CDC Wastewater Data for Measles
- Rows
- 84,537
- Columns
- 38
- Data range
- 000105acf201c01fa0e210e648ab376f to ffffce403249c6ee06e497ea36953ebd
- Auto-refresh
- Re-checks its source every 7 days
- Last refreshed with new data
- 2026-11-17
US site-level measles wastewater detections over time (weekly). Emerging surveillance time-series, government source.
AI analysis
CDC Wastewater Surveillance Dataset — Measles Analysis
1. Dataset Overview & Value
This dataset represents the CDC's National Wastewater Surveillance System (NWSS) applied specifically to measles detection, sourced directly from the CDC's public data API at data.cdc.gov. The single distinct value in dateupdated suggests a static snapshot rather than a rolling feed, so analysts should verify freshness before longitudinal use. With 44,528 weekly site-level records spanning 649 unique monitoring sites across all 50 U.S. states and territories, this is a genuinely rare emerging surveillance time series. Its core value lies in environmental detection of measles viral RNA in wastewater — a signal that can precede clinical case reporting by days to weeks, making it exceptionally well-suited for leading-indicator correlation studies against traditional epidemiological data.
---
2. Data Quality Observations
The dataset carries 127,623 null cells across 38 columns, which is substantial but not uniformly distributed — several columns are essentially unusable as-is. The most problematic is inhibitionadjust at 59.8% null (26,639 rows), meaning PCR inhibition correction is undocumented for the majority of samples, which directly undermines confidence in concentration measurements. pcrtargetmiclin is 49.8% null (22,184 rows) and, critically, has a constant value of 0 with zero standard deviation across its non-null range — rendering it analytically worthless and likely a placeholder column. pcrtargetflowpoplin is 42.7% null (19,009 rows), and flowrate is missing for 18,726 rows (42%), a significant gap given that flow normalization is standard practice for wastewater concentration comparisons. countyfips has 4,720 nulls (~10.6%), which will constrain geographic join operations. The receffpercent column contains a minimum of -1, which is physically impossible for a recovery efficiency percentage and flags likely data entry errors or sentinel values requiring cleaning. Duplicate row counts are pending Parquet phase recomputation and should be confirmed before any aggregation work.
---
3. Key Column Distributions
The central outcome variable, pcrtargetavgconc, tells a striking story: mean=44.7, median=0, Q1=0, Q3=0, skew=32.4 — meaning the overwhelming majority of samples register zero measles RNA detection, with a sparse tail of high-concentration events (max=50,240). This is epidemiologically expected for a rare pathogen but means analysts must treat detection as a binary or rare-event variable rather than a continuous one for most modeling purposes. The related pcrtargetavgconclin (non-null only for ~40,434 rows) shows median=600 and mean=2,556 with skew=3.24, confirming the heavy right tail. populationserved is severely right-skewed (skew=5.27, 5,126 outliers), ranging from 1,000 to 4,000,000 — a few large urban systems dominate the served-population distribution and will need log transformation or stratification. flowrate is similarly extreme (skew=5.97, 3,096 outliers, max=370 MGD), and humfracmicconc is the most extreme column in the dataset (skew=45.7, 3,385 outliers), ranging across six orders of magnitude. Recovery efficiency (receffpercent) has a reasonable median of 74.5% but a max of 956.8% — statistically impossible — confirming calibration anomalies or data errors in a non-trivial slice of records.
---
4. Recommended Join Key Columns
| Column | Role | Notes | |---|---|---| | recordid | Primary unique key | 44,528 distinct values, zero nulls — safe anchor | | site | Site-level aggregation key | 649 sites, zero nulls; join to CDC NWSS site metadata | | countyfips | Geographic join | ~10.6% null; standard FIPS codes enable county-level linkage | | stateterritory | State rollup | 50 distinct, zero nulls; broadest geographic join | | samplecollectdate | Temporal join | 423 distinct dates, zero nulls; weekly cadence enables time-series alignment |
The most robust cross-dataset join strategy is a composite key of countyfips + samplecollectdate, enabling week-county level linkage to clinical, demographic, or mobility datasets — with the caveat that ~10.6% of records will drop out on county-level joins.
---
5. Recommended Companion Datasets for Correlation Analysis
- CDC Measles Case Reports (NNDSS): The most direct pairing — correlating wastewater detection events against reported clinical cases by county and week would quantify the leading-indicator lag and sensitivity of wastewater surveillance, the core scientific question this dataset is designed to answer. - CDC Vaccination Coverage Data (NIS/NIS-Teen): MMR vaccination rates by county or state would allow testing whether lower coverage areas show higher or earlier wastewater detection — a critical public health policy question given recent measles resurgences. - International Travel & Flight Data (e.g., BTS, OAG): Measles is frequently imported; correlating detection sites near major airports or high-travel counties with international arrival volumes could identify importation corridors. - School Enrollment & Exemption Records: State-level vaccine exemption rates and school-age population density would pair naturally with detection frequency to model community transmission risk. - Socioeconomic & Demographic Data (ACS/Census): Population density, housing crowding, and income levels via countyfips joins could reveal structural predictors of wastewater signal strength, controlling for the heavily skewed populationserved distribution.
Columns
- record_id (string)
- site (integer)
- state_territory (string)
- source (string)
- county_fips (integer)
- counties_served (string)
- population_served (integer)
- sample_id (string)
- sample_collect_date (date)
- sample_type (string)
- sample_matrix (string)
- sample_location (string)
- flow_rate (decimal)
- concentration_method (string)
- pasteurized (string)
- pcr_type (string)
- extraction_method (string)
- major_lab_method (integer)
- inhibition_detect (string)
- inhibition_adjust (string)
- ntc_amplify (string)
- pcr_target (string)
- pcr_gene_target_agg (string)
- pcr_target_avg_conc (integer)
- pcr_target_units (string)
- lod_sewage (integer)
- pcr_target_detect (string)
- pcr_target_avg_conc_lin (integer)
- pcr_target_flowpop_lin (decimal)
- pcr_target_mic_lin (integer)
- hum_frac_target_mic (string)
- hum_frac_mic_conc (decimal)
- hum_frac_mic_unit (string)
- rec_eff_percent (decimal)
- rec_eff_target_name (string)
- rec_eff_spike_matrix (string)
- rec_eff_spike_conc (decimal)
- date_updated (datetime)