# Medicare Part D Spending by Drug (PUF), the file's change and CAGR columns named for their window

`drug_silver.vw_fact_drug_spending_part_d_v2` · Drugs and formularies · status live · 71,853 rows

Part D spending by drug and manufacturer per data year, gross of rebates. file_chg_* / file_cagr_* are drug-level constants of the SOURCE FILE, as fractions, repeated on every data_year row: they do not describe the row's year. The change runs DY-1->DY. The CAGR runs from the drug's first year with data (file_cagr_from_year, 2020 for most RY26 drugs, later for drugs new to the market) to DY, over file_cagr_years years; file_cagr_from_year is NULL on 100 data_year-2019 rows whose start year is not held. 'Overall' manufacturer rows are the brand total; filter manufacturer_name or you double count.

**Grain.** One row per (brand_name, generic_name, manufacturer_name, data_year); row for row drug_silver.fact_drug_spending_part_d (71,853 = 71,853 rows, total_spending_usd sums equal; Athena 3e7c390d).

| column | type | meaning |
|---|---|---|
| brand_name | string |  |
| generic_name | string |  |
| manufacturer_name | string |  |
| data_year | int |  |
| total_manufacturers | int |  |
| total_spending_usd | double |  |
| total_dosage_units | double |  |
| total_claims | int |  |
| total_beneficiaries | int |  |
| avg_spend_per_dosage_unit_weighted_usd | double |  |
| avg_spend_per_claim_usd | double |  |
| avg_spend_per_beneficiary_usd | double |  |
| outlier_flag | boolean |  |
| file_data_year_end | int |  |
| file_chg_avg_spend_per_dosage_unit_frac | double |  |
| file_chg_from_year | int |  |
| file_chg_to_year | int |  |
| file_cagr_avg_spend_per_dosage_unit_frac | double |  |
| file_cagr_from_year | int |  |
| file_cagr_to_year | int |  |
| file_cagr_years | int |  |
| observed_period | date |  |
| asserted_at | date |  |
| effective_period_start | date |  |
| effective_period_end | date |  |
| cms_vintage_sequence | int |  |
| parse_revision | int |  |
| parser_artifact_hash | string |  |
| source_url | string |  |
| sha256_content_hash | string |  |
| source_id | string |  |
| is_current | boolean |  |
| bronze_object_key | string |  |
| s3_version_id | string |  |
| vintage_tag | string |  |
| drug_source_version_id | string |  |

Query it: `query_substrate` with `SELECT ... FROM drug_silver.vw_fact_drug_spending_part_d_v2 WHERE ... LIMIT 100`; read `describe_data_source("drug_silver.vw_fact_drug_spending_part_d_v2")` first.