Background
Methods
Study design
Study sites
Sampling and sample size
Data preparation and analysis
State | Sample size | Timing | ||||
---|---|---|---|---|---|---|
Households | Nets | Individuals | Month/year | Rainy season | Months after campaign | |
Sokoto | 1008 | 1271 | 4468 | May 2010 | Dry/early | 5.7 |
Katsina | 1017 | 1532 | 4630 | May 2011 | Dry/early | 6.5 |
Kano | 987 | 1173 | 4642 | Oct 2009 | Late rains | 1.9/5.4a
|
Niger | 1001 | 1280 | 6270 | Jun 2010 | Mid-rains | 6.4 |
Nasarawa | 1015 | 1136 | 5323 | Nov 2011 | Late rains | 10.6 |
Anambra | 1012 | 1781 | 4546 | Nov 2009 | Late rains | 4.1 |
Enugu | 1020 | 1444 | 4644 | Apr 2012 | Dry/early | 13.4 |
Ogun | 952 | 745 | 4373 | Jul 2010 | Mid-rains | 7.2 |
Lagos | 1020 | 937 | 4486 | Jun 2012 | Mid-rains | 9.0 |
Cross river | 1254 | 1316 | 5656 | Jun 2012 | Mid-rains | 8.2/16.5a
|
TOTAL | 10,286 | 12,615 | 49,038 |
General approach to analysis and treatment effect model
Uni- and multi-variable analysis
Results
Exposure to and recall of messages on net hanging and use
Message recall
Determinants of information exposure
Explanatory variablea
| Outcome 1: any exposure | Outcome 2: recall two or more messages | ||||
---|---|---|---|---|---|---|
OR | 95 % CI |
p value | OR | 95 % CI |
p value | |
Net ownership | ||||||
No net | 1.00 | – | – | 1.00 | – | – |
Net but not from campaign | 2.20 | 1.4, 3.5 | 0.001 | 2.12 | 1.2, 3.8 | 0.012 |
Campaign net | 18.02 | 14.5, 22.4 | <0.0001 | 9.62 | 7.6, 12.1 | <0.0001 |
Time since campaign | ||||||
0–3 months | 1.00 | – | – | 1.00 | – | – |
3–6 months | 2.54 | 1.2, 5.4 | 0.015 | 4.87 | 2.0, 11.7 | <0.0001 |
6–9 months | 3.82 | 1.4, 10.7 | 0.11 | 6.89 | 2.3, 20.4 | 0.001 |
9–16 months | 4.73 | 1.8, 12.5 | 0.002 | 9.05 | 3.2, 26.0 | <0.0001 |
Wealth quintiles | ||||||
Poorest | 1.00 | – | – | 1.00 | – | – |
Second | 0.90 | 0.7, 1.2 | 0.42 | 0.87 | 0.7, 1.1 | 0.17 |
Third | 0.83 | 0.6, 1.1 | 0.23 | 0.81 | 0.6, 1.0 | 0.10 |
Fourth | 1.21 | 0.9, 1.6 | 0.20 | 1.12 | 0.9, 1.4 | 0.37 |
Wealthiest | 1.13 | 0.8, 1.5 | 0.04 | 1.10 | 0.8, 1.4 | 0.47 |
Education of head | ||||||
Non-literate | 1.00 | – | – | 1.00 | – | – |
Primary | 1.02 | 0.8, 1.3 | 0.88 | 1.00 | 0.8, 1.2 | 0.98 |
Secondary | 1.23 | 0.9, 1.6 | 0.10 | 1.34 | 0.9, 1.4 | 0.27 |
Tertiary and higher | 1.18 | 0.9, 1.6 | 0.27 | 1.11 | 0.8, 1.5 | 0.48 |
Age of head in years | ||||||
19–29 | 1.00 | – | – | 1.00 | – | – |
30–59 | 1.15 | 0.8, 1.5 | 0.30 | 1.19 | 0.9, 1.6 | 0.23 |
60+ | 1.29 | 0.9, 1.8 | 0.12 | 1.29 | 0.9, 1.7 | 0.092 |
HH with 3 or less people and no children | 0.69 | 0.6, 0.8 | <0.0001 | 0.69 | 0.6, 0.8 | <0.0001 |
HH head is female | 1.20 | 0.9, 1.5 | 0.075 | 1.14 | 0.9, 1.4 | 0.18 |
HH has radio | 1.37 | 1.0, 1.8 | 0.022 | 1.14 | 0.9, 1.4 | 0.18 |
Urban versus rural | 0.98 | 0.7, 1.3 | 0.85 | 0.93 | 0.7, 1.1 | 0.44 |
North versus south | 0.76 | 0.5, 1.2 | 0.20 | 1.34 | 0.9, 2.0 | 0.14 |
Complementarity of information exposure and sources
Variable | Information channel | ||||
---|---|---|---|---|---|
Campaign | Health worker | Mediators | Media | Social networks | |
Has campaign net versus non-campaign nets | OR 32.3
p < 0.001 | OR 6.2
p < 0.001 | OR 3.5
p = 0.007 | OR 2.4
p = 0.005 | OR 1.0
p = 0.99 |
Time since campaigna
| Increasing
p < 0.001 | Increasing
p < 0.001 | Increasing
p < 0.001 | Neutral
p = 0.20 | Decreasing
p < 0.001 |
Wealth quintiles 4/5 versus 1–3 | OR 1.3
p = 0.023 | OR 1.0
p = 0.93 | OR 0.91
p = 0.36 | OR 1.3
p < 0.001 | OR 1.3
p = 0.033 |
Education of head Secondary or higher | OR 1.1
p = 0.50 | OR 1.2
p = 0.033 | OR 0.76
p = 0.011 | OR 1.3
p = 0.008 | OR 0.92
p = 0.50 |
Age of head of HH | |||||
15–29 30–59 60+ | – OR 1.2 OR 1.1
p = 0.53 | – OR 1.1 OR 0.86
p = 0.077 | – OR 1.1 OR 1.5
p = 0.006 | – OR 1.2 OR 1.2
p = 0.43 | – OR 0.75 OR 0.89
p = 0.069 |
Gender of head Female versus male | OR 1.0
p = 0.77 | OR 0.87
p = 0.20 | OR 1.0
p = 0.74 | OR 1.2
p = 0.10 | OR 1.1
p = 0.43 |
Family size Less than 3b versus more | OR 0.91
p = 0.39 | OR 0.67
p < 0.001 | OR 0.85
p = 0.11 | OR 0.84
p = 0.069 | OR 0.66
p = 0.001 |
Residence Urban versus rural | OR 0.88
p = 0.31 | OR 1.2
p = 0.21 | OR 0.82
p = 0.15 | OR 0.85
p = 0.20 | OR 1.7
p = 0.001 |
Region North versus South | OR 0.49
p < 0.001 | OR 0.74
p = 0.15 | OR 3.3
p < 0.001 | OR 1.3
p = 0.17 | OR 0.95
p = 0.77 |
Radio ownership versus none | OR 0.91
p = 0.43 | OR 1.0
p = 0.66 | OR 0.89
p = 31 | OR 2.7
p < 0.001 | OR 1.1
p = 0.52 |
Year 2009 versus 2010–2012c
| OR 3.3
p = 0.004 | OR 0.83
p = 0.66 | OR 2.4
p = 0.076 | OR 1.2
p = 0.62 | OR 0.26
p < 0.001 |
Relationship between information sources and messages recalled
Message | Information channel—adjusted OR (95 % CI) | ||||
---|---|---|---|---|---|
Campaign | Health worker | Mediators | Media | Social networks | |
Use net | 3.4 (2.7, 4.3) | 2.2 (1.7, 2.7) | 3.5 (2.8, 4.6) | 3.3 (2.6, 4.1) | 2.6 (2.0, 3.4) |
Sleep under net every night | 3.3 (2.6, 4.1) | 3.9 (3.1, 4.8) | 3.8 (2.8, 5.3) | 5.3 (4.2, 6.8) | 2.4 (1.8, 3.0) |
Hang net | 4.2 (3.3, 5.2) | 3.5 (2.8, 4.2) | 3.0 (2.3, 3.7) | 1.9 (1.6, 2.4) | 1.7 (1.3, 2.1) |
Nets prevent malaria | 3.0 (2.4, 3.7) | 3.7 (2.9, 4.6) | 3.7 (2.9, 4.7) | 2.8 (2.2, 3.4) | 6.2 (4.8, 7.9) |
Value net | 3.4 (2.5, 4.4) | 2.6 (2.0, 3.5) | 3.0 (2.3, 3.8) | 3.4 (2.6, 4.5) | 2.5 (1.9, 3.4) |
BCC outcome measures
Attitudes towards net use
Knowledge and perceptions/beliefs about net use
Impact of BCC outcomes on net use
Net use by individuals
Explanatory variablea
| Outcome: net use last night | ||
---|---|---|---|
OR | 95 % CI |
p value | |
Confidence to take action on nets | |||
Poor | 1.00 | – | – |
Good | 1.45 | 1.1, 1.9 | 0.006 |
Very good | 1.76 | 1.3, 2.3 | <0.001 |
Excellent | 1.93 | 1.5, 2.5 | <0.001 |
Discussing net use | |||
Discuss versus no discussion | 1.56 | 1.3, 1.9 | <0.001 |
Intention to use net | |||
Use every night versus less frequent | 2.32 | 2.0, 2.7 | <0.001 |
Difficulties in hanging net | |||
Did report versus not | 0.40 | 0.3, 0.6 | <0.001 |
Season | |||
Dry | 1.00 | – | – |
Mid-rain | 1.53 | 1.1, 2.1 | 0.005 |
Late rain | 1.36 | 0.9, 2.1 | 0.18 |
Supply with nets | |||
Less than 1/3 persons | 1.00 | – | – |
1/3 Persons | 2.23 | 1.9, 2.7 | <0.001 |
1/2 Persons or better | 3.38 | 2.9, 4.0 | <0.001 |
Region | |||
North versus south | 1.63 | 1.2, 2.3 | 0.003 |
Residence | |||
Urban versus rural | 0.82 | 0.7, 1.0 | 0.040 |
Age and gender group | |||
Males 15–49 | 1.00 | – | – |
Children under 5 (M and F) | 2.29 | 1.9, 2.7 | <0.001 |
Children 5–14 (M and F) | 1.45 | 1.3, 1.7 | <0.001 |
Women 15–49 not pregnant | 1.70 | 1.5, 1.9 | <0.001 |
Women 15–49 pregnant | 2.13 | 1.7, 2.7 | <0.001 |
Age 50+ (M and F) | 1.31 | 1.1, 1.6 | 0.001 |
Relationship | |||
Is head of household versus other members | 1.57 | 1.4, 1.8 | <0.001 |
Estimation of treatment effects on net hanging and use
BCC outcomes | Outcome: population net use | |||
---|---|---|---|---|
Univariable | Treatment effects model | |||
Estimate (%) | 95 % CI | Adjusted estimate (%) | 95 % CI | |
Confidence to take action on nets | ||||
Poor | 22.9 | 18.6, 27.9 | 27.4 | 25.1, 29.7 |
Good | 32.9 | 29.8, 36.2 | 36.5 | 35.2, 37.8 |
Very good | 41.7 | 38.8, 44.7 | 42.5 | 41.5, 43.6 |
Excellent | 49.7 | 47.4, 52.0 | 44.9 | 44.1, 45.6 |
Treatment effect as difference in use (poor vs. excellent) | 26.8 | 17.4 | 15.0, 19.0 | |
Intention to use net | ||||
Use less than every night | 32.2 | 29.8, 34.6 | 35.1 | 34.3, 35.9 |
Use every night | 54.5 | 52.3, 56.6 | 50.5 | 49.5, 51.4 |
Treatment effect as difference in use | 22.3 | 15.4 | 14.2, 16.6 | |
Discussing net use | ||||
No discussion | 30.6 | 26.6, 34.9 | 34.7 | 33.1, 36.3 |
Discuss | 44.7 | 42.7, 44.6 | 43.1 | 43.0, 44.2 |
Treatment effect as difference in use | 14.1 | 8.4 | 6.7, 10.1 |