frontier

cea.frontier

Incremental cost-effectiveness analysis on the efficiency frontier.

Implements the standard decision-analytic algorithm: order interventions by cost, remove strongly dominated interventions (more costly and no more effective than another), then iteratively remove extendedly dominated interventions (whose ICER exceeds that of the next more effective intervention) until ICERs increase monotonically along the frontier.

Functions

Name	Description
frontier	Interventions on the cost-effectiveness efficiency frontier, cheapest first.
icer_table	Full incremental analysis: dominance, extended dominance, and ICERs.

frontier

cea.frontier.frontier(source, *, effect=None)

Interventions on the cost-effectiveness efficiency frontier, cheapest first.

Example

import pandas as pd from heormodel.cea import frontier means = pd.DataFrame( … {“cost”: [0.0, 10.0, 5.0], “effect”: [0.0, 1.0, -1.0]}, … index=[“A”, “B”, “C”], … ) frontier(means) [‘A’, ‘B’]

icer_table

cea.frontier.icer_table(source, *, effect=None, interval=0.95)

Full incremental analysis: dominance, extended dominance, and ICERs.

Parameters

Name	Type	Description	Default
source	Outcomes | pd.DataFrame	probabilistic Outcomes (means are taken per intervention) or a per-intervention mean table indexed by intervention with columns cost and the effect column.	required
effect	str | None	Effect column name (defaults to the outcomes’ primary effect, or "effect" for plain tables).	None
interval	float | None	Central probability for the uncertainty interval, or None to omit intervals. When source is a probabilistic Outcomes with more than one iteration, each estimate gains _lo/_hi columns holding the two-sided interval across parameter draws (the default 0.95 gives a 95% interval, the 2.5th and 97.5th percentiles). A mean table or a single draw carries no intervals.	0.95

Returns

Name	Type	Description
	pd.DataFrame	DataFrame indexed by intervention, sorted by cost, with columns
	pd.DataFrame	cost, effect, inc_cost, inc_effect, icer and
	pd.DataFrame	status ("ND" on the frontier, "D" strongly dominated,
	pd.DataFrame	"ED" extendedly dominated). Every intervention except the cheapest
	pd.DataFrame	carries an incremental cost and effect against its comparator, the
	pd.DataFrame	cheapest frontier intervention still above it in cost order, so a
	pd.DataFrame	dominated intervention shows the negative incremental effect or excess
	pd.DataFrame	cost that marks it dominated. The ICER is a frontier quantity, filled
	pd.DataFrame	between adjacent frontier interventions and left blank for dominated
	pd.DataFrame	ones and for the cheapest frontier intervention.
	pd.DataFrame	With intervals, each of cost, effect, inc_cost,
	pd.DataFrame	inc_effect and icer is followed by its _lo and _hi
	pd.DataFrame	bounds. Dominance and the frontier are settled once on the mean costs
	pd.DataFrame	and effects; the intervals describe the spread of each measure for that
	pd.DataFrame	fixed frontier. The incremental measures are differences between
	pd.DataFrame	interventions, so their intervals are taken from the paired per-iteration
	pd.DataFrame	difference (cost of the intervention minus cost of its comparator
	pd.DataFrame	in the same draw), not from the separate intervals of the two
	pd.DataFrame	interventions. The ICER interval comes from the per-draw ratio of paired
	pd.DataFrame	incremental cost to paired incremental effect; draws whose incremental
	pd.DataFrame	effect approaches zero make that ratio unstable, so read the incremental
	pd.DataFrame	cost and effect intervals alongside it.

Example

import pandas as pd from heormodel.cea import icer_table means = pd.DataFrame( … {“cost”: [0.0, 100.0, 400.0], “effect”: [0.0, 0.5, 1.0]}, … index=[“A”, “B”, “D”], … ) t = icer_table(means) float(t.loc[“D”, “icer”]) 600.0