Presentation on theme: "Evidence to Policy: Parasites, paragons, and parachutes: evidence for old and new paradigms in malaria vector control Immo Kleinschmidt Short title: Evidence."— Presentation transcript:
Evidence to Policy: Parasites, paragons, and parachutes: evidence for old and new paradigms in malaria vector control Immo Kleinschmidt Short title: Evidence in malaria vector control
Overview and rationale In recent years large reductions in malaria morbidity and mortality have been recorded in many countries where the disease had been endemic These successes have been attributed to comprehensive scaling up of prevention and curative tools Malaria vector control (prevention) tools have formed a sizeable component of the ‘scale up for impact’ This talk will examine : Evidence for efficacy of existing vector control tools Evidence expected of new paradigms of vector control Actual effectiveness of vector control tools in the challenging real world, i.e. Quantifying loss of effectiveness due to partial compliance, changes in vector susceptibility to insecticide etc Attribution of observed impact to the large scale deployment of these tools
Recent progress in reducing malaria burden World Malaria Report 2011: Incidence of malaria globally has reduced by 17% since 2000 Number of cases fell least in countries with highest incidence 43 out of 99 countries recorded reductions of >50% in reported malaria cases between 2000 and 2010. Of these, 8 are in the WHO-Africa region (Algeria, Botswana, Cape Verde, Namibia, Rwanda, Sao Tome an Principe, South Africa, Swaziland and Zanzibar) Malaria-specific mortality rates have reduced by 26% There were an estimated 216 million episodes of malaria in 2010, 174 million were in the African Region There were an estimated 655 000 malaria deaths in 2010, of which 91% were in Africa
Recent scale up of intervention tools: Insecticide treated nets (ITN) World Malaria Report 2011: A total of 82 countries (38 are in the African Region) distribute ITNs free of charge. Number of ITNs delivered by manufacturers increased from 5.6 million in 2004 to 145 million in 2010 in sub- Saharan Africa In sub-Saharan Africa households in malaria risk areas with at least one insecticide-treated net (ITN) rose from 3% in 2000 to 50% in 2011 96% of persons with access to an ITN actually used it according to household surveys
Recent scale up of intervention tools: Indoor Residual Spraying (IRS) World Malaria Report 2011: A total of 185 million people were protected by IRS in 2010, representing 6% of the global population at risk. Households protected by indoor residual spraying (IRS) rose from <5% in 2005 to 11% in 2010 In sub-Saharan Africa the number of people protected by IRS increased from 10 million in 2005 to 81 million in 2010 (11% of the population at risk)
What evidence is there for the effectiveness of these prevention tools? Evidence that is acceptable for policy recommendations generally requires: Randomised clinical trials (the gold standard of evidence) In the case of methods that function at individual as well as community level cluster randomised trials are required i.e. Randomisation of whole communities to intervention or control Some vector control tools have been effectively used long before cluster randomised trials had been invented
Parachute approach to evidence based medicine 1 Parachutes have never been evaluated through trials yet we do not doubt their effectiveness Smith & Pell (2003) 2 wrote: “We conducted a systematic review of randomised controlled trials to determine whether parachutes are effective in preventing major trauma. Finding: The basis for parachute use is purely observational, and its apparent efficacy could potentially be explained by a “healthy cohort” effect” Conclusion “everyone might benefit if the most radical protagonists of evidence based medicine organised and participated in a double blind, randomised, placebo controlled, crossover trial of the parachute” Do some vector control methods fall into the parachute category? 1 Potts, Prata, Walsh and Grossman. Parachute approach to evidence based medicine. BMJ 2006;333:701–3 2 Smith G, Pell JP. Parachute use to prevent death and major trauma related to gravitational challenge: systematic review of randomised controlled trials. BMJ 2003;327:1459-61.
1. Evidence to policy: The insecticide treated net (ITN)
Long lasting insecticidal net (LLIN) Person sleeping under net is protected during peak biting hours of anopheles mosquito In seeking a blood meal from the human under the net, the mosquito is repelled or killed when coming into contact with the insecticide The net prevents the mosquito from transmitting parasites to humans, and vice versa Only pyrethroid insecticides are considered safe for use on nets Photo courtesy of Wikepedia
Example: Evidence for ITN efficacy Binka, F.N., Kubaje, A., Adjuik, M., Williams, L.A., Lengeler, C., Maude, G.H., Armah, G.E., Kajihara, B., Adiamah, J.H. and Smith, P.G. (1996). Impact of permethrin impregnated bednets on child mortality in Kassena-Nankana district, Ghana: a randomized controlled trial. Tropical Medicine and International Health 1, 147–154. 96 geographical areas (clusters) 48 randomly selected to households receiving permethrin impregnated bednets, re-impregnated every 6 months longitudinal demographic surveillance system was used to record births, deaths and migrations bednets were associated with 17% reduction in all-cause mortality in children aged 6 months to 4 years (RR=0.83; 95% CI 0.69-1.00; P=0.05)
Cochrane review of ITN/LLIN efficacy Fourteen cluster randomised and eight individually randomised controlled trials were included in a Cochrane meta-analysis 1 Results demonstrated strong evidence of protective efficacy on several endpoints child mortality: 17% (95% CI 10% to 24%) uncomplicated malaria episodes: 50% (areas of stable malaria); 62% (areas of unstable malaria) severe malaria: 45% (95%CI 20 to 63) parasite prevalence: 13% Therefore evidence of effect is undisputed Good evidence → policy at global and national levels 1 Lengeler C. Insecticide-treated bed nets and curtains for preventing malaria. Cochrane Database of Systematic Reviews 2004, Issue 2
WHO position on the use of LLINs for malaria prevention (2007 ) The WHO Global Malaria Programme recommends: distribute free or highly subsidized LLINs; achieve full LLIN coverage; develop and implement locally appropriate communication and advocacy strategies to promote effective use of LLINs; and implement strategies to sustain high levels of LLIN coverage
2. Evidence to policy: Indoor Residual Spraying (IRS)
Indoor Residual Spraying (IRS) Insecticide applied to interior wall of houses Anopheles mosquitoes rest on interior surface of house after taking a blood meal By making contact with insecticide on walls they are killed thus preventing transmission of the parasite to a another human host Spraying has to be repeated once or twice a year to ensure active ingredient on walls Currently four classes of insecticide are allowed for IRS vector control
IRS effectiveness No large experimental trials Cochrane review 1 concluded “the number of high-quality trials are too few to quantify the size of effect in different transmission settings”. Only studies with >60% coverage were included, but effect of coverage not assessed →No gold standard evidence for IRS ‘Convincing’ effectiveness shown from studies using either 1.Comparison with historical controls, or 2.Comparison between compliers and non-compliers 1 Pluess B, Tanser FC, Lengeler C, Sharp BL. Indoor residual spraying for preventing malaria. Cochrane Database of Systematic Reviews 2010, Issue 4. Art. No.: CD006657. DOI: 10.1002/14651858.CD006657.pub2.
IRS effectiveness example Map of the LSDI malaria control zones and location of sentinel sites. Sharp et al, 2007. Lebombo Spatial Development Initiative (LSDI) in southern Mozambique, Swaziland and KwaZulu Natal, 1999-2005: Impact of IRS by comparison of infection prevalence before and after the intervention and between zones
LSDI (Southern Mozambique): prevalence monitored over seven years Sharp BL, Kleinschmidt I, Streat E, Maharaj R, Barnes KI, Durrheim DN, Ridl FC, Morris N, Seocharan I, Kunene S, La Grange JJP, Mthembu DJ, Maartens F, Martin CL, Barreto A. Seven Years Of Regional Malaria Control Collaboration - Mozambique, South Africa And Swaziland. Am. J. Trop. Med. Hyg., 2007, 76(1): 42–47 Protective effectiveness= 26% reduction in risk of infection per spray round (relative risk 0.74 per spray year) First spray round zone 1 First spray round zone 2 First spray round zone 3
Compelling evidence of IRS effectiveness also provided by entomology...
Evidence for IRS has led to policy recommendation 2006 WHO Position Statement was entitled “Use of indoor residual spraying for scaling up global malaria control and elimination” “Effective implementation of IRS with DDT or other recommended insecticides should be a central part of national malaria control strategies where this intervention is appropriate.”
3. Evidence to policy: New paradigms in malaria vector control Examples: 1.Repellents 2.Attractants 3.Resistance breaking mechanisms for nets and IRS 4.Wall linings 5.Genetically modified mosquito releases Will need to demonstrate evidence of epidemiological impact in cluster randomised trials at least as good as the standard set by ITNs
4. Evidence of loss of effectiveness in real life
LLINs and IRS in the real world: what evidence that they are still effective Nets: How much protection can we expect from the less than perfect net inervention in practice? Nets are not used consistently Acquire holes (≠long lasting) May be less effective if mosquitoes become resistant to pyrethroid insecticide IRS May not last long enough on walls Not applied properly in enough houses “The spread of insecticide resistance,... is a major threat for vector control programmes” (World Malaria Report 2011)
Evidence of a problem: Condition of nets, Bioko 2011 Net condition Percent ITNs/LLINs Intact (no holes) 23 up to 3 finger sized holes, no large holes 3 >3 finger sized holes, no large holes 2 one large hole 3 2 large holes 2 >2 large holes 5 Untreated nets Intact (no holes) 39 With holes 23 Total 100
Adjusted OR (95% CI) 1 Protection from net No net/untreated with holes 1 LLIN/ITN with big holes 2 0.95 (0.71-1.28) Untreated with no holes 0.85 (0.71-1.01) LLIN/ITN with small holes 0.65 (0.54-0.79) LLIN/ITN with no holes 0.65 (0.55-0.77) 1 Adjusted for IRS of house, IRS and ITN coverage of community, age, SES 2 Large holes are those that can fit a D sized torch battery Evidence of loss of effectiveness: Prevalence of infection with malarial parasites, by net use and condition of net § (Bioko and EG mainland) § Rehman, A. M., Coleman, M., Schwabe, C., Baltazar, G., Matias, A., Roncon Gomes, I., Yellott, L., Aragon, C., Nseng Nchama, G., Mzilahowa, T., Rowland, M. & Kleinschmidt, I. How much does malaria vector control quality matter: the epidemiological impact of holed nets and inadequate indoor residual spraying. PLoS One, 2011, 6, e19205.
LLIN effectiveness in the presence of mosquitoes that are resistant to insecticide? There is little evidence of real loss of epidemiological effectiveness in the face of resistance Lack of evidence ≠ evidence of no effect Currently large study is underway to assess the impact of insecticide resistance on vector control effectiveness This will not be ‘clinical trial’ evidence since the exposure cannot be randomly assigned Instead an ecological study design will link insecticide resistance status of mosquitoes (the exposure) with malaria incidence in cohorts of children (the outcome) in the same area Statistical power gained from large numbers of areas (clusters) in five countries
IRS: loss of effectiveness in the real world Due to the effect of inadequate coverage inadequate insecticide residual in long transmission season vector resistance
Crude OR (95% CI)Adjusted OR 1 (95% CI) IRS of the household Unsprayed house11 Sprayed in last 6 months0.5 (0.4-0.7)0.9 (0.8-1.0) Site spray coverage 0-19.9%11 20-49.9%0.3 (0.1-0.6)0.9 (0.6-1.4) 50-79.9%0.2 (0.1-0.3)1.0 (0.7-1.5) 80-100%0.2 (0.0-0.5)0.5 (0.3-0. 9) 1 Results adjusted for use of nets, age, SES, year of survey, IRS of house and IRS coverage of site Bioko and Equatorial Guinea 2009-2010: prevalence of infection in children 1 to<15 years and IRS coverage
Prevalence of infection, % (N) Adjusted Odds Ratio95% CIp-value Median time since last spray (months) 318 (1483) 1.35 per month since spray [1.02,1.79] 0.034 421 (1921) 528 (505) Bioko 2011: Effect of short residual of insecticide on prevalence of infection in children 1 1 unpublished data
Impact of pyrethroid resistance in A. funestus, KwaZulu Natal, South Africa R Maharaj, D J Mthembu, B L Sharp. S Afr Med J 2005; 95: 871-874. 1.1995: IRS with pyrethroid introduced 2.1999: Resistance to pyrethroid discovered in A. funestus. 3.March 2000: Re- introduction of IRS with DDT. 4.Feb. 2001: Introduction of ACTs 1 2 3 4
5. Attribution of health impact to large scale vector control interventions How do we know impact is due to interventions?
Effectiveness on a large scale: Can we attribute impact to country-wide interventions? Outcome measure: change in all-cause mortality in children as proxy for change in malaria specific mortality 1 “Before and after” assessment using plausibility evaluation based on causal pathway: Evidence of large change in coverage of effective intervention → evidence of change in malaria related morbidity → change in all-cause child mortality, taking account of contextual determinants of child survival 1 Rowe AK et al; Roll Back Malaria Monitoring and Evaluation Reference Group. Viewpoint: evaluating the impact of malaria control efforts on mortality in sub-Saharan Africa. Trop Med Int Health. 2007 Dec; 12(12):1524-39.Rowe AK et al; Roll Back Malaria Monitoring and Evaluation Reference Group
Bioko: 2004 to 2008 2000-20042005-2008 Scale-up Protected by IRS or ITN use<5%>80% Malaria first line treatmentChloroquine (failing) ACT (effective) Malaria related indicators in <5y Infection prevalence42% (2004) 18% (2008) Anaemia, Hb<8g/dL15% (2004) 2% (2008)
Bioko: Change in all cause under 5 mortality Plausibility argument was strengthened by subsequent serological analysis which showed that changes in seroconversion rate corresponded to changes in u5m rates by geographical area 1999-20042004 to 2008Hazard Ratio Mortality by age 5 per 1000 live births [95% CI] 152 [122 - 186] 55 [38 to 77] 0.34 [0.23-0.49]
Summary and Conclusion New vector control paradigms require CRTs with epidemiological outcomes with evidence of a standard at least as good as that set by ITN trials Effectiveness of old paradigms (few trials, but much programmatic evidence) – parachute approach? Loss of effectiveness due to partial-compliance, poor application or insecticide resistance requires quantification (operational research; observational study designs) Attribution of large scale interventions to health impact assessed by more indirect approaches such as the before-after plausibility (unless stepped wedge introductions of interventions are possible) Quality of evidence High Low