Association of adherence to epilepsy quality standards with seizure control

Objective: We assessed the relationship between adherence to epilepsy quality measures (EQM) and seizure control over 2-3 years in a retrospective cohort study. Methods: 6150 patients were identified at two large academic medical centers with a primary or secondary diagnosis of epilepsy, were 18-85 years old and seen in outpatient general neurology or epilepsy units between June 2011 and May 2014. Patients were included if: their initial visit was between June 2011 and June 2012, treatment was with ≥1 anti-seizure drug, there was ≥1 visit per year during the timeframe, and seizure frequency was documented at initial and final visits, yielding 162 patients/1055 visits from which socio-demographic, clinical and care quality data were abstracted. Quality care was assessed as (1) percent adherence to up to 8 eligible EQM, and (2) defect-free care (DFC: adherence to all eligible EQM). Seizure control (SC) was defined as ≥50% reduction in average seizures/month between initial and final visits. Chi-square and t-test compared care quality with seizure control. Logistic regression was used to assess the relationships between SC, quality of care and subspecialist involvement. Results: Care quality, reflected by documentation of seizure frequency, addressing therapeutic interventions, and referral to a comprehensive epilepsy center, all exceeded 80% adherence. Care quality as reflected by documentation of seizure type, etiology or syndrome; assessment of side effects, counseling about epilepsy safety and women's issues, and screening for psychiatric disorders ranged from 40 to 57%. Mean EQM adherence across all applicable measures was associated with greater seizure control (p= 0.0098). DFC was low (=8%) and did not covary with seizure control (p= 0.55). The SC and non-SC groups only differed on epilepsy etiology (p= 0.04). Exploratory analysis showed that mean quality scores are associated with seizure control (OR. = 4.9 [1.3-18.5], p=0.017) while controlling for the effect of subspecialty involvement as a possible confounding variable. Conclusions: Average quality of care but not defect-free care was associated with seizure control in this retrospective cohort.