I am trying to model the time until some event occurs for individuals observed over a 24 month period.

For about 75% of people, no event occurs. For 15% of people, we know exact time of the event. For the other 10%, we only know a time window in which it occurs. So some sample "survival times" might be:

>24 
2.5 
>24 
5.0 
0 to 6
6 to 12 
>24 
18 to 24 

and so on. Can this type of data be accommodated within "standard" survival models? If so, some hint at how to construct the Surv() object in R and how to pass it to survreg would be appreciated.