Systematic Review of Human Immunodeficiency Virus (HIV) Knowledge Measurement Instruments Used on the Arabian Peninsula

Introduction:  In 1984, the Kingdom of Saudi Arabia (KSA) began surveillance for human immunodeficiency virus (HIV) incidence and prevalence. In October 1991, the Saudi Minister of Education directed all educational regions and school health units to implement AIDS education in secondary schools.  Until this point, the level of public knowledge about acquired immune deficiency syndrome (AIDS) was generally not addressed in the Kingdom of Saudi Arabia (KSA) for several reasons. At the time, AIDS was an emerging disease with a relatively low prevalence in the KSA compared to other countries, and public discussion of it was taboo. However, beginning with the World AIDS Day in December 1992, measuring the KSA public’s knowledge of AIDS (and later human immunodeficiency virus (HIV)) became more important to the KSA government, spurring the development of instruments to measure AIDS and HIV knowledge in addition to related topics, such as attitudes toward infected individuals.

Today, the highest rate of HIV prevalence in KSA has been consistently found among the non-drug using heterosexual population.  However, no culturally-appropriate standardized questionnaire has been developed to measure HIV prevention knowledge in this population. Evidence exists that married Saudi women are especially at higher risk for infection, but lack knowledge of HIV modes of transmission and underestimate their personal risk of becoming infected. 

Purpose:  To: 1) enumerate appropriate HIV/AIDS knowledge questionnaires that could inform a current study of HIV/AIDS knowledge in the KSA population (especially women), and 2) analyze them for their strengths and weaknesses in terms of reliability, validity, and cultural compatibility.

Methods:  To select appropriate studies and instruments about HIV/AIDS knowledge to review, the following criteria were applied: 1) Must include discussion of a quantitative survey instrument used, 2) quantitative survey must be given either as an interview or self-reported questionnaire, 3) survey must include questions about knowledge HIV or AIDS in specific, but can include questions about other constructs (e.g., attitudes), and can include questions about other STIs, 4) survey must be given in either English or Arabic, but can be given in additional languages, 5) survey must be done in a population in the KSA or other countries bordering the KSA on the Arabian Peninsula, including: Kuwait, Oman, United Arab Emirates (UAE), and Qatar. Google Scholar and Google Web were utilized to search for and identify articles in all of the following databases:  PubMed, PLoS, WHO publications, UN publications, news, and other peer-reviewed publication databases.  Search terms used include the following:  Saudi Arabia, HIV, “HIV knowledge”, questionnaire, instrument, survey, AIDS. Articles were assess manually for meeting criteria, and articles that met the study criteria selected for data collection, which consisted of the following: primary author, year of publication, population studied (in terms of gender as well as other characteristics), total sample completing the instrument, the response rate (if reported), how the population was sampled, description of the questionnaire used, language of the questionnaire, and results of any reliability or validity studies reported.

Results:  The initial search for articles identified 4,410 results. Most studies did not qualify because they did not concern the population of interest. A smaller proportion of studies were disqualified because they did not include a survey, or they did not study HIV or AIDS knowledge. Ultimately, 16 articles met the criteria for this study.

Of these, 10 (63%) were conducted in a KSA population, and the rest were in neighboring countries (Kuwait n=2, Yemen n=2, UAE n=1, and Oman n=1). A majority of the articles studied students of primary, secondary, or post-secondary schools (n=9, 56%), while others studied certain occupational cohorts, such as physicians and bus drivers (n=3) and others studied the general population (n=3). Five studies included only men, while the other 11 included both sexes. Only one study was done on a clinical population. A response rate was either unavailable or not reported for over half the studies (n=9, 56%), but all studies that did report a response rate reported rates of  least 80%, and three studies even achieved 100%.  A variety of sampling approaches were used, the most common being some type of cluster sampling (n=6, 38%). Sampling approaches were unavailable or not reported for three studies.

Information about the measurement instruments used in these studies was unavailable or not reported for four studies (25%), and the language of the questionnaire was unavailable or not reported for seven studies (44%). Six studies (38%) used a measurement instrument based on a questionnaire developed by a health agency, such as the World Health Organization (WHO) Knowledge, Attitudes, Beliefs, and Practices (KABP) survey. Two employed a questionnaire previously used in non-Arabic populations, and two developed their own questionnaire for the study. Eight studies (50%) reported administering the questionnaire in Arabic only. One study translated it into multiple languages to accommodate its multi-lingual population. Information about the language of the questionnaire was unavailable or not reported for seven studies (44%). 

Of the 16 studies, only one commented on validity studies, and two commented on reliabilities studies. However, descriptions of reliability and validity studies as reported in the articles were not clear, in that they did not provide a description of how reliability or validity were assessed or evaluated in the study or previous studies.

Discussion:  This review found 16 HIV/AIDS knowledge questionnaires that could inform a current study of the KSA population. The most important weakness identified in the studies reviewed was the appropriateness of the population subject to the HIV or AIDS knowledge questionnaire. Given the culture of the KSA as described earlier, where most young people do not experience sexual intercourse until after an arranged, legal marriage in accordance with the Saudi expression of the Islamic religion, it is unclear as to why over half the studies reviewed (n=9, 56%) surveyed students, as this would likely constitute a low risk population for sexual transmission of HIV.

Next, it is interesting to note that despite an avoidance of gender-mixing in the Arab world, an instrument targeting women or men only with respect to HIV and AIDS prevention knowledge was not developed during this time span. In fact, 5 of the 16 studies reviewed did not even include women, and no attempt was made to create gender-specific instruments when surveying men or mixed groups.

In addition to concerns over the appropriateness of particular surveyed populations, criticism can be leveraged regarding the choice of instruments used in these studies.  It is reasonable that six studies selected questionnaires developed by health agencies, but two studies used questionnaires based on non-Arabic populations, two used questionnaires that the authors developed for the study (and therefore, they had not been tested or reported on previously), and information about the questionnaire used was not available for 4 studies.  This means that in 50% of the studies reviewed, the questionnaire was likely of low quality.

None of the instruments used were provided the results of reliability or validity studies, so evaluating their quality is difficult. Unfortunately, most of the HIV and AIDS knowledge instruments developed by health agencies, such as the WHO-KABP, are inappropriate for even high-risk populations in the KSA. The WHO-KABP was used in four studies and includes questions about HIV and AIDS pertinent to transmission by intravenous drug use (IDU) and homosexual activity. These questions are inappropriate for the KSA’s population, given its low HIV prevalence, low rates of IDU and homosexuality, and KSA’s strict expression of Islam.

Conclusion:  In summary, 16 papers were reviewed that included HIV knowledge measurement in the KSA or bordering populations. On the positive side, all studies report high response rates, but other features of these reports indicate challenges. First, the absence of a culturally-specific KSA instrument has resulted in the inability to reliably and accurately measure HIV knowledge in KSA populations.  Next, studies examined employed a variety of instruments, but they generally did not report reliability and validity studies, so their relative quality could not be compared. Further, gender-specific instruments were not developed, although transmission patterns in the KSA suggest that transmission modes are potentially much more gender-specific than in other countries. Finally, these articles generally reported on low-risk populations in the KSA. The KSA’s public health goals should more specifically focus on measuring and improving knowledge in high-risk populations such as married women—an option currently limited by commonly available measurement instruments.