single-item · dataset v0.9.0 contested: Convergent and discriminant validity, Test-retest reliability

Work Ability Score (WAS), single-item

Licence verified: not yet verified · literature last reviewed: 2026-09-01 · grades last confirmed: 2026-09-03 · rubric v1.6

In plain English

Work Ability Score (WAS), single-item is a single item. What it claims to measure: Global self-assessed current work ability relative to lifetime best. Licence for an employer or vendor: could not be verified against the steward's current terms; no class is asserted.

The published evidence is moderate for convergent and discriminant validity and criterion validity against organisational outcomes (absence, turnover, performance); weak for test-retest reliability and responsiveness to change. Criterion validity against a reference standard does not apply to this construct: no reference standard exists for it, so nothing is graded and nothing is missing. The evidence base for convergent and discriminant validity and test-retest reliability is contested in the published literature.

Structural validity: n/aConvergent validity: ModerateCriterion (reference standard): AbsentCriterion (organisational): ModerateInternal consistency: n/aTest-retest: LowInvariance: n/aResponsiveness: LowLicence: unverified

Grades summarise the published evidence for this instrument on its own terms; they are not comparable across instruments, and this page makes no recommendation. Whether an instrument fits your workforce is a judgement this registry informs but cannot make. Full evidence, with citations, below. This summary is generated from the record's data, not written by hand.

Who graded this. Every grade in this registry was assigned by one rater employed by the steward (1 rater, 0 independent), with AI assistance in literature retrieval and drafting. Grades and statuses are single-rater and frozen from first publication; they will not move until two named psychometric raters who are not employees of the steward have joined. Until then, automated sweeps add citations and flag cells for review; corrections of fact are made in public; no grade changes. Why, and how to volunteer.

Identity

Version: Single item: current work ability rated against lifetime best on a 0 to 10 numeric scale. It is item 1 (the first dimension) of the Work Ability Index; also termed the WAI single-item or WAI-1.

Structure: 1 item (0 to 10 scale)

Original citation: As item 1 of the WAI: Tuomi K, Ilmarinen J, Jahkola A, Katajarinne L, Tulkki A. Work Ability Index, FIOH, Helsinki (2nd revised edn 1998). The single-item usage was formalised and named 'Work Ability score' by El Fassi 2013.

Steward / publisher: Finnish Institute of Occupational Health (FIOH / Tyoterveyslaitos), as originator of the parent WAI; German distribution via WAI-Netzwerk / BAuA.

unverifiedlicence unverified for employer or vendor use. No primary steward licence page could be read at verification; no class is asserted.
Licence status: not yet verified against current steward terms. The registry's licence-currency rule requires the licence to be confirmed against the steward's current distribution terms; a primary steward licence page could not be reached at the last verification attempt. The originating steward's site (Finnish Institute of Occupational Health, FIOH / Tyoterveyslaitos, ttl.fi) and the German national steward page (BAuA, baua.de) both returned server-side 403 (Forbidden) responses to automated retrieval, so their current terms could not be read. The only reachable steward-network page, the German WAI-Netzwerk portal waiplus.online, is a commercial consultancy site marketing a trademarked survey service ('waiplus') and does not state the licence terms of the underlying WAI questionnaire. Search-result listings (not read as page content) indicate FIOH sells an official user manual ('WAI, How to use the Work Ability Index questionnaire', Rautio and Michelsen) through its bookshop and that the questionnaire is widely described in the literature as freely usable for occupational-health screening and research, but no current primary-source licence terms were confirmed. 'WAI' is presented as a trademark of FIOH. No open-content licence (Creative Commons or equivalent) should be assumed; practitioners must confirm current terms with the relevant national steward before use.
Verification attempted not verified in the latest review pass, FIOH (ttl.fi) and BAuA (baua.de) steward pages returned server-side 403 blocks to automated retrieval; the reachable WAI-Netzwerk portal (waiplus.online) markets a commercial service and does not state questionnaire licence terms; steward source: Attempted primary steward sources: FIOH https://www.ttl.fi/en (403 Forbidden), BAuA https://www.baua.de/EN/Topics/Prevention/Mental-health/Work-Ability-Index (403 Forbidden), and the reachable but non-licence WAI-Netzwerk consultancy portal https://www.waiplus.online/ (retrieved 2026-07-12; markets the trademarked 'waiplus' survey service, no questionnaire licence statement). Founding papers and reviews were NOT used as the licence source; because no primary steward licence page could be read, no licence status is asserted.
Steward page as read, Internet Archive: archived copy 1 · archived copy 2 · archived copy 3

Constructs claimed

Global self-assessed current work ability relative to lifetime best. Intended as an efficient proxy for overall work ability suitable for surveys and repeated monitoring where the full WAI is impractical.

Relations to other records

A relation is published fact about the literature, with the study it rests on. It never says which form an implementer should use.

Evidence

Deployment context caveat. Single-item measure; scale-level psychometrics (internal consistency, factor structure) are category errors and marked Not-applicable. Validity evidence spans occupational-health surveillance and clinical rehabilitation populations; findings from clinical samples (chronic pain, physical disability, low back pain) describe the clinical origin/validation and must not be read as evidence for routine workplace deployment; the cells that rest on them are flagged indirect. No UK-specific validation or norms were located in the latest review pass. (applies to every property below)

Structural validity Not applicable (category difference)untestedevidence form: canonical

absence type: category-error the property does not apply to this construct

state: not_applicable · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03

Not applicable: a single item has no internal factor structure to evaluate.

Convergent and discriminant validity Moderatecontestedevidence form: canonical

direct workers and blue-collar workers; also women on long-term sick leave, online and chronic low back pain samples (flag basis: workers; male blue-collar workers; women on long-term sick leave; chronic low back pain patients)

state: assessed · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03 · under rubric v1.0: indirectness indirect → direct (re-read before first publication, correction C-0004)

The WAS converges strongly with the full WAI, its parent instrument. El Fassi and colleagues reported a Spearman correlation of rs=0.63 between WAS and total WAI in 12,389 workers (El Fassi 2013), and Ahlstrom and colleagues found a very strong association between the single item and the full WAI in women on long-term sick leave (Ahlstrom 2010). The WAS explained about 44 per cent of the variance in WHOQOL-BREF quality of life, almost identical to the 46 per cent from the full WAI (Mokarami 2021). Face validity is supported by a moderate correlation with objectively measured heart-rate reserve in male blue-collar workers (R=-0.33, p=0.005), though not in females (Gupta 2014). In clinical samples convergent validity is weaker: only moderate positive construct validity in a Brazilian online study (Pucci 2024), and in chronic low back pain patients only 7 of 10 (WAS) predefined hypotheses were met, judged insufficient (Boekel 2022).

Sub-grades (evidence differs by subgroup):

  • {"subgroup": "vs full WAI in occupational samples", "grade": "Moderate", "note": "consistent moderate-to-strong convergence (rs around 0.63)"}
  • {"subgroup": "clinical pain/disability samples", "grade": "Low", "note": "construct-validity hypotheses only partly met"}

Criterion validity: reference standard Absent (the property does not apply to this construct)untestedevidence form: canonical

absence type: category-error the property does not apply to this construct

state: assessed_absent · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03 · under rubric v1.1: grade Moderate → Absent; status contested → untested; evidence state assessed → assessed_absent; indirectness direct → unflagged (re-read before first publication, correction C-0005) · under rubric v1.0: indirectness indirect → direct (re-read before first publication, correction C-0004)

The natural reference standard for the WAS is the full multi-item WAI, against which it converges moderately (rs=0.63) (El Fassi 2013) and which it approaches in explained variance for health-related quality of life (Mokarami 2021). Against genuine diagnostic reference standards there is no evidence, and none exists for work ability generally. Where head-to-head against the WAI for risk stratification, the single item is inferior: it poorly discriminated disability-pension risk (AUC 0.67) versus the full WAI (AUC 0.78) and showed miscalibration in male construction workers (Roelen 2014).

Criterion validity: organisational Moderatewell-establishedevidence form: canonical

direct employees, municipal employees, construction workers and nurses; also women on long-term sick leave and a general-population sample (flag basis: Finnish employees; Finnish municipal employees; male construction workers; European NEXT study of nurses; Swedish general population)

state: assessed · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03 · under rubric v1.0: indirectness indirect → direct (re-read before first publication, correction C-0004)

The WAS predicts work outcomes but generally less well than the full WAI. In 11,131 Finnish employees a poor WAS carried a hazard ratio of 9.84 (95% CI 6.68 to 14.49) for register-based disability pension and an incidence rate ratio of 3.08 (95% CI 2.19 to 4.32) for accumulated long-term sickness absence days versus good/excellent WAS (Kinnunen 2017). In Finnish municipal employees poor versus good/excellent WAS gave a disability-pension hazard ratio of 3.4 (95% CI 3.0 to 3.8), lower than the full WAI's 5.0 (Jaaskelainen 2016). Directly compared in 11,537 male construction workers, the WAS was associated with disability pension (OR 0.72 per point, 95% CI 0.66 to 0.78) but discriminated poorly (AUC 0.67) and was miscalibrated, leading the authors to recommend the full WAI for screening (Roelen 2014). Among women on long-term sick leave the single item and full WAI showed similar predictive value for degree of sick leave (Ahlstrom 2010), and the single 'current work ability vs lifetime best' item was one of three WAI items exceeding AUC 0.70 for long-term sickness absence in the Swedish general population (Lundin 2017). WAI-1 and the two-item WAI-2 predicted burnout and intention to leave the profession in the large European NEXT study of nurses (Ebener 2019).

Internal consistency Not applicable (category difference)untestedevidence form: canonical

absence type: category-error the property does not apply to this construct

state: not_applicable · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03

Not applicable: internal consistency (Cronbach's alpha / omega) cannot be computed for a single item.

Test-retest reliability Lowcontestedevidence form: mixed

direct sick-listed workers (flag basis: sick-listed workers)

state: assessed · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03 · under rubric v1.2: evidence form canonical → mixed (re-read before first publication, correction C-0006) · under rubric v1.1: indirectness indirect → direct (re-read before first publication, correction C-0005)

CoefficientTypeIntervalSamplePopulationEvidence form
0.89 (95% CI 0.76 to 0.96)ICC2 to 4 weeks21 (stable subgroup, after excluding 1 outlier)Dutch vocational-rehabilitation patients with physical disability (spinal cord injury, acquired brain injury, neuromuscular disease)canonical
0.89 (95% CI 0.77 to 0.94)ICCadmission retest within vocational rehabilitation34Dutch sick-listed workers with chronic musculoskeletal paincanonical
0.52 (general work ability, 0 to 10)ICC7 days104Dutch workers (general working population)derivative

Test-retest reliability of the WAS is contested and interval/sample-dependent. Two vocational-rehabilitation studies report good ICCs around 0.89 (Stienstra 2021; van Dinter 2025), but both rest on very small stable subgroups (n=34 and n=21). A general-working-population study of a closely analogous single 0 to 10 general work-ability item found only moderate reliability (ICC 0.52 over 7 days) (van Schaaijk 2018). The single-item retest coefficient is therefore not uniformly high, and the strong clinical-sample ICCs should not be generalised to routine workplace surveillance.

Measurement invariance Not applicable (category difference)untestedevidence form: canonical

absence type: category-error the property does not apply to this construct

state: not_applicable · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03

No formal measurement-invariance analysis is possible or reported for the single-item WAS as a stand-alone measure; invariance is a multi-item concept. Cross-group score differences (for example by age or sex) are reported in WAI studies but do not constitute invariance evidence.

Responsiveness and MIC Lowthinevidence form: canonical

direct sick-listed workers (flag basis: sick-listed workers)

state: assessed · rubric v1.6 · literature as of 2026-09-01 · grade confirmed 2026-09-03 · under rubric v1.1: indirectness indirect → direct (re-read before first publication, correction C-0005)

The WAS is responsive to change in clinical rehabilitation, with anchor-based minimal important change estimates. In sick-listed workers with chronic musculoskeletal pain the WAS was responsive (AUC 0.70) with a minimal clinically important change around 1.5 points and baseline-dependent MICs (Stienstra 2021). In chronic low back pain the WAS was responsive (AUC 0.70), MIC about 1.5 points, smallest detectable change about 4.9 points, though its construct validity in that sample was judged insufficient (Boekel 2022). The Brazilian online study likewise reported good-to-excellent test-retest and construct validity for the single item (Pucci 2024). Responsiveness evidence is thus concentrated in clinical populations.

Populations, languages and norms

As the first WAI item, the WAS inherits the WAI's wide language coverage and is used in occupational surveys (Finnish, Swedish, Dutch, Belgian/Luxembourg, Iranian, Brazilian) and increasingly in vocational rehabilitation. Because the 0 to 10 scale is intuitive it is popular for repeated monitoring. No dedicated UK norms were located in the latest review pass.

Criticisms and controversies

The central controversy is whether a single item can substitute for the full WAI. Evidence is mixed: the WAS converges moderately with the WAI and predicts disability pension and sickness absence, but it is a consistently weaker discriminator (for example AUC 0.67 vs 0.78 for disability pension) and can be miscalibrated, so several authors recommend the full index for individual-level screening while accepting the single item for population surveillance. Test-retest reliability is not uniformly high once general-worker (rather than clinical) samples are considered. Much of the strongest reliability and responsiveness evidence comes from clinical rehabilitation cohorts, which is indirect for routine workplace use.

References (14)

  1. Ahlstrom L, Grimby-Ekman A, Hagberg M, Dellve L (2010). The work ability index and single-item question: associations with sick leave, symptoms, and health--a prospective study of women on long-term sick leave. https://doi.org/10.5271/sjweh.2917
  2. El Fassi M, Bocquet V, Majery N, Lair M, Couffignal S, Mairiaux P (2013). Work ability assessment in a worker population: comparison and determinants of Work Ability Index and Work Ability score. https://doi.org/10.1186/1471-2458-13-305
  3. Gupta N, Jensen B, Søgaard K, Carneiro I, Christiansen C, Hanisch C, Holtermann A (2014). Face validity of the single work ability item: comparison with objectively measured heart rate reserve over several days. https://doi.org/10.3390/ijerph110505333
  4. Roelen C, van Rhenen W, Groothoff J, van der Klink J, Twisk J, Heymans M (2014). Work ability as prognostic risk marker of disability pension: single-item work ability score versus multi-item work ability index. https://doi.org/10.5271/sjweh.3428
  5. Jääskeläinen A, Kausto J, Seitsamo J, Ojajärvi A, Nygård C, Arjas E, Leino-Arjas P (2016). Work ability index and perceived work ability as predictors of disability pension: a prospective study among Finnish municipal employees. https://doi.org/10.5271/sjweh.3598
  6. Lundin A, Leijon O, Vaez M, Hallgren M, Torgén M (2017). Predictive validity of the Work Ability Index and its individual items in the general population. https://doi.org/10.1177/1403494817702759
  7. Kinnunen U, Nätti J (2017). Work ability score and future work ability as predictors of register-based disability pension and long-term sickness absence: A three-year follow-up study. https://doi.org/10.1177/1403494817745190
  8. van Schaaijk A, Nieuwenhuijsen K, Frings-Dresen M, Sluiter J (2018). Reproducibility of work ability and work functioning instruments. https://doi.org/10.1093/occmed/kqy010
  9. Ebener M, Hasselhorn H (2019). Validation of Short Measures of Work Ability for Research and Employee Surveys. https://doi.org/10.3390/ijerph16183386
  10. Mokarami H, Cousins R, Kalteh H (2021). Comparison of the work ability index and the work ability score for predicting health-related quality of life. https://doi.org/10.1007/s00420-021-01740-9
  11. Stienstra M, Edelaar M, Fritz B, Reneman M (2021). Measurement Properties of the Work Ability Score in Sick-Listed Workers with Chronic Musculoskeletal Pain. https://doi.org/10.1007/s10926-021-09982-7
  12. Boekel I, Dutmer A, Schiphorst Preuper H, Reneman M (2022). Validation of the work ability index-single item and the pain disability index-work item in patients with chronic low back pain. https://doi.org/10.1007/s00586-022-07109-x
  13. Pucci R, da Silva A, Padula R (2024). Factorial analysis of the Brazilian-Portuguese version of the Work Ability Index, reproducibility and validity of the single item and the short version for online application. https://doi.org/10.1016/j.bjpt.2024.101060
  14. van Dinter R, Jenks A, Roels E, Post M, Reneman M (2025). Test-Retest Reliability and Agreement of the Work Ability Index-Single Item in Persons With Physical Disabilities. https://doi.org/10.1016/j.apmr.2024.10.018

Record notes

Overall confidence: as a single-item proxy the WAS has well-established organisational predictive validity (Moderate, consistently weaker than the full WAI) and Moderate but contested convergence with its parent instrument; test-retest is Low/contested (good ICCs only in small clinical subgroups, moderate in general workers); responsiveness is Low and confined to clinical samples. Scale-level properties (internal consistency, structural validity, measurement invariance) are correctly Not-applicable rather than Absent per rule 4. Under rubric 1.0 every applicable grade was flagged indirect because no UK validation or norms were located; under rubric 1.1 the occupational-cohort evidence is direct and only the cells resting on clinical rehabilitation samples (test-retest, responsiveness) stay indirect (correction C-0004). The schema v0.2 single-item handling worked cleanly; the one nuance recorded honestly is that one retest coefficient (van Schaaijk) is a derivative general single-item work-ability appraisal rather than the exact WAI-1 wording, tagged as derivative evidence_form in the structured test-retest sub-object. Licence could not be verified against a primary steward page in the latest review pass because FIOH (ttl.fi) and BAuA (baua.de) returned server-side 403 blocks and the reachable WAI-Netzwerk portal does not state questionnaire terms; per rule 7 the licence is recorded as not verified in the latest review pass rather than asserted from literature.