Indications for consideration
The core group defined what symptoms, past medical history and relevant
diagnostic tests might influence the decision regarding whether biologic
prescription might be appropriate. Their responses were used to form a
series of matrixes that categorise patients according to these key
indications. Specific combinations of associated comorbidities and
previous treatment were considered, alongside wider statements
concerning the delivery of biologics. One hundred clinical scenarios
were considered, stratified by the presence of co-morbid asthma, N-ERD
or AFRS, the number of previous surgeries (or if a patient was unfit
surgery), and the use of INCS or OCS (or if a patient was unable to use
them). Other characteristics such as baseline symptom scores, the
importance of radiological imaging and the extent of disease on
radiological imaging were also considered.
Consensus process We used the RAND/UCLA methodology with a multi-step process
(www.rand.org)(10). Our expert panel undertook a 2 round modified Delphi
process of ranking and classifying appropriateness of different
investigations and treatment options. Using a 9-point Linkert scale, for
each indication the panellists scored whether a treatment was either:
- Not recommended/ Inappropriate; should not be prescribed for the
indication described within the NHS, based on current evidence base
and costs (scored 1 to 3) ,
- Uncertain; (scored 4 to 6),
- Recommended / Appropriate; should be prescribed for the indication
described within the NHS, based on current evidence base and costs
(scored 7 to 9)
Free text comments were encouraged if greater context was required, if
the question was ambiguous or if anything had been overlooked.
The final recommendation was based on the median ranking scores collated
from each clinical scenario provided that there was consensus. Consensus
was defined as the requirement for more than 70% of responses to fall
into the category defined by the median, and when this score fell in
either recommended or not recommended, less than 15% of responses were
scored as the opposite.
The scores at the end of round 1 were analysed and presented back to the
group, with the distribution of scores summarised for each question. The
panel was then asked to repeat the scoring for any items where consensus
has not been reached having considered if they wished to revise their
previous score.
Round 2 scores were evaluated in the same way. When the median fell
between 4-6, or if the median fell in either 1-3 or 6-9 and the
definition of consensus was not met, no recommendation was made, and the
use of biologics for the given indication was considered uncertain.