Assessing Long-Term Care Needs in France: The GIR System, Eligibility and Access to Support

An older person in France may appear independent during a short conversation while struggling every morning to wash, dress, prepare food or remember medication. Another person may walk around their home safely but become disorientated outside it. A spouse may quietly compensate for increasing dependency until their own health begins to deteriorate. The purpose of assessment is to turn those different realities into a sufficiently accurate picture of loss of autonomy, eligibility and support need.

France does this through a combination of professional assessment, the national AGGIR framework and wider consideration of the person’s circumstances. Across the France Ageing, Long-Term Care & Community Support Knowledge Hub, this assessment architecture is fundamental because it determines who can access the allocation personnalisée d’autonomie (APA), how much support can be included in a plan, and when a change in circumstances should trigger reassessment.

The apparent simplicity of six GIR levels can be misleading. GIR is important, but it is not a complete care plan. At home, the departmental medico-social team must also consider how the person lives, what support already exists, the suitability of the home, the role of proches aidants and what interventions may preserve autonomy. In an EHPAD, GIR has a different operational purpose, informing dependency assessment within an institutional setting.

The central policy challenge is therefore not merely to classify people consistently. It is to ensure that assessment captures real life, distinguishes eligibility from broader need and leads to support that can actually be delivered.

AGGIR Provides a National Language for Loss of Autonomy

France’s principal framework for classifying loss of autonomy among older people is the grille AGGIR: Autonomie Gérontologie Groupe Iso-Ressources. It produces a GIR, or groupe iso-ressources, ranging from GIR 1, representing the greatest loss of autonomy, to GIR 6, representing the lowest level.

The framework creates a nationally recognisable method of determining whether a person falls within the dependency levels relevant to APA. People assessed as GIR 1, GIR 2, GIR 3 or GIR 4 can qualify for APA if the other eligibility conditions are met. GIR 5 and GIR 6 do not give entitlement to APA, although other forms of assistance may be available, including support from retirement funds or other local arrangements.

This national classification has obvious governance value. Without a common framework, two départements might define dependency using substantially different concepts. AGGIR establishes a shared baseline.

However, consistency does not mean that assessment is mechanical. The tool is completed by professionals and depends on observing and understanding how the person actually performs everyday activities.

The AGGIR Grid Looks Beyond Whether a Task Is Technically Possible

AGGIR contains 17 variables. Ten are described as discriminating variables because they are used in the calculation of GIR. Seven additional illustrative variables contribute useful information about the person’s wider situation but do not themselves determine the GIR.

The ten discriminating variables cover areas such as coherence, orientation, personal washing, dressing, eating, continence-related hygiene, transfers, movement inside and outside the home and communication at a distance.

The seven illustrative variables include management of personal affairs, cooking, domestic tasks, transport, shopping, following treatment and leisure activities.

Each activity is not judged through a simple yes-or-no question. The evaluator considers whether the person performs it spontaneously, completely, usually and correctly. The distinction matters. Someone may technically be able to dress themselves but only after repeated prompting. Another person may walk indoors but not recognise hazards or know where they are going.

This helps explain why person-centred planning for older people should not begin with a diagnosis alone. Two people with the same medical condition can have very different levels of functional autonomy and different support needs.

GIR 1 to GIR 6 Represent Different Patterns of Dependency

The GIR categories are designed to group people with broadly comparable levels of loss of autonomy, but they should not be interpreted as six identical care packages.

GIR 1 includes people with the most profound loss of autonomy, often requiring continuous presence because of very significant physical and cognitive impairment. GIR 2 includes people with severe dependency, including some who are largely confined to bed or a chair and others whose cognitive function is substantially impaired despite retaining mobility.

GIR 3 generally reflects people who retain some mental and locomotor autonomy but require assistance several times each day for personal activities. GIR 4 covers people who may retain considerable independence in some areas but require help with transfers, personal care or meals.

GIR 5 and GIR 6 describe lower levels of dependency and do not qualify for APA. Yet this does not mean that people in those groups have no needs. Someone in GIR 5 may still require domestic assistance or support with particular activities. Someone in GIR 6 may be autonomous in the core AGGIR activities while still experiencing isolation, housing difficulty or emerging frailty.

This distinction is important because eligibility thresholds can create a false binary between “dependent” and “independent”. In practice, ageing occurs along a continuum.

Eligibility for APA Requires More Than a GIR Score

GIR is central to APA eligibility, but it is not the only requirement. The person must normally be aged 60 or over, live in France on a stable and regular basis and meet the relevant loss-of-autonomy threshold.

APA is not restricted to people below a fixed income threshold. Resources affect the person’s financial contribution rather than removing eligibility altogether. This separates the assessment of need from the calculation of what the person contributes towards the support plan.

At home, the application is made through the département, commonly using the national application route for aides à l’autonomie à domicile. Once a complete application has been received, the departmental process moves from administrative eligibility into professional assessment.

The person’s needs should therefore be understood through two linked decisions:

• whether their circumstances and GIR make them eligible for APA;
• what support should be included in the personalised plan if they are eligible;
• what financial contribution applies according to their resources; and
• whether the proposed services are actually available in the local area.

The last point is not part of the GIR calculation, but it is critical to effective access.

Assessment at Home Is a Multidimensional Exercise

For APA at home, a professional from the département’s équipe médico-sociale arranges a visit. The purpose is broader than completing AGGIR.

The French system uses a multidimensional assessment reference framework covering the situation and needs of the older person and their close carers. This means the assessment should take account of daily functioning, living conditions, existing support, the home environment, preferences and the role of the person’s entourage.

That broader view is essential. A GIR 4 individual living with a capable spouse in an adapted flat may have a very different practical situation from a GIR 4 individual living alone in an isolated house with stairs, no nearby family and limited transport.

The functional classification can be similar while the plan required to maintain autonomy is very different.

This is why support planning and review should be treated as a distinct stage after classification. GIR establishes one part of the evidence; professional judgement converts that evidence into a practical response.

Operational Scenario: The Same GIR Does Not Mean the Same Plan

Two women in their early eighties are both assessed as GIR 4.

The first lives with her husband in an accessible apartment close to shops, a pharmacy and public transport. He prepares most meals and can assist with some domestic tasks, although he cannot safely help with bathing or transfers.

The second woman lives alone in a rural property. Her daughter lives more than an hour away. The bathroom is upstairs, she has stopped driving and she has difficulty preparing meals safely.

The GIR classification may be identical because their functional limitations fall within the same broad category. Their support needs are not.

The first plan may concentrate on personal assistance and maintaining her husband’s capacity to continue supporting her without injury or exhaustion. The second may require a broader combination of personal assistance, meal support, equipment, home adaptation, tele-assistance and coordination with local services.

A purely score-led system could overlook these differences. A multidimensional approach recognises that autonomy is shaped by the interaction between the person and their environment.

This is also why organisations examining comparable assessment models can use the Positive Risk-Taking Planner to structure thinking about independence, proportionate support and risk. It is not a French assessment instrument and does not determine GIR or APA eligibility, but it can help leaders examine whether support protects people without unnecessarily removing autonomy.

The Plan d’Aide Turns Assessment Into an Operational Offer

If the person is assessed as GIR 1 to GIR 4 and meets the other APA conditions, the département proposes a plan d’aide. This sets out the support that APA can finance and the associated cost.

The plan can include home assistance, meal delivery, equipment, tele-assistance, temporary forms of support and other interventions that help the person remain at home. The precise content depends on assessed need and the applicable rules.

Importantly, the person is not simply presented with an immutable administrative decision. The proposal can be considered, accepted or challenged within the process, and a modified proposal can be requested.

This introduces an element of choice and negotiation into the pathway. Assessment is therefore not simply something done to the person. It should create a basis for discussion about how they want to live and what support is proportionate.

The strongest plans also distinguish between doing things for somebody and enabling them to retain capability. Assistance with dressing, for example, may be necessary, but doing every part of the task automatically could accelerate dependency if the person can safely continue performing some elements independently.

The Assessor Must Also See the Carer

France’s multidimensional assessment framework explicitly recognises proches aidants. This matters because the sustainability of home support can depend heavily on what family and other close carers contribute.

An older person may appear to manage because their spouse prepares every meal, organises medication, accompanies them outside and responds during the night. Without understanding that support, an assessor could significantly underestimate the underlying level of dependency.

Carer assessment is therefore both a person-centred and a system-governance issue. The assessor needs to understand whether the carer’s contribution is sustainable, whether they require information or respite and whether they can realistically continue providing the same level of assistance.

This connects with family partnership and carer support. Recognising family contribution should increase support for carers, not legitimise unlimited reliance on them.

Operational Scenario: A GIR Assessment Changes When the Carer Is Made Visible

An 87-year-old man with cognitive impairment lives with his wife. During the assessment visit he is neatly dressed, walks around the home and answers several questions confidently. A superficial view might suggest that he needs only modest assistance.

Further discussion reveals that his wife selects his clothes, prompts every stage of dressing, prepares all food, supervises medication, prevents him from leaving the house alone and wakes several times each night because he becomes disorientated.

The assessment now looks different. His observable presentation depended on substantial invisible support.

The assessor also learns that his wife has arthritis and has recently missed her own medical appointments because she cannot leave him unsupervised. The question is no longer only which GIR best reflects his functional autonomy. The plan also needs to consider whether the current arrangement can continue safely.

If his wife is regarded simply as available household capacity, risk remains hidden. If her role is formally recognised, respite, additional home support or other interventions can be considered before the arrangement breaks down.

This is a central feature of credible assessment: understanding what enables the current level of functioning, not simply observing the person at one moment in time.

GIR Is a Decision Tool, Not a Complete Description of a Person

Any classification system risks becoming more powerful than intended. GIR is useful because it creates consistency, but it compresses a complex life into one of six groups.

It does not by itself capture every factor that matters to quality of life. Loneliness, anxiety, housing insecurity, cultural identity, grief, transport problems or digital exclusion may materially affect a person without altering the GIR calculation directly.

Nor does GIR provide a detailed clinical assessment. Medical diagnosis, rehabilitation need and nursing requirements remain matters for relevant health professionals.

The strongest use of AGGIR is therefore disciplined rather than expansive: use it for what it is designed to do, then combine it with wider assessment evidence.

This is particularly important in dementia. Cognitive impairment may fluctuate, be masked during a brief encounter or affect safety in ways that are not obvious from mobility alone. Assessment must consider the person’s actual ability to act safely and consistently rather than relying only on physical capability.

Assessment in an EHPAD Works Differently

When an older person lives in an EHPAD or an unité de soins de longue durée, GIR assessment is undertaken within the institutional setting rather than through the departmental home-visit process.

The médecin coordonnateur, working with the care team, evaluates the resident using AGGIR. The assessment is generally undertaken after admission once the team has enough information to understand the resident’s everyday functioning, and it can be revised if their situation changes.

In most of France, the resident’s GIR continues to influence the dependency tariff and APA in establishment. However, the position is different in the 23 départements participating in the EHPAD financing experiment introduced from July 2025. In those territories, the separate APA-in-establishment mechanism has been removed as part of the experiment combining care and dependency funding.

This distinction illustrates an important principle within the French system: the assessment method can remain nationally recognisable even when the financing consequences differ according to an authorised territorial experiment.

GIR also contributes to broader EHPAD resource analysis through measures such as the GIR moyen pondéré, which reflects the overall dependency profile of residents. At organisational level, this matters because increasing resident dependency should influence staffing, clinical oversight, equipment and operational planning.

Assessment Must Be Repeated When Need Changes

Loss of autonomy is not static. Some older people deteriorate gradually; others experience abrupt change following infection, stroke, a fall or hospital admission. Some recover function after rehabilitation. Carer circumstances can also change even when the older person’s condition remains broadly stable.

The French APA framework therefore allows reassessment when circumstances evolve. At home, the person or their representative can request a review of needs and the plan. In an establishment, the resident’s GIR can also be reassessed as their condition changes.

The operational challenge lies in identifying change early enough. A review system that depends solely on scheduled administrative cycles may miss deterioration occurring between formal assessments.

Home-care workers, nurses, family members, GPs and other professionals may notice emerging signs first: repeated falls, missed medication, new confusion, weight loss, reduced mobility or escalating carer exhaustion.

This is why decision-making and escalation need to connect frontline observation with the departmental review process. Information about changing need has limited value if nobody knows who should act on it.

Operational Scenario: Hospital Discharge Makes the Previous GIR Obsolete

An 81-year-old man receiving modest support at home is admitted to hospital after pneumonia. Before admission he walked independently and required help mainly with domestic activities. Following several weeks in hospital, he is weaker, needs assistance with transfers and is anxious about falling.

His existing support arrangement reflected his previous level of functioning. Simply restarting the same plan after discharge would ignore the change.

The hospital team needs to consider what is required for safe transition, while community professionals and the département may need to reassess his longer-term needs. Rehabilitation is also important because some of the deterioration may be reversible.

A poor response would assume that a higher level of dependency is permanent and immediately design indefinite support around it. An equally poor response would assume he will recover and provide insufficient help.

The stronger approach combines temporary support, rehabilitation, risk management and reassessment. The person can receive enough assistance to return home safely while retaining the opportunity to regain function.

This aligns with hospital discharge and admission avoidance for older people: assessment should follow the person’s actual trajectory rather than treating a pre-hospital support plan as permanently valid.

GIR 5 and GIR 6 Expose the Boundary Between Eligibility and Prevention

One of the most important policy boundaries in the French system sits between GIR 4 and GIR 5. People in GIR 5 or GIR 6 do not qualify for APA, yet they may still need help.

Retirement funds can offer support for eligible people with lower levels of dependency, and départements, municipalities or other organisations may provide domestic assistance, prevention and community services under different rules.

This boundary matters because people in GIR 5 may be at a stage where relatively modest intervention has significant preventive value. Help with domestic tasks, transport, nutrition, falls prevention or home adaptation may preserve independence and delay progression towards greater dependency.

Assessment should therefore avoid turning an APA refusal into a dead end. The practical question is not merely “does this person qualify for APA?” but “if not, what route is appropriate?”

The Service public départemental de l’autonomie is particularly relevant here because one of its core ambitions is to reduce the experience of people being passed between institutional doors. A person whose needs fall outside one benefit should still be oriented towards other relevant support.

Access Depends on Understanding the System

Formal eligibility is meaningful only if people can navigate the application process. Older people and carers may approach the system following bereavement, hospitalisation, cognitive decline or crisis. Administrative complexity therefore has direct consequences for equity.

Applications increasingly include digital routes, but postal and human support remain important. Digital transformation can simplify information exchange and reduce repeated administrative work, yet it can also disadvantage people who lack internet access, confidence or equipment.

This is why digitalisation should align with digital inclusion and reducing exclusion. An efficient online form is not an accessible system if the people most likely to need support cannot use it.

The SPDA provides an important strategic response by seeking clearer information, orientation and coordinated pathways across territories. Its success should ultimately be measured through user experience as well as organisational implementation.

SI-APA Could Strengthen Consistency and Visibility

The development of SI-APA is one of the most important digital reforms connected with assessment. The intention is to equip departmental teams managing APA at home with a more common information system and enable older people to make applications through a more consistent digital environment.

The ambition extends beyond administrative efficiency. A common information architecture can support greater equity by reducing unjustified differences in practice and making national and departmental patterns more visible.

Better data could allow leaders to examine application volumes, processing times, GIR distributions, support plans, reassessment activity and whether services authorised through those plans are actually delivered.

This creates a potentially valuable relationship between data quality, metrics and performance dashboards and assessment governance. Variation is not automatically problematic, but it should be explainable.

Organisations considering comparable digital assessment systems can use the Digital Transformation Readiness Assessment to examine governance, infrastructure, cyber resilience, workforce adoption and implementation capacity. It does not replicate SI-APA or French data requirements, but it highlights why digital transformation requires organisational readiness as well as technology.

Territorial Variation Is Most Important When It Changes Outcomes

France’s départements operate within national legislation but differ in population, geography, provider supply, workforce and administrative infrastructure. Some variation in assessment practice and support planning is therefore inevitable.

The important governance question is whether variation reflects legitimate local context or inconsistent treatment.

A rural département may need to construct a support plan differently because services are more geographically dispersed. That does not automatically mean the assessment is inequitable. But if two people with very similar needs receive materially different levels of effective support simply because one territory has poorer processes or persistent capacity shortages, national equity becomes harder to defend.

The CNSA’s role in harmonising practice and strengthening data is important for this reason. National oversight should not remove professional judgement; it should create sufficient visibility to identify unexplained differences.

Organisations examining similar assurance challenges can use the Quality Dashboard Builder to structure measures across access, timeliness, capacity, review and outcomes. In France, any such approach would remain supplementary to the country’s own statutory and departmental systems.

Assessment Quality Depends on Workforce Capability

An assessment framework is only as reliable as the professionals applying it. Evaluators need to understand AGGIR, conduct sensitive conversations, recognise cognitive and functional variation and translate findings into a realistic plan.

This requires more than technical training. Older people may minimise difficulties because they fear losing independence. Family members may disagree about the level of need. Cognitive impairment may affect insight. Cultural expectations can influence what people are comfortable disclosing. Homes can reveal risks that would never appear on a written form.

Professional judgement therefore remains central even within a standardised framework.

The same is true in EHPADs. Consistent GIR assessment affects not only individual classification but aggregate measures of resident dependency. Inaccurate assessment can distort the relationship between need, staffing and resources.

This connects with wider workforce capability and practice competence in older people’s services. Assessment is not an administrative sideline; it is a skilled intervention that shapes access to substantial public support.

Assessment Should Preserve Autonomy Rather Than Reward Dependency

A dependency-based benefit inevitably creates a conceptual risk: the more dependent somebody appears, the greater the potential support entitlement.

That does not mean the French system encourages dependency, but it does make outcome-focused practice important. Support should respond fully to genuine need while still preserving the person’s remaining abilities.

An individual who can safely prepare part of a meal, wash their face or choose their clothing should not automatically lose those activities because support is available. Conversely, promoting independence should never become a pretext for withholding assistance that is genuinely required.

The right balance is proportionate support. Assessment should identify what the person cannot do, what they can still do, what they could regain and which risks can be managed without unnecessarily restricting choice.

This aligns with positive risk-taking and risk enablement. Autonomy means accepting that ordinary life involves some risk while distinguishing that from avoidable harm.

Operational Scenario: Safety and Autonomy Pull in Different Directions

A woman with mild cognitive impairment lives alone and wants to continue preparing her evening meal. Her daughter believes this is unsafe after finding a pan burnt on the hob.

A simplistic assessment could produce one of two extremes: conclude that she can still cook because she physically performs the task, or conclude that cooking should stop entirely because an incident has occurred.

A stronger assessment explores the context. Was the incident isolated? Does she reliably turn appliances off? Does she recognise danger? Could equipment, meal preparation support or a different cooking method reduce risk? What matters to her about continuing the activity?

The plan may ultimately include assistance at particular times while preserving some involvement in food preparation. Technology may support safety, but only if it is acceptable and proportionate.

The assessment has therefore done more than classify dependency. It has helped define the boundary between necessary support and unnecessary restriction.

This is important because older people’s rights are affected not only when support is absent but also when support becomes overly controlling.

Governance Should Follow the Journey From Application to Outcome

Assessment systems can appear successful if leaders look only at completed applications. A stronger governance model follows the entire pathway.

Relevant questions include whether people can find the service, whether assessments occur within appropriate timescales, whether GIR decisions are consistent, whether plans reflect multidimensional need, whether carers are considered, whether authorised services begin and whether reassessment occurs when circumstances change.

The evidence set should also include complaints, disputes, unsuccessful applications, service-capacity gaps and differences between authorised and delivered support.

This is where the Governance Maturity Assessment offers a useful general framework for organisations examining accountability and escalation. It cannot judge French APA decisions, but it can help leaders test whether information from individual cases reaches the level capable of correcting recurring system problems.

A département repeatedly seeing GIR 4 plans that cannot be delivered should not treat each case as unrelated. A pattern is operational intelligence.

International Learning From the French Assessment Model

France’s AGGIR and APA architecture reflects its own social-protection and territorial arrangements, so the model cannot be transplanted directly into countries using different eligibility systems.

Several principles are nevertheless internationally relevant.

First, a national functional framework can improve consistency while leaving room for professional judgement. The transferable lesson is not the specific GIR categories but the value of a shared language for dependency.

Second, classification and support planning should remain distinct. A score can inform entitlement, but real-life need is shaped by housing, carers, geography and personal goals.

Third, family support needs to be assessed rather than assumed. A person may appear more independent precisely because somebody else is compensating for their loss of autonomy.

Fourth, people below a formal eligibility threshold still need navigation and prevention. A good system does not make one benefit the only doorway to support.

Finally, assessment governance should measure implementation. The strongest evidence is not that an assessment occurred, but that it led to an appropriate response and was revisited when circumstances changed.

Conclusion

France’s GIR system provides a clear national method for classifying loss of autonomy, but the quality of long-term care assessment depends on what happens around that classification. AGGIR establishes whether an older person falls within GIR 1 to 4 for APA eligibility, yet the département’s wider assessment must still understand the person’s home, preferences, carers, risks, existing support and potential to retain or regain independence.

The distinction is critical. Two people with the same GIR can need very different support. A person classified as GIR 5 may be ineligible for APA while still requiring preventive assistance. A support plan can be correctly authorised but practically ineffective if no provider has capacity. A family carer can make dependency appear lower than it really is until their own capacity changes.

France’s strongest direction is therefore to treat assessment as a continuing pathway rather than a one-off classification event. The SPDA can make entry and navigation more coherent; SI-APA can improve consistency and data visibility; multidisciplinary assessment can make plans more realistic; and reassessment can respond when health, function or family circumstances change.

Ultimately, assessment succeeds when it does more than allocate a GIR. It should convert the lived reality of an older person into proportionate support, preserve remaining autonomy, recognise the contribution and limits of carers, and give territorial leaders enough evidence to see where formal eligibility is not becoming effective access.