Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to create a restricted cubic spline model adjusted for a categorical variable and with overlaid curves?

    Hello everyone I have been trying to create an adjusted cubic spline model with overlaid curves. I want to create a spline adjusted for a categorical/binary variable "Disease" that is 0=No, 1=Yes, and I want to overlay the two spline curves for each respective variable. Would appreciate some help. Thank you!.
    "Diameter" is the a numerical variable that stands for vessel diameter, "rupture" is a binary variable 0=No and 1=Yes, and "years" is also a numeric variable for years of follow-up. The syntax I use for the unadjusted spline is below,:

    ************************************************** ********
    mkspline spl_diameter=diameter, cubic nknots(3)
    mat knots = r(knots)

    stset years, failure(rupture==1)
    stcox spl_diameter*

    xbrcspline spl_diameter, values(10(10)60) ref(0) eform matknots(knots)

    sort diameter

    levelsof diameter
    xbrcspline spl_diameter, values(`r(levels)') ref(0) matknots(knots) eform gen(diameter_splines hratio lb ub) level(95)

    ************************************************** *********
    ***Plot Spline***
    twoway (line hratio lb ub diameter_splines, lp(1 - -) lc(red blue blue) ) ///
    if inrange(diameter,0,58) , ///
    scheme(s1mono) legend(off) ///
    ylabel(0(10)40, angle(horiz) format(%2.1fc) ) ///
    xlabel(0(10)60) ///
    ytitle("Hazards Ratio") ///
    xtitle("diameter")
    ************************************************** *********
Working...
X