Which tool is better for assessing hypoglycemia awareness, the Clarke questionnaire or the Gold score?

Clarke Questionnaire is Superior for Predicting Severe Hypoglycemia

The Clarke questionnaire should be your primary tool for assessing hypoglycemia awareness, as it demonstrates the strongest predictive validity for severe hypoglycemia and superior diagnostic performance compared to the Gold score 1.

Evidence-Based Comparison

Predictive Validity for Severe Hypoglycemia

The most recent high-quality prospective cohort study (2026) directly compared both methods in 286 Japanese adults with type 1 diabetes and found:

• Clarke method: Adjusted hazard ratio of 8.27 (95% CI: 3.43-20.0, p<0.001) for predicting severe hypoglycemia 1
• Gold method: Adjusted hazard ratio of only 1.90 (95% CI: 0.67-5.36, p=0.227) - not statistically significant 1

This represents a more than 4-fold stronger association with severe hypoglycemia outcomes for Clarke versus Gold.

Diagnostic Performance Metrics

The Clarke questionnaire demonstrated 1:

• Sensitivity: 65.2%
• Specificity: 84.8%
• Negative predictive value: 96.5% (excellent for ruling out risk)
• Positive likelihood ratio: 4.29
• Overall diagnostic validity: 83.2%

These metrics substantially exceed those of the Gold score for identifying patients at genuine risk of severe hypoglycemia.

Guideline Recommendations

Both major diabetes guidelines acknowledge both tools as "commonly used validated scores" with a threshold ≥4 indicating hypoglycemia unawareness 2. However, the 2015 Diabetes Care guidelines do not differentiate between them, as the comparative validity data emerged later 2.

Important Caveats

Limited Agreement Between Methods

A critical pitfall: The two questionnaires identify different populations in approximately 60% of cases 3. The overlap between Gold ≥4 and Clarke ≥4 is only 40% overall (45% in type 1 diabetes, 37% in type 2 diabetes) 3. This means:

• Gold may miss high-risk patients that Clarke identifies
• Patients classified as having impaired awareness by Gold alone had only 9% severe hypoglycemia rate
• Patients identified by Clarke alone had 53% severe hypoglycemia rate 3

What They Actually Measure

The questionnaires assess different, independent aspects of hypoglycemia awareness 3, 4:

• Gold: Focuses more on self-perceived awareness status (single question format)
• Clarke: Captures broader symptom recognition patterns and behavioral responses (8-question format)

The correlation between questionnaire scores and actual epinephrine/symptom responses during experimental hypoglycemia is only moderate (r = -0.50 to -0.59) 4, meaning 32% of subjects may be classified inconsistently between methods 4.

Clinical Application Algorithm

1. Screen all insulin-treated patients with type 1 diabetes using the Clarke questionnaire at routine visits 2
2. Use Clarke score ≥4 as your threshold for impaired awareness 2, 1
3. If Clarke ≥4: Patient has 8-fold increased risk of severe hypoglycemia - initiate intensive intervention 1
4. Consider adding Gold only if you want to capture patients with partial symptom loss who may not yet have elevated severe hypoglycemia risk 4

Pediatric Considerations

In children ≥9 years old, the Clarke questionnaire has higher specificity and superior prediction of clinically significant hypoglycemia compared to Gold 5. The Gold questionnaire in pediatrics tends to overdiagnose impaired awareness (41% prevalence) versus Clarke (22% prevalence), with Clarke-positive patients showing higher rates of actual severe hypoglycemia 5.

For children <9 years, neither questionnaire has been adequately validated 5.