Minimal Insomnia Symptom Scale (MISS)

Structural validity Lowthinevidence form: canonical

indirect all structural evidence is from Swedish general-population, elderly and cardiac-arrest samples, none UK, none occupational

Structural evidence is thin and comes from a small number of Swedish studies. The founding study reported satisfactory basic psychometric properties for the 3-item scale in a general-population sample aged 20 to 64 (Broman 2008). Rasch-model analyses in adult (n=1075) and elderly (n=548) Swedish samples found the data generally met Rasch requirements and the scale could separate distinct groups, with differential item functioning by age but not gender (Westergren 2015). A later polytomous Rasch evaluation in 269 cardiac-arrest survivors found acceptable model fit and targeting with no disordered thresholds and no differential item functioning by age, sex or arrest location (Hellstrom 2025). With only three items, factor-analytic dimensionality is essentially not separately evaluable, and evidence rests on Rasch fit rather than EFA/CFA; the literature is small.

Convergent and discriminant validity Very lowthinevidence form: canonical

indirect no quantified convergent/discriminant coefficients located; what exists is from Swedish samples

Little formal convergent/discriminant validity evidence was located. The founding paper reported validity consistent with the scale distinguishing insomnia cases, and the instrument is presented as measuring the same insomnia-symptom construct as longer scales, but explicit correlations with established comparator sleep measures (for example the ISI or Pittsburgh Sleep Quality Index) with reported magnitudes were not identified in the retrieved validation studies (Broman 2008; Westergren 2015). Convergent and discriminant validity should therefore be regarded as largely untested with quantified magnitudes.

Criterion validity: reference standard Lowthinevidence form: canonical

indirect criterion evidence is from Swedish community, elderly and clinical-survivor samples against self-report/ICD-10 research criteria; no UK or occupational validation and unsettled cut-offs

This is the MISS's best-evidenced property, though still modest in volume. In the founding study, ROC analysis showed the MISS could distinguish subjects with clinical insomnia defined by ICD-10 research criteria in a general-population sample (Broman 2008). In an elderly Swedish sample (n=548), against self-reported insomnia criteria, an optimal cut-off of >=7 gave sensitivity 0.93 and specificity 0.84 (positive/negative predictive values 0.256/0.995), with reliability 0.81 (Hellstrom 2010). A combined analysis of adult and elderly samples suggested different cut-offs by age group (>=6 for adults, >=7 for the elderly), while noting the optimal clinical cut-score is not settled (Westergren 2015). Among cardiac-arrest survivors an optimal cut-off of >=6 was suggested (Hellstrom 2025). Reference standards are self-report or ICD-10 research criteria rather than full clinical interview, and cut-offs are unsettled.

Criterion validity: organisational Absent (a finding about the literature)untestedevidence form: canonical

indirect no organisational-outcome evidence located in any population

No evidence was located validating the MISS against any organisational or work outcome (sickness absence, turnover, performance, or diagnosed conditions in a work context). The retrieved literature is confined to general-population, elderly and clinical-survivor screening. The absence is total and is the finding: the MISS has no demonstrated relationship to work-relevant endpoints.

Internal consistency Lowthinevidence form: canonical

indirect reliability estimates are all from Swedish samples; none from UK working adults

Reliability estimates are acceptable but few. The founding study reported satisfactory reliability (Broman 2008). Reliability was 0.81 in the elderly Swedish sample, with corrected item-total correlations of 0.64 to 0.70 (Hellstrom 2010). Rasch-based reliability was reported as acceptable in cardiac-arrest survivors (Hellstrom 2025). Estimates cluster around the low-0.8 range, adequate for a 3-item screen but drawn from a handful of Swedish studies.

Test-retest reliability Absent (a finding about the literature)untested

indirect no repeated-administration reliability located in any population

No test-retest (repeated-administration) reliability coefficient for the MISS was located in the latest review pass. The reliability figures reported in the literature (for example 0.81 in the elderly sample) are internal-consistency estimates, not stability over time, and must not be read as test-retest. The absence of any test-retest evidence is itself the finding and is a material gap, particularly for any use of the MISS as a repeated or monitoring measure.

Measurement invariance Very lowthinevidence form: canonical

indirect only Rasch DIF within Swedish samples; age DIF present; no occupational or UK evidence

Invariance evidence is limited to differential item functioning (DIF) analyses within the Rasch studies. DIF was found by age but not gender across adult and elderly Swedish samples, with the practical recommendation that a >=6 cut-off allows reasonable comparison between adults and the elderly while raw-score comparisons across age should be made cautiously (Westergren 2015). In cardiac-arrest survivors no DIF was found for age, sex or place of arrest (Hellstrom 2025). No formal multi-group configural/metric/scalar CFA invariance testing, no cross-language invariance, and no occupational or UK invariance evidence was located.

Sub-grades (evidence differs by subgroup):

• {"subgroup": "across age", "grade": "Low", "note": "DIF by age reported (Westergren 2015); caution comparing raw scores across age"}
• {"subgroup": "across sex", "grade": "Low", "note": "no DIF by gender in the Swedish and cardiac-arrest analyses"}
• {"subgroup": "across occupation", "grade": "Absent", "note": "no occupational-group evidence located"}

Responsiveness and MIC Absent (a finding about the literature)untestedevidence form: canonical

indirect no responsiveness/MIC evidence located in any population

No responsiveness or minimal important change evidence was located. The MISS was developed and validated as an epidemiological and screening instrument, not as a change/outcome measure, and no study establishing sensitivity to change or a minimal important difference was retrieved (Broman 2008; Westergren 2015). This is a genuine gap: the MISS should not be assumed responsive to intervention.