Osteoporosis – a health condition that weakens bones – has become increasingly common among both women and men after the age of 50, leading to bone-related morbidities, elevated mortality, and growing health care cost. Major clinical guidelines recommend pharmaceutical interventions to prevent osteoporosis-related adverse outcomes. The justification of this recommendation in part lies in the growing evidence from randomized-controlled trials for the effectiveness of osteoporosis medications. While clinical trials are considered the method of choice to test for effective treatment, they may not fully capture real-life effects occurring during routine care. Moreover, many clinical trials lack the long-run perspective necessary to evaluate treatments. This study focuses on an analysis method – called regression discontinuity or “RD” – that aims to determine if patients profit from a given treatment. RD achieves this by exploiting a threshold (cut-off) in the T-score for bone mineral density that influences whether or not someone receives osteoporosis medications. By “zooming in” around the threshold, it can be assumed that people whose T-score for bone density lies just above and just below the threshold are similar to one another and can, thus, be compared in the same fashion as treated and untreated patients in a clinical trial. Using the CPRD, we will assess the long-term (up to 10 years) benefits and adverse side effects of osteoporosis medications commonly recommended by major clinical guidelines. Given the growing prevalence of osteoporosis, findings from this study will have direct implications for clinical practice.
This study seeks to evaluate the benefits and adverse side effects of osteoporosis medications among primary care patients in the UK using the regression discontinuity (RD) method. The RD design can be used to assess the causal effect of osteoporosis medications on health outcomes by taking advantage of a decision rule used to determine patients’ eligibility for pharmaceutical treatment of osteoporosis. Specifically, we aim to (1) determine whether or not there is a substantial change in the probability of using osteoporosis medications between patients whose T-score for bone mineral density falls just below and above an arbitrary cut-off point as published in major clinical guidelines, (2) evaluate if patient characteristics are balanced within a small bandwidth surrounding the threshold, and (3) investigate whether associations between osteoporosis medications and patient outcomes are robust to different choices of bandwidth around the threshold. Patient outcomes include future bone density T-scores, probability/number of major osteoporosis fracture events, probability of any adverse side effects (e.g. musculoskeletal pain, osteonecrosis of the jaw, atypical thigh bone fracture, ulcer, and hypocalcemia), and all-cause mortality. We hypothesize that the use of osteoporosis medications decreases future T-scores for bone density, the probability/number of fractures, and mortality. We will estimate “fuzzy” RD models using a local linear regression and triangular weights to avoid overfitting the data and to give more influence to observations close to the threshold. In addition, we will use a mean squared error (MSE) optimal bandwidth that is empirically derived. We assess the sensitivity of the results using alternative bandwidths (e.g. bandwidths that are 50%, 75%, 125%, and 150% of the empirically derived bandwidth). Our findings will provide valuable insights into the real-life effects of osteoporosis medications on health, the unintended effects associated with these medications, and the heterogenous treatment effects by detailed patient subgroups.
Primary outcomes (measured within five years from an index date): future bone density T-scores.
Secondary outcomes (measured within one year, five years, and ten years from an index date): Probability/number of major osteoporosis fracture events (e.g. vertebrae, humerus, wrist, femur, and hip fractures), all-cause mortality, probability of any osteoporosis-related unintended adverse side effects (including but not limited to musculoskeletal pain, osteonecrosis of the jaw, atypical thigh bone fracture, ulcer, and hypocalcemia); probability of at least one hospitalization for osteoporosis-related adverse side effects; number of hospitalizations for osteoporosis-related adverse side effects; probability of at least one all-cause emergency hospitalization; number of all-cause emergency hospitalizations; probability of at least one fracture-related hospitalizations; and number of fracture-related hospitalizations; probability of at least one osteoporosis-related hospitalizations; and number of osteoporosis-related hospitalizations. Other secondary outcomes include future bone density T-scores (measured within one year and ten years from an index date)
Till Bärnighausen - Chief Investigator - University of Heidelberg
Duy Do - Corresponding Applicant - University of Heidelberg
Anant Jani - Collaborator - University of Oxford
Christian Bommer - Collaborator - University of Heidelberg
Julia Lemp - Collaborator - University of Heidelberg
Justine Davies - Collaborator - University of Birmingham
Michaela Theilmann - Collaborator - University of Heidelberg
Pascal Geldsetzer - Collaborator - University of Heidelberg
2011 Rural-Urban Classification at LSOA level;HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation