Intervention Review

You have free access to this content

Insecticide-treated bed nets and curtains for preventing malaria

  1. Christian Lengeler*

Editorial Group: Cochrane Infectious Diseases Group

Published Online: 19 APR 2004

Assessed as up-to-date: 18 JAN 2004

DOI: 10.1002/14651858.CD000363.pub2


How to Cite

Lengeler C. Insecticide-treated bed nets and curtains for preventing malaria. Cochrane Database of Systematic Reviews 2004, Issue 2. Art. No.: CD000363. DOI: 10.1002/14651858.CD000363.pub2.

Author Information

  1. Swiss Tropical Institute, Public Health and Epidemiology, Basel, Switzerland

*Christian Lengeler, Public Health and Epidemiology, Swiss Tropical Institute, Basel, 4002, Switzerland. Christian.Lengeler@unibas.ch.

Publication History

  1. Publication Status: Edited (no change to conclusions)
  2. Published Online: 19 APR 2004

SEARCH

 
Characteristics of included studies [ordered by study ID]
Burkina Faso (Habluetzel)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: groups of villages (8 pairs of "clusters" (on average 10 villages) formed on the basis of baseline mortality and geographic similarity).

Number of units: 8:8.

Length of follow up: 24 months.

Mortality was monitored by village reporters and yearly census. A cross-sectional morbidity survey was conducted once, at the peak of the transmission season in September 1995 (n = 800 in 84 villages). All surveys were community-based.


ParticipantsNumber of participants: 16,540.

Inclusion criteria: children aged 0 to 59 months living in the area (newborns were excluded from the analysis).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated curtains on windows, door, and eaves; target dose of 1000 mg/m2; every house used for sleeping in the intervention clusters fitted with the curtains and re-treated every 6 months.

Control: no curtains.


Outcomes1. Overall mortality (1 to 59 months).
2. Prevalence of parasitaemia (any).
3. Prevalence of high parasitaemia (> 5000 trophozoites per µl).
4. Anaemia (mean haemoglobin in g/dl).


NotesStudy location: Oubritenga Province, 30 km north of Ouagadougou, in a rural area.

EIR: 300 to 500.

Malaria endemicity: holoendemic.

Baseline parasite rate in children 6 to 59 months: 85%.

Main vectors: Anopheles gambiae s.l. and A. funestus.

Plasmodium vivax malaria: 0%.

Dropout rate unknown, but immigration/emigration rates were low (2% per year).

Access to health care considered poor.

Cameroon (Moyou-Somo)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (20 households were chosen in each "quartier" (methods not stated)).

Number of units: 20:20.

Length of follow up: 12 months.

Monitoring from January to December 1992.

Overall survey completion rate 75%.

Repeated cross-sectional surveys carried out in February, April, June, August, October, and December 1992 (on average, n = 361, 75.2% of the group).


ParticipantsNumber of (randomized) participants: approximately 480 children aged 0 to 15 years from 20 households.

Inclusion criteria: people living in 2 neighbourhoods.


InterventionsIntervention: deltamethrin-treated bed nets; target dose 25 mg/m2; nets treated in January 1992 and re-treated in August 1992.

Control: no bed nets; < 20% usage.


Outcomes1. Prevalence of any parasitaemia (repeated measure).
2. Splenomegaly (Hackett 1 to 5).


NotesStudy location: Kumba (South-West Province), Cameroon.

EIR: 10 to 20.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children aged 0 to 15 years: 30.2 to 52.5%.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: 0%.

Access to health care was likely to be good.

Colombia (Kroeger)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (22 villages were paired according to size, geographic location, net coverage, and malaria incidence at baseline; within each pair 1 village within each pair was then randomized to receive the intervention).

Number of units: 11:11.

Length of follow up: 12 months.

Single cross-sectional survey carried out during the peak of the malaria season in February to March 1992.


ParticipantsNumber of participants: 4632 participants took part in the cross-sectional survey (high percentage of total).

Inclusion criteria: inhabitants of the 22 trial communities.


InterventionsIntervention: lambdacyhalothrin treatment of existing bed nets; target dose 10 to 30 mg/m2; net treatment in September and November 1991 (nearly 60% of all existing nets were treated at least once); sales and promotion of bed nets, and free net treatment.

Control: untreated bed nets; 96% usage rate).


Outcomes1. Period-prevalence (last two weeks or last four months) of reported "malaria episodes" assessed during the peak of the malaria season (March to April 1992).

Outcome measures similar to Ecuador (Kroeger).


NotesStudy location: lower Rio San Juan, Departamente Choco on the Pacific Coast, Colombia.

EIR: < 1.

Malaria endemicity: hypoendemic.

Baseline parasite rate in the whole population and spleen rate in children aged 2 to 9 years: below 5%.

Main vector: Anopheles nevai.

Plasmodium vivax malaria: 31% of all episodes; no distinction made between P. falciparum and P. vivax in the analysis.

Usage rate was high (96% of families with at least one net).

Access to health care was likely to be good.

Ecuador (Kroeger)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (14 villages were paired according to size, geographic location, net coverage, and malaria incidence at baseline; 1 village within each pair then randomized to receive the intervention).

Number of units: 7:7.

Length of follow up: 17 months.

Single cross-sectional survey carried out during the peak of the malaria season in March to April 1992.


ParticipantsNumber of participants: 2450 participants took part in the cross-sectional survey (high percentage of total).

Inclusion criteria: inhabitants of the 14 trial communities.


InterventionsIntervention: permethrin treatment of existing bed nets; target dose 200 mg/m2; high usage rate high (93% of families with at least 1 net); net treatment in October and December 1991 (6 and 4 months before the evaluation); nearly 80% of all existing nets were treated at least once; sales and promotion of bed nets, and free net treatment.

Control: untreated bed nets; > 90% usage rate.


Outcomes1. Period-prevalence (last 2 weeks or last 4 months) of reported "malaria episodes" assessed during the peak of the malaria season (March to April 1992).

Although no systematic parasitological confirmation was done, quality control procedures ensured adequate accuracy. According to a pilot phase, about 88% to 96% of the self-diagnoses were based on the same criteria as health professionals. In addition, time trends were compared to those obtained from routine data.


NotesStudy location: Canton Muisne, on the northern Coast, Ecuador.

EIR: < 1.

Malaria endemicity: hypoendemic.

Baseline parasite rate in the whole population and spleen rate in children aged 2 to 9 years: < 5%.

Main vector: Anopheles albimanus.

Plasmodium vivax malaria: 51% of all episodes; no distinction could be made between episodes due to P. falciparum or P. vivax in the analysis.

Access to health care was likely to be good.

Gambia (D'Alessandro)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (52 pairs of villages formed on the basis of size, after stratification by 5 geographical areas).

Number of units: 58:52.

Length of follow up: 12 months.

Dropout rate unknown, but immigration/emigration rates were low (< 5% per year).

Mortality monitored by village reporters and yearly census. Morbidity surveys were conducted once, at the peak of the transmission season in October (n = 1500 in 50 villages). All surveys were community-based.


ParticipantsInclusion criteria: children aged 0 to 9 years and living in the area were eligible at the start, but later the analysis was restricted to children aged 1 to 59 months (n = 25,000).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: treatment of existing bed nets in the frame of a national programme; target dose 200 mg/m2 permethrin; impregnation done by village health workers with the assistance of other community members and under the supervision of community health nurses; re-treatment was not done during the 1 year follow-up period since the transmission season lasts only about 4 months.

Control: untreated bed nets.

Usage rate around 70% in both intervention and control areas (varied between 50% and 90% according to the area).


Outcomes1. Overall mortality (1 to 59 months).
2. Prevalence of parasitaemia (any).
3. Prevalence of high parasitaemia (> 5000 trophozoites per µl).
4. Anaemia (mean packed cell volume).
5. Prevalence of splenomegaly (1 to 5 Hackett).
6. Impact on nutritional status (weight-for-age, weight-for-height).


NotesStudy location: 5 distinct areas spread over the whole of The Gambia (all rural areas).

EIR: 1 to 10.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children 12 to 59 months: 39%.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: very low; not taken into account for analysis.

Access to health care moderately easy.

Gambia (Snow I)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (allocation of 110 compounds was done randomly after stratification by 3 levels of "spleen rate": no child with enlarged spleen in household, one child, more than one child).

Number of units: 60:50.

Length of follow up: 4 months.

Morbidity rates monitored longitudinally by weekly home visits during 4 months in the peak transmission season (July 1985 to November 1985). A blood slide was made if the child had an axillary temperature of at least 37.5 °C, or if the mother reported that the child had fever during the last 3 days. Success rate for weekly visits was 97%. Overall dropout rates were 8% in the treatment group and 12% in the control group. Single cross-sectional morbidity survey conducted at the end of the transmission season in November 1985 (n = 275). All surveys were community-based.


ParticipantsNumber of eligible participants: 580.

Number of randomized participants: 389 (67%).

Inclusion criteria: children aged 1 to 9 years living in the village.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin treatment of existing bed nets; target dose 500 mg/m2; usage rate was very high before the trial (98%); nets not re-treated because of the short duration of the trial.

Control: dilute crystal violet solution (placebo treatment) used to treat control nets; 98% usage rate.


Outcomes1. Incidence of mild clinical episodes (children aged 1 to 9 years).
2. Prevalence of any parasitaemia.
3. Prevalence of high parasitaemia (> 5000 parasites/µl).
4. Prevalence of anaemia (mean packed cell volume).


NotesStudy location: village of Katchang, on the north bank of the Gambia River, Gambia.

EIR: 10.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children 1 to 9 years: 8.6% in the low season and 43.1% in the peak season.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: 0%.

Access to health care was considered poor.

Gambia (Snow II)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (allocation of 16 villages was done randomly after stratification by previous net provision and location with regard to a river).

Number of units: 7:9.

Length of follow up: 4 months.

Morbidity rates monitored longitudinally by weekly home visits during 4 months in the peak transmission season (July 1987 to November 1987). Blood slide made if the child had an axillary temperature of at least 37.5 C. Mothers also asked about the well-being of their child on the day of the interview. Completion rate for weekly visits was 95%. Overall dropout rates were 11% in both treatment groups. Single cross-sectional morbidity survey was conducted at the end of the transmission season in November 1985 (n = 422). All surveys were community-based.


ParticipantsNumber of eligible participants: 491.

Number of randomized participants: 454 (92%).

Inclusion criteria: children aged 1 to 9 years living in the village.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin treatment of existing bed nets; target dose 500 mg/m2; usage rate was very high before the trial (> 95%); nets not re-treated because of the short duration of the trial.

Control: dilute milk in water solution (placebo treatment) used to treat control nets; > 95% usage rate.


Outcomes1. Incidence of mild clinical episodes (children aged 1 to 9 years).
2. Prevalence of any parasitaemia.
3. Prevalence of high parasitaemia (> 5000 parasite/µl).
4. Prevalence of anaemia (mean packed cell volume).
5. Prevalence of splenomegaly (Hackett 1 to 5).


NotesStudy location: 16 Fula villages, on the north bank of the Gambia River, west of Farafenni, Gambia.

EIR: 10.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children 1 to 9 years: 25.9% in the low season and 37.3% in the peak season.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: 0%.

Access to health care was considered poor.

Ghana (Binka)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (allocation of 96 "clusters" was done randomly (public ballot) after stratification by 10 chiefdoms).

Number of units: 48:48.

Length of follow up: 24 months (July 1993 to June 1995).

Dropout rate unknown, but immigration/emigration rates were low (< 5% per year).

Mortality was monitored by village reporters and 4-monthly censuses (rolling census). A cross-sectional morbidity survey was conducted twice, in June 1994 (n = 2799) and at the peak of the transmission season in October 1994 (n = 3788). All surveys were community-based.


ParticipantsNumber of participants: 19,900.

Inclusion criteria: children aged 0 to 59 months living in the area (newborns were excluded from the analysis).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated bed nets; target dose 500 mg/m2; enough bed nets distributed to protect both children and the adults; nets re-treated every 6 months.

Control: no bed nets; 4% usage (very low).

No co-intervention at the time of the trial.


Outcomes1. Overall mortality (1 to 59 months).
2. Prevalence of parasitaemia (any).
3. Prevalence of high parasitaemia (> 4000 trophozoites per µl).
4. Anaemia (mean haemoglobin in g/dl).


NotesStudy location: rural area in the Kassena-Namkana, in the Upper East Region of Ghana.

EIR: 100 to 300.

Malaria endemicity: holoendemic.

Baseline parasite rate in children 6 to 59 months: 85 to 94% in the peak season, with strong seasonal fluctuation.

Main vectors: Anopheles gambiae s.l. and A. funestus.

Plasmodium vivax malaria: < 2% (not taken into account in the analysis).
Access to health care poor.

Iran (Zaim I)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (random allocation of 13 villages (10 intervention, 3 control) from a list of eco-epidemiologically homogenous villages).

Number of units: 10:3.

Length of follow up: 8 months.

Morbidity rates monitored longitudinally by passive case detection (high access to health care) as well as home visits every 10 days. Monitoring from April to November 1995, covering the 2 peaks in transmission (April to May and September to October). Blood slide was made for every person reporting with symptoms compatible with malaria; every positive slide labelled a "malaria case" and no differentiation between Plasmodium falciparum and P. vivax malaria made in the analysis. All surveys were community-based.


ParticipantsNumber of participants: 6507.

Inclusion criteria: persons living in the village.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: cyfluthrin treatment of existing cotton bed nets through health workers supervised by the researchers; target dose 40 mg/m2; usage rate very high before the trial (nearly every family reported to have at least 1 net).

Control: untreated bed nets; usage rate not specified but very high.

Co-intervention: residual spraying with propoxur (2 g/m2) stopped 7 months before start of the trial. As a result, mosquito population unlikely to be "natural" at the start of the trial.


Outcomes1. Incidence of mild clinical episodes (all ages).


NotesStudy location: 13 villages in Ghasseregahnd (Baluchistan) in Iran.

EIR: very low.

Malaria transmission: unstable, with 30 to 50 infections per 1000 inhabitants per year.

Main vectors: Anopheles culicifacies and A. pulcherrimus.

Plasmodium vivax malaria: 25% to 63% (mean = 53%) of all cases.

Ivory Coast (Henry)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (allocation of 8 villages by paired randomization).

Number of units: 4:4

Length of follow up: 12 months.

Evaluation by cross-sectional surveys and by active case surveillance.


ParticipantsNumber of participants: 432.

Inclusion criteria: children aged 0 to 59 months.


InterventionsIntervention: lambdacyhalothrin-treated nets; target dose 15 mg/m2; high usage rate; n = 216.

Control: no nets; n = 216.


Outcomes1. Prevalence of parasitaemia, anaemia and incidence of clinical episodes.
2. Anaemia.
3. Incidence of clinical episodes.


NotesStudy location: 8 villages around the town of Korhogo, in northern Ivory Coast.

EIR: 55.

Baseline prevalence rate in small children: 69%.

Plasmodium vivax malaria: no information available.

Kenya (Nevill)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (random allocation of 56 "clusters" (of ˜1000 participants each) after stratification by 3 geographical areas).

Number of units: 28:28.

Length of follow up: 24 months (July 1993 to June 1995).

Dropout rate unknown, but immigration/ emigration rates were low for young children.

Mortality monitored by village reporters and 6-monthly censuses. Cross-sectional morbidity surveys were conducted in infants only (1 to 12 months) after peak of the transmission season in August 1994 (n = 443), January 1995 (n = 540), and March 1995 (n = 496). Monitoring system also was set up at Kilifi District hospital to register all admissions with severe malaria disease. All surveys were community-based.


ParticipantsNumber of participants: 11,000.

Inclusion criteria: children aged 0 to 4 years living in the area (newborns were excluded from the analysis).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated bed nets; target dose 500 mg/m2; enough distributed to protect all children; nets re-treated every 6 months.

Control: no bed nets; 6% usage (very low).


Outcomes1. Overall mortality (1 to 59 months).
2. Incidence of admission with severe malaria disease at the district hospital (1 to 59 months). Case definition: children with Plasmodium falciparum parasitaemia and no other obvious cause of disease; for cerebral malaria: coma or prostration or multiple seizures; severe malaria anaemia was defined as < 5.1 g/dl haemoglobin with more than 10,000 parasites per µl.
3. Prevalence of parasitaemia in infants aged 9 to 12 months (any).
4. Impact on anthropometric parameters (weight-for-age and mid-upper arm circumference).


NotesStudy location: in a rural area in Kilifi District on the Kenyan Coast.

EIR: 10 to 30.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children 1 to 9 years: 49% in the peak season, with seasonal fluctuation.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: 0%.

Access to health care is good and over 10% of all children under 5 years are admitted per year.

Kenya (Phillips-Howard)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (allocation of 221 villages by open lottery).

Number of units: 113:108.

Length of follow up: 24 months.

Mortality was monitored by a full demographic system, a birth cohort study, and cross-sectional surveys.


ParticipantsNumber of participants:
18,500.

Inclusion criteria: children aged 0 to 59 months; (newborns were excluded from the analysis).


InterventionsIntervention: permethrin-treated polyester bed nets; target dose 500 mg/m2; usage rate very high 66% during last night.

Control: no nets.


Outcomes1. Overall mortality.
2. Clinical incidence.
3. Parasite prevalence.
4. Anaemia.
5. Anthropometric measurements.


NotesStudy location: Asembo and Gem areas of Siaya District, western Kenya.

EIR: 60 to 300 (high).

Plasmodium falciparum parasite rate in young children: 88%.

Plasmodium vivax malaria: no information available.

Kenya (Sexton)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (105 families, each with at least one child < 5 years of age were selected randomly from two villages and then allocated randomly to 1 of 3 groups: treated bed nets, treated curtains, or control).

Number of units: 35:35.

Length of follow up: 4 months.

Re-infection rates after radical treatment with sulfadoxine-pyrimethamine (Fansidar) monitored longitudinally by weekly home visits during 4 months in the low transmission season (August 1988 to November 1988). Blood slide made at each visit. In addition, clinical episodes (mainly fever and chills) were recorded twice per week. Participants reporting fever or a history of fever since the last visit had their axillary temperature taken. Completion rate for weekly visits was around 60%. Overall dropout rates were 3% in the bed net group and 0% in the 2 other groups. All surveys were community-based.


ParticipantsNumber of participants: 477.

Inclusion criteria: persons living in the villages (primary analysis was for all ages).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention 1: permethrin-treated bed nets; target dose 500 mg/m2; usage rate very low before the trial (9%); nets not re-treated because of short duration of the trial; n = 154.

Intervention 2: permethrin-treated curtains (eaves, door, windows); target dose 500 mg/m2; usage rate very low before the trial (9%); nets not re-treated because of short duration of the trial; n = 167.

Control group: no bed nets, no curtains; maximum 9% usage rate; n = 156.


Outcomes1. Incidence of reported fever (all ages).

Results for treated bed nets and treated curtains were not significantly different and were therefore pooled ("intervention group").


NotesStudy location: 2 villages in western Kenya (52 km from Kisumu).

EIR: 300.

Malaria endemicity: holoendemic.

Baseline parasite rate in children < 5 years: 87.4%, with little seasonal fluctuation.

Main vector: Anopheles gambiae s.l.

Plasmodium vivax malaria: 0%.

Access to health care was not very good, but there was a high use of antimalarials.

Madagascar (Rabarison)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (91 households (n = 501)).

Number of units: 46:45.

Length of follow up: 15 months.

Overall dropout rates were 15% in the bed net group and 13% in the control group.

Follow up through passive case detection at the Institut Pasteur dispensary set up in the study area. Clinics were held daily and every participant had an axillary temperature taken and a blood slide made. Case of malaria was defined as a temperature of at least 37.5 C and a Plasmodium falciparum parasitaemia of at least 1500 parasites per µl. Monitoring carried out in February to July 1993 and in January to June 1994 (total 12 months) during the high transmission season. All surveys were community-based.


ParticipantsNumber of participants: 244 people lived in intervention houses, and 257 in control houses.

Inclusion criteria: persons living in 1 town area were eligible (primary analysis was for all ages).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: deltamethrin-treated curtains (door, windows); target dose 25 mg/m2; nets re-treated before each transmission season.

Control group: untreated curtains.

No information available on usage rates.


Outcomes1. Incidence of malaria episodes (all ages + children aged 0 to 9 years).


NotesStudy location: town of Ankazobe (100 km from Antananarivo, at 1300m altitude) in Madagascar.

EIR: 2, very seasonal transmission.

Malaria endemicity: mesoendemic.

Main vector: Anopheles funestus.

Plasmodium vivax malaria: 0%.

Access to health care was good.

Nicaragua (Kroeger)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (20 villages were paired according to size, socioeconomic conditions, and malaria incidence at baseline; 1 village within each pair then randomized to receive the intervention).

Number of units: 10:10.

Length of follow up: 4 months.

For the evaluation, 1 cross-sectional survey carried out during the peak of the malaria season in 1996.


ParticipantsNumber of participants: 5260 individuals took part in the cross-sectional survey (high percentage of total).

Inclusion criteria: inhabitants of the 20 trial communities.


InterventionsIntervention: lambdacyhalothrin treatment of existing bed nets; target dose 13 mg/m2; 75% usage rate (high); sales and promotion of bed nets, and free net treatment.

Control group: no nets; (< 25% usage rate of untreated nets).


Outcomes1. Period-prevalence (last 2 weeks or last 4 months) of reported "malaria episodes" assessed during the peak of the malaria season.

Outcome measures similar to Ecuador (Kroeger).


NotesStudy location: El Viejo Municipio, Department of Chinandega, North East Nicaragua (Pacific coast).

EIR: well below 1.

Malaria endemicity: hypoendemic.

Baseline parasite rate in the whole population: 8%.

Main vector: Anopheles albimanus.

Plasmodium vivax malaria: virtually all infections due to P. vivax.

Access to health care was likely to be good.

Pakistan (Rowland)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (random allocation of 192 households with 2792 individuals of all ages after a first random selection of 10% of all households from a census list; the aim of this procedure was to measure the impact of treated nets in a condition of low net usage).

Number of units: 173:186.

Length of follow up: 6 months.

Morbidity rates monitored longitudinally by passive case detection in a project clinic. Blood slide made for all suspected malaria cases; each positive blood slide was a case. Monitoring was from June to December 1991, covering the main transmission period. Overall completion rate was 97%. A single cross-sectional survey was carried out in December 1991 to January 1992.


ParticipantsNumber of participants: 2792 (all ages).

Inclusion criteria: chosen from 2 Afghan refugee camps: Baghicha and Kagan.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated polyester bed nets; target dose 500 mg/m2; 2% usage rate before the trial (very low).

Control group: no bed nets; < 2% usage rate.


Outcomes1. Incidence of mild clinical episodes (all ages) for both Plasmodium falciparum and P. vivax.
2. Prevalence of any parasitaemia (P. falciparum and P. vivax).


NotesStudy location: Mardan District, North West Frontier Province, North West Pakistan.

EIR: low.

Malaria transmission is unstable in the area, with 22% of individuals reporting having had malaria in the past year.

Parasite rates: 2.4% for P. falciparum and 10.9% for P. vivax.

Main vectors: Anopheles culicifacies and A. stephensi.

Plasmodium vivax malaria: 77% of all cases (kept separate in analysis).

Good access to health care.

Peru Amazon (Kroeger)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (36 communities were paired according to size, geographic location, net coverage, and malaria incidence at baseline; 1 village within each pair was then randomized to receive the intervention).

Number of units: 18:18.

Length of follow up: 17 months.

For the evaluation, one cross-sectional survey was carried out in April 1992.


ParticipantsNumber of participants: 5709 individuals took part in the cross-sectional survey (high percentage of total).

Inclusion criteria: inhabitants of the 36 trial communities.


InterventionsIntervention: permethrin treatment of existing bed nets; target dose 200 mg/m2; usage rate very high (95% of families with at least one net); net treatment in November 1991 and January 1992; nearly 61% of all existing nets treated at least once; free bed net treatment (sales were not necessary because of the high usage rate).

Control group: untreated bed nets; 95% usage rate.


OutcomesPeriod-prevalence (last two weeks or last four months) of reported "malaria episodes" assessed in April 1992.

Outcome measures similar to Ecuador (Kroeger).


NotesStudy location: Tambopata District, Madre de Dios Department in the Amazonas region of Peru.

EIR: < 1, little seasonality.

Malaria endemicity: hypoendemic.

Baseline parasite rate in the whole population and spleen rate in children aged 2 to 9 years: < 5%.

Main vectors: Anopheles evansae and A. nunetzovari.

Plasmodium vivax malaria: 100%.

Access to health care was likely to be good.

Peru Coast (Kroeger)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (12 villages were paired according to size, geographic location, net coverage, and malaria incidence at baseline; 1 village within each pair was then randomized to receive the intervention).

Number of units: 6:6.

Length of follow up: 29 months.

2 cross-sectional surveys carried out during the peak of the malaria season in June to July 1992 and 1993.


ParticipantsNumber of participants: 6941 individuals took part in the 2 cross-sectional surveys (high percentage of total).

Inclusion criteria: inhabitants of the 12 trial communities.


InterventionsIntervention: lambdacyhalothrin (first year; target dose 10 mg/m2) or permethrin (second year; target dose 500 mg/m2) treatment of existing bed nets; moderate usage rate (63% of families with at least 1 net); net treatment in January and March 1992 and 1993; nearly 67% of all existing nets treated at least once; sales and promotion of bed nets, and free net treatment.

Control group: untreated bed nets; 63% usage rate.


Outcomes1. Period-prevalence (last 2 weeks or last 4 months) of reported "malaria episodes" assessed during the peak of the malaria season (June to July 1992/1993).

Outcome measures similar to Ecuador (Kroeger).


NotesStudy location: Communidad de Catacaos, Piura Department, northern Peru on the Pacific Coast.

EIR: < 1.

Malaria endemicity: hypoendemic.

Baseline parasite rate in the whole population and spleen rate in children aged 2 to 9 years: < 5%.

Main vector: Anopheles albimanus.

Plasmodium vivax malaria: 100%.

Access to health care was likely to be good.

Sierra Leone (Marbiah)

MethodsStudy design: cluster randomized controlled trial.

Unit of allocation: village (17 villages were paired according to size, altitude, climate, and presence of a health centre; 1 village in each pair was then randomized to the intervention; children were also randomized individually to either chemoprophylaxis with pyrimethamine/dapsone (Maloprim) or placebo - my analysis focused on the placebo group in order to exclude the effect of chemoprophylaxis).

Number of units: 9:9.

Length of follow up: 12 months.

Overall dropout rates were 17% in the bed net group and 18% in the control group.

Follow up through weekly visits to all study children. A short questionnaire was administered to the mother, and the temperature of the child was recorded. Blood slide made if the child was reported to have been ill during the last 7 days or if the temperature was at least 37.5 C; case of malaria recorded if the slide revealed a parasitaemia of at least 2000 parasites per µl (children under 2 years) or at least 5000 parasites per µl (children aged 2 to 6 years). Monitoring from July 1992 to June 1993.

A cross-sectional survey was carried out in March 1993.

All surveys were community-based.


ParticipantsNumber of participants randomized: 920 treated nets (n = 470) or no nets (n = 450).

Inclusion criteria: children aged 3 months to 6 years.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: lambdacyhalothrin-treated bed nets; target dose 10 mg/m2.

Control group: no bed nets; very low usage rate.

In addition, children were randomized individually to either chemoprophylaxis with pyrimethamine/dapsone (Maloprim) or placebo.


Outcomes1. Incidence of malaria episodes (children aged 3 months to 6 years).
2. Prevalence of anaemia (mean packed cell volume).
3. Prevalence of splenomegaly (Hackett 1 to 5).


NotesStudy location: 17 villages near the town of Bo, Sierra Leone.

EIR: 35.

Malaria endemicity: hyperendemic.

Baseline parasite rate in children aged 1 to 5 years: 49.2%.

Main vector: Anopheles gambiae.

Plasmodium vivax malaria: 0%.

Access to health care was considered poor.

Tanzania (Fraser-Hurt)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: individual (random allocation of 120 children aged 5 to 24 months from an existing village list).

Number of units: 120 children.

Length of follow up: 6 months.


ParticipantsNumber of participants: 120.

Inclusion criteria: children aged 5 to 24 months.


InterventionsIntervention: permethrin-treated polyester bed nets; target dose 500 mg/m2; 90% usage rate (very high).

Control group: no nets.


Outcomes1. Parasitaemia.
2. Haemoglobin.
3. Multiplicity of infections measured during repeated cross-sectional survey.


NotesStudy location: Kiberege village, Kilombero District, Tanzania.

EIR: high (around 300 per year).

Plasmodium falciparum prevalence rate in this age group: 60%.

Main vectors: Anopheles gambiae s.l. and A. funestus

Plasmodium vivax malaria: no information available.

Thailand (Kamol-R)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: household (random allocation of 54 households with 270 adults after stratifying for malaria endemicity).

Number of units: 26:28.

Length of follow up: 8 months.

Morbidity rates monitored longitudinally by weekly follow up at which blood slides were taken systematically; each positive blood slide was a case. Monitoring from November 1987 to July 1988 (35 weeks) covering the main transmission period. Completion rates were 96 and 97%. Differentiation made between Plasmodium falciparum and P. vivax malaria (40% of all cases).


ParticipantsNumber of participants: 261.

Inclusion criteria: adult migrant workers (male:female ratio was 1.4).

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated nylon bed nets; target dose 500 mg/m2; approximately 87% usage rate before trial.

Control group: untreated bed nets; > 95% usage rate.


Outcomes1. Incidence of mild clinical episodes (adults) for both P. falciparum and P. vivax.


NotesStudy location: Bothong District, Chonburi Province (rural) in eastern Thailand.

EIR: low.

Malaria transmission is unstable in the area.

Main vector: Anopheles dirus.

Plasmodium vivax malaria: 43% of all cases (kept separately in analysis).

Good access to health care.

Thailand (Luxemberger)

MethodsStudy design: individual randomized controlled trial.

Unit of allocation: individual (random allocation of 350 children aged 4 to 15 years from an existing list of all school children).

Number of units: 175:175.

Length of follow up: 7 months.

Morbidity rates monitored longitudinally by passive case detection (high access to health care) as well as through the identification of school absentees who were brought to the dispensary for examination. Monitoring from August 1990 to February 1991, covering 1 of the 2 peaks in transmission (December to January). Blood slide made for every person reporting with a febrile illness compatible with malaria; every positive slide labelled a "malaria case"; differentiation made between Plasmodium falciparum and P. vivax malaria (30% of all cases). 2 cross-sectional surveys conducted at 3 and 6 months (92% participation rate).


ParticipantsNumber of participants: 318.

Inclusion criteria: children aged 4 to 15 years.

Exclusion criteria: no explicit exclusion criteria except absence of written consent.


InterventionsIntervention: permethrin-treated cotton bed nets; target dose 500 mg/m2; approximately 70% usage rate before trial.

Control group: untreated bed nets; > 95% usage rate.

Co-intervention: 22% use of treated nets at baseline.


Outcomes1. Incidence of mild clinical episodes (5 to 14 years) for both P. falciparum and P. vivax.
2. Prevalence of any parasitaemia.
3. Prevalence of splenomegaly.


NotesStudy location: Shoklo (Karen) refugee camp in northern Thailand.

EIR: low.

Malaria transmission is unstable in the area, with 800 infections/1000 inhabitants/ year in that age group.

Main vectors: Anopheles dirus and A. minimus (likely main vectors).

Plasmodium vivax malaria: 30% of all cases (kept separate in analysis).

Good access to health care.

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Afghanistan(Rowland)Treated chaddar and top sheets, not nets or curtains.

Benin (Akogbeto)Non-randomized allocation of 2 areas within 1 large village.

Brazil (Santos)Non-randomized allocation of 60 households in 2 villages.

Burkina (Carnevale)Non-randomized allocation of 2 areas within 1 village.

Burkina (Pietra)Non-randomized allocation of 2 areas within 1 village.

Burkina F (Procacci)Non-randomized allocation of 2 clusters within 1 village.

Cambodia (Chheang)Non-randomized allocation of 2 “blocks” of each 2 hamlets.

Cameroon (LeGoff)No contemporaneous control group; before-after assessment.

China (Cheng Hailu)Non-randomized allocation of 20 villages.

China (Li)No proper control group but comparison of users and non-users; before-after comparison.

China (Luo Dapeng)Non-randomized allocation of 5 villages.

China (Wu Neng I)Non-randomized allocation of 3 townships.

China (Wu Neng II)Non-randomized allocation of 2 villages.

China (Yuyi station)Non-randomized allocation of 3 villages.

Ecuador (Yepez)Non-randomized allocation of 2 villages.

Gambia (Alonso)Non-randomized allocation of 70 villages

Guatemala (Richards)Non-randomized allocation of 3 villages; a further 100 households in 2 additional villages allocated randomly to treated bed nets or no bed nets.

Guinea-B. (Jaenson)Non-randomized controlled trial; and mechanism of allocation not clear.

India (Banerjee)Military personnel and not general population.

India (Das)Non-randomized allocation of 3 villages.

India (Jana-Kara)Non-randomized allocation of 12 villages.

India (Yadav I)Non-randomized allocation of 6 villages.

India (Yadav II)Non-randomized allocation of 10 villages.

India (Yadav III)Non-randomized allocation of 5 villages.

Indonesia (Nalim)Non-randomized allocation of 4 villages.

Iran (Zaim II)Non-randomized allocation of 5 villages.

Irian Jaya (Sutanto)Non-randomized allocation of 2 villages.

Ivory Coast(Doannio)Non-randomized allocation of 2 areas in 1 large village.

Kenya (Beach)Non-randomized allocation of 3 villages blocks.

Kenya (Macintyre)Treatment of bed sheets ("shukas"), not sheets or curtains.

Kenya (Mutinga)Non-randomized allocation of 3 villages.

Kenya (Oloo I)Non-randomized allocation of 20 houses.

Kenya (Oloo II)Non-randomized allocation of 2 villages.

Malawi (Rubardt)Non-randomized allocation of 12 villages.

Malaysia (Hii I)Non-randomized allocation of 6 villages.

Malaysia (Hii II)Non-randomized allocation of 22 villages.

Mali (Doumbo)Non-randomized allocation of 2 villages.

Mali (Ranque)Non-randomized allocation of only 10 households.

Mozambique (Crook)Non-randomized allocation of 2 areas within part of Maputo (the capital city).

Myanmar (Lwin)Non-randomized allocation of 2 areas within 1 township.

Nepal (Sherchand)Non-randomized allocation of 5 village development committees.

Nigeria (Brieger)Non-randomized allocation of 12 village clusters (into 4 treatment arms).

Papua NG (Graves)Non-randomized allocation of 8 paired villages.

Philippines(Quilala)Allocation “by chance” of the intervention to 6 villages.

Senegal (Faye)Non-randomized allocation of 2 villages.

Solomon (Hii)Non-randomized allocation of 2 zones.

Solomon (Kere I)Non-randomized allocation of 2 zones.

Solomon (Kere II)Non-randomized allocation of 3 areas.

Sudan (El Tayeb)Non-randomized allocation of only 2 villages.

Tanzania (Lyimo)Non-randomized allocation of only 4 villages.

Tanzania (Maxwell)Non-randomized allocation of control villages.

Tanzania (Njau)Non-randomized allocation of 368 households in 1 large village.

Tanzania (Njunwa)Non-randomized allocation of 4 villages.

Tanzania (Premji)Non-randomized allocation of 7 villages in 2 blocks.

Tanzania (Stich)Non-randomized allocation of 2 villages (2 phases, 3 years apart, in a cross-over design).

Vietnam (Dang)Allocation “by chance” of the intervention to 200 workers:

Vietnam (IMPE)Non-randomized allocation of 2 villages.

Vietnam (Nguyen)Non-randomized allocation of 13 hamlets.

Zaire (Karch)Non-randomized allocation of 3 villages.

 
Comparison 1. Insecticide-treated nets versus all controls

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Child mortality from all causes (relative rate)5149221Relative rate (Fixed, 95% CI)0.82 [0.76, 0.89]

    1.1 Controls with no nets
4124369Relative rate (Fixed, 95% CI)0.83 [0.76, 0.90]

    1.2 Controls using untreated nets
124852Relative rate (Fixed, 95% CI)0.77 [0.63, 0.95]

 2 Child mortality from all causes (risk difference)5149221Risk difference (RD) (Fixed, 95% CI)-5.53 [-7.67, -3.39]

    2.1 Controls with no nets
4124369Risk difference (RD) (Fixed, 95% CI)-5.52 [-7.88, -3.16]

    2.2 Controls with untreated nets
124852Risk difference (RD) (Fixed, 95% CI)-5.6 [-10.70, -0.50]