To the Editor: The National Prescribing Service recently reviewed drug interaction information in a range of prescribing and dispensing software systems and reference sources, including the Therapeutic Goods Administration (TGA)-approved product information (PI). Our findings support those of Stockigt1 regarding the quality of information in PI.

An expert panel (see Acknowledgements) assessed 40 pairs of potentially interacting drugs (20 clinically important and 20 minor interactions) by examining the “Interactions” section of 80 PI monographs. The panel assessed both detection and content, using a range of parameters considered to be important in clinical decision making. For the clinically important drug interactions, these included useful management information (defined as sufficient information to proceed without looking elsewhere), pharmacological mechanism of interaction, clinical effects of the interaction, and time frame (onset, duration).

For clinically important interactions, 38 of 40 PI monographs listed the potential interaction; however, there were considerable shortcomings in the amount and quality of information provided. Only five of 38 monographs (13%) provided useful information about management. The mechanism was described in 15 monographs (39%) and clinical effects in 19 (50%); time frame was mentioned in only one (3%). In addition, there were many inconsistencies between any two monographs for a particular drug pair — not surprising given that the information is usually developed by different drug manufacturers at different times.

The results of our study suggest that in many cases the PI for a drug does not provide useful information about potential drug interactions. For decision support to be useful to health practitioners, information must be sufficiently detailed and relevant. Poor quality, irrelevant or inconsistent information is at best unhelpful or a waste of time, and at worst may prompt an inappropriate management decision, potentially compromising patient care. While epidemiological evidence for drug interactions is limited, good quality information and practical advice is available in a number of specialised drug interactions reference sources. This issue will be discussed in more detail with the full results of our study, which will be submitted for publication later this year.