MBSAQIP Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program - onetomapanalytics/Meta_Data GitHub Wiki

MBSAQIP - Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program

General description

  1. Database primary purpose - To provide researchers at participating MBSAQIP sites with a data resource they can use to investigate and advance the quality of care delivered to the metabolic and bariatric surgical patient through the analysis of cases captured by MBSAQIP
  2. Overall data type - Health outcomes
  3. Dataset type - Cross-sectional
  4. Data source - Registry
  5. Data level - Patient level
  6. Geographic location of the data collection sites - United States and Canada
  7. Sponsor, manager, or home institution - American College of Surgeons (ACS) and the American Society for Metabolic and Bariatric Surgery (ASMBS)
  8. Date range - 2015 - 2020
  9. Dates - Date of birth, hospital admission, procedure, discharge, complication, Emergency Department (ED) visit, and death
  10. Clinical areas of interest - Bariatric operative procedures
  11. Number of records - 2020: 168,568 cases submitted by 885 centers; 2019: 206,570 cases submitted by 868 centers; 2018: 204,837 cases submitted by 854 centers; 2017: 200,374 cases submitted by 832 centers; 2016: 186,772 cases submitted by 791 centers; and 2015: 168,903 cases submitted by 742 centers (1)
  12. Variables that are uniquely present in this dataset - The MBSAQIP PUF is the largest, bariatric-specific, clinical dataset in the country and serves as an invaluable resource to investigators looking to answer important clinical questions in this field (1)
  13. Database caveats and limitations - (1) The sex and race distributions are representative of the national surgery patient population, however, only patients over 10 are available for assessment. PUF de-identifies patients above 80 years old (age is set to missing with an indicator variable included to identify patients over the age of 80). (2) Absolute dates were removed to comply with HIPAA. Date of surgery has been restricted to merely the year. Some dates (hospitalization, lab tests, etc.) were recoded as durations (e.g., Date of Admission and Date of Discharge are recoded into Days to Discharge from Hospital Admit). (3) To comply with the Participation Agreement (PA) that is agreed to between the ACS and participating centers, facility identifiers, as well as geographic information regarding the case, have been removed. (4) While many risk factors are tracked, preventative measures are not recorded, which can lead to an underestimation of the risk of certain conditions when such measures are routinely taken before surgery. (5) The data are submitted from centers that are participating in the MBSAQIP and may not represent a statistically valid nationally representative sample. (6) Many patients do not receive all possible preoperative laboratory tests, so some of these variables have a high percentage of missing values (10% to 70%, depending on the tests). This high percentage of missing data can make it problematic to use these variables in a traditional logistic regression model as well as in many other types of analysis.
  14. Other - The following procedures would not meet metabolic or bariatric inclusion criteria: cancer cases (patient admitted to the hospital and has an included procedure to address cancer), trauma cases (any patient admitted to the hospital and has an included procedure to address a traumatic injury), the patient is under 10 years of age, multiple MBSAQIP assessed cases within 30 days (any patient who had an MBSAQIP assessed procedure entered within the previous 30 days at the center, the additional metabolic or bariatric procedure performed within 30 days is only entered as a reoperation or intervention. Only one MBSAQIP procedure can be entered into the data registry per patient, per 30 days, for a center).

Applicable methods

  1. Association methods, such as univariate analysis (2), bivariate analyses (3, 4), multivariate regression (5, 6), logistic regression (7)
  2. Subgroup analysis (8)
  3. Propensity scores (9, 10)

High-impact designs

  • MBSAQIP practical guide to surgical data sets (12)

  • Evaluate the rates of use and efficacy of stent placement for postoperative leak management (13)

  • Racial disparities in physician's decision-making (14)

  • Assess perioperative risk factors (15)

  • Determine disparities between U.S. adolescents class II and III obesity trends and bariatric surgery utilization (16)

  • Compare different techniques on outcomes (17, 18, 19, 20)

  • Evaluate the performance of bariatric surgery prior to and after the implementation of MBSAQIP (21)

Data dictionary

To access the MBSAQIP data dictionary, click here

Variable categories

  1. Patient demographics (e.g., date of birth to calculate patient’s age, sex, race, and ethnicity)
  2. Patient characteristics (e.g., height, weight, highest recorded weight, functional health status, smoker, patient history)
  3. Hospital discharge records (e.g., primary discharge diagnosis, dates of admission and discharge, discharge destination)
  4. Diagnosis (admitting, principal, and "other" diagnosis code)
  5. Procedure (e.g., type, CPT code, initial description, medical specialist)
  6. Occurrences