How should I interpret continuous glucose monitoring (CGM) data?

How to Interpret Continuous Glucose Monitoring Data

Interpret CGM data using the standardized Ambulatory Glucose Profile (AGP) report with at least 14 days of data showing ≥70% sensor wear time, focusing on 10 core metrics with specific clinical targets: Time in Range (TIR) >70%, Time Below Range (TBR) <4%, Time Above Range (TAR) <25%, and coefficient of variation (CV) ≤36%. 1

Data Adequacy Requirements

Before interpreting any CGM data, verify you have sufficient information:

• Minimum 14 days of CGM wear with ≥70% active sensor time (this correlates strongly with 3-month glucose patterns) 1
• For patients with highly variable control or investigating hypoglycemia, consider requiring 4 weeks of data 1

Critical pitfall: Interpreting inadequate data leads to unreliable conclusions. Always check data sufficiency first.

The 10 Core Metrics for Clinical Practice

The 2019 International Consensus streamlined 14 original metrics down to these essential 10 1:

Primary Metrics to Review:

1. Number of days CGM worn (target: 14 days)
2. Percentage of time CGM active (target: ≥70%)
3. Mean glucose
4. Glucose Management Indicator (GMI) - replaces estimated A1C
5. Glycemic variability (coefficient of variation) - target ≤36%

Time in Ranges (The Most Actionable Data):

1. Time Above Range Level 2: >250 mg/dL (>13.9 mmol/L)
2. Time Above Range Level 1: 181-250 mg/dL (10.1-13.9 mmol/L)
3. Time in Range (TIR): 70-180 mg/dL (3.9-10.0 mmol/L) - this is your primary target
4. Time Below Range Level 1: 54-69 mg/dL (3.0-3.8 mmol/L)
5. Time Below Range Level 2: <54 mg/dL (<3.0 mmol/L)

Specific Clinical Targets

For Adults with Type 1 or Type 2 Diabetes:

• TIR: >70% (>16 hours 48 minutes per day in 70-180 mg/dL range) 1
• TBR: <4% (<1 hour below 70 mg/dL) 1
• TAR: <25% (<6 hours above 180 mg/dL) 1
• CV: ≤36% (some evidence suggests <33% provides additional hypoglycemia protection for insulin/sulfonylurea users) 1

For Older/High-Risk Patients:

• TIR: >50% (>12 hours in range) - less stringent target
• TBR: <1% (<15 minutes below 70 mg/dL) - much stricter hypoglycemia avoidance
• TAR: <10% (<2 hours 24 minutes above 250 mg/dL only) 1

Important nuance: Each 5% increase in TIR correlates with clinically significant benefits for diabetes complications 1. This means even incremental improvements matter.

Systematic Interpretation Algorithm

Step 1: Verify Data Quality

• Check days worn and % active time first
• Reject interpretation if <70% active over <14 days

Step 2: Assess Hypoglycemia Risk (Safety First)

• Look at TBR metrics immediately - this determines safety of any intensification
• TBR <54 mg/dL is most critical (severe hypoglycemia risk)
• If TBR >4%, address hypoglycemia before targeting hyperglycemia

Step 3: Evaluate Glycemic Variability

• CV >36% indicates excessive variability - suggests erratic patterns requiring investigation
• High CV with adequate TIR may indicate compensating highs and lows
• High CV increases hypoglycemia risk even with acceptable mean glucose

Step 4: Assess Time in Range

• TIR is your primary efficacy metric
• Compare to 70% target (or 50% for high-risk patients)
• TIR directly correlates with microvascular complications from DCCT data 1

Step 5: Characterize Hyperglycemia Pattern

• Distinguish Level 1 (181-250) vs Level 2 (>250) hyperglycemia
• Level 2 requires more aggressive intervention
• Look at TAR distribution across 24-hour periods to identify patterns

Step 6: Review AGP Visual Pattern

• The AGP report standardizes visualization across all CGM systems 1
• Identify consistent daily patterns (fasting, post-meal, overnight)
• Look for specific meal-related excursions
• Assess glucose recovery patterns after meals

Critical Clinical Pitfalls

CGM complements but does not replace A1C: A1C remains the only prospectively validated tool for long-term complication risk. Use CGM to understand the patterns behind the A1C, not as a replacement 1. However, A1C can be inaccurate in anemia, hemoglobinopathies, iron deficiency, and pregnancy 1.

Sensor lag in dynamic situations: CGM measures interstitial glucose with a delay during rapid changes. Don't over-interpret single rapid excursions 1.

Technology requires active engagement: CGM only improves outcomes when patients wear it consistently and act on the data 2. Discontinuation causes partial reversal of benefits 2.

Personalization is essential but targets are not optional: While individual circumstances matter, the consensus targets represent evidence-based thresholds. Adjust targets for older/high-risk patients using the specific alternative targets provided, not arbitrary individualization 1.

Integration with Treatment Decisions

For patients on basal insulin without prandial coverage: CGM data from randomized trials support using these metrics for safer treatment adjustments and glycemic assessment 2, 3.

For non-insulin-treated type 2 diabetes: CGM helps identify fasting vs postprandial hyperglycemia patterns, supports lifestyle discussions, and clarifies A1C-glucose discordance, though evidence is stronger for glycemic metrics than long-term outcomes 3.

Each metric drives specific interventions: TBR excess requires reducing insulin/sulfonylureas; TAR excess with low CV suggests consistent hyperglycemia requiring intensification; TAR excess with high CV suggests erratic patterns requiring pattern management before intensification 4, 5, 6.