Author + information
- Andrea Messori, PharmD∗ (, )
- Valeria Fadda, PharmD,
- Dario Maratea, PharmD and
- Sabrina Trippoli, PharmD
- ↵∗HTA Unit, Area Vasta Centro Toscana, Regional Health System, Via Guimaraes 9-11, 59100 Prato, Italy
In the comparison made between same-day discharge and routine overnight observation by Brayton et al. (1), the best level of evidence was that derived from randomized controlled trial (RCT) studies. With regard to the composite outcome of death, myocardial infarction, or target lesion revascularization, the analysis of 7 RCTs showed no significant differences between the 2 approaches (pooled odds ratio: 0.90; 95% confidence interval: 0.43 to 1.87; p = 0.78). While this result supports the conclusion of no proven difference, the main question is whether or not the available data support a conclusion in terms of proof of no difference (or proof of noninferiority) rather than the mere demonstration of no proof of difference.
One problem in exploring these methodological questions is that the degree of consensus on how noninferiority meta-analysis can be conducted is still modest (2). On the other hand, the advantages of trial sequential analysis (TSA) are being recognized increasingly not only for handling questions of superiority (3,4) but also with regard to those of noninferiority (3–5); in fact, TSA aims at classifying each meta-analysis into one of only 4 categories (superiority, inferiority, futility/noninferiority, inconclusive result).
We applied TSA to the same 7 RCTs examined by Brayton et al. (1). Our assumptions included 2-sided testing, type 1 error of 5%, and a power of 80%. With respect to the above-mentioned composite outcome, the intervention effect was set at a relative risk reduction (RRR) of 50% or 33%. The expected absolute event rate in the controls was 7.6% (i.e., the cumulative arithmetic rate in the control groups of the 7 RCTs). The main result of TSA was expressed through the graph of a cumulative z-curve; the boundaries in this graph for concluding superiority or inferiority or futility were determined according to the O'Brien–Fleming alpha-spending function. All calculations were carried out using a specific statistical software (User Manual for TSA, TSA, Copenhagen Trial Unit 2011).
Our results are shown in Figure 1. The number of events recorded in the RCTs proved to be insufficient to construct the boundaries of futility in both analyses; in addition, the statistical procedure incorporated only 6 RCTs because one trial (characterized by zero-event frequency in both arms) was uninformative according to the TSA statistical algorithm.
Our results indicate that current information from RCTs does not allow us to draw any firm conclusion about the outcome comparison between the two approaches (i.e., “inconclusive result” of TSA). In fact, while the overall number of patients enrolled in the 6 trials was 2,555, our TSA estimated that the optimal information size would be 10,752 patients (assuming RRR = 50%) or 27,243 patients (assuming RRR = 33%).
In summary, the number of patients studied in the RCTs presently available is only one-fourth or one-tenth in comparison with the ideal sample size required to draw a firm conclusion. Therefore, the comparison between the 2 discharge strategies remains open.
- American College of Cardiology Foundation
- Brayton K.M.,
- Patel V.G.,
- Stave C.,
- de Lemos J.A.,
- Kumbhani D.J.
- Liberati A.,
- D'Amico R.