Association Between Alendronate and All-Cause Mortality and Cardiovascular Mortality Among Hip Fracture: An Alternative Explanation
Bibliographic record
Abstract
To the Editor: Sing and colleagues1 observed that individuals on alendronate (ALN) treatment had a 67% reduction in the risk of cardiovascular (CVD) mortality and 45% reduction in the risk of myocardial infarction (MI). These remarkable effect sizes deserve more explanations, and here we would like to offer an alternative interpretation relating to bias. In the study, the treated group included patients who had been on more than one anti-osteoporosis medication, and these patients were analyzed as being censored at the time of switching medication. These patients contributed to the time of follow-up in the treated group (ie, the denominator of the mortality rate), but they did not experience any event of interest because they were censored at the switching time (ie, the numerator was 0). It is not possible to examine the magnitude and direction of this bias because of no detail on how many patients had been on multiple medications. However, in the Canadian Multicentre Osteoporosis Study, we found that as many as 40% of treated patients had been given more than one anti-osteoporosis medication during the study period2 and that including these treated patients would substantially inflate the effect size. Thus, we expect that if the censoring data were handled more sensibly, the treatment effect would be lower than reported. We consider that a more sensible question to ask is that given the data at hand, what is the chance that ALN reduces CVD mortality risk at the level of clinical significance? This question can be addressed by a Bayesian analysis,3 which takes into account prior knowledge of the effect size and existing data. Two recent meta-analyses4-6 suggest that bisphosphonates reduced the risk of mortality by 1% to 20% (Table 1). Using results of previous meta-analyses as prior information and given the authors’ data, our analysis suggests that there is a very low probability (almost 0) that ALN reduces all-cause mortality risk or CVD mortality risk by more than 50% (Table 1). Moreover, there is less than one-third chance that ALN reduces all-cause mortality risk or CVD mortality risk by more than 30%. Even with the most optimistic prior information, there is less than 90% chance that ALN is associated with a 15% or greater reduction in CVD mortality and all-cause mortality risks. In summary, our interpretation of Sing and colleagues’ data is that the large observed risk reduction in mortality associated with alendronate could be attributable to the treatment of censoring data. When considered in relation to existing data, Sing and colleagues’ data suggest that in hip fracture patients, alendronate is statistically associated with reduced mortality risk, but the magnitude of association is likely modest.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.018 | 0.100 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.007 | 0.001 |
| Research integrity | 0.011 | 0.017 |
| Insufficient payload (model declined to judge) | 0.005 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".