Socio-demographic characteristics of the caregivers
As shown in Table 1, all 350 caregivers interviewed completed the survey with a nearly equal number of respondents from WAU and WAR. Most of them were less than 30 year-old mothers. 70.0% of caregivers were married, while 19.4% were single, and most caregivers were unemployed younger mothers. The total unemployed reached 46.3% including 33 younger students. 27.1% of them had no educational experience and only 64.3% had at least a secondary level of education. But all of them had at least one child under-5 living with them and 54.6% had at least one more child above-5.
Knowledge
99.1% of the caregivers had heard about malaria, and no significant difference of malaria related information sources (P = 0.858) was found between the caregivers from WAR and WAU (Additional file 3: Table S1). Most of them got malaria related knowledge from health workers/facilities (89.0%), followed by radio (60.8%), television (17.6%) and printed materials such as billboards/handbills (8.1%). Besides, 26.8% got information from other sources such as the community sensitization and announcements about malaria, their friends/peers and also neighbors, and only 1.4% were informed through the internet/social media, even only 1.7% of those residing in WAU where the use of internet/social media was expected to be high, just a little higher than those living in WAR (1.1%) (Fig. 2).
As displayed in Fig. 3 and Additional file 3: Table S1, the percent of respondents that did not know malaria can be prevented, cured and lead to death were 18.0%, 3.4%, and 8.9%, respectively. Among them, only 1.2% of urban-based caregivers were unaware of lethality of malaria, whereas the number for rural-based caregivers was over four times higher (5.6%) (P = 0.05). When asked about the symptoms of malaria, the most common responses were fever (71.1%), followed by vomiting (42.9%), loss of appetite (30.9%), and body and joint pains (26.9%), with only a few responses for headaches (9.1%). 58.6% of them stated other signs and symptoms they knew such as weakness and dizziness, dark colored urine, yellow/pale/white eyes, etc. Approximately 10.0% didn’t know that malaria parasites can be transmitted through mosquito bites. This study revealed that, misconceptions about the causes and mode of transmission of malaria still prevailed among the caregivers, as 0.3% of them stated that eating too much and having close contacts with a person who has malaria can cause malaria. Also, 4.0% of them stated other misconceptions about the cause of malaria such as eating too many oranges, not washing hands regularly, cold and flu, dirty hands and feet, flies, drinking too many sweet/soft drinks, etc. While 86.6% of the caregivers stated that bushes/dirty places were convenient resting/breeding places for mosquitoes, followed by stagnant water (51.1%), and dark places/sheds (11.7%).
Attitudes
As summarized in Fig. 3 and Additional file 4: Table S2, although 90.0% of respondents believed that malaria is a very serious health problem, still 3.1% said “No”, 5.4% “Not sure” and 1.4% “Don’t know”. 77.7% said that they and their child(ren) do sleep under a bed net, while the rural-based caregivers(85.3%) showed significantly more positive attitudes than the urban (69.9%) (P < 0.001). Explaining the lack of bed net for their children, 46.2% (46/78) of them mentioned that they could not afford it, 32.1% said not readily available, 6.4% did not like sleeping under a bed net, 3.8% do not think it’s important, and 11.5% stated other reasons such as the statements that they and their child(ren) had allergic reactions or sweat a lot whenever they slept under a bed net (Additional file 4: Table S2, Fig. 4a). Moreover, when asked about how often your child(ren) used the bed net, 85.7% stated always and 14.3% sometimes, but 21.7% did not think the bed net was treated with insecticide. Meanwhile, 88.0% of the respondents thought that malaria-transmitted mosquitoes bite mostly at night. 65.1% stated that they think that Artemisinin-based Combination Therapies (ACTs) were the best treatment for malaria, but 32.0% said “Don’t know”, and 1.4% also shared other opinions such as injections, intravenous solutions, etc.
Practices
As shown in Fig. 3 and Additional file 5: Table S3, although 77.1% of respondents would go to a hospital or clinic when their child(ren) had a fever, 19.4% still preferred self-medicating first at home, and 3.4% went first to a pharmacy. As stated by 93.4% of the caregivers, the most important factor influencing their action to seek medical help was the condition of the child, followed by perceived cost involved (3.7%) and time availability (1.7%). Furthermore, only 54.6% of the caregivers would seek medical attention for their febrile child(ren) within 24 hours, whilst 45.1% usually wait for 2-5 days. Meanwhile, when an antimalarial medication had been prescribed, only 71.4% of caregivers mentioned administration of a complete treatment course, but 12.9% stop administration as soon as the child began to show improvement. There was a significantly higher rate among the urban-based caregivers (79.8%) who performed a complete treatment course than among the rural ones (63.3%) (P < 0.001).
Regarding the use of protective measures against malaria-transmitted mosquitoes, the most common practices employed by 61.1% of respondents slept under an insecticides treated nets (ITNs), the rural-based caregivers had a significantly higher rate of ITNs use than the urban (70.1% vs. 52.0%, P < 0.01) (Fig. 4a). But we observed that misuse of bet nets such as using for doing pond fishing (Fig. 4b) and covering backyard gardens (Fig. 4c) is still common especially in rural and provincial areas of Sierra Leone. Other common selections for protection included measures of using mosquito repellant (40.9%), regularly cleaning surroundings of the house (31.7%), wearing protective clothing (25.1%), and using insecticide spray (24.9%). Only a few caregivers said that they got rid of stagnant water (8.3%, 29/350), and cleared bushes around the house (3.4%, 12/350), while 11 (6.4%, 11/173) of them residing in WAU and only 1 from WAR (0.6%, 1/177). Also, 6.0% of caregivers mentioned other protective measures such as closing the windows and doors early before dark, fixing a mesh on window and door frames to keep mosquitoes out, etc (Fig. 3-4, Additional file 5: Table S3).
Correlations among knowledge, attitudes and practices
Inferred by the spearman correlation test (Table 2), a significant positive correlation was found between knowledge-attitudes (rs = 0.13, P< 0.05) and attitudes-practices (rs = 0.45, P < 0.001). Further univariate logistic regression analysis indicated that improved knowledge might positively affect attitudes (OR = 1.98; 95%CI = 1.21-3.25), while positive attitudes might lead to good practices (OR = 7.91; 95%CI = 4.33-15.54).
Effects of socio-demographic characteristics on KAPs
Through univariate analysis, the association of socio-demographic and malaria-related KAPs were demonstrated by comparing with referred groups (Table 3). The age, religion, educational background and occupation were found to correlate with knowledge. The caregivers in the >30 year-old group tended to have higher knowledge than the 15-20 year-old group (OR = 3.03, 95%CI = 1.35-7.28). In comparison to the Muslim religion, Caregivers’ Christian faith was positively associated with knowledge (OR = 2.20, 95%CI = 1.29-3.86). The caregivers with secondary education had higher knowledge than the group with no education (OR = 1.94, 95%CI = 1.15-3.28) . Meanwhile, comparing with the unemployed caregivers, students knew more malaria-related information (OR = 3.04, 95%CI = 1.20-9.33), but represented a negative practice (OR = 0.22, 95%CI = 0.07-0.56). As for attitudes, age, educational background and district (urban or rural) were inferred to be associated. A positive association with attitudes was found in the respondents with a university education (OR = 4.46, 95%CI = 1.42-19.73), and the rural-based caregivers (OR = 1.83, 95%CI = 1.14-2.96), but a negative association was found in the caregivers of the 21-25 year-old group (OR = 0.39, 95%CI = 0.19-0.76).
Furthermore, independent predictors of KAPs were inferred by a multivariate logistic regression analysis comparing with referred groups (Table 4). The age, religion, education and occupation were found to be independent predictors for malaria-related knowledge. The groups of 26-30 year-old (OR = 2.14, 95%CI = 1.02-4.55) and >30 year-old (OR = 4.83, 95%CI = 1.95-12.69), the caregivers with a secondary education (OR = 2.34, 95%CI = 1.32-4.19), the Christians (OR = 2.28, 95%CI = 1.30-4.14), and the students (OR = 2.98, 95%CI = 1.12-9.48) were more likely to have a higher knowledge about malaria. Meanwhile, district, education, age and relationship were found as the independent predictors for malaria-related attitudes. The rural-based caregivers (OR = 1.97, 95%CI = 1.20-3.28), the group with a higher education (university level, OR = 5.53, 95%CI = 1.67-25.54; secondary level, OR = 1.86, 95%CI = 1.05-3.30) were more likely to have a positive attitude towards malaria, while caregivers in the 21-25 year-old group (OR = 0.40, 95%CI = 0.19-0.79) and fathers (OR = 0.19, 95%CI = 0.04-0.80) were more likely to have negative malaria-related attitudes. Meanwhile the relationship, occupation, and marital status were taken as independent predictors for malaria-related practices. The mothers seemed more likely to have better practices towards malaria control and prevention as compared to the fathers (OR = 0.19, 95%CI = 0.04-0.80). The caregivers who were students (OR = 0.11, 95%CI = 0.01-0.55) or married (OR = 0.53, 95%CI = 0.29-0.97) or in a consensual union (OR = 0.30, 95%CI = 0.12-0.74) were more likely to have bad malaria-related practices.