Iraj Daizadeh1. 1. Global Regulatory Affairs, Pfizer, Inc., 610 Main Street, Cambridge, MA, 02139, USA. iraj.daizadeh@pfizer.com.
Abstract
BACKGROUND: The number of original and supplemental ANDAs, BLAs, NDAs, and Biosimilars FDA drug/biologic approvals (Approvals) has risen dramatically in the recent years, incidentally, so has the number of issued FDA guidances (Guidances). It is hypothesized that if the structures of the two timeseries are similar and/or concomitantly co-evolving, then there is a relationship between the two variables that may be worthy of further investigation. METHODS: Structural breakpoint (SBP) and cointegration (CI) analyses are used to provide insights into the relatedness of the two timeseries (Approvals, Guidances). Various descriptive statistics (e.g., nonparametric correlation testing, decomposition, unit testing, stationarity, and maximum order of integration) were also performed to better understand the nature of the timeseries understudy. RESULTS: Structural breaks were identified with the following dates: Approvals (1983, 1989, 1996, 2004, and 2012) and Guidances (1995 and 2012). Approvals and Guidances were (medium) correlative, nonstationary, and cointegrated with a maximum order of integration of one (I(1)). Descriptive statistical markers suggest additional similarities (e.g., seasonal variation) between the two timeseries. CONCLUSIONS: To the author's knowledge, this is the first work to empirically investigate Guidances and their relationship with Approvals. The similarity in the structure of the timeseries (e.g., seasonal variation, SBPs and CI) suggests a deeper relationship between Guidances and Approvals, including the existence of a "long-run" equilibrium (wherein one or more exogenous factors restrain the divergence) between the two variables. This work offers an exciting opportunity for further research into the processes influencing the rates of Approvals and Guidances. A discussion on the limitations of the approach is also presented.
BACKGROUND: The number of original and supplemental ANDAs, BLAs, NDAs, and Biosimilars FDA drug/biologic approvals (Approvals) has risen dramatically in the recent years, incidentally, so has the number of issued FDA guidances (Guidances). It is hypothesized that if the structures of the two timeseries are similar and/or concomitantly co-evolving, then there is a relationship between the two variables that may be worthy of further investigation. METHODS: Structural breakpoint (SBP) and cointegration (CI) analyses are used to provide insights into the relatedness of the two timeseries (Approvals, Guidances). Various descriptive statistics (e.g., nonparametric correlation testing, decomposition, unit testing, stationarity, and maximum order of integration) were also performed to better understand the nature of the timeseries understudy. RESULTS: Structural breaks were identified with the following dates: Approvals (1983, 1989, 1996, 2004, and 2012) and Guidances (1995 and 2012). Approvals and Guidances were (medium) correlative, nonstationary, and cointegrated with a maximum order of integration of one (I(1)). Descriptive statistical markers suggest additional similarities (e.g., seasonal variation) between the two timeseries. CONCLUSIONS: To the author's knowledge, this is the first work to empirically investigate Guidances and their relationship with Approvals. The similarity in the structure of the timeseries (e.g., seasonal variation, SBPs and CI) suggests a deeper relationship between Guidances and Approvals, including the existence of a "long-run" equilibrium (wherein one or more exogenous factors restrain the divergence) between the two variables. This work offers an exciting opportunity for further research into the processes influencing the rates of Approvals and Guidances. A discussion on the limitations of the approach is also presented.