TY - JOUR
AU - Belot, Aurelien
AB -  Dear Sir or Madam, We have read with great interest a recent article by Crowther and Lambert . In this article, the authors cited one of our previous works and wrote (Section 2.2): Previous work by Remontet et al. used numerical integration, but used quadratic splines, limited to only two knots, with no restriction on the splines . Herein, we would like to draw the attention of the reader to this incorrect reading of our method. Indeed, the method we proposed ( , pages 2216–2217) enables the use of any type of regression spline including restricted splines (of any order and with any number of knots). For example, throughout the article, we used cubic splines (e.g., Equation 6, Figure 2, Table II or Figure 3). We used quadratic splines for comparison with the RSurv function , which handles only quadratic splines ( page 2218). In fact, in our approach, any parametric function, which is linear with respect to the parameters, could be used: this is the case of regression splines, restricted splines, fractional polynomials and so on. Thus, all these functions may be used to describe the log excess mortality baseline rate as well as the functional forms 
TI - Framework and optimisation procedure for flexible parametric survival models
JF - Statistics in Medicine
DO - 10.1002/sim.6489
DA - 2015-11-10
UR - https://www.deepdyve.com/lp/wiley/framework-and-optimisation-procedure-for-flexible-parametric-survival-zoawUSz8h0
SP - 3376
EP - 3377
VL - 34
IS - 25
DP - DeepDyve
ER -