S3R,
Data is collected by all the engine OEMs and sent to the airframe manufacturers; it is then combined with airframe failures (those that are not directly related to the core engine). The aggregated data is then distributed, as information, to (some) operators and as a more detailed analysis to the State's CAA.
In general, most engine/airframe combinations will meet a reliability rate of 1:100,000 hours. In recent years some manufacturers have achieved better than 1:1,000,000 hours for some of their types. On the other hand, some manufacturers have experienced issues and have not maintained their 1:100,000 reliability rate.
The 1:100,000 is the (engine/airframe) reliability rate that is used by some States to qualify their operators for the use of 'exposure' (where a failure would result in a hazardous or catastrophic event). When this target rate is not achieved, the manufacturers make a great effort to establish the cause and rectify it. To smooth the rate, a moving window of 5 years (data) is used - this removes problems due to skewing (because there are so few failures). However, as has been seen recently, failures tend to bunch; this calls for expeditious action as soon as a trend is noted.
The system used by a number of States (soon to be adopted by ICAO) allows failures to be washed out of the data when faults are established and rectification put into place.
In answer to your question: there does not appear to be a centralised system of data aggregation; however, it can be assumed that the 1:100,000 reliability is achieved unless information is present to the contrary. It should not be assumed that a turbine installed on one helicopter type achieves the same reliability as one installed on another; experience has shown that failures due to the installation are more prevalent.
Jim