MPOG Research Standard Data

MPOG Research Standard Data is a standard collection of MPOG phenotypes and commonly used concepts for MPOG research projects. It is designed to provide investigators with a reliable starting point by focusing on data that are widely available, consistently mapped, and well–validated across participating MPOG institutions.

Standard Data for MPOG Research

Includes the following:

  • A recommended starting cohort based on the Research Standard – Modern Active US Sites phenotype 
  • Curated phenotypes with standardized clinical definitions
  • Commonly used concepts with broad institutional representation that have demonstrated reliability across MPOG sites 

See more details about each type of included data in the dropdowns below.

The recommended starting cohort for most MPOG studies is the Research Standard – Modern Active US Sites phenotype. This phenotype is designed to identify a contemporary, high-quality research population by including cases that:  

  • Meet the Intraoperative Research Standard:

    • Has valid Anesthesia Start and End

    • Patient in room time should be valid if exists

    • Anesthesia Duration must be at least 10 minutes for GA cases

    • Anesthesia Duration must be at least 5 minutes for non-GA cases

    • Age must be present

    • Sex must not be unknown

    • ASA Class must not be Missing, Invalid, or Conflicting

    • At least one valid blood pressure value must be present

    • At least one intraoperative medication was administered 

  • Occurred at an MPOG Institution within the United States (for which associated concepts such as Common Procedural Terminology [CPT] codes are available) 
  • Occurred within the most recent six-year period 
  • Came from an institution contributing at least 1,000 eligible cases during each included calendar year of the past six years 

Additional details are available in the MPOG Data Library – Phenotype Browser (Note: Access to detailed phenotype logic may require an MPOG Data Library login and completion of an attestation).

MPOG Phenotypes are strongly encouraged because they provide standardized definitions for common patient characteristics, exposures, procedures, and outcomes. The underlying phenotype logic is continuously validated across institutions to enable improved interoperability across MPOG sites with heterogenous electronic health record systems and data documentation patterns.  

Available phenotypes are listed in the MPOG Data Library – Phenotype Browser. Here you will find the accompanying definitions, enumerations, and logic for each phenotype (Note: Access to detailed phenotype logic may require an MPOG Data Library login and completion of an attestation). 

When a concept is available as an MPOG phenotype (e.g., patient height, weight, sex) use of the phenotype is preferred. However, with careful curation by a study team, certain additional MPOG data concepts are amenable to analysis within a research project due to their standard definitions and availability across a majority of active MPOG sites contributing data. Such concepts include: 

  • Intraoperative Medications (if concept documented by >50% of institutions) 
  • Laboratory or Testing Observations (if concept documented by >50% of institutions) 
  • Physiologic Observations (if concept documented by >50% of institutions) 
  • Surgeon and Anesthesiologist Staffing Type 

Available concepts are listed in the MPOG Data Library – Concept Browser. Here you will find the number of occurrences, cases and institutions the concept is available for as well as accompanying definitions and return columns. By reviewing the “institution count” column, it can be determined if a particular concept (within the concept types listed above) is documented by a majority (i.e. >50%) of institutions. 

MPOG Research Standard Data is intended to support the majority of MPOG research projects. Projects that require a more specific starting cohort or less commonly captured MPOG data may still be feasible using MPOG Extended Data, but should be carefully evaluated for availability, completeness, site representation, and cross-institutional consistency. Research teams should anticipate substantial additional effort for data curation and validation of data within MPOG Extended Data before beginning primary analyses.  

For any inquiries please email
mpog-research@med.umich.edu