Networked control systems are often subject to limited communication resources. By only communicating output measurements when needed, event-triggered control is an adequate method to reduce the usage of communication resources while retaining desired closed-loop performance. In this work, a novel event-triggered control (ETC) strategy for a class of nonlinear feedback systems is proposed that can simultaneously guarantee a finite L-p- gain and a strictly positive lower bound on the inter-event times. The new ETC scheme can be synthesized in an output-based and/or decentralized form, takes the specific medium access protocols into account, and is robust to ( variable) transmission delays by design. Interestingly, in contrast with the majority of existing event-generators that only use static conditions, the newly proposed event-triggering conditions are based on dynamic elements, which has several advantages including larger average inter-event times. The developed theory leads to families of event-triggered controllers that correspond to different tradeoffs between ( minimum and average) inter-event times, maximum allowable delays and L-p- gains. A linear and a nonlinear numerical example will illustrate all the benefits of this new dynamic ETC scheme.