By Lena Enqvist*
In Sweden, mental illness has become the leading cause of sick leave, most often involving so-called common mental disorders such as depression, anxiety, and stress-related conditions.[1] Despite being widespread, these disorders pose particular challenges within the sickness insurance system, especially where medical and legal uncertainties intersect. One reason is that eligibility criteria for sickness benefits (sjukpenning) are built on the assumption that medical, social, and other factors affecting work capacity can be separated and clearly defined. In cases of mental illness, however, such distinctions are rarely straightforward, since causal links and boundaries are often blurred.
Medical and Legal Uncertainties
From a medical perspective, psychiatric disorders often lack the clear biomarkers and test results available in many somatic conditions. Diagnostics rest on symptom clusters rather than underlying mechanisms, and the same diagnosis can manifest very differently across individuals, sometimes making prognosis and assessment of work capacity a challenging task.[2]
From a legal perspective, difficulties stem from the fact that a diagnosis alone is insufficient for entitlement to sickness benefits in the Swedish Social Insurance Code. Coverage is broad, but it only applies where the illness leads to a reduction in work capacity of at least 25 percent. Benefits may then be granted in fixed steps of 25, 50, 75, or 100 percent.[3] These criteria assume that it is possible to separate the degree of capacity loss attributable to illness from labour market, social, or personal factors, and that both the extent and causation of the reduction can be determined. In psychiatric cases, however, that assumption is often fragile.
Disease and Reduced Work Capacity as Interlinked Criteria
In the eligibility assessment, the fundamental criteria of “disease” and “reduced work capacity” in the Swedish Social Insurance Code are thus central.[4] The concept of “disease” rests on a medical foundation, with possible extension to states that, in ordinary language, deviate from the normal life process.[5] Another settled principle is that the loss of working capacity is to be assessed individually, and not against a hypothetical healthy or advantaged person.[6] In psychiatric cases this is crucial as one person with coping strategies and social support may remain able to work, while another with the same diagnosis but lacking such resources may not. As will be seen, such variations expose the tension between the requirement of individual assessment and the tendency to subject work capacity assessments to greater standardisation.
Unlike “disease,” the legal notion of “work capacity” in the Social Security Code is not tied to an external scientific taxonomy.[7] Instead, it sets criteria that establish shifting reference points for assessment, which vary with the length of the individual’s sick leave. Without detailing these reference points, the group of work tasks that the individual’s working ability is assessed against broadens over time – from the person’s own job to the abstract labour market.[8] As a result, both the required degree of incapacity increases and the assessment becomes more abstract. This is most pronounced at the final stage, as a main rule after 180 days of sick leave, when capacity is tested against “normally occurring” work.[9] This category includes ordinary jobs where the worker’s capacity can be used fully or nearly fully, with normal performance demands and little or no accommodation for functional limitations or medical problems. A person deemed able to perform such work is not entitled to sickness benefit, regardless of job availability.[10] The Supreme Administrative Court has clarified that this assessment requires reality-based, case-by-case evaluations that take account of labour-market conditions and their development.[11] With regards to psychiatric diagnoses, however, a particularly difficult but still unresolved issue is the tolerance level for impairments of mental function – such as stress sensitivity, slower pace, irritability, or lack of initiative – that is compatible with such “normally occurring” work.
High Reliance on Extra-legal Tools for Work Ability Assessments
Although “disease” and “work capacity” are formally distinct criteria, they are closely intertwined – medically, since it is often difficult to disentangle influencing factors or to quantify the degree of reduced capacity, and legally, since the causation requirement ontologically links the disease to the loss of work capacity within the assessment framework. The Swedish Social Insurance Code, however, provides little guidance on how such assessments should be conducted or how causality should be established in individual cases. To compensate, the Agency has introduced extra-legal tools that supply greater detail and standardisation in work capacity assessments. While this may promote consistency, it simultaneously restricts individualised evaluation, creating a tension with the variability of disease trajectories and leaving persistent evidentiary difficulties.
One such tool for streamlining the work capacity assessment that the Swedish Social Insurance Agency (Försäkringskassan) has used since 2009, is the so-called DFA chain.[12] The chain lacks a normative legal basis and is described as a methodological tool for case officers to assess and link the diagnosis, D, to a functional impairment, F, which, in turn, leads to a concrete activity limitation, A. In cases of mental illness, however, the line between impairment and activity limitations is often hard to draw, since the impairment typically becomes visible in everyday life and work – features that clinical tests and assessments rarely capture.[13] After criticism of the agency’s heavy emphasis on the DFA-chain in its decision-making practice, the Agency did clarify in its guidance to administrators that a strict logical link between D, F and A is not required by law.[14] While this signalled an easing of some of the rigidity in the Agency’s assessment practice, the DFA chain nevertheless continues to serve as the default framework in eligibility assessments.
Another tool for streamlining work capacity assessments within sickness benefits is the Swedish National Board of Health and Welfare’s (Socialstyrelsen) insurance-medicine guidelines, which build on the DFA chain and advise physicians on typical sick-leave durations by diagnosis.[15] The guidance for psychiatric diagnoses, however, is uneven, as not all conditions are covered by guidelines.
A 2018 Swedish National Audit Office (Riksrevisionen) review, which examined the decision supports specifically for psychiatric conditions, also found that they were applied inconsistently and relied on a diagnosis-specific format poorly adapted to comorbidity, where the primary diagnosis often changes over time.[16] While the current decision supports in use now typically acknowledges comorbidity for psychiatric diagnosis, it rarely offers operational guidance on how interacting conditions should modify recommended sick-leave durations. In Sweden, the guidelines for exhaustion syndrome have been particularly debated, not least because Sweden alone chose in 2005 to include the diagnosis in its nationally modified ICD-10-SE, thereby giving it tailored decision support in sickness certification.[17] This move illustrates how the National Board of Health and Welfare sought to capture stress-related conditions that otherwise risked falling outside established psychiatric categories. The diagnosis as well as the decision support will, however, be removed in 2028 when Sweden adopts the new ICD-11. Individuals currently certified with exhaustion syndrome will then be reclassified and assessed under other supports, and the Board has been tasked with developing new knowledge and decision supports to address the situation created by the removal of the diagnosis.[18] While extra-legal, the guideline alterations that will follow from the reclassification may thus shape entitlement assessments, as differences in the content and recommendations of decision supports can influence both how physicians prepare medical documentation and how the causal link between disease and reduced work capacity is evaluated. How far-reaching these effects will be, however, remains yet to be seen.
Fluctuating Capacity Loss and Inflexible Benefit Criteria
Mental conditions vary not only in the degree to which they reduce work capacity, but also in how stable that reduction remains over time. Stress, sleep, treatment, and work environment can shift both the level and the persistence of impairment, and because recovery and rest are often integral to treatment, the actual extent of reduced capacity can be difficult to assess. This variability, however, clashes with the legal design of sickness benefits, which are paid per day and presume a constant degree of incapacity throughout the compensation period.[19] This static arrangement has long restricted the scope for flexible work scheduling in cases of part-time sick leave. Since the 1980s, only narrow exceptions were allowed.[20] A 2022 amendment has, however, expanded the scope for uneven distribution of working hours, now permitted if it does not hinder return to work and no longer requires medical prescription.[21] While this new possible exception does not allow full day-to-day flexibility, it adapts the system to conditions with dynamic impact and softens its temporal rigidity.
A further mitigating feature is that both actual incapacity (i.e. the disease directly prevents work) and therapeutic incapacity (i.e. work should be avoided to protect recovery) are compensable.[22] This adds some flexibility where impairment is hard to objectify, yet the evidentiary demand to specify and causally link disease to reduced capacity remains. The core problem of how the individual can substantiate entitlement under conditions of medical uncertainty is therefore unchanged.
Evidentiary Challenges
The tension between fluctuating illness trajectories, individual effects on working capacity, and a rigid benefit design is particularly apparent in questions of proof. The individual claiming sickness benefits must establish both the existence of the disease and the resulting loss of capacity, with the medical certificate serving as the principal, and often only, evidence (while more extensive documentation is typically required as the absence continues).[23] While the burden of proof has long been debated within the sickness-benefit system, the approach developed by the Swedish Social Insurance Agency in psychiatric cases has attracted particular criticism. Here, the Agency had started to combine the heavy reliance on the DFA chain with a practice of requiring so-called “objective” medical findings for psychiatric diagnoses, which made it especially difficult for claimants to substantiate their cases. As criticism intensified and centred on the fact that the Agency had, in effect, introduced a condition not grounded in the Social Insurance Code, the practice was abandoned in 2019.[24]
In 2023, the Swedish Supreme Administrative Court clarified that entitlement in psychiatric cases does not presuppose findings beyond the patient’s own account – thus rejecting any requirement of objective medical findings. Instead, the Court emphasised that the decisive element is the physician’s professional evaluation of that account and any accompanying observations. At the same time, however, it also signalled that evidentiary demands may increase as sick leave continues.[25] Taken together, the judgment can thus be read as drawing a line between the domains of medicine and law. The physician’s role is to provide a professional assessment, while the law determines its evidentiary value. Yet in practice, this division is less clear. The evaluation of medical evidence cannot be fully separated from medical assessment, since a meaningful evidentiary appraisal requires medical knowledge and interpretive capacity. This challenge is particularly acute in psychiatric cases, where the symptoms that define the diagnosis are often the very features that impair work capacity. The result is a persistent difficulty for the legal system, which must adjudicate claims in a space where medical uncertainty and legal evaluation are inseparably intertwined.
Conclusions
Work capacity assessments in psychiatric illness expose a structural weakness in the Swedish sickness-benefit system. While the Code distinguishes between “disease” and “reduced capacity,” in practice these criteria converge. Psychiatry rarely offers clear causal boundaries, while the law presumes that they can be neatly isolated and graded. Extra-legal instruments such as the DFA chain and decision-support tools have not solved this problem, but have been used in ways that risk transforming methodological aids into de facto legal standards, shifting the evidentiary burden onto the insured, and narrowing the scope for individualised assessment. In turn, such practices risk running counter to the principle that the insured is to be assessed in their existing condition, whereby the same diagnosis may affect work capacity differently for different persons.
Reforms such as the 2022 allowance for uneven work distribution illustrate adaptations of the framework that acknowledge variations in how reduced work capacity may manifest, yet much of it still reflects a static model ill-suited to fluctuating psychiatric conditions. The 2023 Supreme Administrative Court judgment also shows that, in the absence of clear biomarkers, the emphasis shifts toward the physician’s professional assessment of the individual’s account rather than a demand for objective findings. The evidentiary dilemma, however, still underscores the difficulty of keeping medical assessment and legal evaluation apart. Psychiatric symptoms are often the very features that impair work, making objective separation elusive. The legal framework applies equally to physical and mental illness, yet the demand for precise differentiation and causation is especially ill-suited where knowledge is incomplete and influences overlap. If internal guidelines are treated as binding norms, however, decisions risk becoming template-driven and the burdens of proof unreasonable in atypical cases.
The task of law is therefore not to resolve medical uncertainty, but to manage it. This requires clarifying which elements are legally relevant, setting proportionate evidentiary thresholds, and safeguarding space for clinical judgment. Only then can the system strike a balance between fairness to the insured and the legitimacy of legal decision-making.
* Lena Enqvist is an Associate Professor of Law at Umeå University, Sweden. This blog post is based on the article ”Att bedöma arbetsförmåga vid psykisk ohälsa – när det komplexa blir ännu mer komplext”, published in Nordisk socialrättslig tidskrift No 43-44 2025 p. 41-80. The article can be found open access on https://www.lawpub.se/artikel/10.53292/2d91fddb.d8d925b0
[1] Swedish Social Insurance Agency. Tema psykisk ohälsa, https://www.forsakringskassan.se/statistik-och-analys/tema-psykisk-ohalsa; Myndigheten för arbetsmiljökunskap. Riktlinjer för psykisk hälsa på arbetsplatsen, Version 2, 2024, p. 11.
[2] Danielsson, O. Psykiatri: När dagens diagnoser inte räcker till, Medicinsk Vetenskap, No 2, 2023; Official Government Report SOU 2021:6, p. 168 ff.
[3] Chapter 27, Sections 2 and 4 Swedish Social Insurance Code (2010:110) (Socialförsäkringsbalk), SFB.
[4] Chapter 27, Section 25 SFB.
[5] Official Government Report SOU 1944:15, p. 162; Government Bill Prop. 1994/95:147, p. 19 f.; Supreme Administrative Court RÅ 2009:102 (I och II) and Supreme Administrative Court HFD 2023 ref. 57; Vahlne Westerhäll, L. Thorpenberg, S. & Jonasson, M. Läkarintyget i sjukförsäkringsprocessen: styrning, legitimitet och bevisning, Santérus, 2009, p. 96 and 99.
[6] Official Government Report SOU 2009:89, p. 117 ff.
[7] Vahlne Westerhäll, L. Medicinska och försäkringsrättsliga sjukdoms- och arbetsförmågebegrepp – faktiska respektive normativa kriterier, in Dahlin et al (eds.) Festskrift till Elisabeth Rynning. Integritet och rättssäkerhet inom och bortom den medicinska rätten, Iustus, 2023, p 399.
[8] Chapter 27, Sections 46-49 SFB.
[9] Chapter 27, Section 47 SFB.
[10] Supreme Administrative Court RÅ 2008 ref. 15.
[11] Supreme Administrative Court HFD 2018 ref. 51.
[12] Official Government Report SOU 2009:89, p. 189 ff.; Swedish National Audit Office. Bedömning av arbetsförmåga vid psykisk ohälsa – en process med stora utmaningar, RiR 2018:11, 2018 (a), p. 22; Swedish Social Insurance Agency. Sjukpenning, rehabilitering och rehabiliteringsersättning, vägledning 2015:1, Version 19, 2024, p. 270.
[13] Swedish National Audit Office. 2018(a), p. 49 ff.
[14] Swedish Social Insurance Agency. Domsnytt 2019:014.
[15] Swedish National Board of Health and Welfare. Försäkringsmedicinskt beslutsstöd, https://forsakringsmedicin.socialstyrelsen.
se/beslutsstod-for-diagnoser/, see headline ”Så ska stödet användas”.
[16] Swedish National Audit Office. Försäkringsmedicinskt beslutsstöd – ett stöd för Försäkringskassan
vid psykisk ohälsa? RiR 2018:22, p. 55 ff.
[17] Swedish National Board of Health and Welfare. Frågor och svar: Utmattningssyndrom i ICD-11, https://www.socialstyrelsen.se/statistik-och-data/klassifikationer-och-koder/icd-11/.
[18] Ibid.
[19] Chapter 27, Sections 20 and 46(3) SFB.
[20] FÖD 1986:11; Supreme Administrative Court HFD 2011 ref. 30.
[21] Chapter 27 Section 46(3) SFB; Government Bill Prop. 2021/22:1, Utgiftsområde 10, p. 82.
[22] Government Bills Prop. 178/1953, p. 181 f. and 312/1946, p. 221; Official Government Report SOU 1944:15, p. 20.
[23] Eg. Government Bill Prop. 2002/03:89, p. 22 ff.; Official Government Report SOU 2023:48, p. 13 ff.; Vahlne Westerhäll, L, Thorpenberg, S. & Jonasson M. 2009.
[24] Swedish Social Insurance Agency. Domsnytt 2019:014.
[25] Supreme Administrative Court HFD 2022 ref. 47.
