Re: The relation between total joint arthroplasty and risk for serious cardiovascular events in patients with moderate-severe osteoarthritis: propensity score matched landmark analysis
To the Editor,
In response to Ravi et al reply to our Letter-to-Editor, we want to comment the following:
1. The analysis we conducted on the Health in Men Study is epidemiological. In such epidemiological studies, all subjects are assumed to be at risk of developing a disease, or undergoing a procedure such as TJR. The inclusion of all 12,000 men to construct our propensity score is accurate and is based on the above assumption.
2. Propensity scores can be constructed using any multivariable regression models. The appropriateness of the selected model depends on the outcome investigated. (1)
3. Osteoarthritis in our cohort was not defined on self-report as Ravi et al wrongly assumed. This diagnosis was based on what was recorded in hospital records.
4. Last, and perhaps most importantly, Ravi et al conclusion that undergoing an arthroplasty may protect from having a cardiovascular event has no plausible biological explanation. Similarly, their analysis does not meet many other important conditions that inform causality. (2)
1. Guo S, Fraser WM. Propensity Score Analysis. Statistical Methods and Applications. 2011, SAGE Publications, Inc.
2. Hill AB. The Environment and Disease: Association or Causation? Proc R Soc Med 1965; 58:295-300.
Dr George Mnatzaganian and Professor Janet E Hiller
Competing interests: No competing interests