Spread data to create unique rows

I have a data frame that I want to reshape so that for each medid, there is only one row. I know that I can use spread to accomplish this, but it's not looking at all how I want to look and I just can't wrap my head around the concept when I read blogs, tutorials, etc. and I'm just not sure where I'm going amiss.

My data currently looks like

   medid year dose drug
1    101 2001  500    A
2    101 2002  600    A
3    101 2003  750    A
4    101 2004  550    A
5    102 2001  300    B
6    102 2002  330    B
7    102 2003  350    B
8    102 2004  390    B
9    103 2001  100    C
10   103 2002   NA    C
11   103 2003  250    C
12   103 2004  125    C

but I want it to look like

  medid  dose.2001 dose.2002 dose.2003 dose.2004 drug.2001 drug.2002 drug.2003 drug.2004
1   101       500       600       750       550         A         A         A         A
2   102       300       330       350       390         B         B         B         B
3   103       100        NA       250       125         C         C         C         C

I've tried to use spread() but it doesn't give me unique row for each GEOID, and it just isn't looking how I want.

df <- data.frame(medid=c(101, 101, 101, 101, 102, 102, 102, 102, 103, 103, 103, 103),
                 year=c(2001, 2002, 2003, 2004, 2001, 2002, 2003, 2004, 2001, 2002, 2003, 2004),
                 dose=c(500, 600, 750, 550, 300, 330, 350, 390, 100, NA, 250, 125),
                 drug=c("A", "A", "A", "A", "B", "B", "B", "B", "C", "C", "C", "C"))
df %>% spread(year, medid)