Reported Measles Case Rate by US State
- Rows
- 5,678
- Columns
- 4
- Data range
- 1900-01-01 to 2022-01-01
- Auto-refresh
- Re-checks its source every 30 days
US state-level measles case rates time series for detailed geographic analysis.
AI analysis
Data Analysis: Measles Case Rate by US State
1. Dataset Overview
This dataset, sourced from Our World in Data's grapher API (ourworldindata.org), provides a long-format time series of measles case rates per 100,000 population across US states, spanning an unusually wide range—1900 to 2022. This 122-year window is notable because it captures the pre-vaccine era, the introduction of the measles vaccine (1963), and modern outbreak dynamics, making it valuable for studying long-term public health trends, vaccine policy impacts, and outbreak clustering. As an OWID-hosted dataset, it's likely aggregated from CDC/NNDSS historical records, lending it credibility for epidemiological correlation work (e.g., against vaccination rates, school exemption policies, or population density).
2. Data Quality Observations
The dataset is clean in terms of core measurement data—zero nulls in Year, Entity, and the case rate column itself. However, the Code column is severely deficient, with 5,562 nulls out of 5,678 rows (98.0%)—essentially unusable as-is, and its single distinct non-null value suggests it may be a vestigial or improperly populated ISO/FIPS code field rather than a true state identifier. Since nulls total exactly 5,562 across all columns, this confirms Code is the sole source of missingness—a clean signal that the analytical columns are reliable. Duplicate-row counts are pending recomputation (Phase D), so referential integrity (e.g., one row per state-year) should be verified before merging.
3. Key Column Distributions
The Measles case rate column is the analytical heart of this dataset and shows extreme right-skew (skewness = 3.76), consistent with epidemic-style data: most state-years have low or near-zero rates (median = 8.48, Q1 = 0.054) while a long tail of outbreak years drives the mean (140.93) far above the median. The gap between Q3 (151.3) and the max (3,399.56) reflects severe outbreak spikes—pre-vaccine era surges almost certainly—with 741 flagged outliers (~13% of rows). This distribution shape means analysts should consider log-transformation before regression or correlation work, and should segment by era (pre-1963 vs. post-1963) since pooling both regimes will distort trend analysis.
4. Recommended Join Keys
- Entity (State name) + Year: This composite key is the natural join structure for state-year panel data and should be used to merge with other OWID or CDC datasets that share the same Entity/Year convention. - Avoid using Code as a join key given its 98% null rate—if state-level ISO/FIPS codes are needed for mapping, source them externally and join via Entity name instead.
5. Suggested Pairings for Correlation Analysis
This dataset would pair well with: (1) state-level MMR vaccination coverage and exemption rates (kindergarten immunization data, CDC/state health departments) to study vaccine-preventable disease dynamics; (2) population density and birth rate data to explore herd immunity thresholds; (3) other OWID infectious disease datasets (e.g., pertussis, rubella case rates) sharing the same Entity/Year schema for multi-disease outbreak comparison; and (4) socioeconomic/policy datasets (state health spending, school vaccination mandate enactment dates) to test policy-intervention effects on the post-1963 decline and any modern resurgence years.
Columns
- Year (date)
- Entity (string)
- Code (string)
- Measles case rate per 100,000 population (decimal)