Disclosure-safe two-by-two comorbidity (person co-occurrence)

A thin wrapper over ds.omop.crosstab that cross-tabulates the person-level presence/absence of two concepts as a 2x2 table, using the SAME server-side omopCrossTabDS machinery (primary + iterative complementary suppression, no exact margins). Because counting and suppression happen server-side, suppressed cells cannot be backed out from margins client-side.

Usage

ds.omop.comorbidity(
  conceptA,
  conceptB,
  tableA = "condition_occurrence",
  tableB = tableA,
  ...
)

Arguments

conceptA: Integer; the first concept ID (row presence axis).
conceptB: Integer; the second concept ID (column presence axis).
tableA: Character; the CDM table holding conceptA (default: "condition_occurrence").
tableB: Character; the CDM table holding conceptB (default: same as tableA).
...: Additional arguments forwarded to ds.omop.crosstab (e.g., cohort_table, scope, pooling_policy, symbol, conns, execute).

Value

A dsomop_result object (see ds.omop.crosstab).

Comorbidity is descriptive, not inferential

This reports raw person-overlap between two conditions. It is NOT an adjusted measure of association and does not control for confounders such as age. For a genuine multivariable comorbidity model, route to ds.glm.

Examples

if (FALSE) { # \dontrun{
cm <- ds.omop.comorbidity(316866, 201826)  # hypertension x type 2 diabetes
cm$per_site
} # }