Obviously you cant as there is no independent source. Obviously this direct entry must be detailed in the approved protocol as per ICH E6 R2 6.4.9. Caution should be exercised in using this method of direct entry into the CRF. It is preferable for there to be an independent source from which data are transcribed into the CRF. If there is no independent source than there should be a rationale for this method. Importantly the investigator must have , at all times, a local contemporaneous original copy of the data they have generated , under their control (see EMA IWG webpages Q&As) . There is a very good blog by the MHRA on whether your electronic devices are actually eCRFs (for instance eDiaries). >
Your Monitoring Plan (E6 R2 1.64; 5.18.6.e; 5.18.7) should details what data will be SDV /SDR. Using a Risk Proportionate approach to clinical Trials will help you create a Risk Assessment plan and from that a monitoring plan.