Get concept prevalence for a table — ds.omop.concept.prevalence • dsOMOPClient

Retrieves the most frequent concepts in the specified OMOP CDM table, ranked by person count or record count. Results are disclosure-controlled on the server side (small cells are suppressed) and returned as a dsomop_result with per-site and optionally pooled data. Pooling sums counts across servers and re-ranks.

Usage

ds.omop.concept.prevalence(
  table = NULL,
  concept_col = NULL,
  metric = "persons",
  top_n = 50,
  cohort_table = NULL,
  window = NULL,
  offset = 0L,
  global = FALSE,
  cohort = NULL,
  scope = c("per_site", "pooled"),
  pooling_policy = "strict",
  symbol = "omop",
  conns = NULL,
  execute = TRUE
)

Arguments

table: Character; the CDM table name (e.g., "condition_occurrence", "drug_exposure").
concept_col: Character; the concept column name, or NULL for automatic detection based on the table's standard concept column (default: NULL).
metric: Character; "persons" (default) to rank by distinct person count, or "records" to rank by total record count.
top_n: Integer; number of top concepts to return (default: 50).
cohort_table: Character; name of a server-side cohort temp table to restrict the analysis to a specific cohort (default: NULL).
window: List with start and end date strings (ISO 8601) for temporal filtering, or NULL for no date restriction (default: NULL).
offset: Integer; number of ranked concepts to skip for pagination (default: 0).
global: Logical; if TRUE, rank concepts across all supported clinical tables rather than only table (default: FALSE).
cohort: Cohort reference (a dsomop_cohort_handle, a cohort_definition_id, or a server-side cohort table name), or NULL. Takes precedence over cohort_table.
scope: Character; "per_site" (default) or "pooled".
pooling_policy: Character; "strict" (default) requires all servers to succeed, "pooled_only_ok" allows partial results.
symbol: Character; the session symbol (default: "omop").
conns: DSI connection object(s) or NULL to use the session default.
execute: Logical; if FALSE, return a dry-run result containing only the generated call code (default: TRUE).

Value

A dsomop_result object with $per_site (named list of data frames with columns concept_id, concept_name, count_value, etc.), $pooled (combined data frame when scope is "pooled", otherwise NULL), and $meta.

Examples

if (FALSE) { # \dontrun{
result <- ds.omop.concept.prevalence("condition_occurrence")
head(result$per_site$server1)

pooled <- ds.omop.concept.prevalence("drug_exposure",
                                      metric = "records",
                                      scope = "pooled")
pooled$pooled
} # }