Medical cyber-physical systems (MCPS) have been appeared as a possible approach for detecting and mitigating human errors. In this context, a MCPS in which medical and non-medical devices are connected could promptly detect potential risky or wrong procedures that deviate from standardized approaches. In this work, such a paradigm has been expanded to prototypical diagnostic devices. A distributed platform of such instruments is proposed in order to avoid misdiagnosis: diagnostic instrument send data to a message-oriented middleware, thus allowing technicians and doctors to remotely access test-related information for further processing and for creating an historical database. A Broker has been used for collecting data and distribute them to cloud database and to users. Two of the most diffused protocols, i.e., MQTT and AMQP, have been considered. An experimental setup has been developed to verify performance; time-related metrics confirm that the proposed approach has an end-to-end delay on few hundreds of milliseconds even for geographical scale networks. © 2019 IEEE.