Demographic Characteristics of Participants
In drawing participants, care was taken to select individuals engaged in occupations that are
highly vulnerable to HIV and AIDS. These included, the motor bike riders, bar attendants,
CSWs, long distant drivers and those for public service vehicles. Primarily, the study targeted
individuals aged above 18 years. This age cut-off requirement was necessary to avoid seeking
consent from participants’ parents or guardians. Plus, the study recruited participants from a
group known to be disproportionately affected by HIV as key informants. The group chosen
was the CSWs because sex is the main pathway through which HIV is transmitted. Lastly,
the study also interviewed HIV experts—medical professionals in charge of the HIV
programme or treating HIV patients at the KTRH.
In total, 407 participants were recruited: 385 from the general population, responded to a
questionnaire, 20 from among CSWs, were engaged in a FGD, and lastly, 2 experts who
participated in an interview to provide insights into the HIV situation in the County.
Despite all participants returning the completed questionnaires, 19 were found to have
been completed unsatisfactorly as they failed to satisfy the inclusion criteria. Effectively
then, 388 out of 407 or 95.3% of the recruitees were considered as participants in the
study. Table 4.1 shows a summary of the demographic characteristics of the participants.
Table Participants Demographic Characteristics
Type of Participant Characteristic Number Total %
Type of Participants
From the general population 385 407 94.6
Commercial Sex Workers 20 407 4.9
HIV Experts 2 407 0.5
Gender
Male 213 387 55.0
Female 174 387 45.0
Participants’ Ages
Below 18 Years 4 368 1.1
Between 18 – 24 172 368 46.7
Between 25 – 49 176 368 47.8
Over 49 Years 16 368 4.3
Educational Level
College level and Above 214 386 55.4
Secondary 145 386 37.6
Primary 27 386 7.0
Level of Income Earnings Per Month
Less than Kshs. 24,000 226 342 66.0
Between Kshs. 24,000 to 99,000 97 342 28.4
Over Kshs. 99,000 14 342 5.6
Marital Status
Never Married (Singles) 196 386 50.8
In a marital relationship (Married) 166 386 43.0
Separated or Divorced 24 386 6.2
Table shows that out of 407 participants recruited, 388 completed the questionnaire
satisfactorily. However, the number that provided their demographic characteristics
ranged between 342(88.1%) and 387(99.7%). Further, Table 4.1 shows that although the
study targeted individuals who were above 18 years old, four were found to be underage.
These minors belonged to the CSWs category. Despite the threatening risks faced by this
category, the young girls have probably been driven to continue to trade sex, as a result
of coming from poverty stricken settings.
In addition, Table shows that more females 213(55%) compared to 174(45%) males
participated in the study. This compares favourably with the target population consisting
of 42% males against 58% females. Also, majority (93.2%) of the participants surveyed
were educated beyond secondary school level. However, the majority of the CSWs were
secondary school dropouts. For this study, educational attainment is important as it is the
foundation of a participant understanding and appreciating the risks of infection and the
benefits of testing for HIV.
Table 4.1 shows that the item answered by the least number of participants (342) touched
on participants' income. This may have been so; perhaps for two reasons: due to the
sensitivity of the information or lack of conceptualization of the term especially for those
not in formal employment. Participants employed in the informal sectors, find the
concept of income rather difficult to conceptualize since they do not earn a constant
amount for their daily wages. However, the results indicate that majority of those
surveyed earn less than ten dollars per day given the exchange rate. Earnings are
important considering the amount required for drugs as well as food to fight the virus.
In sum, the rate of return of the questionnaires for this survey was impressive considering
that surveys that rely on questionnaires have a return rate of as low as 20% (Wilkinson &
Birmingham, 2003). To have achieved these results, the study adopted a strategy of
collecting questionnaires immediately after participants completed responding to them. In
that way, participants lacked the opportunity of placing the questionnaire aside and
perhaps misplacing it, forgetting it or lacking time to complete it altogether.
Lastly, marital status is an important demographic characteristic that has the potential to
either aggravate or alleviate the HIV situation. Figure 4.1 depicts the status of the
participants’ marital status.
Separated/
Divorced, 24
single, 196
Married, 166
Figure Participants’ Marital Status
Figure 4.1, shows that majority of the study participants surveyed were single (never
married), followed by the married—currently in a marital relationship. However, the third
and last group were those who either have separated or divorced. This type of
categorisation is useful considering that HIV has been mis-characterised as a singles
problem leading to marital unfaithfulness; and thus, causing a resurgence of the disease
incidence.
Key risk Factors Influencing Increased HIV Prevalence
The first question that this study sought to answer was: ‘which risk factors play a key role in
influencing the HIV prevalence in Kisii central sub-county?' This question is pertinent since
factors play differing roles in influencing the HIV prevalence; some are more potent than
others. Also, there is a higher likelihood of avoiding factors known to exert more
influence on the HIV prevalence than those with less influence.
The answer to this question lies in establishing risk factors playing a key role in influencing
increased prevalence. One way of approaching this issue is to make all risk factors to be
study variables to be investigated. But, from an analytical point of view, such an approach
is likely to complicate matters, since each variable has a unique behaviour depending on
the context. Consequently, this study routed for an approach rating the perceived
contribution of commonly known risk factors influencing the HIV prevalence.
Despite there being an abundance of factors and circumstances responsible for increased
HIV prevalence, thirteen seem to be the most commonly cited in literature. These are:
practising outdated cultural values; mother-to-child transmission; increased homosexual
behaviour; high migration of PLWHA; non-observance of religious teachings; lack of HIV
policies; infected persons living longer; intergenerational sex; high unemployment rates;
negative attitudes towards HIV testing; increased alcohol use and substance abuse;
increased numbers of commercial sex workers and high poverty levels.
Participants were asked to rate the extent to which each of the 13 factors is perceived to
influence the prevalence of HIV using a five-point scale; where one represents the
highest influence and five the least influence. However, in scoring the responses for
analysis, factors with high perceived contribution were scored five while factor with the
least perceived contribution was scored one. The scores were entered into the SPSS
computer programme and each factor's contribution to the total HIV prevalence
determined by performing a factor analysis using the principal component Analysis
method for extraction since the principal axis factoring method was not appropriate.
Table 4.2 displays the results of the analysis.
Table
Factor components, Eigen Value and HIV prevalence
Component Eigen Values % Prevalence Cumulative %
1 2.870 22.080 22.080
2 2.009 15.451 37.531
3 1.088 8.370 45.901
4 1.070 8.228 54.129
5 .985 7.577 61.707
6 .882 6.783 68.490
7 .778 5.988 74.477
8 .750 5.772 80.250
9 .649 4.995 85.245
10 .585 4.502 89.747
11 .548 4.218 93.965
12 .428 3.294 97.259
13 .356 2.741 100.000
Table 4.2 shows that based on the pattern matrix, 8 variables present the best chance of
explaining over eighty percent (80.3%) of the cumulative HIV prevalence. However, the
table shows that over fifty percent (54.1%) of the HIV prevalence can be predicted by four
factors. This means that the other four factors can be used to explain slightly less than 30% of
the HIV prevalence. An alternative way of viewing the same information is to plot it on
a scree-plot, assigning ‘components’ on the X-axis and Eigen Values on the Y-axis. Figure 4.2 is a
scree plot showing the contribution by the variables.
Figure shows that the most prominent change in slope occurs after component number three,
which is the “elbow” of the scree plot. Thus, based on the screen plot, it can be argued that
the first three variables are key to explaining the factors influencing increased HIV
prevalence. These account for approximately fifty percent (45.9%) of the total prevalence.
However, it does not look logical to regard the first three factors alone because one is
disregarding more than half (54.1%), though contributed in small volumes, of what is being
included. Consequently, to fully account for key variables that can be used to explain the
HIV prevalence, the study needed to include more factors. Table lists in order of importance
the perceived influence each factor contributes towards the HIV prevalence.
Table Ranking order of variables explaining cumulative prevalence
Component
1 2 3 4
High poverty levels .727 -.042 .377 .044
High unemployment .725 -.061 .450 -.039
Increased Alcohol and Substance use .671 -.242 -.197 -.324
Increased uptake of CS work .641 -.285 -.361 -.039
Poor HIV testing attitudes .588 .215 -.280 .202
Intergenerational sex .535 .288 -.059 .084
Increased homosexual behaviour -.207 .664 .048 -.205
High migration of HIV+ individuals .075 .601 .000 -.178
Lack of HIV Policies .312 .535 -.129 .036
Non-observance of religious teachings .082 -.460 .435 .357
Practising outdated cultural values .288 .437 .334 -.261
Increased MTC HIV Transmission -.214 .394 .338 .319
PLWHIV living longer .203 .298 -.241 .735
A four component extraction using the Principal Component Analysis.
Table shows the order of variables in explaining cumulative prevalence. Top on the list is
high poverty levels; followed closely with high unemployment and increased alcohol and
substance abuse. In the middle lies increased uptake of commercial sex work, poor HIV
testing attitudes, intergenerational sex, and increased homosexual behaviours. The factors
that lie at the bottom are high migration of HIV infected individuals, lack of HIV
policies, non-observance of religious teachings, practising outdated cultural values,
Increased Mother to Child HIV transmissions, and PLWHIV living longer.
Table 4.3 shows that out of the 13 factors investigated, three were perceived to be playing a
leading role in influencing HIV prevalence. These were, poverty – measured using levels of
income earned while HIV status was determined indirectly on the basis of whether
participants belonged to a social support group or not. This is from the general
knowledge that usually those who test positive for HIV have a more compelling reason to
join a social support group while those not, often do not joint a support group. However,
the weakness with this approach is that some participants may test positive and abstain
from joining a support group, especially when living in denial.
Besides poverty being the leading factor influencing HIV prevalence; it was also found to
be significantly associated with age, gender, marital status, educational attainment, and
testing for HIV. This implies that poverty and HIV/AIDS have a more complex
relationship than probably contemplated. In certain instances, wealth rather than poverty
may be the cause for contracting HIV. This study found that 75 or 22% of the study
participants could be considered HIV positive while the rest, HIV negative. Table 4.4
shows a cross tabulation between HIV testers and Participant’s Income.
