Temporal Structure

A measure’s temporal structure describes two things about the measure. First, it describes the type of observation denominating the measure, i.e., is the measure denominated by events like admissions, by entities like patients, or intervals like member-months? Second, temporal structure describes how denominator observations are assigned to reporting segments when measures are created.

Population measures are denominated by events or entities, while performance measures can be denominated by events, entities, or intervals.

Temporal Structure: Event

An event-denominated measure’s observations represent a single moment in time (e.g., hospital admissions, claims, lab results; etc.).

An example of an event-denominated population measure is one that counts the number of hospital admissions. In this example, the hospital admissions are the denominator observations because the admissions happened at single moments in time.

A 30-Day Readmission measure is an example of an event-denominated performance measure. In this example, the hospital discharges are the denominator observations because the discharges happened at single moments in time.

Event-denominated measures (both population and performance), use the event temporal structure. This temporal structure calculates measure results based on the number of denominator events that occurred within a reporting period. The number of denominator events allocated to a reporting period is based on the denominator event’s effective date (e.g., admission date).

In this diagram, each green square represents a single denominator event occurring during the reporting period.

Each green circle represents the total number of denominator events occurring during the reporting period.

The event temporal structure places denominator events into specific reporting periods by specifying the event’s date (e.g., admission date) is the measure’s case effective date.

Temporal Structure: Entity

An entity-denominated measure’s observations represent a person, place, or thing that for analytic purposes do not represent a moment in time or a period of time but are considered to have a continuous presence (e.g., patients).

An example of an entity-denominated population measure is one that counts the number of active plan members. In this example, the members are the denominator observations.

A measure of the percentage of patients who saw their attributed primary care physician in the last 90-days is an example of an entity-denominated performance measure. In this example, the patients are the denominator observations.

Entity-denominated measures, which use the entity temporal structure, calculate measure results based on the number of entities meeting the denominator’s inclusion criteria as of a given snapshot date. The number of entities meeting the inclusion criteria is determined by assessing the state of each entity as of the period that includes the snapshot date. The period’s start and end date are the measure’s case effective dates.

In the diagram, each green square represents a single denominator entity qualifying for inclusion during the reporting period.

Each green circle represents the total number of denominator entities qualifying for inclusion during the reporting period.

The entity temporal structure checks whether entities qualify for inclusion in a measure’s denominator on specific, recurring snapshot dates (e.g., the first of every quarter) aligned with reporting periods.

Temporal Structure: Interval

An interval-denominated measure’s observations represent fixed periods of time (e.g., member-months).

For example, a measure of a plan's per member per month spending is denominated by the sum of membership periods during a given month.

Interval-denominated measures, which use the interval temporal structure, calculate measure results based on the sum of the value of denominator interval contributions for each reporting period. The value of denominator contributions is determined by calculating the difference between the measure’s case effective start and end dates.

In the diagram, each green rectangle represents the number of days a patient’s timeline met the denominator inclusion criteria.

Each green circle represents the total number of days qualifying for inclusion during the reporting period.

The timeline period temporal structure calculates each reporting period’s denominator contribution, e.g., member-months, based on the total number of days associated with all cases during the reporting period.

Intervals that overlap reporting periods contribute days to both reporting periods. For example, Patient 2’s first interval overlaps Q1 and Q2. This interval contributes 90 days to Q1 and 5 days to Q2.

Temporal Structure: Static

Static measures include all cases associated with an object irrespective of a report's date range. This makes this measure type useful for validation and reporting objects that contain aggregate information. However, static measure’s lack of temporal information makes them unsuitable for most performance measurement exercises.