Re: Venous thrombosis in users of non-oral hormonal contraception: follow-up study, Denmark 2001-10
Thank you to Julie M Chandler et al. (JMC) for their rapid response1 to our new cohort study on hormonal contraception and venous thrombosis2.
JMC expresses two concerns, the first about adjusting for length of use, and the second about missing adjustment for known risk factors of venous thrombosis, such as BMI, family disposition and in some studies smoking.
Length of use.
Since 1994 all prescriptions for hormonal contraception have been collected electronically from all pharmacies in Denmark by law, together with a personal identification number for each patient. From this National prescription registry we have access to daily updated information about exposure to hormonal contraception for all Danish citizens. This database thus provides not only more complete, but also more reliable exposure information than any other data source for the study of hormonal contraception.
Our study from the Danish National Prescription Registry thus included 1.6 million women of reproductive age. The study window was limited to the ten-year period 2001-2010, but we had exposure information back to 1995. This information allowed us to categorize women into the appropriate length of use group as of January 1, 2001, and to adjust our estimates for length of use. Although combined oral contraceptives (COC) with levonorgestrel has been used for several decades, there are still new users of this product group. Likewise, many users of NuvaRing® have been using other types of COC before their switch to NuvaRing®. Any implication that NuvaRing® users are all new, naive users of hormonal contraception and all levonorgestrel users are the opposite is therefore incorrect.
In addition to adjusting for length of use, we also stratified our analysis into different duration of use categories. Also for the long term users (>4 years) we found significant differences in risk of VTE between users of NuvaRing® and COC with levonorgestrel. Therefore the authors’ contention, that we did not take length of use into account, is incorrect. On the contrary, we did control for length of use, and probably did so better and more precisely than done in the majority of other studies.
It is always a good thing to adjust for other risk factors in epidemiological studies. It is, however, important to discriminate between risk factors and confounders. Only missing adjustment for confounders will bias an estimate. Whether a certain risk factor is a confounder depends on the distribution of this risk factor among users and non-users of hormonal contraception and between users of different types of hormonal contraception. Although body weight, family disposition and smoking are risk factors, these risk factors have been found to equally distributed between users and non-users and between users of different types of hormonal contraception in studies with access to this information3-8. This observation has been documented in Denmark, The Netherlands, Germany, Great Britain, Israel and in the USA. Therefore our lack of control for these risk factors is unlikely to have biased our results. And in studies with access to these risk factors, adjustment did not change the estimates.
Studies by Dinger.
We acknowledge that the company sponsored studies by Jürgen Dinger et al. did not found any difference in risk of venous thrombosis between users of COC with levonorgestrel versus users of COC with third generation progestogens (desogestrel or gestodene), or between users of COC with levonorgestrel and users of COC with drospirenone. Several methodological circumstances, however, explain these results. That Dinger et al. do not find any difference in risk of venous thrombosis between users NuvaRing® and users of COC could be due to the same methodological problems, as in his previous studies. It does not appear from the response from JMC whether users of COC with third generation progestogens are included in the comparison group of uses of COC in the new study by Dinger. This is crucial, because these COC confer the same risk of venous thrombosis as NuvaRing®. We look forward to the publication of these results to see whether inclusion of women with previous thrombosis, and pregnant and puerperal women in the control population again bias the results by Dinger et al.
None of the concerns expressed by JMC are likely to have biased the results in the large Danish historical cohort study.
1. Chandler JM, Rekers H, Kaplan K, Goldstein B. Re. Venous thrombosis in users of non-oral hormonal contraception: Follow-up study, Denmark 2001-10. BMJ 2012; http://www.bmj.com/content/344/bmj.e2990/rr/585117
2. Lidegaard Ø, Nielsen LH, Skovlund CW, Løkkegaard E. Venous thrombosis in users of non-oral hormonal contraception: follow-up study, Denmark 2001-10, BMJ 2012;344:e2990.
3. Lidegaard Ø, Edström B, Kreiner S. Oral contraceptives and venous thromboembolism. A five-year national case-control study. Contraception 2002; 65: 187-96.
4. Vlieg AVH, Helmerhorst FM, Vandenbroucke JP, Doggen CJ, Rosendaal FR. The venous thrombotic risk of oral contraceptives, effects of estrogen dose and progestagen type: results of the MEGA case-control study. BMJ. 2009; 339: b2921.
5. Parkin L, Sharples K, Hernandez RK, Jick SS. Risk of venous thromboembolism in users of oral contraceptives containing drospirenone or levonorgestrel: Nested case-control study based on UK General Practice Research Database. BMJ 2011: 340: d2139.
6. Jick SS, Hernandez RK. Risk of non-fatal venous thromboembolism in women using oral contraceptives containing drospirenone compared with women using oral contraceptives containing levonorgestrel: case-control study using United States claims data. BMJ 2011; 340: d2151.
7. Food and Drug Adminidsration, Office of surveillance and epidemiology. Combined hormonal contraceptives (CHCs) and the risk of cardiovascular disease endpoints. FDA 2011: http://www.fda.gov/downloads/Drugs/DrugSafety/UCM277384.pdf.
8. Gronich N, Lavi I, Rennert G. Higther risk of venous thrombosis associated with drospirenone-containing oral contraceptives: a population-based cohort study. CMAJ 2011; DOI:10.1503/cmaj.110463.
Competing interests: The author has within the last three years received honoraria for speeches in pharmacoepidemiological issues, including fees from Bayer Pharma Denmark, MSD Denmark, and Theramex, Monaco, and has been expert witness for plaintiff in a legal US case in 2011.