1. Study Design - as a function of
questions
• Is there any added impact from adding interventions together as apposed to one intervention on its own? • USID questions behind combination study design:
– In an area of high coverage but limited usage how much transmission reduction can be expected by adding in IRS? – In areas where IRS is being conducted how much added
transmission reduction can be added by usage of LLINs either with children under 5s or pregnant women?
– In areas where one or more IRS rounds have reduced transmission, can LLIN introduction sustain these gains if IRS is withdrawn?
• IRS coverage of >80% in studies is close to
operational realities
• LLIN coverage that is often in the range of
30-40% coverage
• Studies have shown that LLIN coverage much
above 80% there is much less added protective
gain.
• Need to consider quality and physical durability
of LLINs are the key issues.
• LLIN residual insecticidal life is generally
proven.
• LLIN physical durability is less than previously
assumed.
• These key questions. Say “we are doing
ongoing vector control with LLINs or with
IRS and want studies to tell us of any
added values, or not, of combining another
intervention.
“Effectiveness” of LLINs and IRS is now a
bigger issue than “efficacy” for studies.
•To answer this well studies need to consider the following potential tools individually or in
combinations: •IRS
•LL IRS •DL
•LLIN
DL is a different starting point as it is new and not
currently incorporated within ongoing malaria control programmes but may answer some of the limitations seen with LLINs and IRS.
Do the SOFT factors significantly affect
the effectiveness of LLINs and IRS:
• Also are there other approaches that can improve “effectiveness” of LLINs and/or IRS?
• Studies may need to include measurement of the efficacy of a ‘package” of interventions, i.e.. tools and IEC and usage instruction/validation, rather than just the intervention tools.
• Does combination of nets and IRS alone result in increased impact in comparison with nets + IRS + IEC/hang up. Or do we need to simply ensure that all combination studies include delivery criteria of IRS/LLIN IEC and LLIN hang up.
Does the sequence of intervention affect
impact:
• Does starting with LLINs and adding IRS
or starting with IRS and adding LLINs:
Could study this with four arms:
• LLIN then add in IRS
• LLIN
• IRS then add in LLIN
• IRS
Study type:
•
Generally a Cluster Randomised
Design is likely to be the best feasible
approach:
• Challenge: Is malaria control success
Future Challenge:
• We will need to study what “tools” are
needed to mop up the last cases and
maintain low transmission. Will today’s
tools be suitable when high transmission
areas become low and epidemic?
Summary re study design: RCT
•
Introducing 2nd Vector control tools in
an area with ongoing VC tool
•
Test “absolute” (=efficacy) impact of
LLIN vs IRS vs combination
•
Testing a new VC tool vs an existing
tool.
•
*considering the soft factor such as
2. Duration:
• All phase 3 combination studies should ideally be a 2 years and there may be some advantages in 3 years studies in order to ensure we have a clear idea of durability and do not make assumptions about a tools durability. These will be more expensive but can be designed as studies with different phases:
• Phase 1: over a defined period (probably 2 years) adequate to give defined IMPACT INDICATORS. At the end of this study phase an analysis of those key outputs should be released to provide he evidence base to help advice programmatic usage. • Phase 2: over an extended period of years to monitor other
factors such as durability / impact of durability on usage / and a minimum monitoring of entomological/epidemiological indicators to advise on how impact changes over time (and provide for
Sample size and power
• All trials need to be of adequate size and power to generate significant results
• There will be a trade off between size an duration • Cluster balance, size and number will vary with
setting.
• In areas of lower transmission - as a rule of thumb anything less than 10 per arm may not show small differences (superiority) with combinations and
probably 15 is a reasonable guide for studies. • Areas of very high transmission like DR Congo,
Liberia etc cluster requirements are fewer and smaller
Where are combination trials are
needed:
•
Where currently planned and feasible
•
Where Arabiensis is predominant as current
tools may not impact on the vector much.
•
Its coverage is expanding, especially as
An. G
reduces
•
Can we reduce vector man contact with other
tools.
• Within the same area, the same tools may control one species more effectively than another.
• If we control a leading species, will transmission
continue via another species. Will it increase over time as other vectors become more efficient? Generally
studies should include these end points.
• Need to run combination studies in Gambiense areas • Need to run combination studies in Funestus areas
• Studies to measure the value of combinations of tools in areas where vectors (same species or multiple
species) bite indoors or outdoors (but may rest indoors)
Studies - Resistance
• If used properly can IRS be used as a resistance
management tool in combination with LLINs or other tools / insecticide classes? (Hemingway, E. Mexico, old study). Future reliance on IRS + LLINs in
combination would be a very expensive route for resistance management.
• Can combinations of insecticides (class or mode of action) on material weaves manage resistance? This should be explored and would be a more efficient and cheaper route if feasible.
• Agri. Industry needs to be consulted further to see if they have other resistance management planning