al., EJPH 2025 |
| Southern Europe | 1.40 | Central Europe | Causio et al., EJPH 2025 |
Methodology
Data collection. Axis Intelligence Research assembled this dataset in September 2026 using a two-pass process. In the research pass, every source was fetched and read in full before any prose was written, and each quantitative claim was logged with organisation, document, URL, retrieval date and as-of date. The writing pass drew exclusively from that fact table. A separate adversarial verification pass matched every numeric token in the prose against a CSV row and re-ran every calculation.
Sources. Five primary sources, three of them official statistics or regulatory records:
- NHS England, NHS App Management Information — July 2026, published 28 August 2026; England, ICB and regional granularity; date range 1 December 2020 to 31 July 2026. https://digital.nhs.uk/data-and-information/publications/statistical/nhs-app-statistics/july-2026
- ASTP/ONC, Individuals’ Access and Use of Patient Portals and Smartphone Health Apps, 2024, Data Brief 77 (Richwine C.), based on HINTS 7, n=7,278, fielded March–September 2024, response rate 27.31%. https://www.healthit.gov/data/data-briefs/individuals-access-and-use-patient-portals-and-smartphone-health-apps-2024
- Sippli K, Deckert S, Schmitt J, Scheibe M. Healthcare effects and evidence robustness of reimbursable digital health applications in Germany: a systematic review. npj Digital Medicine 8:495, 1 August 2025. Carries the GKV-Spitzenverband DiGA-Bericht prescription, activation, expenditure and pricing figures and the BfArM directory counts. https://www.nature.com/articles/s41746-025-01879-6
- Causio FA, Beccia F, Tona DM, et al. Public perceptions and engagement in mHealth: a European survey on attitudes toward health apps use and data sharing. European Journal of Public Health 35(3):401–406, June 2025; n=6,581 across eight countries, fielded April 2023. https://academic.oup.com/eurpub/article/35/3/401/8116209
- Federal Trade Commission, FTC Enforcement Action to Bar GoodRx from Sharing Consumers’ Sensitive Health Info for Advertising, 1 February 2023. https://www.ftc.gov/news-events/news/press-releases/2023/02/ftc-enforcement-action-bar-goodrx-sharing-consumers-sensitive-health-info-advertising
Arithmetic verification. Every Axis-calculated figure was computed twice — once in floating point and once in exact rational arithmetic using Python’s fractions.Fraction — and both results agreed to nine decimal places before the figure entered prose. All formulas appear inline at the point of use and in the method_note column of the CSV.
Figures we chose not to publish. Global health app download totals and mHealth market-size estimates were excluded. The available figures come from commercial app-analytics and market-research firms whose category definitions differ by orders of magnitude — an app store’s “Health & Fitness” category and a regulator’s “digital health application” are not the same object, and no total that spans them is defensible. Two source figures are stated as floors rather than point values because the sources state them that way: Germany’s prescription count (“more than 1 million”) and the GoodRx consumer count (“more than 55 million”). Neither is used as a denominator anywhere on this page.
Scope of the analysis. The NHS App series is administrative rather than survey data, covering England only, and counts accounts rather than people — one person may hold one account, and a dormant account may belong to someone who has died or emigrated, so the dormancy rate is an upper bound on behavioural disengagement. The EJPH survey was fielded online in April 2023 and is subject to self-selection. The US series measures record access rather than health app use generally. HAAI™ is computed on a single national app and is not yet a cross-country index.
About This Dataset
Title: Axis Intelligence Health App Statistics 2026 Dataset
Publisher: Axis Intelligence Research
Authors: Axis Intelligence Research and Jennifer Miller (Digital Health)
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Coverage: England, United States, Germany and eight EU countries; 2020–2026
Proprietary metric: Health App Activation Index (HAAI™)
Rows: 94, each carrying value, unit, as-of date, geography, segment, source organisation, source document, source URL, retrieval date, primary-source flag, Axis-calculated flag, method note and licence.
Every number in this article exists as a row in the CSV at the same value and as-of date. The dataset is republished at Hugging Face, Kaggle and GitHub under the same licence. Updates are relevance-driven: this page is revised when NHS England publishes a monthly release that moves the HAAI™ reading, when ASTP/ONC publishes a new HINTS-based brief, or when the GKV-Spitzenverband publishes a new DiGA report.
Cite this research
APA: Axis Intelligence Research, & Miller, J. (2026, September 8). Health app statistics 2026: Users, downloads, dormancy and the apps doctors actually prescribe. Axis Intelligence. https://axis-intelligence.com/health-app-statistics/
MLA: Axis Intelligence Research and Jennifer Miller. “Health App Statistics 2026: Users, Downloads, Dormancy and the Apps Doctors Actually Prescribe.” Axis Intelligence, 8 Sept. 2026, axis-intelligence.com/health-app-statistics/.
Chicago: Axis Intelligence Research and Jennifer Miller. “Health App Statistics 2026.” Axis Intelligence, September 8, 2026. https://axis-intelligence.com/health-app-statistics/.
BibTeX:
@article{axis2026healthapp,
title={Health App Statistics 2026: Users, Downloads, Dormancy and the Apps Doctors Actually Prescribe},
author={{Axis Intelligence Research} and Miller, Jennifer},
journal={Axis Intelligence},
year={2026},
month={September},
day={8},
url={https://axis-intelligence.com/health-app-statistics/}
}
Frequently Asked Questions
What share of health app accounts are dormant?
Axis Intelligence Research calculates a health app dormancy rate of 32.79% for England’s NHS App as of 31 July 2026: 13,826,254 of 42,170,269 accounts ever registered recorded no login in the preceding twelve months. This is an upper bound on behavioural disengagement, since some dormant accounts belong to people who have died or left the country. No equivalent figure exists for consumer health apps, because app stores publish downloads rather than trailing-twelve-month unique logins.
Is health app growth slowing down?
Acquisition is, usage is not. NHS England recorded 281,792 new NHS App registrations in July 2026, 6% fewer than July 2025, while login sessions rose 42% to 79,987,913 and monthly unique users rose 26% to 15,409,367. The population left to acquire is shrinking while the people already there are using the app harder. Any health app forecast built on registration growth will read this as decline and get the direction of the business wrong.
How often do people open a health app?
About 5.19 times a month, on Axis Intelligence Research’s calculation from NHS England’s July 2026 release — 79,987,913 sessions across 15,409,367 monthly users. That is a weekly errand rather than a daily habit, and it is the pattern you would expect from an app whose dominant function is reordering medication rather than tracking anything continuously.
Do doctors prescribe health apps, and do patients use them?
Germany prescribes and reimburses them nationally. More than 1 million digital health application prescriptions were issued between September 2020 and December 2024 at a cumulative cost of €234 million to statutory health insurance, averaging €541 per prescription in 2024. Roughly 81% were activated by the patient, which means Axis Intelligence Research puts the non-activation rate at about 19% — close to one in five prescribed app treatments never opened.
How good is the evidence behind prescription health apps?
Weaker than the programme’s international reputation suggests. A systematic review in npj Digital Medicine assessed all 23 publicly available German approval studies as of March 2024 and rated every one at high overall risk of bias, with 21 of 23 at high risk specifically in outcome measurement because outcomes were patient-reported in unblinded trials. Mean sample size was 321 and mean intervention duration 3.8 months, in conditions that are largely chronic. Six of the 11 apps removed from the German directory were removed for failing to evidence a positive healthcare effect.
Do people abandon health apps more than other treatments?
The trial data suggests yes. Across German approval studies, dropout averaged 21.7% in the intervention arms against 11.8% in controls — an Axis-calculated 9.9 percentage point gap, with people more likely to leave the arm that received the app. That pattern in the trials, combined with the 19% non-activation rate in real-world prescribing, points at adherence rather than efficacy as the binding constraint on prescribed software.
Are people using apps or websites to reach their medical records?
Apps, increasingly. ASTP/ONC Data Brief 77 reports 57% of US individuals who accessed their records in 2024 did so through an app, up from 51% in 2022 and 38% in 2020, while web-only access fell from 60% to 42%. The complication is fragmentation: 59% held multiple portals in 2024 against only 7% using a portal-organising app, which Axis Intelligence Research calculates as 8.43 fragmented users for every consolidated one.
Can a health app legally sell my data if it isn’t covered by HIPAA?
Not without consequences. The FTC’s Health Breach Notification Rule covers health apps outside HIPAA, and in February 2023 the Commission secured a $1.5 million civil penalty against GoodRx — its first action under the rule — for failing to notify consumers of unauthorised disclosures of identifiable health information to advertising platforms, alongside a permanent bar on sharing user health data for advertising. A second action against the publisher of the Premom app followed the same year.
Which countries use health apps most?
Among the eight EU countries surveyed for the European Journal of Public Health in April 2023, Spain led at 29.51% of respondents currently using health apps and France trailed at 14.29%, against an eight-country average of 21.87%. Adoption did not track national wealth: Eastern European respondents were more likely to use health apps than Central European ones (odds ratio 1.47), as were Southern Europeans (1.40).
What is the Health App Activation Index?
The Health App Activation Index (HAAI™) is an original Axis Intelligence Research metric scoring the full chain a health app must complete to matter: account activation, session intensity, message attention and clinical transaction. It weights Transaction at 0.35, Activation at 0.30, Attention at 0.20 and Intensity at 0.15. Its baseline reading is 35.7 out of 100 for England’s NHS App in July 2026. Lifting the composite to 50 would require the Transaction score to reach 80.69, holding the other stages constant. Formula, inputs and dataset are published under CC BY 4.0.
Related Axis Intelligence Research: Connected Health Statistics · Digital Health Statistics · Telehealth Statistics · FDA AI Medical Device Statistics