Steenland MW, Rodriguez MI, Cohen JL. Changes in the supply duration of combined oral contraception during the coronavirus disease 2019 (COVID-19) pandemic. Obstet Gynecol 2022;139.
The authors provided this information as a supplement to their article.
©2022 American College of Obstetricians and Gynecologists. Page 1 of 2
Appendix 1.
Power calculation:
We used the following methods to conduct a power calculation for the study outcome. We create a simulated dataset for the study outcome by taking a random draw from a binomial distribution for each week in the study. The total number of trials in each weekly binomial distribution was equal to the total number of claims for first prescription pick-ups in that week and the success probability was equal to the percent of first pick-ups in that week. We then simulated outcome data by taking random draws from the baseline distribution to create a dataset containing 108 weekly observations for receipt of a >1 month COC supply. We repeated this simulation 1,000 times, with the simulated data containing the sample size of the analysis (n=108 weeks) and the interrupted time series analytic strategy described in the study’s methods. We find that I have sufficient power at alpha <0.05 to calculate a minimal detectable effect size of 0.017 percentage points for the weekly change in trend, and 0.7 percentage points for the immediate change in level.
Table 1: Percent of p-values <0.05 out of 1,000 simulations for an increase in the trend in the study outcome (baseline=47.58)
0.010 0.011 0.012 0.013 0.014 0.015 0.016 0.017
0.0 0.7 3.9 20.3 41.5 72.4 91.8 98.0
Table 2: Percent of p-values <0.05 out of 1,000 simulations for an increase in the level of the study outcome at the start of the pandemic (baseline=47.58)
0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 2.4 21.9 64.5 93.8 99.4 100
Steenland MW, Rodriguez MI, Cohen JL. Changes in the supply duration of combined oral contraception during the coronavirus disease 2019 (COVID-19) pandemic. Obstet Gynecol 2022;139.
The authors provided this information as a supplement to their article.
©2022 American College of Obstetricians and Gynecologists. Page 2 of 2
Supplement Figure: Percent of first pick-ups for combined oral contraception with a supply of more than one month, May 2019–December 2020.
Notes: Each plot shows a trend line based on predicted values from two linear regression models (one line before the start of the COVID-19 pandemic, from May 1, 2019 through March 17, 2020, and a second line starting the week after the pandemic was declared a national emergency, from March 18 2020 through December 31, 2020. Vertical dotted lines at March 17, 2020 represent the first week after the pandemic was declared a national emergency.