SEARCH

SEARCH BY CITATION

Keywords:

  •  abortion;
  • HIV infection;
  • reproduction;
  • women

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

Objectives

The aim of the study was to investigate whether HIV diagnosis affected reproductive planning over time and to assess independent predictors of abortion overall and following HIV diagnosis.

Methods

Donne con Infezione da HIV (DIDI) is an Italian multicentre study based on a questionnaire survey carried out in 585 HIV-positive women between November 2010 and February 2011. The incidence and predictors of abortion were measured by person-years analysis and Poisson regression.

Results

The crude incidence rate of abortion was 18.8 [95% confidence interval (CI) 16.5–21.4] per 1000 person-years of follow-up (PYFU). Compared with women who terminated their pregnancy before HIV diagnosis, women who terminated their pregnancy after HIV diagnosis but before 1990 showed a 2.56-fold (95% CI 1.41–4.65) higher risk. During 1990–1999 and 2000–2010, HIV diagnosis was not significantly associated with outcome [adjusted rate ratio (ARR) 0.93 (95% CI 0.55–1.59) and ARR 0.69 (95% CI 0.32–1.48), respectively]. Age [ARR 0.96 (95% CI 0.94–0.99) per 1 year older] and injecting drug use [ARR 1.38 (95% CI 0.98–1.94)] were found to be predictors of abortion overall. After HIV diagnosis, being on combination antiretroviral therapy [ARR 0.54 (95% CI 0.28–1.02)], monthly income < €800 [ARR 1.76 (95% CI 0.99–3.12)], younger age [ARR 0.95 (95% CI 0.91–1.00) per 1 year older] and fear of vertical transmission [ARR 1.95 (95% CI 1.04–3.67)] were found to be independently associated with abortion.

Conclusions

We observed a higher incidence of abortion compared with data available for the general Italian population. Awareness of HIV diagnosis was predictive of abortion only in the 1980s. Women with HIV infection are still worried about vertical HIV transmission. Interventions promoting HIV screening among women who plan to have an abortion and informative counselling on motherhood planning in the setting of HIV care are needed.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

World-wide, the majority of HIV-infected women are of reproductive age [1]. Several studies, primarily focused on pregnancy outcome, have tried to assess the rates of induced abortion among women with HIV infection in industrialized countries [2-6]. In recent years, seropositive women who have conceived have seemed to be more likely to continue their pregnancies. This decision has probably been influenced by the implementation of measures to reduce mother-to-child transmission (MTCT) [7-9] and the improvement in survival driven by highly active antiretroviral therapy (HAART) [10]. However, most of the studies focusing on reproductive choices in HIV-infected women were conducted before 2002 [2-4]. Studies published more recently [5, 6] addressed the proportion of pregnancies ending in termination and the characteristics associated with abortion, but did not allow estimation of the incidence rate or the investigation of possible time trends. Diagnosis of HIV infection might have a significant impact on a woman's decision whether to carry a pregnancy to term. This is particularly true in developing countries, where women who are aware of their HIV status are less likely to want and to have a child following infection diagnosis [11, 12]. Few data are available on the impact of HIV on reproductive decision-making in the HAART era in high-income countries. Further, no recent studies have investigated whether women living with HIV, when unaware of their infection, should be considered at higher risk of abortion compared with the general population. A European study conducted in 2000 [3] revealed that the number of induced abortions was high before HIV diagnosis and that it significantly increased thereafter.

To provide more contemporary insights, we assessed, through self-report, the incidence of induced abortion in the context of HIV infection by calendar year. In particular, we measured the time trends of induced abortion in women living with HIV, distinguishing two periods, one before and one after HIV diagnosis. The possibility of an interaction between awareness of HIV infection and calendar period was formally tested. Moreover, we investigated independent predictors of induced abortion overall and following HIV diagnosis.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

The DIDI study

Donne con Infezione da HIV (DIDI) is an Italian multicentre study based on a questionnaire survey carried out in 585 HIV-positive women between November 2010 and February 2011. Health care workers administered the anonymous, in-depth questionnaire to all women aged 18 years or older, with a fair understanding of the Italian language, followed at 16 Italian infectious diseases centres. Women were approached at their routine follow-up visits. Written informed consent was obtained after local human subjects committees’ approval. The items covered in the questionnaire included the woman's sociodemographic characteristics, and data on recreational drug use, smoking, spiritual and religious attitudes, sexual and gynaecological health, the desire for motherhood, partnership, HIV disclosure, physical and mental health, and adherence to combination antiretroviral therapy (cART). In particular, women were asked to report the number of previous abortions, miscarriages and pregnancies, their age at the event, the number of children and their relative ages, and the number of children infected with HIV and their relative ages. Data on baseline HIV staging and viro-immunological parameters, antiretroviral drug experience, including the start and stop date for each drug, coinfection with hepatitis viruses, and other sexually transmitted diseases were available from the patients’ records.

Induced abortion in Italy

Abortion in Italy became legal in May 1978, when women were allowed to terminate a pregnancy on demand during the first 90 days of pregnancy. Women are eligible to request an abortion for health, economic or social reasons, including the circumstances under which conception occurred. Abortions are performed free of charge in public hospitals or in private clinics authorized by the regional health authorities. The law also allows termination in the second trimester of pregnancy, but only when the life of the woman would be at risk if the pregnancy were carried to term or when the fetus has genetic or other serious malformations which would put the mother at risk of serious psychological or physical consequences. Although the law only permits pregnancy termination for women at least 18 years old, it also includes provisions for women younger than 18, who can request the intervention of a judge when the legal tutor refuses the intervention, or there are reasons to exclude the legal tutor from the process.

Statistical analyses

For the purpose of this study, abortion was defined as the induced termination of pregnancy. Spontaneous abortion, also known as miscarriage, was not considered. Women who reported at least one abortion were compared with women who did not in terms of general and HIV-related characteristics using χ2 and Wilcoxon tests where appropriate. The following variables were analysed: age at enrolment, citizenship (migrant vs. native Italian), education level (primary school vs. high school/university), monthly salary (cut-off €800), age at first sexual intercourse (cut-off 15 years), total number of pregnancies (none vs. at least one pregnancy), number of children with HIV infection (none vs. at least one child with HIV infection), age at HIV diagnosis, calendar year of HIV diagnosis, mode of HIV transmission [injecting drug use (IDU) vs. sexually transmitted], CD4 count nadir, CD4 count at enrolment, Centers for Disease Control and Prevention (CDC) stage (A/B vs. C), and current use of cART.

Incidence and predictors of first abortion over total follow-up and after HIV diagnosis

Person-years analyses were conducted to assess the time to occurrence of the first induced abortion. Incidence rates of first abortion were determined using the number of women at risk for pregnancy. Women were considered at risk for abortion from 14 to 49 years of age. For person-years of follow-up (PYFU) calculations, the baseline was the date of the 14th birthday, and data were censored at the date of the first abortion, the woman's 50th birthday, or completion of the questionnaire, whichever occurred first.

Overall first abortion incidence rates were calculated according to whether the women were aware of an HIV diagnosis, current age, current calendar year, age at first sexual intercourse (fitted as a binary covariate with a cut-off at 15 years), whether they had had at least one previous pregnancy and demographics. Poisson regression was used for multivariable analysis to identify the predictors of first induced abortion. In addition, we tested for the presence of an interaction between the awareness of HIV infection and the calendar period, to investigate whether known HIV infection may have a different impact on abortion rate over time.

Analysis of the incidence and predictors of first abortion for the period after HIV diagnosis was also carried out. For this analysis, PYFUs were calculated using as baseline the date of the first HIV-positive test, while data censoring remained the same as above. Incidence rates were calculated according to whether women were diagnosed with HIV during pregnancy and, if they were aware of their HIV infection, whether they were currently on cART (women not taking cART were those who for whatever reason were off therapy, including those on a treatment interruption), and according to patient-reported fear of vertical transmission or of con-natal malformations, the self-reported negative impact of HIV on motherhood desire, HIV disclosure, whether they had had at least one previous pregnancy, age at first sexual intercourse, current calendar period and demographics. Poisson regression was used for multivariable analysis.

Women with missing date regarding first abortion were excluded from the incidence rate analysis.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

Characteristics of the DIDI study population

Five hundred and eighty-five women participated in the study. The median age of the women at the time of completing the questionnaire was 44 years [interquartile range (IQR) 39–48 years], 70 (11.9%) were migrants, and 111 (18.9%) were infected by IDU. The median time from HIV diagnosis was 13 (IQR 7–19) years, 122 (22.1%) were in CDC stage C, and the median CD4 count nadir was 200 cells/μL (IQR 101–288 cells/μL). The majority (78.8%) were on virologically effective cART, 8.4% were on virologically ineffective cART or on a treatment interruption, and 12.8% were treatment-naïve. At the time of completing the questionnaire, the median CD4 count was 554 cells/μL (IQR 397–727 cells/(L) (Table 1).

Table 1. Characteristics of women who did and did not report abortion
 Population (n = 585)Abortion (n = 242)No abortion (n = 343)P*
  1. cART, combination antiretroviral therapy; CDC, Centers for Disease Control and Prevention; IDU, injecting drug use; IQR, interquartile range.

  2. 2 test for discrete variables and Wilcoxon test for continuous variables.

Sociodemographic characteristics    
Age at questionnaire completion [median (IQR)]44 (39–48)45 (40–49)43 (38–48)0.004
Migrant [n (%)]70 (11.9)26 (10.7)44 (12.8)0.44
Lower education (primary school) [n (%)]233 (40.5)104 (42.9)133 (38.8)0.30
Baseline monthly income < €800 [n (%)]318 (54.4)141 (58.3)177 (51.6)0.11
IDU [n (%)]111 (18.9)65 (26.9)46 (13.4)< 0.0001
Sexual health history    
First sexual intercourse at age ≤ 15 years [n (%)]117 (20)58 (23.9)59 (17.2)0.04
At least one pregnancy [n (%)]321 (54.9)149 (61.6)172 (50.1)0.0006
At least one HIV-infected child [n (%)]18 (3.1)12 (4.9)6 (1.8)0.02
HIV-related variables    
Age at HIV diagnosis (years) [median (IQR)]29 (23–35)28 (23–34)29 (24–35)0.23
Calendar year at HIV diagnosis [median (IQR)]1997 (91–93)1995 (89-02)1999 (92-04)< 0.0001
Years from HIV diagnosis [median (IQR)]13 (7–19)16 (8–21)12 (6–18)<0.0001
Nadir CD4 count (cells/μL) [median (IQR)]200 (101–288)190 (100–275)205 (101–300)0.22
CDC group C [n (%)]122 (22.1)51 (22.8)71 (21.6)0.75
cART treatment at enrolment [n (%)]540 (87.2)207 (85.5)303 (88.3)0.31
CD4 count at enrolment (cells/μL) [median (IQR)]554 (397–727)576 (398–773)549 (396–696)0.27
HIV RNA < 50 copies/ml at enrolment [n (%)]461 (78.8)186 (76.8)275 (80.1)0.50

Overall, 242 (41.4%) women reported at least one abortion. Some of these women reported more than one abortion: two abortions were reported by 72 women, three abortions by 19, and more than three by 30. Table 1 shows sociodemographic, sexual health history and HIV-related variables for women who reported abortion vs. those who did not. Women who reported at least one abortion were significantly older at the time of completing the questionnaire (P = 0.003), more frequently recalled that their first sexual intercourse was at age 15 years or younger (P = 0.04), and more frequently were infected by IDU (P < 0.0001). Compared with women who did not report abortion, those who did had a more remote calendar year of HIV diagnosis (P < 0.0001) and were more likely to have had at least one pregnancy (P = 0006) and to report children with HIV infection (P = 0.02).

In more than half of cases (n = 140; 57.9%), women reported that the first abortion occurred before HIV diagnosis. Sixty women (24.8%) reported abortion after HIV diagnosis and 24 (9.9%) before and after, and in 7.4% of cases this information was missing in the questionnaire.

Analysis of incidence and predictors of first induced abortion over follow-up

Eighteen women were excluded from the analysis because the date of the first abortion was missing in the questionnaire. Overall, 224 abortions were recorded in 567 women who contributed to 11 929 PYFU. Thus, the overall lifetime incidence rate of first abortion in our patient population was 18.8 (95% CI 16.5–21.4) per 1000 PYFU.

The first abortion incidence rate appeared to decrease over time, declining from 25.9 per 1000 PYFU (95% CI 21.7–31.1) before 1990 to 19.1 per 1000 PYFU (95% CI 15.1–24.1) and 9.1 per 1000 PYFU (95% CI 6.5–12.9) in 1990–1999 and 2000–2010, respectively (p for trend < 0.0001). A declining trend in abortion rates was confirmed even after considering separately the time periods before (test for trend P = 0.05) and after HIV diagnosis (test for trend P < 0.0001). In the period before 1990, the incidence of abortion occurring in women after HIV diagnosis was extremely high [67.9 per 1000 PYFU (95% CI 40.2–114.6)] and was almost threefold higher than the incidence rate observed in the same calendar period in women not yet diagnosed with HIV infection [24.1 per 1000 PYFU (95% CI 19.9–29.0)]. Conversely, in the more recent period from 2000 to 2010, the incidence rate of abortion in women after HIV diagnosis was very low [7.8 per 1000 PYFU (95% CI 5.1–11.8)]. Women who acquired HIV by IDU were at high risk of abortion [28.1 per 1000 PYFU (95% CI 21.8–36.2)] (Table 2).

Table 2. Absolute incidence rates of first induced abortion and adjusted rate ratios (RRs) from fitting a multivariable Poisson regression model
FactorNumber of first abortionsPYFURate per 1000 PYFU (95% CI)Adjusted RR (95% CI) P
  1. CI, confidence interval; IDU, injecting drug use; PYFU, person-years of follow-up.

  2. * P for interaction between awareness of HIV infection and calendar period 1990–1999.

  3. P for interaction between awareness of HIV infection and calendar period 2000–2010.

Age     
Per 1 year older   0.96 (0.94–0.99)0.009
Mode of HIV transmission     
Sexual transmission164979216.7 (14.4–19.5)1 
IDU60213728.1 (21.8–36.2)1.38 (0.98–1.94)0.018
Citizenship     
Migrant2001073618.6 (16.2–21.4)1 
Native Italian24119320.1 (13.5–30.0)1.14 (0.73–1.80)0.54
Education     
High school/university130711618.3 (15.4–21.7)1 
Primary94481319.5 (16.0–23.9)1.00 (0.75–1.29)0.92
Baseline monthly salary     
≥ €80095571316.6 (13.6–20.3)1 
< €800129621620.8 (17.5–24.7)1.17 (0.88–1.57)0.26
Age at first sexual intercourse     
> 15 years174972817.9 (15.4–20.8)1 
≤ 15 years50220122.7 (17.2–30.0)1.21 (0.87–1.67)0.23
Number of pregnancies     
< 174299124.7 (19.7–31.1)1 
≥ 1150893816.8 (14.3–19.7)0.93 (0.69–1.26)0.67
Timing of HIV diagnosis and calendar period     
Abortion before HIV diagnosis164758021.6 (18.6–25.2)1 
Before 1990107444924.1 (19.9–29.0)  
1990–199946234719.6 (14.7–26.2)  
2000–20101178414.0 (7.8–25.3)  
Abortion after HIV diagnosis60434813.8 (10.7–17.8)  
Before 19901420667.9 (40.2–114.6)2.56 (1.41–4.65)0.010*
1990–199924131718.2 (12.2–27.2)0.93 (0.55–1.59)0.004
2000–20102228257.8 (5.1–11.8)0.69 (0.32–1.48) 
Total2241192918.8 (16.5–21.4)  

In the multivariable analysis, HIV diagnosis was not associated with abortion [adjusted rate ratio (ARR) 1.22 (95% CI 0.81–1.83); P = 0.32]. However, compared with women who terminated their pregnancy before HIV diagnosis, women who terminated their pregnancy after HIV diagnosis but before 1990 showed a 2.56-fold (95% CI 1.41–4.65) higher risk of abortion. Among those who had terminations in the periods 1990–1999 and 2000–2010, HIV diagnosis did not seem to be significantly associated with the outcome [ARR 0.93 (95% CI 0.55–1.59) and ARR 0.69 (95% CI 0.32–1.48) vs. before HIV diagnosis, respectively]. The P-values for the interaction between HIV diagnosis and calendar period were significant in the adjusted model (ARR of abortion relative to HIV diagnosis in 1990–1999 vs. < 1990, P = 0.010, and in 2000–2010 vs. <1990, P = 0.004). After adjusting for potential confounders, younger age at abortion [ARR 0.96 (95% CI 0.94–0.99) per 1 year older; P = 0.009] and IDU [ARR 1.38 (95% CI 0.98–1.94)] vs. sexually transmitted HIV [ARR 1.38 (95% CI 0.98–1.94); P = 0.01] were independent predictors of abortion.

Analysis of incidence and predictors of first abortion after HIV diagnosis

To evaluate the impact of HIV-related factors on the incidence of first abortion, we then focused on the 60 events that occurred during 4349 PYFU after HIV diagnosis [crude incidence rate 13.8 per 1000 PYFU (95% CI 10.7–17.8)]. We observed a high incidence rate of induced abortion among women who acquired HIV by IDU [23.0 per 1000 PYFU (95% CI 15.5–34.0)] and those who were not on cART and were aware of being HIV-infected before pregnancy [7.6 per 1000 PYFU (95% CI 19.5–39.9)]. Further, women who self-reported a fear of vertical HIV transmission [22.9 per 1000 PYFU (95% CI 15.3–34.2)] or of con-natal malformations [19.7 per 1000 PYFU (95% CI 10.7–35.1)] had high abortion incidence rates. Conversely, a low incidence rate was observed in women who were already aware of their HIV infection and who were on cART at the time of the abortion [8.6 per 1000 PYFU (95% CI 5.7–12.8)] and those who declared a monthly income higher than €800 [9.4 per 1000 PYFU (95% CI 6.1–14.4)]. The abortion incidence rate in migrant women living with HIV was even lower [3.5 per 1000 PYFU (95% CI 0.5–24.8)].

In the multivariable model, the risk of first abortion was significantly lower in more recent calendar years. In fact, compared with the period before 1990, the risk of first abortion was 0.47 (95% CI 0.22–0.99; P = 0.04) in 1990–1999 and 0.37 (95% CI 0.13–1.02; P = 0.05) in 2000–2010. Among women who were aware of their HIV infection before pregnancy, the current use of cART was protective against abortion [receiving vs. not receiving cART, ARR 0.54 (95% CI 0.28–1.04); P = .06]; women who had a diagnosis at pregnancy did not show an increased risk of abortion compared with those who were already aware of their infection and who were off cART [HIV diagnosed during pregnancy vs. already aware of HIV infection and not receiving cART, ARR 0.84 (95% CI 0.37–1.90); P = 0.68]. Fear of vertical transmission was strongly associated with abortion after HIV diagnosis: women who were concerned about infecting the child showed a twofold higher risk of abortion compared with those who were not [ARR 1.90 (95% CI 1.02–3.56); P = 0.04]. Monthly income lower than €800 [ARR 1.76 (95% CI 0.99–3.11); P = 0.05 vs. monthly income ≥ €800] and younger age [per 1 year older, ARR 0.95 (95% CI 0.91–1.00); P = 0.05] were also found to be independent predictors of first abortion after HIV diagnosis. The risk of abortion in women who became sexually active before 15 years of age tended to be higher [ARR 1.65 (95% CI 0.91–2.98); P = 0.09]. The risk of induced abortion did not change according to whether women had previously had at least one pregnancy [ARR 1.13 (95% CI 0.53–2.41); P = 0.73] (Table 3). In three cases during 108 PYFU, a vertically infected child was born. In a further analysis, we did not find a significant difference in the risk of induced abortion among those who had had a previous pregnancy according to the experience of a vertically transmitted infection (data not shown).

Table 3. Absolute incidence rates of first induced abortion after HIV diagnosis and adjusted rate ratios (RRs) from fitting a multivariable Poisson regression model
FactorNumber of first abortionsPYFURate per 1000 PYFU (95% CI)Adjusted RR (95% CI) P
  1. ART, antiretroviral therapy; CI, confidence interval; IDU, injecting drug use; PYFU, person-years of follow-up.

Calendar year     
Before 19901420667.9 (40.2–114.6)1 
1990–199924131718.2 (12.2–27.2)0.47 (0.22–0.99)0.04
2000–20102228257.8 (5.1–11.8)0.37 (0.13–1.02)0.05
Age     
Per 1 year older   0.95 (0.91–1.00)0.05
Mode of HIV transmission     
Sexual transmission35326110.7 (7.7–14.9)1 
IDU25108823.0 (15.5–34.0)1.22 (0.68–2.19)0.50
Citizenship     
Migrant12873.5 (0.5–24.8)1 
Native Italian59406220.1 (13.5–30.0)3.21 (0.42–24.5)0.26
Education     
High school/university35274012.8 (9.2–17.8)1 
Primary25160915.5 (10.5–23.0)1.02 (0.59–1.75)0.92
Baseline monthly salary     
≥ €8002122439.4 (6.1–14.4)1 
< €80039210618.5 (13.5–25.3)1.76 (0.99–3.11)0.05
Age at first sexual intercourse     
> 15 years41346211.8 (8.7–16.1)1 
≤ 15 years1988721.4 (13.7–33.6)1.65 (0.91–2.98)0.09
Number of pregnancies     
< 1940922.0 (11.5–42.3)1 
≥ 151394012.9 (9.8–11.7)1.13 (0.53–2.41)0.73
HIV awareness at pregnancy     
HIV diagnosis at pregnancy854114.7 (7.4–29.5)0.84 (0.37–1.90)0.68
Aware of HIV diagnosis; currently off ART28101527.6 (19.1–39.9)1 
Aware of HIV diagnosis; currently on ART2427938.6 (5.7–12.8)0.54 (0.28–1.04)0.06
Fear of vertical transmission     
No36330410.9 (7.8–15.1)1 
Yes24104522.9 (15.3–34.2)1.90 (1.02–3.56)0.04
Fear of con-natal malformations     
No49378313.1 (9.7–17.1)10.8
Yes1156619.7 (10.7–35.1)0.93 (0.42–2.07)6
HIV disclosure     
No14117811.9 (7.0–20.1)1 
Yes46317114.5 (10.9–19.4)1.55 (0.83–2.89)0.16
Negative impact of HIV on motherhood desire     
No29209213.9 (9.6–19.9)1 
Yes31225713.7 (9.7–19.5)0.86 (0.50–1.50)0.61
Total 60 4349 13.8 (10.7–17.8)   

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

In this cross-sectional survey carried out in Italy from November 2010 to February 2011, more than 40% of interviewed women living with HIV reported at least one induced abortion in her reproductive health history. This unexpectedly high prevalence might be driven by the fact that the median age of the women included in the study was > 40 years and that nearly 20% had a history of drug abuse, which is known to be a factor associated with abortion in the general population [13-16]. Another reason for our finding may be that our study was based on self-report and not chart review or cohort data. The fact that women of all ages were interviewed at their routine visit at the HIV care centre and not when accessing a specific health care service [such as gynaecology or sexually transmitted diseases (STD)] or at a specific time-point may have increased detection rates. In most cases abortion occurred before HIV diagnosis, suggesting that women diagnosed with HIV infection often have a sexual health history that includes multiple, complex and traumatic events. In our study, the high proportion of abortions before HIV diagnosis may also be a result of the fact that women participating in the DIDI study generally received their HIV diagnosis at an advanced stage of disease, with a CD4 count nadir of approximately 200 cell/μL, in their late twenties or thirties.

After specific Italian abortion legislation was enacted in 1978, rates of abortion among the general Italian female population first rose and then declined steadily, from a peak of 16.9 abortions per 1000 women of reproductive age in 1983, to 9.7 in 1996, to 9.6 in 2005 and 8.3 in 2009 [17]. In our study, the rate observed in women not yet diagnosed with HIV infection was 24.1 per 1000 PYFU before 1990 and declined to 19.6 and 14.0 per 1000 PYFU in 1990–1999 and 2000–2010, respectively. Thus, we can conclude that our multicentre population of HIV-positive women displayed a much higher risk of abortion even before the HIV diagnosis, compared with the general population in Italy. In particular, during the last 10 years, they have had a 50% increased risk [17]. This study identifies a need for more effective strategies in the management of women who plan to have an abortion, with particular emphasis on HIV and other sexually transmitted diseases. This may be achieved by establishing routine HIV counselling and testing at the time of the abortion. To date in Italy, HIV and family planning services have been offered separately. From a public health point of view, a high induced abortion rate among HIV-infected and uninfected women is of particular concern, being the result of unprotected sexual intercourse, which carries the danger of HIV acquisition or transmission.

Looking at the period after HIV diagnosis, the incidence rate of induced abortion observed in the years before 1990 among women participating in the study was extremely high, being up to four times that in the general Italian female population during the same calendar period [17]. More positively, in more recent calendar years the incidence of abortion after HIV diagnosis was lower and comparable to that reported for Italian women in general. This finding has several implications. First, it suggests that the impact of HIV infection on the desire to have children and the decision to terminate pregnancy may have changed over time in HIV-positive women. Indeed, the awareness of HIV infection had a significant effect only in the 1980s, when women who knew that they were HIV-infected had a 2.5-fold higher risk of abortion compared with those who were unaware of their serostatus. During the 1990s, the incidences of abortion before and after HIV diagnosis were comparable. However, the incidence in HIV-infected women (either before or after diagnosis) was almost twofold that reported for the Italian HIV-negative population [17], suggesting that, regardless of awareness of infection, women with HIV infection at that time had to be considered a particularly vulnerable group. Hence, our results confirm those of previously published reports indicating that contraception in HIV-infected women is generally suboptimal [18-21]. Many factors may account for unprotected sexual practices among HIV-positive women, including difficulties in negotiating condom use, in particular when they have an HIV-positive partner [20]. Beliefs regarding lower levels of infectivity under antiretroviral therapy are also associated with less condom use. Studies have reported higher levels of unprotected sex among women after antiretroviral treatment initiation, which did not vary with the therapeutic response [21]. More recently, awareness of HIV infection was again found not to be related to the risk of abortion, and the lower incidence of abortion observed among HIV-positive women aware of their status may partially reflect temporal trends in the epidemiology of HIV acquisition, with the progressive substitution of IDU with women who acquired infection through sexual transmission [1, 13-16]. This change in epidemiology in recent years may also explain the lack of an association between mode of HIV transmission and abortion documented when we studied only PYFU after HIV diagnosis. The decrease in the abortion rate in the later HAART era has already been described elsewhere [4], and mainly reflects the better life expectancy of HIV-infected women provided by efficient antiretroviral drugs and the wide availability of MTCT protocols, which has increased positive attitudes towards motherhood. Furthermore, the current use of antiretroviral therapy was protective against abortion, after adjusting for other factors. Data on plasma HIV viral load at induced abortion were not available and we could not discriminate between effective and suboptimal cART. Surprisingly, an HIV diagnosis during pregnancy did not put women at a significantly higher risk of induced abortion in our cohort. Of note, however, is the finding that fear of vertical transmission in our study was strongly associated with the decision to induce abortion, independently of the time period and the use of cART. Women who were concerned about infecting their child had a twofold increased risk of pregnancy termination. This demonstrates that there is still a need to improve preconception counselling and to provide HIV-infected women with detailed information about the efficient measures adopted to prevent MTCT.

This study has a number of limitations. First, abortion rates were calculated based on events that may have occurred some years previously in the personal history of each women, and therefore recall bias cannot be ruled out. Secondly, as abortion rates may differ greatly with respect to population characteristics, such as median age and the prevalence of IDU and of migrant women, caution should be exercised when generalizing from our results. Thirdly, the DIDI study collected data about condom use and contraception, marital status, spirituality/religiosity and family support, but the information refers to the time at which the questionnaire was completed and not the time of the abortion, which might have occurred many years before, and hence their association with induced abortion was not investigated in the present analysis. The same was true for abortions occurring after HIV diagnosis; parameters related to stage of HIV disease were collected from charts at the time of completion of the questionnaire and were not available for the time of the abortion. We assumed that the women's socioeconomic status would not radically change over time and included it in the analysis; this may possibly have resulted in an underestimation of the number of women in the lower stratum. However, the strengths of our study should also be mentioned: the multicentre nature of the study, the high number of interviewed women living with HIV, and the fact that the outcome was self-reported. Further, our study provides important updated information about abortion rates in HIV-infected women and is the first who formally determine whether there is an interaction between awareness of HIV and calendar period.

In conclusion, the high frequency of induced abortion in women who are or will be diagnosed with HIV infection underlines the absolute need to implement HIV screening among women who plan to have an abortion, together with sexual and general health-promoting counselling. Our results indicate that these women may already be HIV-infected, or may have been infected at conception of the terminated pregnancy, or may acquire HIV in the future. Moreover, our study demonstrates that, even now, women who have been living with HIV for a long time and who are receiving cART have a fear of vertical HIV transmission. At the same time, it is well known that unwanted and mistimed pregnancies continue to occur in HIV-positive women, primarily because in sexually active women contraceptive methods are frequently not used [22-26]. Interventions promoting informative counselling on effective contraception, motherhood planning, and the prevention of MTCT are greatly needed in the setting of routine care of HIV-infected women.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

We acknowledge Women for Positive Action (WFPA), a global initiative established in response to the need to address specific concerns of women living and working with HIV. The DIDI Study Group stemmed from the WFPA Italia.

Appendix: the DIDI study group

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References

Study coordinators: Antonella d'Arminio Monforte (Milan) and Adriana Ammassari (Rome).

Study participants: Enza Anzalone (Frosinone), Teresa Bini (Milan), Antonella Castagna (Milan), Anna Maria Cattalan (Rovigo), Gabriella D'Ettorre (Rome), Fiorella Di Sora (Rome), Daniela Francisci (Perugia), Miriam Gargiulo (Naples), Nicoletta Ladisa (Bari), Giuseppina Liuzzi (Rome), Tiziana Quirino (Busto Arsizio), Raffaella Rosso (Genova), Maria Paola Trotta (Rome) and Francesca Vichi (Firenze).

Experts: Antonella Cingolani (Rome) and Rita Murri (Rome).

Statistician and data manager: Paola Cicconi (Milan) and Paola Pierro (Rome).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Appendix: the DIDI study group
  9. References
  • 1
    UNAIDS Report on the global AIDS epidemic2010.
  • 2
    Bongain A, Berrebi A, Marinè-Barjoan E et al. Changing trends in pregnancy outcome among HIV-infected women between 1985–1997 into Southern French Universtity Hospitals. Eur J Obstet Gynecol Reprod Biol 2002; 104: 124128.
  • 3
    van Benthem BH, de Vincenzi I, Delmas MC et al. Pregnancies before and after HIV diagnosis in a European cohort of HIV-infected women. European Study on the Natural History of HIV infection in Women. AIDS 2000; 14: 21712178.
  • 4
    Massad LS, Springer G, Jacobson L et al. Pregnancy rates and predictors of conception, miscarriage and abortion in US women with HIV. AIDS 2004; 18: 281286.
  • 5
    Towsend C, Cortina-Borja M, Peckham C et al. Trends in management and outcome of pregnancies in HIV-infected women in the UK and Ireland, 1990–2006. BJOG 2008; 115: 10781086.
  • 6
    Floridia M, Tamburrini E, Tibaldi C et al. for the Italian Group on Surveillance on Antiretroviral Treatment in Pregnancy. Voluntary pregnancy termination among women with HIV in the HAART era (2002–2008): a case series from a National study. AIDS Care 2010; 22: 5053.
  • 7
    Connor EM, Sperling RS, Gelber R et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. N Engl J Med 1994; 331: 11731180.
  • 8
    de Ruiter A, Mercey D, Anderson J et al. British HIV Association and Children's HIV Association guidelines for the management of HIV infection in pregnant women. HIV Med 2008; 9: 452502.
  • 9
    Marazzi M, Palombi L, Nielsen-Saines K et al. Extended antenatal use of triple antiretroviral therapy for prevention of mother to child transmission of HIV-1 correlates with favorable pregnancy outcomes. AIDS 2011; 25: 16111618.
  • 10
    Ray M, Logan R, Sterne JA et al. The effect of combined antiretroviral therapy on the overall mortality of HIV-infected individuals. AIDS 2010; 24: 123137.
  • 11
    Kannippan S, Jeyapaul MJ, Kalyanwala S. Desire for motherhood: exploring HIV-positive women's desires, intentions and decision-making in attaining motherhood. AIDS Care 2008; 20: 625630.
  • 12
    Nattabi B, Li J, Thompson SC et al. A systematic review of factors influencing fertility desires and intentions among people living with HIV/AIDS: implications for policy and service delivery. AIDS Behav 2009; 13: 949968.
  • 13
    Prager SW, Steinauer JE, Foster DG et al. Risk factors for repeat elective abortion. Am J Obstet Gynecol 2007; 197: 16.
  • 14
    Martino SC, Collins RL, Ellickson PL, Klein DJ. Exploring the link between substance abuse and abortion: the roles of unconventionality and unplanned pregnancy. Perspect Sex Reprod Health 2006; 38: 6675.
  • 15
    Reardon DC, Coleman PK, Cougle JR. Substance use associated with unintended pregnancy outcomes in the National Longitudinal Survey of Youth. Am. J. Drug Alcohol Abuse 2004; 30: 369383.
  • 16
    Rasch V, Gammeltoft T, Knudsen LB et al. Induced abortion in Denmark: effect of socioeconomic situation and country of birth. Eur J Public Health 2008; 18: 144149.
  • 17
    Relazione del Ministero della salute sulla attuazione della legge contenente norme per la tutela sociale della maternità e per l'interruzione di gravidanza (legge 194/78) – Dati preliminari 2009, dati definitivi 2008. Roma 6 Agosto 2010.
  • 18
    Wilson TE, Massad LS, Riester KA et al. Sexual, contraceptive, and drug use behaviors of women with HIV and those at high risk for infection: results from the Women's Interagency HIV Study. AIDS 1999; 13: 591598.
  • 19
    Kline A, Strickler J, Kempf J. Factors associated with pregnancy and pregnancy resolution in HIV seropositive women. Soc Sci Med 1995; 40: 15391547.
  • 20
    Weinhardt LS, Kelly JA, Brondino MJ et al. National Institute of Mental Health Healthy Living Project Team. HIV transmission risk behavior among men and women living with HIV in 4 cities in the United States. J Acquir Immune Defic Syndr 2004; 36: 10571066.
  • 21
    Wilson TE, Gore ME, Greenblatt R et al. Changes in sexual behavior among HIV-infected women after initiation of HAART. Am J Public Health 2004; 94: 11411146.
  • 22
    Stuart GS. Fourteen million women with limited options: HIV/AIDS and highly effective reversible contraception in sub-Saharan Africa. Contraception 2009; 80: 412416.
  • 23
    Blanchard K, Holt K, Bostrom A, van der Straten A et al. Impact of learning HIV status on contraceptive use in the MIRA trial. Fam Plann Reprod Health Care 2011; 37: 204208.
  • 24
    Chibwesha CJ, Li MS, Matoba CK et al. Modern contraceptive and dual method use among HIV-infected women in Lusaka, Zambia. Infect Dis Obstet Gynecol 2011; 2011: 261453.
  • 25
    Jhangri GS, Heys J, Alibhai A et al. Unmet need for effective family planning in HIV-infected individuals: results from a survey in rural Uganda. J Fam Plann Reprod Health Care 2012; 38: 2329.
  • 26
    Johnson KB, Akwara P, Rutstein SO et al. Fertility preferences and the need for contraception among women living with HIV: the basis for a joint action agenda. AIDS 2009; 23 (Suppl 1): S7S17.