I am a PhD student at Swansea University looking at the psychological and health impact of failed dispatcher-assisted CPR on an emotionally close relative following out of hospital cardiac arrest.
I don't know how much research has been done on the impact on relatives from failed CPR out of the hospital. I know in the US there has been considerable research on families who stay in the hospital room when a code is being done on a close relative. There's debates about whether to allow the family to stay in the room or ask them to leave. Generally speaking, families who stay in the room even when the patient dies, do better. They know everything that needed to be done, was done is what they say. That research might be helpful.
There was a recent NEJM article that found found more positive outcomes was associated with family members being present in the room......http://www.nejm.org/doi/full/10.1056/NEJMoa1203366
My opinion is that "failed CPR" requires additional clarification. Are you thinking that if those at the scene tried to respond to their best ability and training but the patient did not (ultimately) survive that CPR failed? Are you thinking that if the family was trained, responded and had available and used an AED but the patient did not (ultimately) survive that this means there is failure? My opinion is that CA/SCA survival is an emergent property co-produced by the interaction of many elements. CPR is necessary but not sufficient. "CPR failure" is too vague for me (personally) because I do not see this as a root cause problem. Scroll through this if interested in a different perspective: http://www.systemswisdom.com/sudden-cardiac-arrest-survival
We've been performing research on CPR much in the way one would plan the saving of a troubled company: we made a list of the assumptions that many people make about CPR and then proceeded to test them, one at a time. We demonstrated that most people cannot perform guideline-compliant chest compressions ("GC3s") on even an average-stiffness chest for ten minutes. Additionally we found that most people negate the benefit of the CPR they perform by not achieving full recoil at the end of each compression. The ASPIRE, RISC, and CIRC trials failed to demonstrate the superiority of mechanical compression devices. We suspect strongly that the CPR performed by humans prior to application of the mechanical device was sufficiently flawed so that the quality of compressions that far into the code (10-15 minutes) was immaterial.
We're gearing up to test this hypothesis, and if you have any insight to share, We'd love to receive it.