UCSF Patient Populations
Patients can belong to more than one population, however, no matter how many populations are selected, or when 'All Patients' is selected, patients are only counted once.
- Primary Care 3YR: patients who are Primary Care empanneled at UCSF in the last 3 years prior to date of dashboard 'data as of' date.
- Specialty Care: patients with at least two visits in a specialty department (non-primary care and non-covid testing or vaccination departments) in the 2 years prior to date of dashboard 'data as of' date.
- Adult Cancer Center: patients with at least two visits in the cancer center (UCSF adult cancer center departments) in the 2 years prior to date of dashboard 'data as of' date.
- Inpatient: patients with at least 1 hospital inpatient encounter in the 2 years prior to date of dashboard 'data as of' date.
- Primary or Specialty: a distinct combination of the populations for Specialty Care and Primary Care 3YR. This means that if patients are part of both populations patient is only counted once.
- All of the above: patients who meet any of the criteria for the populations defined above.
Reference Populations
The American Community Survey (ACS), an initiative within the United States Census Bureau, acts to describe our 5,6 and 9-County Bay Area regional demographic features. The ACS was selected as a best practice since being utilized by UCSF for comparison of employee data.
The ACS methodology does have some distinctions from UCSF Health’s demographic collection methodologies. Specifically, ACS does not share data for race/ethnicity fields that are unknown or for those identifying as Southwest Asian, or North African. Additionally, ACS redistributes those reporting a race/ethnicity of Other into the main categories based on their specific methodology, when providing race/ethnicity data by age group.
In addition to the use of the ACS as a data-source for reference population, the dashboard also leverages Census Bureau county-level data, which has a methodology that better aligns with UCSF-Health’s approach to our derived race/ethnicity variable.
- 9-County Bay Area: Population includes persons who reside in the following counties Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, and Sonoma counties, based on data from the ACS 2020 5-year or Census 2020 data sets.
- 6-County Bay Area: Population includes persons who reside in the following counties Alameda, Contra Costa, Marin, San Francisco, San Mateo, and Santa Clara counties, based on data from the ACS 2020 5-year or Census 2020 data sets.
- 5-County Bay Area: Population includes persons who reside in the following counties Alameda, Contra Costa, Marin, San Francisco, and San Mateo counties, based on data from the ACS 2020 5-year or Census 2020 data sets.
- San Francisco: Population includes persons who reside in the San Francisco county based on data from the ACS 2020 5-year or Census 2020 data sets.
Derived Race-Ethnicity
At UCSF Health, patients are asked to self-identify their race and ethnicity, which are captured and stored in 3 separate data fields (Race, Ethnic Group, and Ethnicity) in APeX. However, when describing our populations and aggregating metrics, we often only reported on 1 of the 3 fields that described race or ethnicity due to the complications added when stratifying by more than 1 field. The Health Equity Division, Data Equity Taskforce, and Health Equity Council at UCSF Health have worked together to develop and maintain a derived database element that combines these 3 race/ethnicity fields using an algorithmic categorization scheme to provide a standard methodology for describing and aggregating our populations across UCSF Health.
For the non-UCSF reference populations, the race-ethnicity category shown is derived from the 2015 American Community Survey (ACS) 5yr data set. We chose to use the 2015 data set due to the fact that ethnicity combined with age and county of residence was no longer captured in the ACS data after 2015. We needed to distinguish if a person in the county identified as Hispanic or Latino to appropriately categorize them with our UCSF derived race-ethnicity categories.
Language Group
At UCSF Health, patients are asked to self-identify their preferred language for communication with the health system. This language is then categorized according to if it is English or not. Those that are not English are categorized as Limited English Proficiency.
For the non-UCSF reference populations, language group represents the language selected as spoke at home in the 2015 American Community Survey (ACS) 5yr data set. We chose to use the 2015 data set due to the fact that ethnicity combined with age and county of residence was no longer captured in the ACS data after 2015. We needed to distinguish if a person in the county identified as Hispanic or Latino to appropriately categorize them with our UCSF data. Therefore, to ensure a consistent population denominator language spoken at home was also derived from this data set.