Excess risk and predictors of fracture/s following bariatric surgery for obese patients in the NHS: a real-world self-controlled case series and cohort study

Study type

• Study topic:
  Disease /health condition
• Study type:
  Non-interventional study
• Scope of the study:
  Assessment of risk minimisation measure implementation or effectiveness

  Disease epidemiology
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

  Other
• Non-interventional study design, other:
  Self-controlled case series

Study drug and medical condition

• Medical condition to be studied:
  Fracture

  Gastric bypass

  Gastric banding

  Gastric stapling

  Bariatric gastric balloon insertion

  Obesity

Population studied

• Short description of the study population:
  All patients aged 18 or above registered in a CPRD-linked HES practice with up to standard (UTS) data available between 01/04/1997 and 31/03/2016 and a BMI of 35 or above, undergoing bariatric surgery according to CPRD or HES records were eligible for Objectives 2 and 3.; Only those sustaining 1 or more fracture/s in the observation period (years before or after bariatric surgery) were included for the SCCS (case-only) analysis (Objective 1).
• Age groups:
  Adults (18 to < 46 years)

  Adults (46 to < 65 years)

  Adults (65 to < 75 years)

  Adults (75 to < 85 years)

  Adults (85 years and over)
• Estimated number of subjects:
  10989

Study design details

• Main study objective:
  Our specific objectives are:1. To study the risk of post-operative fracture risk amongst severely obese NHS patients undergoing bariatric surgery2. To determine key risk factors of post-operative fracture, and to derive a prediction tool the case-finding of high-risk patients at the time of pre-operative bariatric surgery assessment.
• Outcomes:
  Our primary outcome will be any osteoporotic fracture/s (excluding skull, face, or digits) in the 5 years following bariatric surgery, with sensitivity analyses using 0-3 and >3 year periods instead. Our secondary outcomes are 1.major fracture (hip, wrist/forearm, spine, and proximal humerus), and 2.peripheral fracture/s (wrist/forearm, ankle, and proximal humerus) in the same time frames. All the proposed outcomes will be ascertained using previously validated lists of READ codes.
• Data analysis plan:
  Self-controlled case series (SCCS) analysis will be used to study the association between bariatric surgery and fracture risk. IRR and 95% confidence intervals will be calculated for incident events observed within the “exposure” period compared to baseline (“non-exposure”) using the SCCS method and assuming a Poisson distribution. Sensitivity analyses will be conducted where the duration of the exposure period will be set at 0-3 and >3 years (post-surgery) respectively.A retrospective cohort design will be used to estimate the incidence rates and cumulative incidence of fracture in the years after bariatric surgery, and to identify key risk factors associated with an increased fracture risk in this population.For the derivation of a clinical prediction tool for post-operative fracture risk, logistic regression analyses will be conducted, where fracture/s in the post-operative risk period will be considered a binary outcome.