The Discord Outcome Analysis (DOA) as a Reporting Standard at Three Months and Five Years in Randomised Varicose Vein Treatment Trials.

ABSTRACT

OBJECTIVES: Treatment success for chronic superficial venous insufficiency could be defined as an improvement in three domains (i) disease specific quality of life, (ii) clinical severity, (iii) reflux. The aim was to report these at five years using a Venn diagram to profile the outcomes: a discord outcome analysis (DOA). METHODS: Patients (n = 50 patients/legs in each treated group) were randomised to endovenous laser ablation (EVLA) with concurrent phlebectomies vs. ultrasound guided foam sclerotherapy (UGFS). Outcomes were assessed using three domains (i) Aberdeen varicose vein questionnaire (AVVQ), (ii) venous clinical severity score (VCSS), (iii) venous filling index (VFI) of air plethysmography. Change scores were calculated by subtracting the final score after treatment from the baseline score before treatment to quantify the improvement. This was followed by a DOA profile for each patient where a discord was defined as the percentage of patients with a numerical deterioration in one or two domains. RESULTS: The median [interquartile range] follow up was 68 [64-72] months. Follow up in all three domains was EVLA 45/50, UGFS 42/50. On ultrasound examination, GSV occlusion at some point above the knee was 93% for EVLA and 64% for UGFS (p = .001). There was no significant difference in improvement between the two treatment groups in the VCSS and the VFI. However, the EVLA group had a statistically significant AVVQ improvement (p = .004). Using a DOA, only 76% EVLA versus 60% UGFS had success in all three domains. Using improvement thresholds, this reduced to 54% and 39%, respectively. The commonest discord pattern was an improvement in the VCSS and VFI but deterioration in the AVVQ. CONCLUSIONS: A DOA demonstrated that the definition of success is reduced if deterioration in one or two domains is taken into account. A DOA should be considered as a reporting standard for comparative analyses.