Abstract
Background
Recent data on the rate and risk factors of induced abortion among women living with HIV (WLWH) are limited. Our aim was to use Finnish national health register data to 1) determine the nationwide rate of induced abortions of WLWH in Finland during 1987–2019, 2) compare the rates of induced abortions before and after HIV diagnosis over different time periods, 3) determine the factors associated with terminating a pregnancy after HIV diagnosis, and 4) estimate the prevalence of undiagnosed HIV at induced abortions to see whether routine testing should be implemented.
Methods
A retrospective nationwide register study of all WLWH in Finland 1987–2019 (n = 1017). Data from several registers were combined to identify all induced abortions and deliveries of WLWH before and after HIV diagnosis. Factors associated with terminating a pregnancy were assessed with predictive multivariable logistic regression models. The prevalence of undiagnosed HIV at induced abortion was estimated by comparing the induced abortions among WLWH before HIV diagnosis to the number of induced abortions in Finland.
Results
Rate of induced abortions among WLWH decreased from 42.8 to 14.7 abortions/1000 follow-up years from 1987–1997 to 2009–2019, more prominently in abortions after HIV diagnosis. After 1997 being diagnosed with HIV was not associated with an increased risk of terminating a pregnancy. Factors associated with induced abortion in pregnancies that began after HIV diagnosis 1998–2019 were being foreign-born (OR 3.09, 95% CI 1.55–6.19), younger age (OR 0.95 per year, 95% CI 0.90–1.00), previous induced abortions (OR 3.36, 95% CI 1.80–6.28), and previous deliveries (OR 2.13, 95% CI 1.08–4.21). Estimated prevalence of undiagnosed HIV at induced abortion was 0.008–0.029%.
Conclusions
Rate of induced abortions among WLWH has decreased. Family planning should be discussed at every follow-up appointment. Routine testing of HIV at all induced abortions is not cost-effective in Finland due to low prevalence.
Keywords: HIV infections, Induced abortion, HIV testing, Registries, Retrospective studies
Background
Before effective methods to prevent vertical transmission were discovered studies reported an increased proportion of pregnancies ending in induced abortion after HIV diagnosis [1, 2]. Development of antiretroviral treatment (ART) has reduced the risk of vertical transmission in European cohort studies from 15 to 20% in the pre-ART era to 10% with mono-ART and to less than 1% in the current era of combined ART (cART) [3–7]. Although over 80% of the nearly 50,000 women newly diagnosed with HIV in Europe in 2018 were fertile-aged [8], little is known about the rates and risk factors for induced abortions among women living with HIV (WLWH). Most data come from limited questionnaire studies with varying response rates [9–11]. In questionnaire studies on the general population, induced abortions have been significantly underreported [12–14]. The risk factors for WLWH having an induced abortion have included younger age, earlier year of HIV diagnosis, and transmission by intravenous drug use (IDU) [15].
In Finland, an indication is needed for induced abortion but 95% of abortions are performed on social indications [16]. Induced abortion is widely available and inexpensive in public health care. Since HIV testing at an induced abortion is currently not routinely recommended, the prevalence of undiagnosed HIV at the time of induced abortion is unknown.
The aim of the study was to use Finnish national health register data to 1) determine the nationwide rate of induced abortions of WLWH in Finland in 1987–2019, 2) compare the rates of induced abortions before and after HIV diagnosis over different time periods, 3) determine the factors associated with terminating a pregnancy after HIV diagnosis, and 4) estimate the prevalence of undiagnosed HIV at induced abortions to see whether routine testing should be implemented.
Methods
This is a nationwide retrospective study combining data from several national registers. The study population consisted of all women who were diagnosed with or treated for HIV in Finland before 1 January 2020, were fertile-aged (15–49 years) between 1987 and 2019, and had a valid personal identification code (PIC; granted to all citizens and immigrants with a permit to stay at least 1 year). The women were identified from the Finnish HIV Quality of Care Register (FINHIV), which is described in detail elsewhere and consists of all people diagnosed with or treated for HIV in Finland since 1 January 1984 [17]. Sex in the register is reported predominantly based on the PIC which is sex-specific (categories female/male); generally this means sex assigned at birth.
Using the PIC as an identification, data from the FINHIV were combined with two national registers maintained in the Finnish Institute for Health and Welfare (THL): Register on Induced Abortions and Sterilizations, and Medical Birth Register. Reporting induced abortions, sterilizations, and births to their respective registers is mandatory by law, and the electronical information is available since 1983 (induced abortions), 1987 (sterilizations), and 1987 (births). The following variables were collected: from the FINHIV date of birth, date of death, date of HIV diagnosis, transmission route, country of birth, date of immigration, date of emigration, first CD4+ cell count and its date, date of AIDS diagnosis, hepatitis C antibodies (positive/negative), hepatitis B S-antigen (positive/negative), and date of latest contact to HIV care (e.g. date of latest HIV viral load); from the Register on Induced Abortions and Sterilizations date of abortion, gestational age at abortion, indication for abortion, and date of sterilization; from the Medical Birth Register date of delivery and gestational age at birth. Because only data on induced abortions in Finland were available and the immigration and emigration data in the FINHIV are incomplete, the estimated follow-up time was corrected for those not born in Finland with dates of first and latest recorded contact to health care. These were collected from the Care Register for Health Care and Register for Primary Health Care Visits (maintained at THL; earliest data from 1994 and 2011, respectively).
For women born in Finland, the start of the follow-up was defined as 1 January 1987 or the day they turned 15 years of age, whichever later. The end of the follow-up was defined as 31 December 2019, the date of turning 50 years of age or, when applicable, date of death, sterilization or emigration. For women not born in Finland, we used additional cut-off points to correct for the time before immigration and the increased rate of emigration: for the start of follow-up, they consisted of the date of immigration as recorded in the FINHIV or the first recorded health care contact (including the date of first delivery or induced abortion). Similarly, for the end of the follow-up the additional cut-off point was the last recorded date of health care contact. Formation of the study population is depicted in Fig. 1.
Fig. 1.
Formation of the study population. aPersonal Identification Code, given to all citizens or immigrants with permit to stay at least 1 year. bDates of first and last recorded contact to health care were collected from these registers and used as cut-off points for follow-up time for foreign-born women because immigration and emigration data in the FINHIV are incomplete
The date of HIV diagnosis is most often recorded as the date the first positive HIV antigen/antibody test was taken, and there is a delay of approximately 7 days before a positive result is confirmed. Arranging an induced abortion takes a minimum of 3 days. Therefore, an induced abortion was considered to have taken place before the HIV diagnosis (i.e., positive HIV test result could not have affected the decision to have an induced abortion) if the date of the abortion was before the date of HIV diagnosis, or the HIV diagnosis date was no more than 10 days prior to the date of abortion. If the HIV diagnosis date was earlier than 10 days prior to the date of abortion, the abortion was considered to have taken place after the HIV diagnosis.
A predictive multivariable logistic regression model was used to assess the factors associated with terminating a pregnancy, with all pregnancies ending in delivery or induced abortion in 1987–2019 included and clustering by person. An additional analysis focusing on the factors associated with WLWH having an induced abortion was done by constructing another model with time limited to 1998–2019 and to time and pregnancies that began after HIV diagnosis. The models were created by including all available variables and checking for interactions, and the final models were chosen by including variables and interactions terms based on significance in the model construction (p ≤ 0.05) and by comparison of Akaike and Bayesian Information Criteria between models.
To estimate the maximum prevalence of undiagnosed HIV at the time of induced abortion, we examined all induced abortions taking place 0–5 years before HIV diagnosis during 1987–2014. Additionally, we used the CD4-positive lymphocyte count at HIV diagnosis (if measured 0–90 days after HIV diagnosis) to assess whether the women had been infected at the time of the last induced abortion. The thresholds for approximating the time passed since transmission were based on calculations by Lodi et al. [18] where the median times for 25–30-year-old heterosexual women from seroconversion to CD4+ cell count dropping below 500 cells/μL, 350 cells/μL, and 200 cells/μL were 1.63, 5.66, and 10.71 years, respectively. We set the thresholds for probable minimum time as infected at 6 months for those with CD4+ cell count over 500 cells/μL, 1 year for those with CD4+ cell count 350–499 cells/μL, 3 years for those with 200–349 cells/μL, and 5 years for those under 200 cells/μL. The calculated maximum and minimum numbers of induced abortions with undiagnosed HIV were then compared to the total number of induced abortions in the time period as recorded in the Register of Induced Abortions and Sterilizations [19]. We also evaluated the prevalence for some subgroups: women not born in Finland, women with previous induced abortions, and women with previous deliveries. The total of induced abortions performed for these subgroups was estimated from THL open data [19] for previous pregnancies and from articles by Malin et al. [20] and Heino et al. [16] for women not born in Finland.
For statistical analysis, SPSS 27.0 (IBM, Chicago, Illinois, USA) and Stata 17 (StataCorp LLC, College Station, Texas, USA) were used.
Results
The study consists of 1017 women described in Table 1. The total follow-up time was 15,807 years (median 14.5 years, IQR 8.4–22.1), with 6509 years before and 9297 years after HIV diagnosis. A total of 249 women (24.5% of the study population) had 396 induced abortions. The number of abortions per woman ranged from 1 to 6, with 154 (61.8%) having one abortion, 61 (24.5%) having two, and 34 (13.0%) having three or more. The indication for induced abortion was recorded in 386 cases (97.5%); social circumstances for 349 (90.4% of those recorded), under 17 years of age in 13 (3.4%), endangerment of the woman’s health in 11 (2.8%), fetal indication in 6 (1.6%), and other indication in 7 (1.8%).
Table 1.
Characteristics of the study population
| Population | All | Divided by year of HIV diagnosisa | ||||||
|---|---|---|---|---|---|---|---|---|
| 1987–1997 | 1998–2008 | 2009–2019 | ||||||
| Number of women | 1017 | 138 | 13.6% | 413 | 40.6% | 466 | 45.8% | |
| Median age at HIV diagnosis | 30.7 | 29.0 | 30.4 | 32.0 | ||||
| Country of birth | ||||||||
| Finland | 428 | 42.1% | 97 | 70.3% | 205 | 49.6% | 126 | 27.0% |
| Other | 586 | 57.6% | 41 | 29.7% | 208 | 50.4% | 337 | 72.3% |
| Unknown | 3 | 0.3% | 0 | 0 | 3 | 0.6% | ||
| Diagnosed before immigration | 159 | 15.6% | 10 | 7.2% | 45 | 10.9% | 104 | 22.3% |
| Mode of transmission | ||||||||
| Sex | 809 | 79.5% | 118 | 85.5% | 289 | 70.0% | 402 | 86.3% |
| IDUb | 99 | 9.7% | 8 | 5.8% | 81 | 19.6% | 10 | 2.1% |
| Verticalc | 13 | 1.3% | 3 | 2.2% | 8 | 1.9% | 2 | 0.4% |
| Blood transfusionc | 11 | 1.1% | 2 | 1.4% | 2 | 0.5% | 7 | 1.5% |
| Unknown | 85 | 8.4% | 7 | 5.1% | 33 | 8.0% | 45 | 9.7% |
| Hepatitis C antibodies | ||||||||
| Positive | 112 | 11.0% | 6 | 4.3% | 69 | 16.7% | 37 | 7.9% |
| Negative | 647 | 63.6% | 87 | 63.0% | 227 | 55.0% | 333 | 71.5% |
| Unknown | 258 | 25.4% | 45 | 32.6% | 117 | 28.3% | 96 | 20.6% |
| Hepatitis B S-antigen | ||||||||
| Positive | 32 | 3.1% | 2 | 1.4% | 14 | 3.4% | 16 | 3.4% |
| Negative | 728 | 71.6% | 91 | 65.9% | 282 | 68.3% | 355 | 76.2% |
| Unknown | 257 | 25.3% | 45 | 32.6% | 117 | 28.3% | 95 | 20.4% |
| CD4+ cell count at diagnosis availabled | 498 | 49.0% | 5 | 3.6% | 203 | 49.2% | 290 | 62.2% |
| ≥500 | 164 | 32.9% | 1 | 85 | 41.9% | 78 | 26.9% | |
| 350–499 | 90 | 18.1% | 3 | 39 | 19.2% | 48 | 16.6% | |
| 200–349 | 110 | 22.1% | 1 | 42 | 20.7% | 67 | 23.1% | |
| <200 | 134 | 26.9% | 37 | 18.2% | 97 | 33.4% | ||
| AIDSe | 154 | 15.1% | 30 | 21.7% | 61 | 14.8% | 63 | 13.5% |
| AIDS within 90 days of HIV | 72 | 7.1% | 3 | 2.2% | 29 | 7.0% | 40 | 8.6% |
| Deceased during follow-up | 63 | 6.2% | 18 | 13.0% | 33 | 8.0% | 12 | 2.6% |
aIf diagnosed before immigration, year of immigration. 1987–1997 includes 3 women diagnosed 1984–86
bIntravenous drug use
cOnly two of the vertical transmissions and one of the blood transfusion transmissions occurred in Finland
dMeasured in Finland 0–90 days after HIV diagnosis. Percentages reported of those with the CD4+ cell count at diagnosis available
eDiagnosed with an illness characteristic of AIDS (e.g., opportunistic infection). Does not include cases with CD4 + cell count < 200 in the absence of a concurrent AIDS-defining illness
Before HIV diagnosis, 160 women (15.7%) had 240 induced abortions, of which in 14 cases (5.8%) the HIV test was taken 0–10 days before the date of abortion (i.e., the HIV test was likely taken because of the planned abortion). After HIV diagnosis, 109 women (10.7%) had 156 induced abortions, of which in 20 cases (12.8%) the HIV diagnosis was made during the pregnancy based on the gestational age recorded at the time of the induced abortion.
The induced abortion rate during 1987–2019 was 25.1 abortions/1000 follow-up years. It decreased from 42.8 abortions/1000 follow-up years during 1987–1997 to 14.7 abortions/1000 follow-up years during 2009–2019, with the decrease more prominent in induced abortions after diagnosis (Fig. 2).
Fig. 2.
Induced abortion rate in the study population over time by year of follow-up (abortions/1000 follow-up years)
To better assess the effect of the HIV diagnosis on the rate of induced abortion, we also examined separately the 5 years before and after HIV diagnosis for those who had follow-up time both before and after the diagnosis (i.e., diagnosed in Finland aged 15–49 years; n = 689). For those diagnosed in 1987–1997, the rate of abortions was higher after the HIV diagnosis (51.2 compared to 35.4 abortions/1000 follow-up years) while for those diagnosed in 1998–2008 the rate was lower after diagnosis (24.9 compared to 45.0 abortions/1000 follow-up years), and for those diagnosed in 2009–2019, there was little change before and after diagnosis (Fig. 3).
Fig. 3.
Induced abortions/1000 follow-up years 5 years before and after HIV diagnosis by year of diagnosis. Includes only the WLWH with follow-up time both before and after HIV diagnosis (n = 689)
In the multivariable logistic regression model HIV diagnosis during or before pregnancy was associated with a decreased risk of induced abortion in 1998–2008 and 2009–2019 compared to 1987–1997 (Table 2). Increasing age was associated with a decreasing risk. In the model an increasing number of previous abortions or deliveries were both associated with an increased risk of induced abortion in women who were not born in Finland. Transmission route, hepatitis B and hepatitis C were not significant factors in any of the constructed models and are not included in the final models.
Table 2.
Multivariable logistic regression model for induced abortion vs delivery 1987–2019
| Variable | OR | 95% Cl | p-value | |
|---|---|---|---|---|
| lower limit | upper limit | |||
| Country of birth | ||||
| Other (reference Finland) | 1.00 | 0.64 | 1.55 | 0.988 |
| Age at conception (per year) | 0.96 | 0.93 | 0.99 | 0.006 |
| AIDS | ||||
| diagnosed during pregnancy | Too few cases | |||
| diagnosed before pregnancy | 2.61 | 0.97 | 7.03 | 0.057 |
| Number of previous induced abortions | ||||
| Born in Finland | 1.24 | 0.83 | 1.84 | 0.29 |
| Not born in Finland | 1.85 | 1.25 | 2.72 | 0.002 |
| Number of previous deliveries | ||||
| Born in Finland | 0.95 | 0.65 | 1.39 | 0.784 |
| Not born in Finland | 1.95 | 1.18 | 3.21 | 0.009 |
| Year of conception | ||||
| 1987–1997 (ref) | 1 | |||
| 1998–2008 | 1.39 | 0.89 | 2.17 | 0.153 |
| 2009–2019 | 0.67 | 0.32 | 1.42 | 0.300 |
| HIV diagnosis during pregnancya | ||||
| conception in: | ||||
| 1987–1997 | 1.08 | 0.36 | 3.23 | 0.889 |
| 1998–2008 | 0.13 | 0.05 | 0.35 | < 0.001 |
| 2009–2019 | 0.32 | 0.10 | 1.01 | 0.052 |
| HIV diagnosed before pregnancya | ||||
| conception in: | ||||
| 1987–1997 | 1.92 | 0.85 | 4.34 | 0.117 |
| 1998–2008 | 0.30 | 0.17 | 0.55 | < 0.001 |
| 2009–2019 | 0.34 | 0.15 | 0.76 | 0.008 |
| Constant (clustering by woman) | 2.03 | 0.94 | 4.38 | 0.07 |
Variables excluded from the model based on AIC, BIC and statistical non-significance p > 0.05: hepatitis B, hepatitis C, route of transmission. Interaction terms included in the model based on AIC, BIC and statistical significance p < 0.05: Relation of HIV diagnosis to period of conception, relations of country of birth to number of previous induced abortions and deliveries
aDiagnosis considered to be before pregnancy if diagnosis date (sampling date) more than 7 days prior to calculated conception date. If diagnosis date less than 7 days prior to conception date and either more than 10 days before abortion or at less than 23 (gestational) weeks, diagnosis was considered during pregnancy (could affect decision to terminate pregnancy). Consequently in case of giving birth, diagnosis ≥23 weeks was considered not known (= reference)
In analyzing factors associated with the risk of terminating a pregnancy specifically in pregnancies that began after HIV diagnosis in 1998–2019, the results were similar (Table 3). Not being born in Finland and previous induced abortions and pregnancies were associated with an increased risk of terminating a pregnancy, while increasing age was associated with a decreasing risk.
Table 3.
Multivariable logistic regression for factors associated with terminating a pregnancy conceived after HIV diagnosis 1998–2019
| Variable | OR | 95% Cl | p-value | |
|---|---|---|---|---|
| lower limit | upper limit | |||
| Country of birth | ||||
| Other (reference Finland) | 3.09 | 1.55 | 6.19 | 0.001 |
| Age at conception (per year) | 0.95 | 0.90 | 0.997 | 0.037 |
| AIDS | ||||
| diagnosed during pregnancy | Too few cases | |||
| diagnosed before pregnancy | 2.45 | 0.85 | 7.08 | 0.097 |
| Number of previous induced abortionsb | 3.36 | 1.80 | 6.28 | < 0.001 |
| N of prev. i. abortions2 [squared] b | 0.83 | 0.69 | 1.00 | 0.051 |
| Number of previous deliveriesb | 2.13 | 1.08 | 4.21 | 0.03 |
| N of prev. deliveries2 [squared] b | 0.84 | 0.66 | 1.08 | 0.178 |
| Constant (clustering by woman) | 0.27 | 0.06 | 1.33 | 0.108 |
Variables excluded from the model based on AIC, BIC and statistical non-significance p > 0.05: hepatitis B, hepatitis C and route of transmission
bCalculated increase in OR until 3–4 previous induced abortions and 2–3 previous deliveries after which the OR decreases
There were 68 women who had 88 induced abortions during 1987–2014 within 5 years before HIV diagnosis. During the same years, 304,985 induced abortions were performed in Finland, which makes the maximum prevalence of undiagnosed HIV 0.29/1000 abortions for those years. By using the CD4+ cell count thresholds, we estimated that of the 40 abortions where the woman had a CD4+ cell count at diagnosis available, HIV was undiagnosed at 11 abortions (28%). Assuming this percentage would hold true for the 88 abortions, the minimum prevalence of undiagnosed HIV would be 0.08/1000 induced abortions for 1987–2014 (Table 4).
Table 4.
Analysis for prevalence of undiagnosed HIV at induced abortions
| Assumed maximum time from transmission to symptoms that prompt HIV testing | |||
|---|---|---|---|
| 5 years | 10 years | 15 years | |
| Time considereda | 1987–2014 | 1987–2009 | 1987–2004 |
| Maximum estimate: | |||
| Induced abortions in Finlandb | 304,985 | 253,739 | 200,518 |
| WLWH with i.abortion before HIV diagnosis | 68 | 100 | 110 |
| Number of i.abortions before HIV diagnosis | 88 | 136 | 152 |
| Undiagnosed HIV/1000 induced abortions | 0.29 | 0.54 | 0.76 |
| Minimum estimate based on CD4+ cell count: | |||
| WLWH with CD4+ count at diagnosis availablec | 32 | 50 | 56 |
| Abortions with CD4+ count at diagnosis available | 40 | 65 | 71 |
| WLWH estimated undiagnosed at ≥1 abortiond | 10 | 19 | 20 |
| Abortions as undiagnosed based on CD4 count | 11 | 21 | 22 |
| Undiagnosed HIV/abortions with CD4+ count avail. | 28% | 32% | 31% |
| Undiagnosed HIV/1000 induced abortions | 0.08 | 0.17 | 0.23 |
| Estimate for foreign-born women: | |||
| Induced abortions among foreign-born in Finlande | 24,395 | 17,761 | 12,030 |
| Foreign-born WLWH with abortion | 25 | 28 | 23 |
| Number of abortions among foreign-born WLWH | 29 | 35 | 30 |
| Undiagnosed HIV/1000 induced abortions | 1.19 | 1.97 | 2.49 |
| Estimate for women with previous i. abortion: | |||
| Abortions in Finland with previous i. abortionsb | 97,956 | 79,118 | 60,229 |
| WLWH with prev. Abortion at the time of abortion | 31 | 41 | 43 |
| Abortions of WLWH with prev. abortions | 40 | 54 | 50 |
| Undiagnosed HIV/1000 induced abortions | 0.41 | 0.68 | 0.83 |
| Estimate for women with previous deliveries: | |||
| Abortions in Finland with previous deliveriesb | 150,647 | 125,691 | 100,403 |
| WLWH with abortion and prev. delivery in the data | 22 | 37 | 42 |
| Abortions of WLWH with prev. delivery in the data | 35 | 54 | 57 |
| Undiagnosed HIV/1000 induced abortions | 0.23 | 0.43 | 0.57 |
aTo ensure all who had an induced abortion would have been diagnosed before study end-point
bData from THL open data [19]
cMeasured in Finland 0–90 days after HIV diagnosis
dUsed thresholds based on Lodi et al., [18] CD4+ cell count at diagnosis → estimated time since transmission
for 5 year-limit: ≥500 → 6 months, 350–499 → 1 year, 200–349 → 3 years, < 200 → 5 years
for 10-year-limit: ≥500 → 1 year, 350–499 → 3 years, 200–349 → 8 years, < 200 → 10 years
for 15-year-limit: ≥ 500 → 2 years, 350–499 → 6 years, 200–349 → 12 years, < 200 → 15 years
eCalculated by roughly estimated proportions from [16, 20]: 8% for 5 year limit, 7% for 10 year limit, 6% for 15 year limit
The estimated maximum prevalence for women not born in Finland, women with previous induced abortions, and women with previous deliveries were 1.19, 0.41 and 0.23/1000 induced abortions, respectively. Of those women not born in Finland who had an induced abortion 0–5 years before HIV diagnosis, all but one were born in countries where the prevalence of HIV is considered high enough (> 1%) to warrant opt-out screening of HIV for refugees, asylum seekers and immigrants after arrival in Finland [21, 22].
Because both the time frame for development of symptomatic HIV infection prompting testing and the rate of decline for CD4+ cells are varied and uncertain, we did a sensitivity analysis by changing the time frame and thresholds to 10 and 15 years before HIV diagnosis and the results did not change significantly.
Discussion
Main findings
The rate of induced abortions among WLWH in Finland has decreased from 42.8 to 14.7/1000 follow-up years from 1987 to 1997 to 2009–2019, with the change more pronounced in abortions after HIV diagnosis. In our predictive regression model, after 1997 being diagnosed with HIV was not associated with an increased risk of terminating a pregnancy. Factors associated with terminating a pregnancy among WLWH in our model were younger age, country of birth other than Finland, and having previous induced abortions or deliveries. Estimated prevalence of undiagnosed HIV at induced abortions in Finland 1987–2014 was 0.008–0.029%.
Strengths and limitations
Strengths of the study are nationwide coverage and the use of register data over a long, 33-year follow-up period. The study population comprises all WLWH in Finland since the beginning of the HIV epidemic, and linkage between registers is accurate as the PIC is used as an identification in all health care contacts in Finland. The only WLWH missing from the study population are those with no valid PIC, who are mostly short-time visitors unlikely to contribute many events or much follow-up time in the data. The Register on Induced Abortions and Sterilizations was recently estimated to include 97% of the induced abortions [23]. Since it is easy to access induced abortion in public health care, there should be no underestimation due to clandestine abortions.
The major weakness is that only data from Finland were available. Abortions or deliveries before immigration are not included and for immigrants, the data are included only after receipt of a valid PIC (generally upon a permit to stay exceeding 1 year). The emigration (and to some extent, immigration) dates are incomplete especially for the earlier years of the follow-up. Because our data had no socioeconomic factors we considered relevant for determining reasons for having an induced abortion, the regression models were constructed for predictive aims only and causality of the factors included cannot be inferred from the models. Additionally, the study population consists of only 1017 women and so statistical modelling for some factors with several relevant variables is not accurate enough to examine small associations. Order of previous pregnancy outcomes or time since last pregnancy were not considered in model construction. Association of HIV diagnosis with risk of terminating an on-going pregnancy might be biased by the fact that there has been a national opt-out antenatal screening of all pregnant women planning to deliver since 1998, but HIV testing at induced abortions is not routine.
Interpretation
To our knowledge, this is the first nationwide study on induced abortions among WLWH covering over 30 years. The rates of induced abortions among WLWH both before and after HIV diagnosis are approaching the rate in the Finnish general population (7.7/1000 women of 15–49 years of age in 2019, steadily decreasing from 10.3 in 1987) [24]. The slightly higher rate among WLWH might be explained, besides differences in calculating time at risk, by the higher proportion of WLWH in follow-up at the age of 20–30 years and the higher proportion of immigrants, as they are at higher risk than Finnish-born individuals [16]. The decrease in induced abortion rate was similar to the decrease seen in an Italian cohort from 1980s to 2010 [15] but differed from an US cohort where no change was seen in 1994–1997 to 2006–2012 [10].
The decrease in the rate of induced abortions was more pronounced after HIV diagnosis and after 1997 being diagnosed with HIV was associated with a decreased risk of terminating a pregnancy in our predictive model. This might reflect the trust in diminished risk of vertical transmission of the women and their caregivers, as well as the possibilities of WLWH to discuss their reproductive plans and contraception with their caregivers.
Of the WLWH in Finland, 11% have had an induced abortion after being diagnosed, a proportion similar to some previous studies [15, 25, 26]. Similar to our study, younger age and previous pregnancies have been associated with an increased risk of induced abortion in two previous cohort studies among WLWH in Italy and the US [10, 15]. In contrast to these studies where migration status or ethnicity were not significantly associated with the risk of induced abortion, in our model not being born in Finland was associated with an increased risk. This is however consistent with the Finnish general population, where first-generation immigrants had a higher risk of induced abortion than Finnish-born women in a recent national register study [16].
The estimated prevalence of undiagnosed HIV at induced abortion was 0.08–0.29/1000 abortions (0.008–0.029%) during 1987–2014. A prevalence of 0.1–1.0% is generally considered cost effective for testing, though some have suggested a prevalence as low as 0.05% [27–29]. Therefore, although the prevalence of the most recent years cannot be determined for certain without implementing routine testing, it is reasonable to assume that routine testing of all women having an induced abortion in Finland is not cost-effective if the limit is set at 0.1%. This is in contrast with estimates for some other high-income countries, where the HIV prevalence is higher and routine testing at induced abortions may be recommendable based on the same criteria [30]. However, in Finland the prevalence of undiagnosed HIV is likely higher among women not born in Finland (0.12%) and more specifically, women born in countries with a high HIV prevalence; an induced abortion would therefore be an event that could prompt offering HIV testing to those with pre-existing risk factors, including country of birth. The prevalence of undiagnosed HIV for those with previous induced abortions or deliveries is not high enough to warrant routine testing when using the 0.1% limit.
Conclusions
The rate of induced abortions among WLWH in Finland has decreased markedly in three decades, approaching the rate among the general population. In our model, being diagnosed with HIV was not associated with an increased risk of terminating a pregnancy after effective cART to prevent vertical transmission became available. However, since every woman living with HIV in Finland meets her doctor once or twice a year with the possibility to discuss family plans and contraception, one might assume the rate of induced abortions among WLWH to be even lower. Caregivers in Finland should remember to discuss the wishes of WLWH regarding family planning at every follow-up appointment, and especially with younger women, women with previous pregnancies, and women who were not born in Finland.
Routine testing of all women at induced abortion is not likely to be cost-effective in Finland due to low HIV prevalence, but testing could be offered to women with pre-existing risk factors, such as being born in a country with a high HIV prevalence.
Acknowledgements
Not applicable.
Abbreviations
- WLWH
Women living with HIV
- ART
Antiretroviral treatment
- cART
Combined antiretroviral treatment
- IDU
Intravenous drug use
- PIC
Personal identification code
- FINHIV
Finnish HIV Quality of Care Register
- THL
Finnish Institute for Health and Welfare
Authors’ contributions
All authors took part in planning the study. JO and MG collected and combined the datasets. MM and JO analyzed the data with IA, PK and KL participating in the interpretation of the results. MM wrote the initial draft for the article with significant editing by IA and PK. All authors revised and approved the draft.
Funding
Open Access funding provided by University of Helsinki including Helsinki University Central Hospital. This study was partly funded with a research grant from Gilead Nordic Fellowship Programme. The funder was not involved in planning, conducting or reporting of the study.
Availability of data and materials
The data that support the findings of this study are available from Data Permit Authority Findata that works in conjunction with the Finnish Institute for Health and Welfare but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. For more information on how the data may be requested please contact the corresponding author (mikaela.mutru@helsinki.fi).
Declarations
Ethics approval and consent to participate
The study protocol and methods were approved by the Finnish Institute for Health and Welfare’s Data and Analytics Unit (THL/1572/6.02.00/2019). A review from an ethics committee is not required for research on humans that uses only register data according to Finnish national guidelines (Finnish National Board on Research Integrity TENK: The national ethical principles of research with human participants and ethical review in the human sciences in Finland, https://tenk.fi/sites/tenk.fi/files/Ihmistieteiden_eettisen_ennakkoarvioinnin_ohje_2019.pdf [31]).
For this study, informed consent is deemed unnecessary in accordance with national legislation that permits the use of retrospective health register data for scientific research without informed consent provided that appropriate measures to protect personal data are taken (Act on Secondary Use of Health and Social Data, 552/2019, https://www.finlex.fi/fi/laki/smur/2019/20190552 [32] and Act on the Institute for Health and Welfare, 668/2008, https://www.finlex.fi/fi/laki/ajantasa/2008/20080668 [33]). All methods were performed in accordance with the ethical guidelines of the Finnish National Board on Research Integrity, national legislation and the institutional guidelines of the Finnish Institute for Health and Welfare.
Consent for publication
Not applicable.
Competing interests
The authors declare no conflicts of interest directly related to the study. PK and IA have received personal fees for Advisory Board participation, lectures and travel expenses from GSK, Gilead and Merck.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Mikaela Mutru, Email: mikaela.mutru@helsinki.fi.
Pia Kivelä, Email: pia.kivela@hus.fi.
Jukka Ollgren, Email: jukka.ollgren@thl.fi.
Kirsi Liitsola, Email: kirsi.liitsola@thl.fi.
Mika Gissler, Email: mika.gissler@thl.fi.
Inka Aho, Email: inka.aho@hus.fi.
References
- 1.van Benthem BHB, de Vincenzi I, Delmas MC, Larsen C, van den Hoek A, Prins M. 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(14):2171–2178. doi: 10.1097/00002030-200009290-00014. [DOI] [PubMed] [Google Scholar]
- 2.Hankins C, Tran T, Lapointe N, Canadian Women’s HIV Study Group Sexual behavior and pregnancy outcome in HIV-infected women. Canadian Women’s HIV study group. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;18(5):479–487. doi: 10.1097/00042560-199808150-00010. [DOI] [PubMed] [Google Scholar]
- 3.Thorne C, Boer K, England K, Godfried MH, Newell ML, Mahdavi S, et al. Mode of delivery in HIV-infected pregnant women and prevention of mother-to-child transmission: changing practices in Western Europe. HIV Med. 2010;11(6):368–378. doi: 10.1111/j.1468-1293.2009.00800.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Duong T, Ades AE, Gibb DM, Tookey PA, Masters J. Vertical transmission rates for HIV in the British Isles: estimates based on surveillance data. BMJ. 1999;319:1227–1229. doi: 10.1136/bmj.319.7219.1227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Warszawski J, Tubiana R, le Chenadec J, Blanche S, Teglas JP, Dollfus C, et al. Mother-to-child HIV transmission despite antiretroviral therapy in the ANRS French Perinatal Cohort. AIDS . 2008;22(2):289–99. [DOI] [PubMed]
- 6.Townsend CL, Byrne L, Cortina-Borja M, Thorne C, de Ruiter A, Lyall H, et al. Earlier initiation of ART and further decline in mother-to-child HIV transmission rates, 2000-2011. AIDS. 2014;28(7):1049–1057. doi: 10.1097/QAD.0000000000000212. [DOI] [PubMed] [Google Scholar]
- 7.Aho I, Kivelä P, Kaijomaa M, Surcel HM, Ristola M, Heikinheimo O, et al. Comprehensive nationwide analysis of mother-to-child HIV transmission in Finland from 1983 to 2013. Epidemiol Infect. 2018;146(10):1301–1307. doi: 10.1017/S0950268818001280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mårdh O, Quinten C, Kuchukhidze G, Seguy N, Dara M, Amato-Gauci AJ, et al. HIV among women in the WHO European region – epidemiological trends and predictors of late diagnosis, 2009-2018. Euro Surveill. 2019;24(48):1900696. 10.2807/1560-7917.ES.2019.24.48.1900696. [DOI] [PMC free article] [PubMed]
- 9.Kowalska JD, Pelchen-Matthews A, Ryom L, Losso MH, Trofimova T, Mitsura VM, et al. Prevalence and outcomes of pregnancies in women with HIV over a 20-year period. AIDS. 2021;35(12):2025–2033. doi: 10.1097/QAD.0000000000002954. [DOI] [PubMed] [Google Scholar]
- 10.Haddad LB, Wall KM, Mehta CC, Golub ET, Rahangdale L, Kempf MC, et al. American Journal of Obstetrics and Gynecology. Mosby Inc.; 2017. Trends of and factors associated with live-birth and abortion rates among HIV-positive and HIV-negative women; pp. 71.e1–71.e16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Aebi-Popp K, Mercanti V, Voide C, Nemeth J, Cusini A, Jakopp B, et al. Neglect of attention to reproductive health in women with HIV infection: contraceptive use and unintended pregnancies in the Swiss HIV cohort study. HIV Med. 2018;19(5):339–346. doi: 10.1111/hiv.12582. [DOI] [PubMed] [Google Scholar]
- 12.Jokela S, Lilja E, Kinnunen TI, Gissler M, Castaneda AE, Koponen P. Births and induced abortions among women of Russian, Somali and Kurdish origin, and the general population in Finland –comparison of self-reported and register data. BMC Pregnancy Childbirth. 2018;18(1):296. 10.1186/s12884-018-1931-x. [DOI] [PMC free article] [PubMed]
- 13.Scott RH, Bajos N, Wellings K, Slaymaker E. Comparing reporting of abortions in three nationally representative surveys: methodological and contextual influences. BMJ Sex Reprod Health. 2019;45(3):213–219. doi: 10.1136/bmjsrh-2019-200321. [DOI] [PubMed] [Google Scholar]
- 14.Lindberg L, Kost K, Maddow-Zimet I, Desai S, Zolna M. Abortion reporting in the United States: an assessment of three National Fertility Surveys. Demography. 2020;57(3):899–925. doi: 10.1007/s13524-020-00886-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ammassari A, Cicconi P, Ladisa N, di Sora F, Bini T, Trotta MP, et al. Induced first abortion rates before and after HIV diagnosis: results of an Italian self-administered questionnaire survey carried out in 585 women living with HIV. HIV Med. 2013;14(1):31–39. doi: 10.1111/j.1468-1293.2012.01032.x. [DOI] [PubMed] [Google Scholar]
- 16.Heino AE, Gissler M, Malin M, VÄisÄnen H. Induced abortions by woman’s country of origin in Finland 2001–2014. Scand J Public Health. 2020;48(1):88–95. doi: 10.1177/1403494818812640. [DOI] [PubMed] [Google Scholar]
- 17.Mutru M, Isosomppi S, Aho I, Liitsola K, Brummer-Korvenkontio H, Ollgren J, et al. Finnish HIV Quality of Care Register (FINHIV). BMJ Open. 2022;12(1):e053287. 10.1136/bmjopen-2021-053287. [DOI] [PMC free article] [PubMed]
- 18.Lodi S, Phillips A, Touloumi G, Geskus R, Meyer L, Thiébaut R, et al. Time from human immunodeficiency virus seroconversion to reaching CD4+ cell count thresholds <200, <350, and <500 cells/mm3: assessment of need following changes in treatment guidelines. Clin Infect Dis. 2011;53(8):817–825. doi: 10.1093/cid/cir494. [DOI] [PubMed] [Google Scholar]
- 19.Data cube: Induced abortions, whole country. Finnish Institute for Health and Welfare (THL). 2020. https://sampo.thl.fi/pivot/prod/fi/ab/kokomaa/fact_ab_kokomaa. Accessed 22 Apr 2022.
- 20.Malin M, Gissler M. Induced abortions among immigrant women in Finland. Finnish J Ethnicity Migration. 2008;1(3):2–12. [Google Scholar]
- 21.Finnish Institute for Health and Welfare, Ministry of Social Affairs and Health. Pakolaisten ja turvapaikanhakijoiden infektio-ongelmien ehkäisy [prevention of infection problems among refugees and asylum seekers]. Reports of the Ministry of Social Affairs and Health. 2009. p. 21. https://julkaisut.valtioneuvosto.fi/handle/10024/73989. Accessed 1 June 2022.
- 22.Finnish Institute for Health and Welfare. Maahanmuuttajien tartuntatautien seulonta sekä lasten tuberkuloosi-ja hepatiitti B-rokotukset. 2022. https://thl.fi/documents/533963/6240821/Maaluettelo+17.3.2022.pdf/b32df567-75d4-a3a8-73f0-03d31f658c96?t=1648637058846. Accessed 1 May 2022.
- 23.Heino A, Niinimäki M, Mentula M, Gissler M. How reliable are health registers? Registration of induced abortions and sterilizations in Finland. Inform Health Soc Care. 2018;43(3):310–319. doi: 10.1080/17538157.2017.1297306. [DOI] [PubMed] [Google Scholar]
- 24.Heino A, Gissler M. Statistical report 10/2021. Induced abortions in the Nordic countries 2019. Appendix table 3. Induced abortions in the Nordic countries per 1000 women aged 15–49 years 1970–2017. 2021. https://www.julkari.fi/handle/10024/141156. Accessed 1 June 2022.
- 25.Wessman M, Aho I, Thorsteinsson K, Storgaard M, Johansen IS, Lunding S, et al. Perception of sexuality and fertility in women living with HIV: a questionnaire study from two Nordic countries. J Int AIDS Soc . 2015;18(1):19962. 10.7448/IAS.18.1.19962. [DOI] [PMC free article] [PubMed]
- 26.Hachfeld A, Atkinson A, Calmy A, de Tejada BM, Hasse B, Paioni P, et al. Decrease of condom use in heterosexual couples and its impact on pregnancy rates: the Swiss HIV cohort study (SHCS) HIV Med. 2022;23(1):60–69. doi: 10.1111/hiv.13152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yazdanpanah Y, Sloan CE, Charlois-Ou C, le Vu S, Semaille C, Costagliola D, et al. Routine HIV screening in France: clinical impact and cost-effectiveness. PLoS One. 2010;5(10):e13132. 10.1371/journal.pone.0013132. [DOI] [PMC free article] [PubMed]
- 28.Sanders GD, Bayoumi AM, Sundaram V, Bilir SP, Neukermans CP, Rydzak CE, et al. Cost-effectiveness of screening for HIV in the era of highly active antiretroviral therapy. N Engl J Med. 2005;352(6):570–85. 10.1056/NEJMsa042657. [DOI] [PubMed]
- 29.Paltiel AD, Weinstein MC, Kimmel AD, Seage GR, Losina E, Zhang H, et al. Expanded screening for HIV in the United States-an analysis of cost-effectiveness. N Engl J Med. 2005;352(6):586–95. 10.1056/NEJMsa042088. [DOI] [PubMed]
- 30.Filippidis P, Francini K, Jacot-Guillarmod M, Mathevet P, Lhopitallier L, Cavassini M, et al. HIV testing in termination of pregnancy and colposcopy services: a scoping review. Sexually transmitted infections. NLM (Medline) 2022;98:143–149. doi: 10.1136/sextrans-2021-055111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Finnish National Board on Research Integrity TENK. The national ethical principles of research with human participants and ethical review in the human sciences in Finland. 2019. https://tenk.fi/sites/tenk.fi/files/Ihmistieteiden_eettisen_ennakkoarvioinnin_ohje_2019.pdf. Accessed 4 Nov 2022.
- 32.Laki sosiaali- ja terveystietojen toissijaisesta käytöstä [Act on Secondary Use of Health and Social Data]. 552/2019. 2019. https://www.finlex.fi/fi/laki/smur/2019/20190552. Accessed 4 Nov 2022.
- 33.Laki Terveyden ja hyvinvoinnin laitoksesta [Act on the Institute for Health and Welfare]. 668/2008. 2008. https://www.finlex.fi/fi/laki/ajantasa/2008/20080668. Accessed 4 Nov 2022.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from Data Permit Authority Findata that works in conjunction with the Finnish Institute for Health and Welfare but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. For more information on how the data may be requested please contact the corresponding author (mikaela.mutru@helsinki.fi).