Evaluating the Quality of Malaria-Related Health Information in the Nigerian Internet Context ()
1. Background
Malaria is the leading cause of death and disease in many developing countries and has serious health burden around the world [1] . In 2010, there were 24 million confirmed malaria cases in 106 malaria endemic areas or territories, with the majority of the cases occurring in Sub-Saharan Africa [2] . Specifically, 40% of all deaths caused by malaria can be accounted for in Nigeria, Congo and Dominican Republic [2] .
The devastating disease burden has not gone unnoticed. There has been an increase concentration on malaria control efforts, and the issue is now on the political agenda of several of the world’s wealthiest countries. Funds have become available from the Global Fund to Fight AIDS, Tuberculosis and Malaria, The US President’s Malaria Initiative, the World Bank and bilateral donors, which have not been seen on this scale since the first attempted malaria eradication campaign in the 1950s and 1960s [3] . The Roll Back Malaria Partnership is now coordinating the global fight against malaria effort to reduce mortality, morbidity and economic consequences with major donor foundations, such as the Bill and Melinda Gates Foundation, which have greatly increased financial support for malaria research [3] [4] . Despite the wide scope of funding and concentration on malaria, eradication efforts have not been successful.
For reliable diagnosis of malaria, a combination of the methods of diagnosis is recommended, and training and better quality assurance should be established for the medical, laboratory and community health workers in order to improve malaria diagnosis [5] . Not only would providing open resources to access accurate and relevant information be a valuable resource for health care workers in Africa, but also individuals could become empowered to take control of their own health. The main sources of malaria information in Africa are currently family members, friend, community meetings, health facilities, radio, community health workers and print media [6] . The evidence suggests that social support systems are a valuable preferred resource to receive malaria information. The Internet could become a major resource to disseminate malaria education and information in Africa as accessibility continues to increase.
According to the United Nations Conference on Trade and Development [7] 2011 Information Economy Report, 84 million mobile handsets are already capable of using the Internet, and 7 out of 10 are expected to be Internet-enabled by 2014. Alone, mobile internet penetration in Africa has increased by 155.59 percent in two years [7] . Therefore, this is a pivotal time to evaluate the current state of malaria information on the internet. In return, high quality malaria information on the internet could strengthen the information exchanged through the social support systems, which remains as a preferred source of malaria information. A rapid increase in access and use of the internet exists in Africa. No studies to date have systematically assessed the quality of consumer health information in the internet available to the African user. The purpose of this study is to evaluate the quality of malaria health and treatment information available on the internet provided by the Nigerian context.
2. Methods
A total of 38 websites were included in the final evaluation (Table 1). The websites were found using three search engines: Google, Yahoo! and Bing. In order to retrieve articles as if the searches were conducted in Nigeria, the Local Area Network (LAN) settings were changed to a Nigerian proxy server, with a local Internet Protocol address. To simulate search patterns of novice internet users, the keywords used were limited to “malaria” and “treatment”. Two searches were conducted in each search engine, the first used the keyword “malaria” and the second used the combination “malaria and treatment”. At each search, the top 20 results from each of the 3 search engines were eligible for analysis, as a previous study suggested that users usually do not search beyond those [8] . Therefore, the searches resulted in 120 results. The Google search engine was used as the reference, and any repeated websites from Yahoo and Bing or from the second search were analyzed only once. In addition, links to books, reports, presentations, journal, magazine and newspaper articles were excluded as the target population is the general public and not specialized health care users (Figure 1).
Table 1 . List of evaluated websites.
Figure 1. Flow chart of the search conducted to retrieve websites.
Three raters evaluated the quality of information using the DISCERN [9] instrument criteria. The evaluation was done considering the information available in different sections of the included websites. Links to external addresses, reports and other documents (such as .pdf or .doc) were not included in the evaluation of a specific website. Kendall’s concordance coefficient (W) was calculated to determine the level of agreement among the three raters.
To assign a single quality score for each website, an average between the scores given by each rater was calculated. For each domain, an average score between all websites included in that domain was also calculated. SAS 9.3 (NC, Cary) was used to perform the data analysis.
DISCERN [9] is a standardized set of criteria for judging the quality of health information written for the public. Although DISCERN cannot be used to assess the scientific quality of the information, the instrument is useful to determine the reliability and quality of an information source without recurring to specialist knowledge or other publications and advisers [10] . As the purpose of this study was to evaluate consumer health information from the perspective of a novice Nigerian user, this validated instrument was sufficient to assess the quality of available information. DISCERN offers a list of 16 questions or criteria, each rated through a five-point scale ranging from one (No) to five (Yes) and it indicates if the quality criterion is present, partially present, or not present in the publication [10] . The total score assigned to a website could range from 16, representing the lowest level of quality, to 80, the highest level. Simplified user instructions are available at the website www.discern.org.uk.
3. Results
Among the 38 websites evaluated, the highest inter-rater average score was attributed to the Patient.co.uk website (average = 61), followed by Wikipedia web site and Malaria Site (average = 59) (Figure 2). The “Home Remedies for You” website received the lowest score (average = 18).Most evaluated websites were .com domains (47%, n = 18). Other domains were .org (26%, n = 10), .uk (11%, n = 4), .gov (8%, n = 3), .net (5%, n = 2), and .int (3%, n = 1). The highest average score was given to .co.uk domains (average = 47) while .int had the lowest score (average = 34) (Figure 2).
According to Landis and Koch [11] classification of degrees of agreement, the three raters had moderate (k = 0.4 - 0.6) to substantial agreement (k = 0.61 to 0.8) in all the DISCERN evaluation criteria (Table 2). The raters moderately agreed in 53% (n = 8) of the questions and substantially agreed in 47% (n = 7) of the questions. Responses to Question 2 (website achieves the aims proposed) depended on the response to Question 1 (aims of the website are clearly stated), and if the latter had scored 1 the former should not be considered relevant. Therefore, W could not be achieved for Question 2 using the SAS macro because all three raters had responses missing (R1 = 87%; R2 = 87%; R3 = 36%) due to the irrelevance of the question. The range of agreement was from W = 0.47 (p = 0.02) for Question 6 to W = 0.73 (p < 0.001) for Question 14.
4. Discussion
The quality of information related to malaria treatment available on the internet could be described as fair to medium, from the perspective of the user located in Nigeria. Not surprisingly, .co.uk domains received the highest average score among the three raters. Such result could be due to the DISCERN instrument being developed