Digital Care Applications in Germany: A Benefit Without a Product
Five years of DiPA and no permanent BfArM listing. Why the 2026 reform opens the route, while evidence requirements remain, and where digital care can create realistic value.
Digital care applications, DiPA, have been a statutory benefit since 2021. As of August 2026, not a single application was permanently listed in the BfArM register. This impulse examines what five years of DiPA reveal about the limits of digital solutions to the long-term care workforce gap.
Executive Summary
| Claim | Assessment |
|---|---|
| The entitlement exists, the products do not. | Digital care applications, DiPA, have been a statutory benefit under Section 40a SGB XI since 2021. As of August 2026, not a single application has been permanently listed in the BfArM register. Lindera was rejected in 2024 because the submitted study could not isolate the effect of the app from care services the participants were already receiving. |
| The 2026 reform is a candid admission of failure. | The BEEP Act, promulgated on 29 December 2025 and in force since 1 January 2026, removed the prior necessity assessment, introduced a twelve-month provisional listing against an evaluation concept, permitted parallel price negotiation, extended the definition to products supporting informal carers, and restructured the monthly budget to 40 euros for the application plus up to 30 euros for supporting services. |
| The fiscal envelope is small by construction. | At 25% uptake among the 5.2 million people receiving home care benefits, the application component costs the long-term care funds approximately 624 million euros per year, equivalent to 0.85% of 2025 expenditure. |
| The break-even threshold is demanding. | For a cohort of 100,000 users drawing the full entitlement, the benefit pays for itself only if it keeps roughly 10 in every 100 users out of residential care for a full year at care level 3, or approximately 7 in 100 at care level 4. |
| The workforce gap is an order of magnitude larger than the instrument. | Germany requires 2.15 million nursing staff by 2049 against a projected supply of 1.46 to 1.87 million, a shortfall of 13.0 to 32.1% of required capacity. DiPA address a setting in which no professional full-time equivalents are counted at all. |
| The realistic case for DiPA is narrower and still worth making. | Their defensible contribution lies in delaying escalation, structuring informal care and stabilising carer capacity, not in substituting professional time. |
1. The instrument and what it actually contains
DiPA were created by the Digital Healthcare and Care Modernisation Act, DVPMG, in 2021 as the long-term care counterpart to digital health applications, DiGA. The construction differs in three ways that matter.
| Dimension | DiGA, Section 33a SGB V | DiPA, Section 40a SGB XI |
|---|---|---|
| Payer | Statutory health insurance | Social long-term care insurance, SPV |
| Trigger | Physician or psychotherapist prescription, or fund approval | Application by the insured person to the care fund |
| Objective | Detection, monitoring, treatment or alleviation of disease | Reduction of impairments, prevention of aggravation of care dependency, relief of caring relatives |
| Regulatory status | Medical device, class I or IIa | Not necessarily a medical device |
| Monthly reimbursement | Negotiated price, average 227 euros after negotiation | Up to 40 euros, plus up to 30 euros for supporting services, Section 40b SGB XI |
| Register | BfArM DiGA directory, 74 applications listed to date | BfArM DiPA directory, no permanent listing to date |
The asymmetry in the final two rows is the entire story. A DiGA commands an average negotiated price of 227 euros per month; a DiPA is capped at 40 euros for the software itself. The regulatory burden, until 2026, was broadly comparable: evidence of a nursing benefit, data protection conformity, quality and interoperability requirements, and a necessity assessment by the BfArM.
Every person with care level 1 to 5 cared for at home is eligible. That is the population the arithmetic below rests on.
2. The record: five years, zero listings
| Date | Event |
|---|---|
| June 2021 | DVPMG creates Sections 40a and 40b SGB XI. |
| 2022 | Digital Care Applications Ordinance, DiPAV, establishes the listing procedure. |
| 2024 | Lindera application rejected by the BfArM; the submitted study could not isolate the app's effect from existing care services. |
| 22 / 29 December 2025 | BEEP Act enacted and promulgated, BGBl. I 2025 No. 371. |
| 1 January 2026 | Reformed DiPA framework enters into force. |
| August 2026 | Register still contains no permanently listed application. |
Five years of statutory entitlement have produced a benefit no insured person can claim. This is not slow uptake. It is a benefit whose economics never supported the evidence burden attached to it.
The DiGA comparison sharpens the point. Between September 2020 and December 2025, statutory health insurance spent 401 million euros on 1.6 million digital health application prescriptions, with activations rising from 65,000 in 2021 to 695,000 in 2025. Seventy-four applications have been listed, 58 remain in the catalogue, and 16 were removed for insufficient evidence. The DiGA route is criticised, with justification, for weak evidence and high prices. But it produced a market. The DiPA route produced nothing.
3. What the 2026 reform changed
The BEEP Act addressed each of the identified barriers. Whether it addressed them sufficiently is the open question.
| Element | Until 31 Dec 2025 | From 1 Jan 2026 |
|---|---|---|
| Necessity assessment by the BfArM | Required | Removed |
| Route to listing | Full evidence of nursing benefit required upfront | Provisional listing for up to twelve months against a scientific evaluation concept |
| Price negotiation | Sequential, after listing | Parallel to the listing procedure |
| Monthly budget | 50 euros combined, dynamically adjusted to approximately 53 euros | 40 euros for the application plus up to 30 euros for supporting services |
| Definition of a qualifying application | Must reduce impairments of the care-dependent person | Extended to applications supporting caring relatives or volunteers, including those that improve only the carer's situation. The statutory claim itself remains with the care-dependent person, Section 40a(1a) SGB XI. |
| Evidence standard | Benefit to the care-dependent person | Demonstrated relief of carers accepted as an equivalent alternative |
Two changes are substantive. Provisional twelve-month listing converts an unfinanceable upfront evidence requirement into a staged one, mirroring the mechanism that made DiGA viable. Recognising carer relief as an equivalent endpoint aligns the evidence standard with what these products actually do.
The budget change is weaker than it looks. The application component moved from roughly 53 euros combined to 40 euros dedicated, with the 30 euro supporting-services component sitting alongside it and requiring an ambulatory provider to deliver something. For a manufacturer, addressable revenue per user per month is 40 euros, not 70.
4. Calculation 1: the fiscal envelope
The relevant population is the 5.2 million people receiving long-term care benefits in a home setting in 2025. All figures below assume the 2026 amounts and are stated per year.
Assumptions: 5.2 million eligible; 40 euros per month for the application, 480 euros per year, and 70 euros per month for the full package, 840 euros per year; SPV expenditure 2025 of 73.82 billion euros; SPV deficit projected for 2026 of 3.5 billion euros. Uptake is applied uniformly.
| Uptake | Users | Application only, EUR m | Share of SPV expenditure | Full package, EUR m | Share of SPV expenditure |
|---|---|---|---|---|---|
| 5 % | 260,000 | 125 | 0.17 % | 218 | 0.30 % |
| 10 % | 520,000 | 250 | 0.34 % | 437 | 0.59 % |
| 25 % | 1,300,000 | 624 | 0.85 % | 1,092 | 1.48 % |
| 50 % | 2,600,000 | 1,248 | 1.69 % | 2,184 | 2.96 % |
| 100 % | 5,200,000 | 2,496 | 3.38 % | 4,368 | 5.92 % |
Two readings follow, and they point in opposite directions. For the payer, the instrument is cheap. Even at an ambitious 25% uptake it consumes 0.85% of annual expenditure. Set against a fund that ran a 0.5 billion euro deficit in 2025, faces a projected 3.5 billion euro deficit in 2026, and whose average contribution rate is projected to rise from 3.7 to 5.2% by 2040, DiPA is a rounding error. It cannot destabilise the fund, and it cannot rescue it.
For the manufacturer, the same numbers are the problem. A 40 euro ceiling across a realistically addressable share of the market must fund clinical evaluation, conformity work, information security, integration, distribution to an elderly and digitally heterogeneous population, and ongoing support. That is the calculation behind five years of empty register.
5. Calculation 2: what a DiPA would have to achieve to pay for itself
The economic case for DiPA rests on a single mechanism: keeping people in home care longer and out of residential care. This can be quantified precisely, because the SPV pays defined amounts in both settings.
Assumptions: cohort of 100,000 users drawing the full 840 euro annual entitlement, giving an annual cost of 84.0 million euros. Savings are calculated as the difference between the residential care amount under Section 43 SGB XI and the care allowance under Section 37 SGB XI. Section 43c surcharges are excluded, so the savings are understated and the break-even threshold is correspondingly conservative.
| Care level | Care allowance, EUR/month | Residential amount, EUR/month | Monthly difference | Person-years of residential care to be avoided | Share of cohort |
|---|---|---|---|---|---|
| 2 | 347 | 805 | 458 | 15,284 | 15.3 % |
| 3 | 599 | 1,319 | 720 | 9,722 | 9.7 % |
| 4 | 800 | 1,855 | 1,055 | 6,635 | 6.6 % |
| 5 | 990 | 2,096 | 1,106 | 6,329 | 6.3 % |
The entitlement is fiscally self-financing if, in a cohort of 100,000 users at care level 3, it keeps approximately 9,700 people out of residential care for a full year. That is roughly one in ten users. At care level 4 the threshold falls to about one in fifteen.
This is a demanding but not absurd bar, and it is the number every manufacturer, payer and policymaker in this field should be able to state. It reframes the question: not whether an application is pleasant to use, but whether it demonstrably changes transition rates into institutional care.
A caveat that cuts the other way. The comparison above uses the care allowance as the home-care baseline. If the person instead draws the full ambulatory care services budget under Section 36 SGB XI, home care is not cheaper for the SPV at all. At care level 3, the ambulatory budget of 1,497 euros exceeds the residential amount of 1,319 euros by 178 euros per month; at care level 5, 2,299 euros against 2,096 euros. Only once the Section 43c surcharges are added does residential care become clearly more expensive for the fund. The home-before-residential savings logic therefore holds firmly only against the care-allowance baseline, which is to say against the 3.1 million people cared for exclusively by relatives.
That is precisely the population DiPA are designed for, which is the strongest argument for the instrument. It is also the population least likely to have the digital confidence, device access and support structures these applications assume.
6. Calculation 3: the gap DiPA is asked to help close
| Metric | Value | Source basis |
|---|---|---|
| Nursing staff, 2019 baseline | 1,620,000 | Destatis nursing workforce projection |
| Nursing staff required, 2049 | 2,150,000 | +33 % against 2019 |
| Projected supply 2049, trend variant | 1,870,000 | Gap of 280,000 |
| Projected supply 2049, status quo variant | 1,460,000 | Gap of 690,000 |
| Gap as share of required capacity | 13.0 % to 32.1 % | Calculated |
| Care-dependent persons, 31 Dec 2023 | 5,690,000 | Of whom 4.9 m at home, 3.1 m by relatives only |
| Projected care-dependent persons, 2055 | 6,800,000 to 7,600,000 | Constant and rising care-rate scenarios |
Germany must raise effective care capacity by 13 to 32% by 2049 simply to stand still. That is the denominator against which any single instrument should be judged.
DiPA operate entirely outside the professional workforce. They add no full-time equivalents, do not shorten a shift, and do not reduce a residential facility's minimum staffing requirement. Their only route to the professional gap is indirect: preventing or delaying the escalation that turns an informally cared-for person into a consumer of professional capacity. Calculation 2 quantifies how much of that would be required.
7. The evidence problem
The mechanism DiPA rely on is carer relief. The best available synthesis of that mechanism is a meta-analysis of 14 randomised controlled trials with 1,010 informal caregivers of older adults. It found a pooled standardised mean difference on caregiver burden of −0.13 with a 95% confidence interval of −0.25 to −0.00, heterogeneity of 34% and moderate GRADE quality.
Three observations follow. The effect is real but small, at roughly two thirds of the threshold conventionally regarded as small. The upper bound of the confidence interval touches zero. Subgroup effects were larger where the care recipient had a defined chronic condition, at −0.43, which suggests indication-specific applications outperform generic ones.
This is not a reason to abandon the category, but a reason to be exact about the claim. An effect of this size is undetectable in an underpowered single-arm study, which is the class of evidence submitted and rejected so far. The BfArM's stated ground for the 2024 rejection, that the study could not isolate the application's effect from care services already being received, is the predictable consequence of trying to demonstrate a small effect without an adequate comparator.
The DiGA experience makes the same point from the other end: 16 of 74 listed applications were later removed for insufficient evidence, and list prices of 119 to 2,077 euros settled at an average negotiated 227 euros across the 40 applications with concluded price agreements. A provisional route lowers the entry barrier, not the eventual evidential burden. Anyone entering under the new twelve-month pathway should plan the confirmatory study from day one, powered for the effect size the literature actually supports.
8. Where DiPA can realistically contribute
The critical case above is not an argument against the instrument but against the claim usually attached to it. Stated honestly, the contribution is narrower and still valuable.
| Claim | Assessment | What would have to be shown |
|---|---|---|
| DiPA relieve the professional nursing shortage | Not supportable | Measurable reduction in professional care hours consumed, at population scale |
| DiPA delay transitions into residential care | Plausible, unproven | Transition rates in a controlled comparison; threshold per Calculation 2 |
| DiPA reduce informal carer burden | Supported, small effect | Already shown at SMD −0.13; larger for defined indications |
| DiPA improve structure and safety of informal care | Plausible, under-researched | Falls, medication errors, escalation events as endpoints |
| DiPA extend the duration of informal care capacity | The most valuable claim, least studied | Time to carer breakdown or withdrawal as a primary endpoint |
The final row deserves emphasis. Germany's care system rests on 3.1 million people cared for exclusively by relatives, and the binding constraint there is not the care-dependent person's function but the carer's endurance. An application that measurably postpones carer breakdown would be economically significant even with a modest effect on any clinical scale, because the counterfactual is the most expensive setting in the system. That endpoint is essentially absent from the current evidence base.
9. Implications for manufacturers
| Decision | Recommended position |
|---|---|
| Business model | Do not build a case on 40 euros per month alone. Treat the DiPA listing as a reimbursement anchor within a broader offering to care funds, insurers and providers, not as the revenue base. |
| Evidence strategy | Design the confirmatory study before applying for provisional listing, powered for a small effect, with an active or usual-care comparator that isolates the application from concurrent care services. |
| Primary endpoint | Prefer carer-side endpoints, burden, capacity and time to withdrawal, and transition to residential care, over functional scores of the care-dependent person. |
| Target population | Prioritise defined indications where the literature shows larger effects, over generic care-support propositions. |
| Distribution | Assume low digital confidence in the target population. Onboarding, device provision and human support are product components, not marketing. |
| Regulatory path | Clarify early whether the product is a medical device. A DiGA route may be economically superior for the same technology, at an average negotiated price roughly 5.7 times the DiPA ceiling. |
The last row is the most consequential and the least discussed. For the same software, the DiGA route offers an average negotiated 227 euros per month against a DiPA ceiling of 40 euros. Where an indication-based claim under SGB V is defensible, the economics point clearly in that direction, and the DiPA register will fill only with products for which that alternative does not exist.
Conclusion
Germany created a statutory entitlement to digital care applications in 2021 and, five years on, no one can claim it. The 2026 reform diagnosed the causes correctly and fixed most of them. Whether a 40 euro ceiling can fund the evidence the register still demands is the open question, and it will be answered by whichever manufacturer files first under the new pathway.
The larger point is proportion. Germany faces a shortfall of 280,000 to 690,000 nursing staff by 2049, equivalent to 13 to 32% of required capacity, against a fund already in deficit and an average contribution rate projected to rise from 3.7 to 5.2% by 2040. A benefit whose entire fiscal envelope reaches 5.9% of annual expenditure only at universal uptake is not a solution to that.
What DiPA can plausibly do is protect the informal care arrangements that carry 86% of the caseload, and the 54% cared for by relatives alone, making them more structured, safer and more durable. That claim is smaller, it is measurable, and it is worth several billion euros a year to the system if it holds. It has never been tested properly. The twelve-month provisional route is the first realistic opportunity to test it, and whether the sector uses that window for real evidence or repeats the underpowered studies behind the 2024 rejection will decide whether the register looks any different in 2031 than it does today.
Note on calculations
All figures are computed from the sources listed below. Assumptions are stated with each calculation and are reproduced here for transparency.
1. Eligible population: 5.2 million recipients of SPV benefits in a home-care setting, 2025. Uptake is applied uniformly across care levels; in practice uptake would skew towards lower care levels and younger cohorts, which would lower cost and lower effect.
2. Per-user cost: 40 euros per month, 480 euros per year, for the application; 70 euros per month, 840 euros per year, for the application plus supporting services. Both are ceilings, so all cost figures are upper bounds.
3. Break-even: annual cohort cost divided by the monthly difference between the residential care amount under Section 43 SGB XI and the care allowance under Section 37 SGB XI, converted to person-years. Section 43c surcharges are excluded, which understates the saving and makes the stated break-even share conservative. The calculation assumes the avoided residential episode would otherwise have occurred within the same year.
4. Workforce gap share: shortfall divided by the 2,150,000 staff projected to be required in 2049.
5. 2026 benefit amounts are unchanged from 2025; the next statutory dynamisation is scheduled for 1 January 2028.
Sources: [1] Bundesinstitut für Arzneimittel und Medizinprodukte: Digitale Pflegeanwendungen (DiPA), bfarm.de · [2] Gesetz zur Befugniserweiterung und Entbürokratisierung in der Pflege, enacted 22 December 2025, promulgated 29 December 2025, BGBl. I 2025 No. 371, recht.bund.de · [3] Statistisches Bundesamt: 5.7 million people in need of care at year end 2023, 18 December 2024, destatis.de · [4] Statistisches Bundesamt: At least 280,000 additional nursing staff needed by 2049, 24 January 2024, destatis.de · [5] Sachverständigenrat: Soziale Pflegeversicherung, chapter 4 of the 2026 spring report, sachverstaendigenrat-wirtschaft.de · [6] AOK Presse und Politik: Zahlen und Fakten zur Pflegeversicherung, aok.de · [7] Zhou Y. et al.: Technology-based interventions on burden of older adults' informal caregivers, BMC Geriatrics 2024; 24:398, DOI 10.1186/s12877-024-05018-w · [8] Deutsches Ärzteblatt: Kassen beziffern Kosten für Gesundheits-Apps auf bislang 400 Millionen Euro, 8 April 2026 · [9] QuickBird Medical: DiPA-Leitfaden, updated 20 August 2026 · [10] Diakonie Deutschland: Leistungen der vollstationären Pflege ab 01.01.2026; betanet: DiPA.
This impulse article reflects the state of discussion in September 2026. It does not constitute legal or reimbursement advice. Register status should be verified against the BfArM directory before any commercial decision.
Do you have a concrete project?
Talk to us about your project, no obligation, at eye level.
Schedule a conversation
