Prescription opioids, a type of potent painkiller, have increasingly been used over the past two decades, however, there is little evidence that they are effective for long-term pain. Previous research has found that the use of opioids is associated with an increased risk of bone fracture, however, there are possible limitations to this evidence due to methods used to indicate whether people used opioids, and to account for differences in peoples characteristics. Therefore, the purposes of this study are to:
1. Understand how prescription opioids are used in patients prescribed opioids between 2009 and 2017
2. Discover how many of these new opioid users had a bone fracture between 2009 and 2017, after they were first prescribed opioids
3. Find out how the use of opioids, dose and duration of use affect the risk of having a bone fracture, accounting for changes in how opioids are used over time.
This information will help to understand if opioids contribute to the risk of having a fracture and will support healthcare professionals to safely prescribe opioids to people with pain.
This self-controlled case series study aims to assess the association between prescription opioids and bone fracture using the CPRD GOLD (2009 to 2017) linking with HES. The objectives are to:
1. Quantify individual-level prescribing of opioids in a cohort of incident opioid users
2. Calculate the incidence rate of initial bone fracture outcomes in a cohort of incident opioid users
3. Evaluate the effect of current use, current and cumulative dose, and duration of opioid use on the risk of bone fracture in a cohort of incident opioid users
Patients initiating opioids between 2009 and 2017 will be followed to the occurrence of death, transfer out of the CPRD practice or the end of study. For each new opioid user, the duration and dose of each opioid prescription will be calculated. Their follow-up time will be categorised into periods of exposed time (from the prescription start date to the calculated prescription end date) and unexposed time (during which no opioid is prescribed).Daily dose for each exposed period will be calculated and converted to oral morphine equivalent dose (OMEQ).
Fracture outcomes will be identified from the CPRD and linked HES datasets. The association between prescription opioids and bone fracture will be assessed using self-controlled case series methodology; within-participant comparisons will be made between unexposed time, and exposed time, which will be split into risk periods. Risk periods will be the: 1st week of opioid use, 2nd week of use, 3rd and 4th weeks of use, and remaining time of opioid use. Gaps in exposure following periods of opioid use will be classified as unexposed time, in addition to a one-year unexposed period, prior to opioid initiation.
Poisson regression with adjustment for time-varying covariates, will provide incidence rate ratios and corresponding 95%CIs for risk of fracture during these risk periods compared to unexposed time.
Primary outcomes:
Bone fracture, any
Secondary outcomes: Composite of fracture sites
Osteoporotic fracture; vertebral fracture; fragility fracture; non-vertebral fracture
Secondary outcomes: ICD-10 fracture sites
Fractures of: multiple body regions; spine; upper limb; lower limb; skull and facial bones; neck; rib(s), sternum and thoracic spine; lumbar spine and pelvis; shoulder and upper arm; forearm; wrist and hand level; femur; hip; lower leg, including ankle; foot, except ankle
Roger Knaggs - Chief Investigator - University of Nottingham
Emily Peach - Corresponding Applicant - University of Nottingham
Andrew Cooper - Collaborator - Mundipharma Research Ltd
Fiona Pearce - Collaborator - University of Nottingham
Janice Fuller - Collaborator - Mundipharma Research Ltd
Li-Chia Chen - Collaborator - University of Manchester
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation