I was interviewed yesterday by Michael Woodhead, a health journalist from Australian Doctor regarding my thoughts on the new interesting study by Dinger et al. (2014), on the risk of venous thromboembolism (VTE) from drospirenone containing combined oral contraceptive pill. The article containing my commentary was published today.
My perspective
The following is an except of an e-mail that I sent to Michael:
This is an interesting new study and it seems to have been well designed. It demonstrates results that are somewhat different to some prior results, but consistent with some others. It should be acknowledged that are variations in estimates (of incidence of VTE, and differences between drospirenone and levonorgestrel containing cOCP) in the evidence base.
The majority of the research in this field is in the form of observational (as opposed to experimental, like a randomised trial) data. This includes this study. There is always a possibility of unaccounted confounders and biases. The discussion in this study suggests that some of the results from past studies may have been due to these confounders. This could be the case.
It is also possible that this study may have biases that would favour drospirenone, e.g., prescribers choosing levonorgestrel/non-drospirenone containing cOCP to women they intuitively feel to be at higher risk of VTE, and drospirenone containing cOCP to those they feel to be at low risk. For instance, in table 3 we see that there were generally more smokers/heavy smokers, and people who had diabetes and CVD on levonorgestrel/non-drospirenone. It may be that this “clinical gestalt” is not well measured in the assessment questionnaires.
This is not to say that the results of this study are invalid. Far from it – it provides valuable data on an important question. As before, it does certainly seem to be quite well designed. It doesn’t in my mind, however, represent an overturning of the previous data either.
My take home message? This study suggests that the previous estimates that drospirenone containing cOCP approximately doubles the risk of VTE compared to levonorgestrel may have been over-estimates. This is important as some have suggested that this may have been an under-estimate. I think that the conclusion that the risk from drospirenone for VTE is less than double that of levonorgestrel is very likely to be true. The more assertive claim that drospirenone-OCP carries no higher risk than other cOCP I would hold as suspect.
We do need to consider absolute risks as well for perspective. The absolute risk for VTE of being on any cOCP (compared to non-use) is higher than the difference in risk between different types of cOCP – and all these risks are quite small.
