The recently published GUSTO results [1] led our department of cardiology to request that streptokinase no longer be designated as the formulary-approved thrombolytic agent of choice (with tPA reserved for patients with contraindications to streptokinase). The reduced mortality associated with tPA compared with that of streptokinase in the GUSTO investigation appeared to be related to the aggressive dosing method and the time from the onset of chest pain. The Pharmacy and Therapeutics Committee mandated a 6-month audit (done between 1 October 1993 and 31 March 1994) of all lytic therapy at our institution to establish the standard of practice and to define the eventual formulary status of thrombolytic agents. During this time, 25 patients received thrombolytic therapy for acute myocardial infarction, including 17 patients who received tPA and 8 who received streptokinase.

As a tertiary care teaching institution associated with Albany Medical College (a participating site in GUSTO) and as an institution that promotes itself as the "chest pain experts" on radio and television, we were surprised to find eight different regimens used in our 17 patients receiving tPA. Five different regimens were prescribed in the first nine patients receiving tPA. Other potential undesirable end points, such as inadequacy of heparinization during the first 24 hours after thrombolytic therapy (thought most crucial to prevent reocclusion), were not monitored. In addition, the time from the onset of chest pain to the initiation of thrombolytic therapy averaged about 5 hours for each agent (with large standard deviation), including an average of 1 hour between triage and treatment in the emergency department. On the basis of the GUSTO results, many of our patients may have been treated too late to benefit from the subsequent reduced mortality associated with tPA. The high degree of variability in weight (range, 55 to 167 kg) and age (range, 42 to 88 years) in our 25 patients suggests that standardized dosing of tPA (or heparin) might be inappropriate. Although not a primary focus of our audit, heparinization was initiated with a standard bolus and infusion dose and was not adjusted for age or weight.

We considered the GUSTO regimen (15-mg bolus, 50 mg over 30 minutes, and 35 mg over 1 hour) to be our new gold standard. The other tPA regimens used in our patients were not the original 3-hour administration protocol, nor were they adjusted for weight. In fact, the largest man treated received the smallest dose of tPA (Table 1). With streptokinase, 1.5 million units over an hour was administered to all eight patients with acute myocardial infarction who received that agent.

Most problems appeared to be in the emergency department, where the housestaff were unfamiliar with the GUSTO thrombolytic regimen and where the lack of an updated thrombolytic protocol was evident. Cardiologists approved tPA therapy and were specific about which thrombolytic agent their patients received but often did not specifically recommend a method of administration.

We presented our initial 3-month data on tPA to the emergency department, the coronary care unit physicians, the appropriate quality assurance committees, and the nurse clinicians. The result was a major effort to establish a new protocol and order sheet for both heparin and tPA in an effort to optimize administration. Both are now widely available through computer printouts from anywhere in the institution.

In summary, despite being an investigative site for GUSTO-1 and having several years of experience with thrombolytic therapy, our center's drug use evaluation identified a major problem in tPA prescribing that we feel could minimize any potential advantage of this thrombolytic agent. It also appeared that most of our patients were treated too long after the onset of chest pain to make the choice of tPA or streptokinase relevant. Our pharmacists were forced to delay therapy until an order was clarified, and, in some instances, they also could not correctly describe the GUSTO tPA administration method. As our drug use evaluation continued, we began piloting a thrombolytic order sheet and developed a heparinization order sheet with dosing based on body weight. The continuing audit showed remarkable progress with our thrombolytic prescribing after the order sheet was instituted, and our heparinization has improved with dosing based on body weight.

We caution institutions to investigate how they administer tPA and to interpret the GUSTO results in light of their own demographic data for time to treat, age, type of infarction, and other variables. Based on the tPA regimens our patients received, 8 of 17 patients received an administration regimen that may have been suboptimal (wrong dose, excessive duration of administration, and so forth).