Medical and surgical abortion for women living with HIV

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

Our primary objective is to assess the effectiveness and safety of medical versus surgical abortion among women living with HIV. Our secondary objectives are to: (1) assess abortion outcomes for both methods between women living with HIV and HIV‐negative women and (2) assess abortion outcomes among women living with HIV.

Background

Description of the condition

The World Health Organization (WHO) guidelines for safe abortion recommend medical interventions for abortion with mifepristone and misoprostol or surgical interventions for the termination of pregnancy with vacuum aspiration or dilation and evacuation (the latter for pregnancies greater than 12 to 14 weeks) as safe and effective options (WHO 2012). However, there are no specific clinical considerations stipulated within these guidelines for women living with HIV. Although studies have demonstrated no obstetric surgical postoperative differences following obstetric surgery in women due to HIV sero‐status (Cavasin 2009; Sekirime 2009), women living with HIV could be at greater risk for anaemia and bleeding from surgical abortion. Vomiting from drugs used for medical abortion could also reduce the efficacy of antiretroviral drugs for HIV treatment (de Bruyn 2003). Hospitalisation, haemorrhage requiring blood transfusion, and major operative interventions resulting from abortion complications could also add increased risk, stress, and costs to women undergoing an abortion.

Description of the intervention

We define medical and surgical abortion according to the WHO Safe Abortion Guidelines (WHO 2012):

Medical abortion methods: use of pharmacological drugs to terminate pregnancy. Sometimes the terms 'non‐surgical abortion' or 'medication abortion' are also used.

Surgical abortion methods: use of transcervical procedures for terminating pregnancy, including vacuum aspiration and dilation and evacuation.

We considered medical and surgical abortion, both in the first and second trimester of pregnancy.

How the intervention might work

Manski and colleagues conducted a systematic review of clinical outcomes of surgical and medical abortion among women living with HIV (Manski 2012). Searching for peer reviewed articles published through 2011, they found few studies on surgical abortion and no studies on medical abortion among this population. The studies they did identify mostly covered illegal or non‐specified abortion, which may or may not have met WHO Safe Abortion Guidelines (WHO 2012), but suggested no significant differences in complications by HIV status. They then reviewed data from gynaecological procedures that were similar to or more invasive than abortion and again found no differences in complications by HIV status. Finally, they reviewed data on changes in haemoglobin levels due to blood loss from medical abortion, where anaemia is prevalent, and frequency and duration of vomiting due to medical abortion drugs; both analyses suggested these were relatively minor concerns. The Manski 2012 review concluded, based on the limited available data, that both medical and surgical abortions are safe and appropriate for women living with HIV.

Why it is important to do this review

In order to inform an update of the WHO's Safe Abortion: Technical and Policy Guidance for Health Systems that will include guidance on women living with HIV (WHO 2012), and build upon the global community survey on the sexual and reproductive health priorities of women living with HIV (Salamander Trust 2014), we seek to expand the work in Manski 2012 by (1) updating the search using a rigorous systematic review process (Higgins 2011), (2) searching conference abstracts and trial registries (in addition to peer reviewed articles), (3) searching for articles in languages other than English, and (4) contacting authors to identify unpublished datasets, with information on both abortion outcomes and HIV, that could potentially be used to examine whether medical and surgical abortion are effective and safe for women living with HIV.

Objectives

Our primary objective is to assess the effectiveness and safety of medical versus surgical abortion among women living with HIV. Our secondary objectives are to: (1) assess abortion outcomes for both methods between women living with HIV and HIV‐negative women and (2) assess abortion outcomes among women living with HIV.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomised controlled trials (RCTs), non‐RCTs, and observational (e.g. cohort) studies. Per our objectives, we will include: (1) studies on the effectiveness and safety of medical versus surgical abortion among women living with HIV; (2) studies examining abortion outcomes for both methods between women living with HIV and HIV‐negative women; and (3) studies that examine the health outcomes of abortion among women living with HIV.

Types of participants

Pregnant women living with HIV (with known HIV status) and HIV‐negative pregnant women seeking abortion services, compliant with WHO clinical practice guidelines for safe abortions (WHO 2012). There will be no exclusions.

Types of interventions

Medical and surgical abortion, as defined by the WHO Safe Abortion Guidelines (WHO 2012):

Medical abortion methods: use of pharmacological drugs to terminate pregnancy. Sometimes the terms 'non‐surgical abortion' or 'medication abortion' are also used.

Surgical abortion methods: use of transcervical procedures for terminating pregnancy, including vacuum aspiration and dilation and evacuation.

We considered medical and surgical abortion in both the first and second trimester of pregnancy.

Types of outcome measures

Primary outcomes

1. Efficacy (complete abortion).

2. Serious adverse events: a) death, b) hospitalisation, c) haemorrhage requiring blood transfusion, d) major operative intervention (e.g. laparotomy).

Secondary outcomes

1. Other adverse outcomes and side effects: a) haemorrhage not requiring blood transfusion, b) vomiting, c) infection.

2. Patient satisfaction, assessed using a validated survey questionnaire, such as the Patient Satisfaction Questionnaire‐18 (PSQ‐18) (Marshall 1994), or Treatment Satisfaction Questionnaire for Medication (TSQM) (Atkinson 2004).

Search methods for identification of studies

We will search for all published and unpublished trials and observational studies of medical and surgical abortion among pregnant women living with HIV seeking services for the termination of pregnancy, irrespective of when they received their HIV diagnosis and irrespective of CD4 count or other health status measures, without restrictions based on language or intervention location.

Electronic searches

We will search the following electronic databases for relevant studies.

1. The Cochrane Central Register of Controlled Trials (via CENTRAL Register of Studies Online (CRSO) (from inception to present) (Appendix 1).

2. MEDLINE Ovid (1946 to present) (Appendix 2).

3. Embase Ovid (1980 to present) (Appendix 3).

4. PsycINFO Ovid (1806 to present) (Appendix 4).

5. CINAHL (Cumulative Index to Nursing and Allied Health Literature) Ebsco (from 1961 to present) (Appendix 5).

6. WHO International Clinical Trials Registry Platform (www.who.int/ictrp; to date of search). (Appendix 6)

7. ClinicalTrials.gov (www.clinicaltrials.gov; to date of search). (Appendix 6)

• The MEDLINE search was combined with the Cochrane highly sensitive search strategy for identifying randomized trials which appears in the Cochrane Handbook of Systematic Reviews of Interventions (Version 5.0.2 chapter 6, 6.4.11)

• The EMBASE and CINAHL searches are combined with trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN) http://www.sign.ac.uk/search‐filters.html

• The PsycINFO search filter was developed by Health Information research unit – McMaster University; http://hiru.mcmaster.ca/hiru/HIRU_Hedges_PsycINFO_Strategies.aspx

Our search strategies use combinations of terms for abortion and HIV.

Searching other resources

In order to obtain additional data, we will also conduct secondary reference searching on all included studies, as well as the Manski 2012 review.

We will handsearch the following conference websites for available abstracts: International AIDS Conference (IAC), IAS Conference on HIV Pathogenesis, Treatment, and Prevention (IAS), the Conference on Retroviruses and Opportunistic Infections (CROI), the International Conference on AIDS and STIs in Africa (ICASA), International Federation of Gynecology and Obstetrics (FIGO) World Conference, and the American Congress of Obstetricians and Gynecologists (ACOG).

We will also search for unpublished studies, existing cohort studies and trials that might have collected data on medical or surgical abortion outcomes and HIV status, but had not analysed their data by HIV status. To identify such studies, we will review the reference lists of two systematic reviews on abortion outcomes (Kulier 2011; Say 2005). We will contact the corresponding authors to see if they could disaggregate their data on abortion outcomes by HIV status of participants. We also will ask these experts to identify any additional studies that we may have missed.

Data collection and analysis

Selection of studies

A member of the study team (TP) will screen titles, abstracts, citation information, and descriptor terms of citations initially identified through the search strategy. We will obtain full text articles of all selected potentially eligible abstracts studies. Two review authors (HTS, CEK) will independently examine these full text articles for compliance with the inclusion criteria and determination of final study selection. We will resolve differences through consensus. We will document this selection process through a PRISMA flow chart (Moher 2009).

Data extraction and management

Two review authors (HTS, CEK) will independently extract data using standardised data extraction forms, with differences resolved through consensus. Data extraction forms will collect information on study objectives; location; population characteristics (including time since HIV diagnosis, CD4 count/HIV clinical stage, antiretroviral and other related treatment/medication use, gestational age or trimester, or both, and abortion history); abortion method; study design; sample size; follow‐up periods; loss to follow‐up; analytic approach; outcome measures; number of participants in each comparison group; effect sizes; confidence intervals; significance levels; conclusions; limitations.

We will attempt to contact study authors when insufficient information is presented on methods or results, or both.

Assessment of risk of bias in included studies

For randomised controlled trials (RCTs), two review authors (HTS, CEK) will assess the risk of bias using the Cochrane tool for assessing risk of bias (Higgins 2011). This tool assesses random sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data addressed (attrition bias), and selective reporting (reporting bias). We will assess and classify methodological components of the studies as high or low risk of bias. Examples of study elements that could constitute high risk of bias in this review include: differences in how outcomes were measured or determined, particularly for subjective measures, such as patient satisfaction; non‐reporting of insignificant differences in outcomes between comparison groups (even if measured in the original study); differences in how abortion services were provided, particularly if health providers became aware of the HIV status of their patients from physical markers; and others. We will assess studies that do not explicitly state that abortion procedures followed WHO Safe Abortion Guidelines (WHO 2012), as high risk of other bias. For observational studies using different designs, we will use an adapted version of the Newcastle‐Ottawa scale to assess study quality (Wells 2009).

Measures of treatment effect

We will consider all effect measures reported by individual studies; outcome measures are not part of the inclusion criteria for the review. Measures of treatment effect may include risk ratios, odds ratios, mean difference, prevalence, or simple descriptive statistics. If sufficient, comparable outcome data exist to conduct meta‐analysis for any of the primary or secondary outcomes, we will start our analysis with fixed‐effect models. Since our primary outcomes are biological, we anticipate the same underlying mechanism across studies. However, if we detect high heterogeneity, we will shift to a random‐effects meta‐analysis and potentially present both analyses, to include Mantel‐Haenszel odds ratios and I² statistics to assess heterogeneity. We will follow Cochrane suggestions for thresholds for I² (Higgins 2011):

• 0% to 40% might not be important;

• 30% to 60% may represent moderate heterogeneity;

• 50% to 90% may represent substantial heterogeneity;

• 75% to 100% is considerable heterogeneity.

Unit of analysis issues

If we identify RCTs, we will consider the unit of analysis as 'per woman randomised'. Given the nature of the intervention, we do not anticipate cross‐over designs or cluster‐randomised trials. If individual studies present repeated observations for participants (in this case, multiple pregnancies for an individual woman), we will include only the first instance of termination of pregnancy for each participant and conduct separate analyses on second, third, and fourth pregnancies.

Dealing with missing data

If we identify missing data, we will attempt to contact the study authors to obtain more information. As data allow, we will consider sensitivity analyses to remove studies with significant amounts of missing data. For studies in which data are missing or incomplete, we will report the findings as 'unclear', or 'high risk' if the missing data point to potential risk of bias; for example, if a study has more data missing in one arm than another. We will document reasons for these final judgements.

Assessment of heterogeneity

We will assess heterogeneity in meta‐analyses using the I² statistic.

Assessment of reporting biases

Because detecting and correcting for publication bias and other reporting biases is difficult, we will minimise their potential impact by doing a comprehensive search for both published and unpublished studies and contact authors for datasets that might be used to answer the research questions. We will consider using funnel plots to assess for reporting biases if ten or more eligible studies are included in the review.

Data synthesis

We will analyse data by aim and by outcome, as data allow, using Review Manager software (RevMan 2014). We will conduct a meta‐analysis if a sufficient number of studies (at least 3) address the same research question and present data on sufficiently comparable outcomes. We will combine effect sizes in odds ratios using the Mantel‐Haentzel method in meta‐analysis.

Medical versus surgical abortion in HIV positive women

Any method of abortion in HIV positive versus HIV negative women

Subgroup analysis and investigation of heterogeneity

If sufficient data exist, we will conduct subgroup analyses based on world region and country income group (according to World Bank country classifications of low, lower‐middle, upper‐middle, and high‐income countries) ( World Bank 2017).

Sensitivity analysis

We will conduct sensitivity analyses by removing studies with high risk of bias (including studies with significant amounts of missing data) and test results using a risk ratio instead of an odds ratio.

Overall quality of the body of evidence: 'Summary of findings' table

We will prepare two 'Summary of findings' tables using GRADEpro GDT and Cochrane methods (GRADEpro GDT 2015; Guyatt 2008). The first table will evaluate the overall quality of the body of evidence for the primary and secondary outcomes, including all adverse events, comparing women living with HIV to HIV‐negative women who have had either a medical or surgical abortion. The second table will evaluate the overall quality of the body of evidence for primary and secondary outcomes of medical versus surgical abortions among women living with HIV. We will assess the quality of the evidence using GRADE criteria: risk of bias, consistency of effect, imprecision, indirectness, and publication bias). Two review authors working independently (HTS, CEK), will make judgements about evidence quality (high, moderate, low or very low), and we will resolve disagreements by discussion. We will justify, document, and incorporate judgements into reporting of results for each outcome.

We plan to extract study data, format our comparisons in data and analyses tables, and prepare a 'Summary of findings' table before writing the results and conclusions of our review.

Appendices

Appendix 1. CENTRAL Register of Studies Online (CRSO) search strategy

Web platform

Searched from inception to present

#1 MESH DESCRIPTOR Abortion, Induced EXPLODE ALL TREES

#2 abort*:TI,AB,KY

#3 (terminat* adj3 pregnan*):TI,AB,KY

#4 (menstrual regulation):TI,AB,KY

#5 #1 OR #2 OR #3 OR #4

#6 MESH DESCRIPTOR HIV Infections EXPLODE ALL TREES

#7 (immunodeficiency syndrome*):TI,AB,KY

#8 (human immunodeficiency):TI,AB,KY

#9 (human t cell*):TI,AB,KY

#10 HIV:TI,AB,KY

#11 (immunodeficiency virus*):TI,AB,KY

#12 (acquired immune deficiency):TI,AB,KY

#13 (immun* deficiency syndrome*):TI,AB,KY

#14 (Acquired Immunodeficienc*):TI,AB,KY

#15 (T‐Lymphotropic Virus Type III):TI,AB,KY

#16 AIDS:TI,AB,KY

#17 #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16

#18 #5 AND #17

Appendix 2. MEDLINE search strategy

Ovid platform

Searched from 1946 to present

1 exp Abortion, Legal/ or exp Abortion, Induced/ 2 exp Abortion, Therapeutic/ 3 abort*.tw. 4 (terminat* adj3 pregnan*).tw. 5 menstrual regulation.tw. 6 or/1‐5 7 exp HIV‐2/ or exp HIV/ or exp HIV‐1/ 8 exp Acquired Immunodeficiency Syndrome/ 9 immunodeficiency syndrome*.tw. 10 human immunodeficiency.tw. 11 human t cell leukemia.tw. 12 human t cell lymphotropic.tw. 13 HIV.tw. 14 immunodeficiency virus*.tw. 15 acquired immune deficiency.tw. 16 immun* deficiency syndrome*.tw. 17 Acquired Immunodeficienc*.tw. 18 T‐Lymphotropic Virus Type III.tw. 19 AIDS.tw. 20 or/7‐19 21 exp animals/ not humans.sh. 22 6 and 20 23 22 not 21

Appendix 3. Embase search strategy

Ovid platform

Searched from 1980 to present

1 exp HIV‐2/ or exp HIV/ or exp HIV‐1/ 2 immunodeficiency syndrome*.tw. 3 human immunodeficiency.tw. 4 human t cell leukemia.tw. 5 human t cell lymphotropic.tw. 6 HIV.tw. 7 immunodeficiency virus*.tw. 8 acquired immune deficiency.tw. 9 immun* deficiency syndrome*.tw. 10 Acquired Immunodeficienc*.tw. 11 T‐Lymphotropic Virus Type III.tw. 12 AIDS.tw. 13 exp acquired immune deficiency syndrome/ 14 or/1‐13 15 exp induced abortion/ or exp medical abortion/ or exp abortion/ or exp surgical abortion/ or exp therapeutic abortion/ 16 abort*.tw. 17 (terminat* adj3 pregnan*).tw. 18 menstrual regulation.tw. 19 or/15‐18 20 14 and 19 21 (exp animal/ or animal.hw. or nonhuman/) not (exp human/ or human cell/ or (human or humans).ti.) 22 20 not 21

Appendix 4. PsycINFO search strategy

Ovid platform

Searched from 1806 to present

1 exp INDUCED ABORTION/ 2 abort*.tw. 3 (terminat* adj3 pregnan*).tw. 4 menstrual regulation.tw. 5 or/1‐4 6 exp HIV/ 7 exp AIDS/ 8 immunodeficiency syndrome*.tw. 9 human immunodeficiency.tw. 10 HIV.tw. 11 AIDS.tw. 12 acquired immune deficiency.tw. 13 Acquired Immunodeficienc*.tw. 14 or/6‐13 15 5 and 14

Appendix 5. CINAHL search strategy

Ebsco platform

Searched from 1961 to present

#	Query
S20	S5 AND S19
S19	S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18
S18	TX AIDS
S17	TX T‐Lymphotropic Virus Type III
S16	TX Acquired Immunodeficienc*
S15	TX immun* deficiency syndrome*
S14	TX acquired immune deficiency
S13	TX immunodeficiency virus*
S12	TX HIV
S11	TX human t cell lymphotropic
S10	TX human t cell leukemia
S9	TX human immunodeficiency
S8	TX immunodeficiency syndrome*
S7	(MM "Human Immunodeficiency Virus+") OR (MM "Acquired Immunodeficiency Syndrome")
S6	(MM "HIV Infections+") OR (MM "HIV‐Infected Patients+") OR (MM "HIV‐1")
S5	S1 OR S2 OR S3 OR S4
S4	TX menstrual regulation
S3	TX terminat* N3 pregnan*
S2	TX abort*
S1	(MM "Abortion, Induced+")

Appendix 6. Trials Registries search strategy

From inception to present

Web platforms

“HIV AND abortion”; “AIDS AND abortion”

Contributions of authors

CK, MN and BG developed the initial protocol, HS provided critical feedback on preliminary drafts and all authors approved the final manuscript.

Sources of support

Internal sources

• World Health Organization, Department of Reproductive Health and Research, Switzerland.

  Funding

External sources

• None, Other.

Declarations of interest

HTS, MN, BG and CEK have no conflict of interest to declare. The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of their respective institutions.

New