The review included 112 trials (21,224 participants). Of these 97 had no reporting bias for occurrence of deaths, six were unclear and nine had a reporting bias.
Thirty-three deaths occurred in 19 studies. SUDEP was diagnosed in 20 patients and 14 trials. Eleven deaths met the criteria for definite SUDEP, seven deaths met the criteria for probable SUDEP and two for possible SUDEP (not used in primary analysis). Ten deaths (three SUDEP) occurred in the efficacious antiepileptic drug group, four (three SUDEP) occurred in the non-efficacious antiepileptic drug group and 19 (14 SUDEP) occurred in the placebo group. There was a greater risk of definite or probable SUDEP in patients assigned to placebo rather than anti-epileptic drugs at efficacious doses (OR 0.17, 95% CI 0.05 to 0.57). Results were the same when possible SUDEP was also included.
There was a greater risk of death from all causes in the placebo group compared to the group receiving efficacious doses of antiepileptic drugs (OR 0.37, 0.17 to 0.81).
Pooling data from the non-efficacious antiepileptic drug group and efficacious of placebo groups did not significantly affect the findings. Rates of definite or probable SUDEP per 1,000 person-years were 0.9 (95% CI 0.2 to 0.27) in patients who received efficacious antiepileptic drugs and 6.9 (95% CI 3.8 to 11.6) in those assigned to placebo.