THE USE OF INTERQUARTILE DEVIATION IN ESTABLISHING DELPHI PANEL CONSENSUS- A PRIORITIZATION OF INTRAVENOUS IMMUNOGLOBULIN UTILIZATION
Author(s)
Orange J1, Lennert B2, Rane P3, Eaddy M2
1Texas Children's Hospital, Houston, TX, USA, 2Xcenda, Palm Harbor, FL, USA, 3University of Houston, Houston, TX, USA
To use consensus-building methodologies to prioritize disease states for intravenous immunoglobulin (IVIG) utilization while considering disease severity and alternative therapeutic options. METHODS: A 7-member expert panel independently ranked 50 disease states across 2 domains: (1) Disease severity (DS) (1=immediately life-threatening, 2=life-threatening, 3=life-modifying, 4=other) and (2) the perceived efficacy of therapeutic alternatives (TA) (1=none, 2=low, 3=medium, 4=high). An interquartile deviation of ≤0.5 was used to determine consensus for disease states within each domain. Disease states reaching consensus across both domains were ranked according to a 4x4 algorithmic scale to establish priority. RESULTS: The panel reached consensus on the severity of all diseases states; however, 11 of the 50 disease states did not reach consensus on the availability of alternative therapeutic options. No disease state was designated as being immediately life-threatening without an available alternative therapeutic option (DS1TA1), while 3 disease states (X-linked agammaglobulinemia, common variable immunodeficiency, primary immunodeficiency with absent B-cells) were designated as life-threatening with no therapeutic alternatives (DS2TA1). The priority distribution of disorders based on the algorithm is as follows: DS1TA1=0, DS1TA2=1, DS1TA3=1, DS1TA4=1 DS2TA1=3, DS2TA2=4, DS2TA3=3, DS2TA4=1 DS3TA1=0, DS3TA2=7, DS3TA3=14, DS3TA4=0 DS4TA1=0, DS4TA2=0, DS4TA3=3, DS4TA4=1 CONCLUSIONS: The application of interquartile deviation in establishing consensus across two 4-point Likert scales resulted in prioritizing 80% of disease states where IVIG can be used. Additional consensus-building rounds will be needed to prioritize the remaining disease states.
Conference/Value in Health Info
Value in Health, Vol. 18, No. 7 (November 2015)
Code
PRM220
Topic
Methodological & Statistical Research
Topic Subcategory
Confounding, Selection Bias Correction, Causal Inference
Disease
Multiple Diseases