Sensitivity and Specificity of Geriatric Assessment Tools
The diagnostic accuracy of these geriatric assessment tools varies considerably, with the MMSE showing high sensitivity (81%) and specificity (83%) for moderate dementia, the AD8 demonstrating excellent performance when combined with functional assessments, and the GDS-15 achieving sensitivity of 84.3% and specificity of 73.8% for depression screening. 1, 2, 3
1. Comprehensive Geriatric Assessment (CGA)
Sensitivity: 95%
Specificity: 87%
The CGA demonstrates superior diagnostic accuracy compared to individual screening tools. When used as a comprehensive screening approach, it provides higher sensitivity and specificity than the MoCA alone for identifying mild cognitive impairment 4. The CGA's strength lies in its multidimensional approach, capturing cognitive, functional, nutritional, and physical domains simultaneously.
2. AD8 (Ascertain Dementia 8)
Sensitivity: 89%
Specificity: 81% (when used alone)
Specificity increases to 93% when combined with IADL scale
The AD8 is recommended as a rapid psychometric screening tool and performs particularly well as an informant-based questionnaire 1. Combining the AD8 with the IADL scale significantly improves overall accuracy, with specificity reaching 93% and positive likelihood ratio of 11.74 2. This combination is especially valuable when patients cannot complete direct cognitive testing due to time constraints or lack of cooperation 1.
3. Geriatric Depression Scale (GDS)
GDS-30 (Long Form):
- Sensitivity: 81.9%
- Specificity: 77.7%
GDS-15 (Short Form):
- Sensitivity: 84.3%
- Specificity: 73.8%
GDS-4/5 (Ultrashort Form):
- Sensitivity: 92.5%
- Specificity: 77.2%
The GDS-4/5 ultrashort form is the most efficient option, correctly identifying an additional 25 true positives with only 10 false positives per 100 patients screened, compared to routine clinical judgment 3. The GDS-15 performs comparably to the GDS-30 but with slightly more false positives 3.
Important caveat: In institutionalized elderly populations, the GDS shows reduced sensitivity (47%) and should be interpreted cautiously 5. The accuracy is not significantly influenced by the presence of dementia or severity of medical burden 3, 6.
4. Katz ADL and Lawton IADL Scales
Lawton IADL Scale:
- Sensitivity: 89%
- Specificity: 81% (cutoff score 6/7)
- Area under curve: 0.925
- Positive likelihood ratio: 4.75
The IADL scale demonstrates discriminant power comparable to the MMSE and AD8 for identifying dementia 2. When combined with the AD8, specificity improves dramatically to 93% 2. The Katz ADL and Lawton IADL scales are recommended for rapid functional autonomy screening when used with patient and family member input 1.
5. MMSE (Mini-Mental State Examination)
Sensitivity: 81%
Specificity: 83% (for moderate dementia, cutoff <24)
The MMSE remains the most widely used and best-studied instrument with high sensitivity and specificity for separating moderate dementia from normal cognition 1, 7.
Critical limitations:
- Lacks sensitivity for mild dementia or MCI 1
- Accuracy depends on age and educational level, potentially leading to more false-positives in older people with lower education and false-negatives in younger people with higher education 7
- When used to screen unselected patients, the predictive value of a positive result is only fair 7
Clinical recommendation: Use MoCA instead when mild cognitive impairment is suspected or when MMSE scores are in the "normal" range (24+/30) but clinical suspicion remains 1.
6. MNA-F (Mini Nutritional Assessment - Full)
No specific sensitivity/specificity data provided in the evidence, though it is incorporated as part of the CGA screening battery and shows significant differences between patients with mild cognitive impairment and healthy individuals 4.
7. SARC-F
No sensitivity/specificity data provided in the evidence. This tool was not addressed in the available guidelines or research studies.
8. EASI (Elder Abuse Screening Index)
No sensitivity/specificity data provided in the evidence. This tool was not addressed in the available guidelines or research studies.
9. Caregiver Burden Scale (Zarit Burden Interview)
This is not a diagnostic screening tool but rather an assessment instrument measuring caregiver distress 8. It uses a 22-item questionnaire with scores ranging from 0-88:
- 0-20: Little or no burden
- 21-40: Mild to moderate burden
- 41-60: Moderate to severe burden
- 61-88: Severe burden
The scale has shown mixed results in correlating with other clinical measures such as cognition and sleep 9, but it has demonstrated responsiveness to intervention 9.
Key Clinical Pearls
For cognitive screening: Combine the AD8 with IADL for optimal specificity (93%) when screening for dementia 2. Use MMSE for moderate dementia but switch to MoCA for suspected MCI 1.
For depression screening: The GDS-4/5 ultrashort form offers the best efficiency with administration time under 1 minute versus 2.7 minutes for the GDS-15 3, 10.
For comprehensive assessment: CGA provides superior overall diagnostic accuracy (sensitivity 95%, specificity 87%) compared to individual tools 4.