Objective: The objective of this study was to determine whether pharmacy records of nitrate prescriptions could be used as a marker of angina pectoris.
Method: This study was conducted within the Rotterdam Study, a prospective follow-up study which started in 1991 and included 7983 elderly subjects. During follow-up, 1601 subjects filled a first prescription for a nitrate and later filled at least one other prescription for nitrates according to pharmacy records. After excluding subjects who started using nitrates in 1991 and who had less than one year of medication history, we took a random sample of 78 subjects (10%). We studied discharge and outpatient cardiologist letters and files from general practitioners for additional information on angina pectoris in these subjects, and allocated patients to one of three categories according to the possibility of the initial diagnosis of angina being correct.
Results: From the random sample of 78, additional information was available on 75 subjects. Definite angina pectoris was present in 33, probable angina pectoris in 18, and possible angina pectoris in 19 subjects. Five subjects had no angina pectoris. Therefore, 93% had at least a possible diagnosis while 68% had at least a probable diagnosis of angina pectoris. The positive predictive value of 2 nitrate prescriptions of which at least one was for rescue therapy was 94%.
Conclusions: We conclude that the use of more than one nitrate prescription can be used as a marker for angina pectoris. This marker may be useful in epidemiological studies.