CHAPTER 2 Unsafe Abortion: The Global Public Health Challenge

Unsafe abortion: The global public health challengeand largely inaccessible, or legal but difficult to obtain.Little information is available on abortion practice in thesecircumstances, and abortions tend to be unreported or vastlyunderreported. Moreover, clandestine induced abortionsmay be misreported as spontaneous abortion (miscarriage)[13,14]. The language used to describe induced abortion re flects this ambivalence: terms include "induced miscarriage"(jausse couche provoquee) [15], "menstrual regulation," and"regulation of a delayed or suspended menstruation [16]."In spite of these challenges, estimates of the frequency ofunsafe abortion can be made mainly by using hospital dataon abortion complications or abortion data from surveysand validated against the legal context of induced abortion,contraceptive prevalence, and total fertility rate (the averagenumber of children a woman is likely to have by the end ofher reproductive years).Globally, WHO estimates that some 19 to 20 million un safe abortions occurred each year between 1993 and 2003[7]. This figure has remained relatively constant despite anincrease in contraceptive prevalence during the same period.Although the transition to low fertility with smaller familieshas become a norm in most countries, family planning hasnot been able to entirely meet the need of couples to regu late fertility.Recently published research from sub-Saharan Africa,Southern Asia, and Latin America has improved the preci sion of the estimates. Although the estimate of the globalnumber of unsafe abortions is close to earlier figures, the re gional estimates have changed. For example, the recent es timates for Africa are higher than the previous cautious es timates, better reflecting the actual situation and suggestingthat earlier estimates were too low.Regional differentials in unsafe abortionGlobally, an estimated 1 in 10 pregnancies ended in anunsafe abortion in 2003, giving a ratio of 1 unsafe abortionto about 7 live births [7] (Table 2.3). The unsafe abortionrates or ratios for each region are estimated by dividing thenumber of unsafe abortions in that region by the regionalnumber of all women aged 15 to 44 years or by the regionalnumber of live births, respectively, in the same referenceyear (Box A).Table 2.3 provides the average rates and ratios, thatis, relative to women and to births of all countries of asubregion, region, or globally, whether unsafe abortion isknown to take place (e.g., Kenya) or not (e.g., China) ortakes place in parallel to abortions within the framework ofthe law (e.g., India). However, measures that consider onlythose countries with reported incidence of unsafe abortiondescribe its magnitude more adequately. This approachcorrectly links both numerator (unsafe abortions) anddenominator (number of women or number of live births)to the same set of countries in the region or globally. There fore, Table 2.3 also reports, in parentheses, rates and ratios13Table 2.) Global and regional estimates of annual incidence of unsafeabortion in 2003 (Rates and ratios are calculated for all countries and, inparentheses, only for countries with evidence of unsafe abortion.')Numberrounded bWorld19700000500000Developed countries CDeveloping countries 19200000Least developed4 000 000countriesOther developing15 300 000countries4700 000Sub-Saharan AfricaAfrica5500 000Eastern Africa2300 000Middle Africa600 000Northern Africa1 000 000Southern Africa200 000Western Africa1 500 000Asia c9800 000Eastern Asia cSouth-Central Asia6300 000South-East Asia3 100 000400 000Western AsiaEurope500 000400 000Eastern Europe2 000Northern EuropeSouthern Europe100 000Western EuropeLatin America andthe Caribbean 3900 000Caribbean100 000Central America900 000South America2900 000North AmericaOceania c20 000Incidencerate per 1000women aged75 to 44 yearsIncidenceratio per 700live births14 (22)2 (6)16 (24)15 (20)3 (13)16 (20)251515 (23)3129392622 (23)182811 (20)17 (22)1617201220 (21)181413 (18)1823 (27)8 (13)3 (6)5 (6)0.1 (1)3 (6)1627 (31)7 (10)6 (13)13 (14)0.1 (2)7 (14)29 (30)16 (28)253333 (34)19 (26)2638118Rates, ratios, and percentages are calculated for all countries of eachregion, except Asia (which excludes Japan) and Oceania (whichexcludes Australia and New Zealand). Rates, ratios, and percentages inparentheses were calculated exclusively for countries with evidence ofunsafe abortion. Where the difference between the two calculations wasless than one percentage point, only one figure is shown.b Figures may not exactly add up to totals because of rounding.C Japan, Australia, and New Zealand have been excluded from theregional estimates, but are included in the total for developed countries.o No estimates are shown for regions where the incidence is negligible.arestricted to affected countries (Le., those with evidence ofunsafe abortion), with the number of unsafe abortions,women aged 15 to 44 years and live births referring 10 thesame set of countries. The resultant rates and ratios arehigher than those using all countries, better illustrating the

Chapter 214Box A Measurement IndicatorsAbsolute numbers of unsafe abortions cannot be compared meaningfully across different regions and subregions or over time because of differingsize of populations risk. The choice of a particular descriptive measure is dictated by the purpose of presentation and discussion. The followingstandardized meas.uresare often used for ualriumber ofuns fe abo.rtions per 1,000 women aged 15 to 44 years. This summary measure describesthe lev 1 (l1ewcase ).q u s·afe'abCirtion in a given popula lon .in a specified time interval. It shows how many women of reproductive age (J 5 to44 ye;j )J'ave 1!n.\Jris\lfe bQrtior)per 1,000 in' the same age range during a particular year. Further decomposition of this overall rate by 5-yearage grclupsaIIQW fQrascerialnment of age patterns of unsafe abortion as well as the indic1!tw.tot9/ullsafe abort/OIl rote,which describes the'averagenul11bet'of;u"s fll:1!b0r1ipnsa woman is likely to experience by the end ofherrepro llctive life (generally assllmedat 45 years) if the.current ilge specifi Ja es p rsist:.' un C!fe9})oftioil{dt ::T:tie estimated. annual number of unsafe abortions perl 00 Iiye births. The indlcat6qhows the relative propensity of unsafeabq .9nSqlrn rE!dtoIiY!iibiithS in a population. Byextension, substituting live birth as a proxy for pregnancy, this rneasureroughiy indicatesthe lU ellh06d'thatapregnancywillendin unsafe abortion. rather than a live birth. , . ,,",'. , . -, -.'.:.- ., - .'"."-- .'.''.'Unsafe,apoit;QiJmortality ratio: The estimated annual number of maternal de.aths due to urisafe abortion per 100,000 live Ilirths. This indicator is asubs t onne mati!mal[i1ortality ratio (number of maternal deaths per 100,000 live births) and measures the risk of a woman dying due to unsafeal:lOrtiqn relil!ivetolt)ll,OOO live births.Unsafe iibQlt/o tiis:e"fatblityrate:This measure refers to the estimated number of maternal deaths per 100,000 unsafe abortion procedures; it issometinlesexpr sedper1()OprQcedures. The case-fatality rate shows the mortality riskassClC:iated with unsafe abortion.Percentage of maternal deaths due to unsafe abortion: This measure indicates the estimatednumber of unsafe abortion deaths per 100 maternaldeaths. When maternal 'mortality is relatively low and where other causes of maternal death have already been substantially reduced, a smallnumber of unsafe abOrtion deaths may account for a significant percentage of maternal deaths. This measure is, therefore, not particularlysuitable for comparison, lCspecially across countries with different levels of maternal mortality.severity of the public health problem in the countries of aregion where unsafe abortions occur.Unsafe abortion rates close to 30 per 1,000 women,aged15 to 44 years are seen in both Africa and Latin America andthe Caribbean; however, because of the higher numbers ofbirths, the unsafe abortion ratio for Africa is only half thatfor Latin America (Table 2.3). According to recent estimates,the number of unsafe abortions in South America mayhave reached a peak and begun to decline. If Cuba, whereabortion is legally available upon request, is excluded fromthe calculation, the rate for the Caribbean falls between thatfor Central America (25 per 1,000) and South America (33per 1,000). The range of estimates for Africa is wide: easternAfrica has the highest rate of any subregion, at 39 per 1,000,whereas South Africa has among the lowest, at 18 per 1,000(not counting legal abortions of 5 per 1,000 women). The1996 law liberalizing abortion in South Africa has clearlyreduced the number of unsafe abortions in the subregion.Half of all unsafe abortions take place in Asia; however,. rates and ratios are generally lower. Only in South-EastAsia are rates and ratios similar to those of Africa and LatinAmerica. South-Central Asia has the highest number ofunsafe abortions of any subregion, owing to the sheer sizeof its population.The differences in the estimates based on countries at fiskas compared to all countries in the region (Table 2.3) are par ticularly marked for Asia. When the populous region of east ern Asia (with abortion available upon request) is excludedfrom the denominator, the rate rises from 11 to 20 unsafeabortions per 1,000 women aged 15 to 44 years. This pat tern is also apparent for the Caribbean (28 vs. 16 per 1,000)when Cuba is excluded. On the other hand, the exclusion ofCuba makes little difference for the rates for Latin Americaas a whole (30 v . 29 per 1,000). The differences in South East Asia (27 vs. 23 per'I,OOO) and western Asia (13 vs. 8per 1,000) are the result of excluding Singapore and Viet nam, and Turkey, respectively, from the calculations.The ratio of unsafe abortion generally ranges from 10 to20 unsafe abortions per 100 births (Table 2.3). However,when declining fertility results in fewer and fewer birthswithout an accompanying major shift from unsafe abortionto modern contraceptive uptake, ratios become high. Also,where the motivation is stronger to end an unwanted or un intended pregnancy through abortion rather than unwantedbirth, the ratio would be higher. Such is the case in SouthAmerica (38 per 100), Central America (26 per 100), theCaribbean (26 per 100 for all countries vs. 19 per 100 forcountries at risk) and South-East Asia (31 per 100 for allcountries vs. 27 per 100 for countries at risk).The global figures in Table 2.3 show the full effect ofrestricting the analysis appropriately only to the relevantcountries with evidence of unsafe abortion. The 19.7 mil lion unsafe abortions that occurred worldwide in 2003 cor respond to an unsafe abortion rate of 22 per 1,000 womenaged 15 to 44 years when only countries with unsafe abor tion are considered versus 14 per 1,000 when the rate is

Unsafe abortion: The global public health challengebased on all countries. The respective change in the abor tion ratio is 20 versus 15 per 100 live births. For develop ing countries, the rate increases from 16 to 24 per 1,000women of reproductive age when only countries at risk areconsidered. The few developing countries with liberal abor tion laws and no evidence of unsafe abortion (e.g., China,Cuba, Turkey, and Singapore) all fall in the group of "otherdeveloping countries:' leading to a marked difference.in theincidence rate and ratio. The least developed countries showa high unsafe abortion rate of 25 per 1,000 women.In short, the alternative figures presented in parenthesesin Table 2.3 reveal where unsafe abortion is clearly a majorpublic health concern. These figures are alarming and re quire urgent attention by policy makers and program man agers.Unsafe abortion trends by regionRates and ratios of unsafe abortion vary Widely by region(Fig. 2.1). For the sake of comparability with the previous es timates. the rates are for women aged 15 to 49 years and for; all countries of each region. The comparisons are illustrativeof trends, but 1993 estimates are less credible than 2003 esti mates; for example, the latest research evidence from Africashows higher rates of unsafe abortion than previously be lieved probable. The 2003 estimates more accurately reflectthe current situation in Africa; thus, the increases may beEastern Africa15less accentuated than those indicated in Fig. 2.1. Eastern,middle, and western Africa show separate patterns in a highfertility setting. The rate for eastern Africa is notable, increas ing to more than 35 per 1,000 women aged 15 to 49 years asuse of contraception has remained low (around 20%) in theregion; the ratio has decreased because of a less significantincrease in unsafe abortion than in births. Aside from Africa,the rates mostly show a slow decline while ratios have in creased; however, the trend in ratios is less marked.The interpretation of trends in unsafe abortion ratios isnot straightforward because it is a compOsite index of thedegree of motivation to terminate an unwanted pregnancyby induced abortion as well as the trends in unsafe abortionrelative to live births. With the increasing motivation to reg ulate fertility, the unsafe abortion ratio increases.Notwithstanding the complex relationship between trendsin fertility and trends in unsafe abortion ratios, two mainpatterns emerge (Fig. 2.1). The first is represented by SouthAmerica, and also includes Central America, the Caribbean,and South Africa, where fertility has declined to around 2.5children per woman. South Africa nevertheless is distinct. with legal, safe abortion increasingly replacing unsafe abor tion. However, the case of South America is striking: the un safe abortion ratio is still very high in spite of a rise in theprevalence of modern contraceptives from 50 to 65%, withmore than half of the modern method use attributable toiiiiiiiiiiiii- iTII.IIIlIl ! ,· "Jii;i!i .f,South America 'I1I .!1iJce: : : : : ,1 I Il I 'L! I4\l !fl ! lJ ' Western AfricaSouth-East AsiaNorthern AfricaSouthern AfricaSouth-Central Asia.1I1I IlL.Caribbean.oceania !'l!i.2003.2003 1993Ii 1993Western Asia iii Europeo10203040Number of unsafe abortions per 1,000 women 15-49o10203040Number of unsafe abortions per 100 live birthsFigure 2.1 Trends in unsafe abortion rate (per 1,000 women aged 15 to 49 years) and ratio (per 100 births), 1993 and 2003 (Frorn WHO, 1994 [4],WHO, 2007 [7].)

16Chapter 2Table 2.4 Percent of women using a contraceptive method by type of method used in 2005 and unsafe abortion rate and ratios in 2003. (Unsafeabortion rates and ratios are calculated for all countries of each region.')Unsafe abortion incidenceContraceptive use (% of women in union)Anymethod.,WorldDeveloped countriesDeveloping countriesAfricaEastern AfricaMiddle AfricaNorthern AfricaSouthern AfricaWestern AfricaAsiaEastern AsiaSouth-Central AsiaSouth-East AsiaWestern AsiaEuropeEastern EuropeNorthern EuropeSouthern EuropeWestern EuropeLatin America and the CaribbeanCaribbeanCentral AmericaSouth AmericaNorth AmericaOceaniaAustralia/New nmethodsFemaleand 91826521494995446Rate per1,000 women15 to 44Ratioper 100live births142b16293926221828153b1617201220181413 bllb182383503291625333b101627761307331926384b8 See footnotes and text with Table 2.3.b Japan, Australia, and New Zealand have been excluded from the regional estimates of unsafe abortion, but are included in the total for developedcountries.o No estimates are shown for regions where the incidence is negligible.sterilization to terminate childbearing (Table 2.4). Nonethe less, an unmet need for spacing births appears to be metthrough unsafe abortion. The decline in regional numbersof births is because of the increasing tendency to regulatefertility by either contraceptive use or unsafe abortion. Thespeed of decline in fertility has outstripped the decline in un safe abortion, thus accounting for relatively higher ratios.South-East Asia and South-Central Asia (and to some ex tent western Asia and Oceania) represent the other patternof moderately high fertility of around three children perwoman and less than 50% modern contraceptive methoduse. A moderate decline in the unsafe abortion rate is no ticed with little change in the ratio relative to live births. Thetrend in western Asia is less clear, because available data aregenerally limited.Who is more likely to have an unsafe abortion?All sexually active (including sexually coerced) fertilewomen face some risk of unintended pregnancy and, con sequently, of induced abortion or unwanted birth. Contraryto the commonly held view, most women seeking abor tion are married or live in stable unions and already haveseveral children. Some have an induced abortion to limitfamily size and some to space births [17-22]. Where abor tion is highly restricted, educated affluent women can oftensuccessfully obtain an abortion from a qualified provider,

Unsafe abortion: The global public health challengewhereas poor women or those who have little or no ed ucation lack this option [23,24J. Policy makers and pro gram managers often need to know if certain groups requireparticular attention for prevention of unplanned pregnancyand unsafe abortion. Because of the limited data, however,socioeconomic and demographic differentials in unsafe abor tion by marital status, education, income, work participa tion, type of occupation, urban-rural place of residence, eth nicity, and parity are difficult to document.Contraceptive methods remain inaccessible or limited inchoice for married women in some countries. However, ac cess to contraception is worse for unmarried women, partic ularly adolescents. The age patterns of unsafe abortion revealthese most vulnerable groups. A recent review found thattwo-thirds of unsafe abortions occur among women aged 15to 30 years [25]. More importantly from a public health per spective, 2.5 million, or almost 14%, of all unsafe abortionsin developing countries occur among women younger than20 years of age. Unsafe abortions show a distinct age pattern by region(Fig. 2.2). The proportion of women aged 15 to 19 years inAfrica who have had an unsafe abortion is higher than inany other region; almost 60% of unsafe abortions in Africaoccur among women younger than 25 years old, and almost80% are among women younger than 30 years of age. Thissituation contrasts with Asia, where 30% of unsafe abortionsoccur among women less than 25 years old and 60% areamong women les

many women are unaware that safe and legal abortion is available. Even those who know of its legality may not have access to safe abortion because of poor quality of services and. lor . economic and social constraints. Reports also suggest that unsafe abortions may be increasing in several of the