%)4
803,*/(
1"1&34

%FTJSF
GPS
$IJMESFO
BOE
6ONFU
/FFE
GPS
$POUSBDFQUJPO
BNPOH
)*71PTJUJWF
8PNFO
JO
-FTPUIP

5JN
"EBJS


/P


$%-/'2!0()#
!.$
.BSDI


5IJT
EPDVNFOU
XBT
QSPEVDFE
GPS
SFWJFX
CZ
UIF
6OJUFE
4UBUFT
"HFODZ
GPS

(%!,4(
*OUFSOBUJPOBM
%FWFMPQNFOU
2%3%!2#(
 Desire for Children and Unmet Need for Contraception
among HIV-Positive Women in Lesotho

Tim Adair
Macro International Inc.

March 2007

Corresponding author: Tim Adair, Macro International Inc., 11785 Beltsville Drive, Suite 300, Calverton,
MD, USA 20705. Phone (301) 572 0448, Fax: (301) 572 0994, Email: tim.adair@ orcmacro.com
Abstract
In Lesotho, where the HIV prevalence rate for women is 26.4%, there is high risk for
mother-to-child transmission (MTCT) of HIV. Efforts to prevent MTCT can focus on
reducing the fertility level of HIV-positive women. This paper examines the desire for
children and unmet need for contraception to limit or space births among HIV-positive
women age 15-49 years, using data from the 2004 Lesotho Demographic and Health
Survey. Multivariate analysis of HIV-positive women, 83% of whom have never learned
their HIV status, shows that the desire for children in the future does not differ by
socioeconomic status. Unmet need for contraception is highest among women in the
poorest households. Although these women have lower HIV prevalence, they have higher
potential for MTCT. The multivariate results are similar for both HIV-positive and HIV-
negative women because of low self-awareness of HIV status. Efforts to reduce the level
of MTCT require improved access to family planning services for all women, especially
the poorest, and an increase in HIV testing and counseling.
Background
The southern African country of Lesotho has one of the highest national HIV prevalence
rates in the world. According to the 2004 Lesotho Demographic and Health Survey
(LDHS), 26.4% of women age 15-49 years are HIV positive (MOHSW et al., 2005). A
Total Fertility Rate (TFR) of 3.5 births per woman in Lesotho has caused mother-to-child
transmission (MTCT) of HIV to be of major concern (MOHSW et al., 2005). It is
estimated that the current under-five mortality rate in Lesotho of 123 deaths per 1,000
births would be 71 if there was no AIDS (Population Reference Bureau, 2006). HIV
testing of pregnant women and the distribution of antiretroviral drugs during delivery and
following birth are two primary approaches currently used to prevent MTCT (PMTCT).
Efforts to decrease the level of MTCT can also focus on reducing fertility among HIV-
positive women. Two causes of the level of fertility are the desire for children and unmet
need for contraception to limit or space births. Therefore, this paper utilizes the 2004
LDHS to analyze the factors associated with these two causes among HIV-positive
women, to help identify strategies to reduce their fertility levels and the extent of MTCT
in Lesotho.

Worldwide, 1,800 children under the age of 15 become infected with HIV each day, the
vast majority of whom are in Africa (UNAIDS, 2006). The risk of mother-to-child
transmission occurs during pregnancy, delivery, and breastfeeding. During pregnancy and
labor the risk of transmission is 15% to 30%. Breastfeeding through 18 to 24 months
increases the overall risk to 30% to 45% (De Cock et al., 2000). There are a number of
PMTCT services. HIV testing of pregnant women can identify women who require
antiretroviral therapy and counseling. 1 Antiretroviral therapy for the mother and child
consists of one dose of the drug nevirapine to the mother at delivery and one dose to the
child soon after birth. Counseling on risk reduction of transmission through breastfeeding
may also reduce the risk of mother-to-child transmission 2 ; if there are no appropriate
substitute feeds, however, this method may not be feasible. Despite the risk of MTCT in

1
Promoting voluntary HIV testing and counseling of women that are not pregnant can contribute to
preventing HIV-positive women from becoming pregnant.
2
A report on a study in Malawi indicates that the reduction of mortality among children of HIV-positive
women due to breastfeeding more than offsets the risks of mother-to-child transmission (Taha et al., 2006).

1
Lesotho, only 14% of pregnant women in the region were offered PMTCT services in
2005 (USAID et al., 2004).

The reduction of unintended pregnancies is a particularly important method of PMTCT,

especially in high-prevalence countries like Lesotho. A study found that moderate falls in
unintended pregnancies, ranging from 5.6% to 34.8% by country, can result in the same
number of prevented HIV infections as current use of nevirapine (Sweat et al., 2001).
Furthermore, other analysis found that HIV-positive births can be reduced more by an
increase in contraceptive use at the same level of expenditure as provision of nevirapine
(Reynolds et al., 2005). Contraceptive use can also help HIV-positive women space
births, which can help their health as well as lower the need for PMTCT services
(Rutenberg et al., 2003). Increases in contraceptive use by HIV-positive women can
occur through integration of family planning and HIV services; however, at the time of
the 2004 LDHS, Lesotho’s national HIV policy did not make any mention of family
planning (Strachan et al., 2004). Throughout sub-Saharan Africa, the prevention of
unintended pregnancies has been described as “an undervalued and little-used strategy”
for PMTCT (Reynolds and Wilcher, 2006: 8).

Efforts to reduce MTCT are hampered by the fact that only a small proportion of HIV-
positive women know they are infected. Data from the 2004 LDHS in Table 1 show that
over four-fifths of HIV-positive women had not had prior testing for HIV and learned
their status. 3 Given that the data do not indicate whether the woman was infected at the
time of the test—there is no accurate measure for this in the LDHS—those women who
were tested and learned their status in the past 12 months are more likely to have been
HIV positive at the time of the test than those tested prior to the past 12 months. The
former group comprises only 8.7% of HIV-positive women in Lesotho.

3
Prior HIV test excludes the test conducted as part of the 2004 LDHS. Respondents did not find out the
result of their HIV test in the 2004 LDHS.

2
 Table 1 Percent distribution of HIV-positive women age 15-49 by HIV testing
status, when tested, and whether they learned the results of the test, Lesotho 2004
Prior testing status/when tested/whether learned results % N
Tested in past 12 months and learned status 8.7 64
Tested prior to past 12 months and learned status 8.6 63
Never tested, or tested but did not learn status 82.7 606
Total 100.0 733
Note : Weighted cases. Only women who declared their fertility intentions are included (i.e., does
not include sterilized or infecund women).

For women who know they are HIV positive, the literature suggests various scenarios
regarding their fertility intentions. Some studies have found that even if an HIV-positive
woman is told of her status and is counseled on the risks of MTCT, pregnancy levels
remain high (Heyward et al., 1993; Nebie et al., 2001; Temmerman et al., 1990).
Qualitative research in Côte d’Ivoire shows a strong desire for future childbearing among
women who know they are HIV-positive (Aka-Dago-Akribi et al., 1999). In high-fertility
societies, some women may wish to have a child to conceal their HIV status and avert
suspicion that they are infected. Having children may also provide a sense of normalcy to
family life and an affirmation of health (Rutenberg et al., 2000). In contrast, research in
Kenya shows that HIV-positive women aware of their status are less likely to want to
have a child in the future than HIV-negative women (Reynolds and Wilcher, 2006).
Furthermore, in Uganda a study found that only 7% of HIV-positive women who know
their status want to have a child (Nakayiwa et al., 2006). Overall, the fertility level of
HIV-positive women is generally lower than that of HIV-negative women because they
are more likely to be widowed, divorced, or co-infected with a sexually transmitted
infection (Lewis et al., 2004; Terceira et al., 2003).

If an HIV-positive woman does not want to have a child in the future or if she wants to
space her births, unmet need for contraception may still put her at risk of pregnancy.
While in some studies knowledge of HIV status among infected women resulted in an
increase in contraceptive use, other studies found a lack of persistent use of contraception
beyond one year or no significant difference compared with HIV-negative women (Allen
et al., 1992; Allen et al., 1993; Kamenga et al., 1991; Nebie et al., 2001; Rutenberg and
Baek, 2005). In Uganda, a study found that 73% of women exhibiting behavior that put

3
them at risk of pregnancy did not want any more children (Nakayiwa et al., 2006). In
Lesotho, only 35% of currently married women use a modern contraceptive method,
despite increases in contraceptive use in the late 1990s (MOHSW et al., 2005; Tuoane et
al., 2004).

In sub-Saharan Africa, contraceptive prevalence is more than five times higher among
women in the highest wealth quintile compared with those in the lowest wealth quintile, a
far larger differential than in any other region of the world (UNFPA, 2002). Cost and
accessibility have been identified as barriers to use of family planning services for poor,
rural women (Tuoane et al., 2004). Although the level of HIV prevalence is lower among
women in the poorest wealth quintile (19.6%) than those in the other quintiles, the risk of
MTCT among poor women is of concern because they have a higher level of unmet need
for contraception (MOHSW et al., 2005). Other research in sub-Saharan Africa has found
that use of contraception increases if a woman has previously discussed contraception,
been exposed to mass media about family planning, or approves of family planning
(Gupta et al., 2003; Kayembe et al., 2006; Tawiah, 1997).

Data and Methodology

The desire for children and unmet need for contraception are examined using data from
the 2004 LDHS, a nationally-representative survey with information on both HIV and
family planning. A total of 8,592 households were surveyed, including 7,522 women age
15-49. The survey had a two-stage sample design with 405 clusters in the first stage and
systematic selection of households in the second stage. All women who stayed in the
household the night before the survey were eligible to be interviewed, and HIV testing
was undertaken for all eligible women in every second household in the original sample.
Respondents voluntarily provided blood samples for HIV testing after being informed
about the procedures, confidentiality, and the availability of voluntary counseling and
testing (VCT) services. Three to five drops of blood were collected from a finger on a
filter paper card; the filter paper was dried overnight and taken for laboratory testing.

4
Nineteen percent of the 3,758 eligible women were not tested 4 (12% of eligible women
refused to be tested, 22% of those in urban areas and 8% of those in rural areas). The HIV
data were anonymously linked to the sociodemographic data collected in other
questionnaires, after information that could potentially identify an individual was
destroyed.

Two outcome variables are used in this analysis: the desire for children and unmet need
for contraception. The desire for children is measured by whether a woman would like to
have a child in the future. Women that have not had sex are included in the construction
of the variable; those that cannot get pregnant (i.e., sterilized or infecund) are not
included; and those that are undecided are coded as not wanting a child. No variable in
the LDHS measures the number of children desired in the future.5 Unmet need for
contraception comprises women who have an unmet need for limiting or spacing births.
The unmet need variable is related to fertility preferences: broadly defined, unmet need
for limiting births refers to women who do not want a child in future and are not using a
method of family planning, and unmet need for spacing births refers to women who want
a child in future but not within two years and are not using a method of family
planning. 6, 7 , 8 A more detailed definition can be found in MOHSW et al. (2005: 103).

There is no accurate measure in the LDHS of whether an HIV-positive woman is aware

of her status. Instead, a proxy variable—whether a woman has ever been tested, when,
and if she learned her status—is used (Table 1). Women who were last tested in the past
12 months and learned their status are assumed likely to be aware they are infected,
although they may not be certain. This variable will be used in the analyses in the results
section and may show if women who are most likely to be aware they are HIV positive

4
Analysis of nonresponse for the 2004 LDHS is in Mishra et al. (2007).
5
There is a variable that measures ideal family size; in some cases this figure is less than the number of
liivng children.
6
Women desiring a birth within the next two years, not having sex, or never had sex are not included in the
analysis.
7
An attempt was made to create separate variables for unmet need to limit births and unmet need to space
births. However, the number of cases for unmet need to space births for HIV-positive women is too small
to allow a detailed analysis.
8
There is a variable measuring whether a woman wanted her last birth.

5
have different fertility preferences and different levels of unmet need for contraception
from other women. Although this variable has limitations, and any conclusions will be
limited, it is the best available from the DHS data to measure self-awareness of HIV-
positive status.

The results section initially presents tables on the desire for children and unmet need for
contraception among HIV-positive women based on their knowledge of their status, as
well as among HIV-negative women. The multivariate analyses of desire for children and
unmet need for contraception are conducted on HIV-positive women and, for
comparison, HIV-negative women, to determine if there are any differences in behavior
by infection status. Given the low level of testing among HIV-positive women, even if
awareness of being infected influences a woman’s behavior, the results are expected to be
similar.

The explanatory variables for both multivariate analyses include the socioeconomic and
demographic measures of residence, marital status, household wealth quintile, education,
age, and religion. Another variable measures the woman’s knowledge that MTCT can
occur during pregnancy, delivery, and breastfeeding. 9 For analysis of desire for children,
a variable measuring the number of children still living is also included. The unmet need
models include a variable measuring whether the respondent accessed family planning
messages in at least one type of media in the past month (i.e., newspaper, radio, and
television). Included in these models is a variable measuring whether the respondent
approves of contraception and another variable indicating whether the respondent spoke
to family planning staff in the past 12 months. 10 The multivariate analyses are conducted
using logistic regression in Stata 8.1 (StataCorp, 2003). HIV weights are applied in the
regression and the standard errors are adjusted for the cluster design of the sample. The
odds ratio, which is the exponential of the coefficient in the model, is also presented.

9
An attempt was made to incorporate MTCT knowledge into the ever learned HIV status variable;
however, the category sizes were too small to allow adequate analysis.
10
The variable of whether the respondent spoke to family planning staff comprises women who were either
visited by a family planning officer or visited a health facility and spoke to about family planning.

6
Results
HIV-positive women in Lesotho who have learned their HIV status have a slightly lower
desire for children than those who have not learned their status. Table 2 shows that the
desire for children among HIV-positive women age 15-49 is 37.8% for those last tested
in the past 12 months and learned their status, 32.5% for those last tested prior to the past
12 months and learned their status, and 39.4% for those who have never learned their
status. However, these differences are not significant. There is no clear evidence of an
association between knowledge of a woman’s HIV status and desire for children.
Furthermore, a sizable proportion of infected women (38.7%) intend to have a child, thus
there is considerable potential for mother-to-child transmission in Lesotho. The
proportion of HIV-negative women who want to have a child is higher (44.4%).

Table 2 Percentage of women age 15-49 who want a child in the future, by HIV
status, when tested, and whether they learned the results, Lesotho 2004
Wants child
HIV status/when tested/whether learned results in the future N
Positive, last tested in past 12 months and learned status 37.8 64
Positive, last tested prior to the past 12 months and learned
32.5 63
status
Positive and never learned status 39.4 606
All positive women 38.7 733
All negative women 44.4 2,093
Total 42.7 2,826
Note: Weighted cases. A chi-square test was undertaken of desire for a child for all
HIV-positive women versus HIV-negative women and is significant (p<0.05). A chi-
square test was also undertaken of desire for a child by testing status of HIV-positive
women but is not significant (p<0.05).
Source: MOHSW et al., 2005

Table 3 shows that unmet need for contraception is lower among HIV-positive women
who were last tested in the past 12 months and learned their status (21.6%) than among
those who were last tested prior to the past 12 months and learned their status (29.2%)
and those who have never learned their status (32.5%). However, this difference is not
significant. HIV-negative women have significantly higher unmet need for

7
 Table 3 Percentage of women age 15-49 with an unmet need for contraception to limit or space
births, by HIV status, when tested, and whether they learned the results, Lesotho 2004
Unmet
HIV status/when tested/whether learned results Need (%) N
Positive, last tested in past 12 months and learned status 21.6 31
Positive, last tested prior to past 12 months and learned status 29.2 43
Positive and never learned status 32.5 331
All positive women 31.3 405
All negative women 44.3 1,042
Total 40.6 1,447
Note: Weighted cases. A chi-square test was undertaken of unmet need for contraception for all HIV-
positive women versus HIV-negative women and is significant (p<0.01). A chi-square test was also
undertaken of unmet need for contraception by testing status of HIV-positive women but is not
significant (p<0.05).
Source: MOHSW et al. (2005)

contraception (44.3%). Table 4 shows that among both HIV-positive and HIV-negative
women, injection and the pill are the most popular contraceptive methods. Only 6.3% of
HIV-positive women age 15-49 in Lesotho use a condom, thus contraceptive use is
contributing little to the prevention of transmission of the virus to a serodiscordant
partner.

Table 4 Percent distribution of women who are HIV positive and women who
are HIV negative by use of specific contraceptive methods, Lesotho 2004
HIV-positive women HIV-negative women
Current contraceptive
% N % N
method
Not using 64.0 510 73.9 1,641
Pill 10.9 87 5.8 130
IUD 0.8 6 1.3 29
Injections 14.3 114 9.5 212
Condom 6.2 49 5.6 123
Female sterilization 2.2 17 2.3 51
Other 1.7 13 1.5 34
Total 100.0 798 100.0 2,221
Note: Weighted cases.
Source: MOHSW et al. (2005)

Table 5 shows the univariate statistics for the variables included in the multivariate
analyses.

8
Table 5 Univariate statistics for variables included in multivariate analyses of HIV-positive women and HIV-
negative women age 15-49, Lesotho 2004
HIV-positive women HIV-negative women
Variable % N % N
Place of residence
Urban 30.3 232 22.0 461
Rural 69.7 533 78.0 1,633
Marital status
Never married 17.6 135 37.0 775
Currently married 54.4 417 53.5 1,121
Formerly married 27.9 214 9.5 198
Household wealth quintile
Lowest 10.9 84 16.1 336
Second 19.4 148 18.4 385
Middle 17.8 136 18.2 381
Fourth 22.0 168 20.9 439
Highest 29.9 229 26.4 553
Highest education level
None/primary incomplete 33.0 252 34.9 730
Primary complete 27.3 209 26.6 557
Secondary incomplete+ 39.7 304 38.5 807
Age
15-19 years 6.8 52 29.9 626
20-24 years 19.0 146 21.5 449
25-34 years 41.5 318 23.1 484
35+ years 32.7 250 25.5 534
Religion
Roman Catholic 42.4 323 45.0 940
Lesotho Evangelical 22.2 170 20.9 437
Other 35.4 270 34.1 713
Knowledge of MTCT
No 29.0 213 31.0 602
Yes 71.0 521 69.0 1,340
Ever learned HIV status
No 82.7 606
Yes, before last 12 months 8.6 63
Yes, within last 12 months 8.7 64
Children still living1
0 16.4 126 34.6 725
1 27.0 207 20.0 419
2+ 56.6 433 45.4 950
Heard of FP in the media2
No 61.7 259 64.7 674
Yes 38.3 160 35.3 368
Spoke to FP staff2
No 85.2 356 83.7 869
Yes 14.8 62 16.3 170
Approve of contraception2
No/don’t know 10.1 42 15.3 159
Yes 89.9 377 84.7 883
Total 100.0 766 100.0 2,094
Note: Weighted cases. For HIV positives, there are 3 missing cases for religion, 32 missing cases for knowledge of MTCT and 33
missing cases for ever learned HIV status. For HIV negatives, there are 4 missing cases for religion, 150 missing cases for
knowledge of MTCT, and 3 missing cases for visited by FP worker.
1
Includes current pregnancy.
2
Only used in analysis of unmet need, so only based on 419 cases for HIV positives and 1,042 cases for HIV negatives.
Source: MOHSW et al. (2005)

9
Table 6 presents the bivariate statistics for the two multivariate analyses undertaken.
There is no major difference in the bivariate relationship for each outcome variable
according to HIV status of the woman. A significantly higher proportion of both HIV-
positive and HIV-negative women who are younger or have fewer living children want a
child in future than other women. A higher proportion of married HIV-positive women
want a child in future than women who are never married or formerly married, while for
HIV-negative women those who are never married are most likely to want a child in the
future. Other variables, including knowledge of MTCT, household wealth quintile, and
education, have no significant bivariate relationship with the desire for children for HIV-
positive women. Unmet need for contraception to limit or space births has a bivariate
relationship with a number of variables. There is significantly greater unmet need among
HIV-positive women who live in rural areas, are currently married, live in poorer
households, have less education, are older, have heard of family planning in the media,
and approve of contraception. HIV-positive women with knowledge of MTCT have
higher unmet need, the opposite of the finding for HIV-negative women. Unmet need is
lower among women who have spoken to family planning staff, irrespective of HIV
status.

10
Table 6 Bivariate statistics for variables included in multivariate analyses of HIV-positive women
and HIV-negative women age 15-49, Lesotho 2004
Wants child in future Has unmet need for contraception
HIV HIV HIV HIV
Variable positive negative positive negative
Place of residence
Urban 38.5 42.9 19.9 27.1
Rural 38.2 44.8 * 36.8 ** 49.2 **
Marital status
Never married 36.9 51.3 15.5 16.7
Currently married 47.1 45.0 38.7 50.3
Formerly married 21.8 ** 13.9 ** 17.5 ** 23.9 **
Household wealth quintile
Lowest 31.8 47.1 56.1 68.4
Second 38.6 40.2 32.3 60.8
Middle 34.2 47.2 38.2 42.7
Fourth 43.8 44.0 30.9 40.8
Highest 38.8 44.0 21.8 ** 22.8 **
Highest education level
None/primary incomplete 41.5 41.4 40.1 54.0
Primary complete 35.8 43.7 35.9 53.6
Secondary incomplete+ 37.3 47.6 23.8 ** 29.6 **
Age
15-19 years 71.8 55.5 18.9 46.7
20-24 years 52.6 60.7 32.5 41.8
25-34 years 38.5 44.4 22.1 37.7
35+ years 22.7 ** 17.7 ** 45.2 ** 51.9
Religion
Roman Catholic 39.6 44.5 35.0 46.8
Lesotho Evangelical 32.8 42.1 22.5 41.0
Other 40.2 45.8 33.7 43.1
Knowledge of MTCT
No 36.6 49.8 21.1 48.2
Yes 39.6 42.3 ** 35.9 * 41.1 *
Ever learned HIV status
No 39.4 - 32.5 -
Yes, before last 12 months 32.5 29.2
Yes, within last 12 months 37.8 21.6
Children still living
0 76.8 63.4 - -
1 52.4 52.9
2+ 20.4 ** 26.2 **
Heard of FP in the media
No - - 33.9 49.2
Yes 27.5 * 35.5
Spoke to FP staff
No - - 32.4 45.3
Yes 26.3 38.2
Approve of contraception
No/don’t know - - 70.5 72.0
Yes 29.2 ** 39.4 **
Note: weighted cases.
*p<0.05 **p<0.01
Source: MOHSW et al. (2005)

11
The multivariate results in Table 7 show that factors associated with wanting to give birth
in the future have a similar relationship for HIV-positive and HIV-negative women in
Lesotho. Marital status and the number of children still living are the strongest
determinants of whether a woman in wants to give birth in future. A currently married
HIV-positive woman is almost 14 times more likely than a never-married woman to want
to have a child, controlling for other factors. Both currently married (odds ratio=17.5) and
formerly married (odds ratio=4.1) HIV-negative women are more likely to want a child
than never-married women. There is a strong inverse relationship between future
childbearing desires and the number of children still living for both HIV-positive and
HIV-negative women. HIV-positive women age 35 and over are significantly less likely
to want a child in future compared with those age 15-19. However, no relationship is
found for any other age group. For HIV-negative women age 20-34 there is an increased
desire for children compared with women in their teens.

12
 Table 7 Multivariate analysis of desire for children in the future, HIV-positive and
HIV-negative women age 15-49, Lesotho 2004
HIV-positive women HIV-negative women
Odds Odds
Variable ratio Z ratio Z
Place of residence
Urban Ref Ref
Rural 1.262 0.83 1.565 1.95
Marital status
Never married Ref Ref
Currently married 13.765 ** 5.46 17.460 ** 8.65
Formerly married 2.639 1.87 4.058 ** 3.35
Household wealth quintile
Lowest Ref Ref
Second 1.076 0.18 0.843 -0.86
Middle 1.225 0.51 0.929 -0.32
Fourth 1.420 0.88 0.852 -0.72
Highest 1.521 1.09 0.889 -0.48
Highest education level
None/primary incomplete Ref Ref
Primary complete 0.603 -1.77 1.035 0.22
Secondary incomplete+ 0.629 -1.82 1.032 0.18
Age
15-19 years Ref Ref
20-24 years 0.603 -0.99 1.918 ** 3.35
25-34 years 0.478 -1.46 1.643 * 2.16
35+ years 0.292 * -2.21 0.555 * -2.25
Religion
Roman Catholic Ref Ref
Lesotho Evangelical 0.585 -1.86 0.731 * -2.10
Other 0.930 -0.29 1.085 0.57
Children still living
0 Ref Ref
1 0.136 ** -4.63 0.066 ** -8.69
2+ 0.022 ** -7.38 0.016 ** -10.33
Knowledge of MTCT
No Ref Ref
Yes 1.865 * 2.39 0.951 -0.36
Ever learned HIV status
No Ref -
Yes, before last 12 months 1.067 0.17
Yes, within last 12 months 1.215 0.46
No. cases (unweighted) 715 1,924
*p<0.05 **p<0.01
Source: MOHSW et al. (2005)

For HIV-positive women, there is no relationship between the desire for children and if
and when they have learned their HIV status. Knowledge of MTCT is significantly
associated with increased likelihood of wanting to give birth in the future for HIV-
positive women, but there is no association for HIV-negative women. The desire for

13
children is not related to household wealth or education, irrespective of HIV status, nor is
it related to religion for HIV-positive women.

Table 8 presents the multivariate results for unmet need for contraception, which again
show little difference between HIV-positive and HIV-negative women. There is a strong
relationship between wealth quintile and unmet need among both HIV-positive and HIV-
negative women. A woman whose household is in the fourth quintile (HIV positive odds
ratio=0.30, HIV negative odds ratio=0.40) or in the highest quintile (HIV positive odds
ratio=0.22, HIV negative odds ratio=0.19) is much less likely to have an unmet need for
contraception compared with those in the lowest quintile. HIV-positive women in the
second quintile and HIV-negative women in the middle quintile are also significantly less
likely to have unmet need for contraception. Such a disadvantage for the poorest women
has been seen elsewhere (UNFPA, 2002). Currently married women have greater unmet
need for contraception (HIV positive odds ratio=3.28, HIV negative odds ratio=6.00)
than never-married women. HIV-positive women age 35 and above are far more likely to
have unmet need for contraception compared with teenage women, while HIV-negative
women age 20-34 have less unmet need.

14
 Table 8 Multivariate analysis of unmet need for contraception to limit or space
births, HIV-positive and HIV-negative women age 15-49, Lesotho 2004
HIV-positive women HIV-negative women
Odds Odds
Variable ratio Z ratio Z
Place of residence
Urban Ref Ref
Rural 1.405 0.90 0.929 -0.27
Marital status
Never married Ref Ref
Currently married 3.281 * 2.31 5.996 ** 4.85
Formerly married 0.812 -0.31 1.742 1.05
Household wealth quintile
Lowest Ref Ref
Second 0.155 ** -3.17 0.813 -0.83
Middle 0.439 -1.42 0.339 ** -3.89
Fourth 0.296 * -2.09 0.396 ** -3.06
Highest 0.215 * -2.44 0.189 ** -4.81
Highest education level
None/primary incomplete Ref Ref
Primary complete 1.124 0.30 1.336 1.42
Secondary incomplete+ 0.726 -0.86 0.681 -1.80
Age
15-19 years Ref Ref
20-24 years 2.671 1.35 0.432 ** -2.65
25-34 years 1.730 0.74 0.418 ** -2.61
35+ years 5.585 * 2.38 0.919 -0.25
Religion
Roman Catholic Ref Ref
Lesotho Evangelical 0.679 -0.98 1.019 0.07
Other 0.918 -0.26 0.859 -0.76
Knowledge of MTCT
No Ref Ref
Yes 1.696 1.56 0.663 * -2.18
Ever learned HIV status
No Ref -
Yes, before last 12 months 1.446 0.73
Yes, within last 12 months 1.650 -0.76
Heard of FP in the media
No Ref Ref
Yes 1.162 0.45 0.997 -0.02
Spoke to FP staff
No Ref Ref
Yes 0.662 -0.95 0.691 -1.60
Approve of contraception
No/don’t know Ref Ref
Yes 0.277 ** -2.82 0.302 ** -4.05
No. cases (unweighted) 394 975
*p<0.05 **p<0.01
Source: MOHSW et al. (2005)

Having ever learned HIV status and, if so, when, has no relationship with the desire for
children among HIV-positive women. Knowledge of MTCT is not significant for HIV-

15
positive women, but it decreases the likelihood of unmet need for HIV-negative women.
Unmet need has no association with having heard of family planning in the media or
having spoken to family planning staff; however, unmet need is lower if an HIV-positive
woman approves of contraception.

Discussion
In Lesotho, which has one of the highest HIV prevalence rates in the world and a total
fertility rate (TFR) of 3.5 births per woman, reducing the level of mother-to-child
transmission of HIV is of critical importance. Knowledge of the factors associated with
the desire for children and unmet need for contraception among HIV-positive women can
help identify strategies to reduce fertility levels and, hence, mother-to-child transmission.
The desire for children among HIV-positive women is highest among those who are
currently married, older women, and those who do not have any children. These factors
are likely to be associated with fertility preferences in most contexts. However, the desire
for children does not differ by socioeconomic status, except for religion, perhaps because
the variable does not measure how many children a woman wants. Those with knowledge
of mother-to-child transmission are more likely to want children in future, but it is
difficult to interpret this finding because most of these woman are not aware they are
HIV positive.

The variable that measures whether HIV-positive women are aware of their status has no
bivariate or multivariate relationship with desire for children. There is, therefore, no
conclusive evidence that HIV-positive women are changing their fertility desires based
on knowing their HIV status. However, interpretation of this finding needs to take into
consideration the limitations of this variable. The factors associated with the desire for
children differ only slightly between HIV-positive and HIV-negative women. This result
is not surprising because even if there is clear evidence that women who know they are
infected with HIV limit future births, they are too small a proportion of the population of
HIV-positive women to have an overall impact.

An important finding is that HIV-positive women in the lowest wealth quintile are most

16
likely to have an unmet need for contraception to limit or space births. Although the
poorest women in Lesotho are the least likely to be HIV positive, their low level of
contraceptive use (56.1% have an unmet need) and the importance of contraceptive use
for prevention of mother-to-child transmission means their children are at greater risk of
becoming infected. It is essential that access to family planning services is increased, to
enable all HIV-positive women to “make informed reproductive choices” (Reynolds and
Wilcher, 2006:8). Such improvements can occur through integration of family planning
and HIV services. This will enable women who test positive for HIV and want to limit or
space their births to have better access to contraception, and to reduce the level of
mother-to-child transmission. Integration of HIV services, including voluntary counseling
and testing, into family planning services will increase the proportion of HIV-positive
women who are aware of their status and will educate them about the benefits of
contraceptive use as a means of preventing mother-to-child transmission.

Other findings from the analysis of unmet need for contraception show that HIV-positive
women have less unmet need for contraception (to limit or space births) compared with
HIV-negative women, but they are less likely to have learned their HIV status than HIV-
negative women. In the multivariate analysis, “ever learning HIV status” is not related to
unmet need among HIV-positive women. Again, data limitations prevent major
conclusions based on this variable. The lower level of unmet need among HIV-positive
women compared with HIV-negative women may be explained by their higher wealth
status. Of the other factors, the strongest relationship for HIV-positive women is for
married status; currently married women are at greater risk of unmet need for
contraception than never-married women. This result suggests that never-married women
and their partners are more likely to use contraception because they are not in a formal
union and want to guard against pregnancy.

An HIV-positive woman’s approval of contraception is found to decrease her unmet

need; this suggests efforts to promote the understanding of the benefits of contraception
will help women who wish to limit or space births. However, learning of family planning
in the media or speaking to family planning staff in the past 12 months is not related to

17
unmet need. This finding indicates that messages describing the benefits of family
planning need to be understood by the intended recipients.

In general, the predictive power of the factors associated with unmet need for
contraception is similar for HIV-positive and HIV-negative women, as in the case of
desire for children. Even if awareness of being infected did alter women’s behavior, these
women comprise too small a proportion of the population of HIV-positive women to
substantially affect the overall results of those infected. An increased level of testing,
with a focus on counseling on mother-to-child transmission, is needed to increase the
proportion of HIV-positive women who are aware of their status.

With more than one-quarter of women age 15-49 in Lesotho HIV positive and a
substantial number of excess deaths among children under five resulting from AIDS, it is
crucial that the level of mother-to-child transmission is reduced. In addition to increasing
HIV testing and counseling, strengthening family planning services for HIV-positive
women is needed to reduce the level of unmet need for contraception and the level of
unwanted fertility among these women.

18
REFERENCES

Aka-Dago-Akribi, H., A. Desgrees Du Lou, P. Msellati, R. Dossou, and C. Welffens-Ekra. 1999.

Issues surrounding reproductive choice for women living with HIV in Abidjan, Côte
d’Ivoire. Reproductive Health Matters 7(13): 20-29.

Allen, S., J. Tice, P. Van De Perre, A. Serufilira, E. Hudes, F. Nsengumuremyi, J. Bogaerts, C.

Lindan, and S. Hulley. 1992. Effect of serotesting with counseling on condom use and
seroconversion among HIV discordant couples in Africa. British Medical Journal
304(6842): 1605-1609.

Allen, S., A. Sreufilira, V. Gruber, S. Kegeles, P. Van De Perre, M. Carael, and T.J. Coates. 1993.
Pregnancy and contraceptive use among urban Rwandan women after HIV testing and
counseling. American Journal of Public Health 83(5): 705-710.

De Cock, K.M., M.G. Fowler, E. Mercier, I. De Vincenzi, J. Saba, R. Hoff, D.J. Alnwick, M.
Rogers, and N. Shaffer. 2000. Prevention of mother-to-child HIV transmission in resource-
poor countries: Translating research into policy and practice. Journal of the American
Medical Association 283(9): 1175-1182.

Gupta, N., C. Katenda, and R. Bessinger. 2003. Associations of mass media exposure with family
planning attitudes and practices in Uganda. Studies in Family Planning 34(1): 19-31.

Heyward, W.L., V.L. Batter, M. Malulu, N. Mbuyi, L. Mbu, M.E. St Louis, M. Kamenga, and
R.W. Ryder. 1993. Impact of HIV counseling and testing among childbearing women in
Kinshasa, Zaire. AIDS 7(12): 1633-1637.

Kamenga, M., R.W. Ryder, M. Jingu, N. Mbuyi, L. Mbu, F. Behets, C. Brown, and W.L.
Heyward. 1991. Evidence of marked sexual behavior change associated with low HIV-1
seroconversion in 149 married couples with discordant HIV-1 serostatus: Experience at an
HIV counseling center in Zaire. AIDS 5(1): 61-67.

Kayembe, P.K., A.B. Fatuma, M.A. Mapatano, and T. Mambu. 2006. Prevalence and
determinants of the use of modern contraceptive methods in Kinshsa, Democratic Republic
of Congo. Contraception 74(5): 400-406.

Lewis, J.J., C. Ronsmans, A. Ezeh, and S. Gregson. 2004. The population impact of HIV on
fertility in sub-Saharan Africa. AIDS 18(2 Suppl.): S35-S43.

Ministry of Health and Social Welfare (MOHSW) [LESOTHO], Bureau of Statistics (BOS)
[LESOTHO] and ORC Macro. 2005. Lesotho Demographic and Health Survey 2004.
Calverton, Maryland, USA: MOHSW, BOS and ORC Macro.

Nakayiwa, S., B. Abang, L. Packel, J. Lifshay, D.W. Purcell, R. King, E. Ezati, J. Mermin, A.
Coutinho, and R. Bunnell. 2006. Desire for children and pregnancy risk behavior among
HIV-infected mean and women in Uganda. AIDS and Behavior 10(4 Suppl.): 95-104.

Nebie, Y., N. Meda, V. Leroy, L. Mandelbrot, S. Yaro, I. Sombie, M. Cartoux, S. Tiendrebeogo,

B. Dao, A. Ouangre, B. Nacro, P. Fao, O. Ky-Zerbo, P. Van De Pierre, and F. Dabis. 2001.

19
 Sexual and reproductive life of women informed of their HIV seropositivity: A prospective
cohort study in Burkina Faso. Journal of Acquired Immune Deficiency Syndrome 28(4):
367-372.

Population Reference Bureau (PRB). 2006. How HIV and AIDS affect populations. BRIDGE.
Washington, D.C.: Population Reference Bureau.

Reynolds, H.W., M.J. Steiner, and W. Cates, Jr. 2005. Contraception’s proved potential to fight
HIV. Sexually Transmitted Infections 81(2): 184-185.

Reynolds, H.W., and R. Wilcher. 2006. Best kept secret in PMTCT: Contraception to avert
unintended pregnancies. AIDSLink (97): 8-9.

Rutenberg, N., and C. Baek. 2005. Field experiences integrating family planning into programs to
prevent mother-to-child transmission of HIV. Studies in Family Planning 36(3): 235-245.

Rutenberg, N., A.E. Biddlecom, and F.A.D. Kanoa. 2000. Reproductive decision-making in the
context of HIV and AIDS: A qualitative study in Ndola, Zambia. International Family
Planning Perspectives 26(3): 124-130.

Rutenberg, N., M. Siwale, C. Kankasa, R. Mduati, D.M. Ngacha, J. Oyieke, S. Geibel, and S.
Kalibala. 2003. Family planning and PMTCT services: Examining interrelationships,
strengthening linkages. Horizons Research Summary. Washington, D.C.: Population
Council.

Stata Corp. 2003. Stata 8.1. College Station, Texas, USA: Stata Corp.

Strachan, M., A. Kwateng-Addo, K. Hardee, S. Subramaniam, N. Judice, and K. Agarwal. 2004.

An analysis of family planning content in HIV/AIDS, VCT, and PMTCT policies in 16
countries. Policy Working Paper Series No. 9. Washington, D.C.: POLICY.

Sweat, M.D., K.R. O’Reilly, G.P. Schmid, J. Denison, and I. De Zoysa. 2004. Cost-effectiveness
of nevirapine to prevent mother-to-child HIV transmission in eight African countries. AIDS
18(12): 1661-1671.

Taha, T.E., N.I. Kumwenda, D.R. Hoover, G. Kafulafula, S.A. Fiscus, C. Nkhoma, S. Chen, and
R.L. Broadhead. 2006. The impact of breastfeeding on the health of HIV-positive mothers
and their children in sub-Saharan Africa. Bulletin of the World Health Organization 84(7):
546-554.

Tawiah, E.O. 1997. Factors affecting contraceptive use in Ghana. Journal of Biosocial Science
29(2): 141-149.

Temmerman, M., S. Moses, D. Kiragu, S. Fusallah, I.A. Wamola, and P. Piot. 1990. Impact of
single session post-partum counseling of HIV infected women on their subsequent
reproductive behavior. AIDS Care 2(3): 247-252.

Terceira, N., S. Gregson, B. Zaba, and P.R. Mason. 2003. The contribution of HIV to fertility
decline in rural Zimbabwe, 1985-2000. Population Studies 57(2): 149-164.

Tuoane, M., N.J. Madise, and I. Diamond. 2004. Provision of family planning services in

20
 Lesotho. International Family Planning Perspectives 30(2): 77-86.

UNAIDS. 2006. 2006 report on the global AIDS epidemic. Geneva, Switzerland: UNAIDS.

UNFPA. 2002. State of world population 2002: People, poverty and possibilities. Geneva,
Switzerland: UNFPA.

USAID, UNAIDS, WHO, UNICEF and the Policy Project. 2006. Coverage of selected services
for HIV/AIDS prevention, care and support in low and middle income countries in 2003.
Washington, DC.: USAID, UNAIDS, WHO, UNICEF and the POLICY Project.

21