Quality of life following early medical or surgical abortion

1. Introduction 

Effects of the abortion experience on women have been controversial. Overall, the effectiveness and safety of early abortion using medical or surgical techniques is well established [1], [2], [3]. Major complications occur in less than 1% of cases, and safety is greatest when abortion is performed at early gestational ages. Many studies have examined the psychological effects of abortion and a consensus has emerged that these are minimal [4], [5], and can be best understood within the framework of a normal stress response. There is, however, little information regarding short-term quality of life following an abortion. In particular, there is scant information to compare short-term differences in quality of life between medical abortion using mifepristone versus early surgical abortion techniques.

The acceptability of medical abortion as a new technique has been demonstrated in a large sample of US women [6]. In-depth interviews with medical abortion patients showed that the desirability of the method is based on its noninvasive nature, privacy, support and the comfort of the home environment rather than the symptom profile [7]. More recent results confirm these findings about women’s selection of medical abortion, and indicate that a sense of control is an important factor for women selecting medical abortion techniques [8].

Comparisons of medical and surgical abortions have focused on satisfaction and symptoms. Concordance between patients’ expectations of side effects and their experience was an important factor in overall satisfaction [9]. Most comparative studies have been conducted outside the US [9], [10], [11], [12], and results may not be applicable to the US because of differences in medical culture. Also, abortion regimens in these studies were somewhat different from current U.S. medical and surgical abortion practices, and different outcome measures were considered. Both a US study using contemporary approaches to medical and surgical abortion and a British study found that bleeding after early surgical abortion is substantially shorter in duration than bleeding after medical abortion [12], [13]. In studies where surgical abortion was performed under general anesthesia, medical abortion patients reported more pain [10]. These and similar findings could indicate that patients undergoing surgical terminations have fewer physical side effects. In this prospective study we sought to assess the quality of life over 1 month among women undergoing medical and surgical abortions with local anesthesia in a single US practice setting.

2. Materials and methods 

This was a prospective observational cohort study intended to characterize the quality of life among women undergoing early abortion. The setting was a faculty practice where all abortion care was provided in the office by a single attending physician in conjunction with medical students, residents and nurse practitioners. All women aged 18 or older seen for termination of pregnancy prior to 9 weeks gestational age based on sonographic dating were eligible for the study. Recruitment took place for 5 months in 2000 during 1 day per week when both medical abortion and surgical abortion were available to all patients. Prospective patients received telephone information and counseling about both abortion procedures prior to their appointment and had the option of either medical or surgical abortion. Office billing practices during the study deliberately avoided creating any financial incentive for choosing between these approaches.

At the first office visit, all patients underwent a routine medical history, social history and physical examination including transvaginal sonogram. After this assessment and a final decision by the patient regarding which type of abortion she preferred, we obtained informed consent for the abortion itself. This study took place prior to the US Food and Drug Administration (USFDA) approval of mifepristone, and all subjects who chose a mifepristone abortion were enrolled in an Institutional Review Board (IRB)-approved prospective study of the usefulness of sonography during medical abortion care [14]. All medical abortions were provided using a regimen of 200 mg mifepristone given orally in the office at the first visit followed in 24–48 h by a self-administered 800 μg vaginal dose of misoprostol at home. Surgical abortions were provided in the office during the first visit via manual vacuum aspiration after oral analgesia with ibuprofen and a paracervical block. One week later, all patients underwent an interval history, physical examination, repeat transvaginal ultrasound and contraceptive counseling.

We invited English-speaking women to participate in the IRB-approved quality-of-life study at the end of the first visit when all treatment decisions were completed. Those who underwent surgical abortion were invited and enrolled about 1/2 h after the procedure. Those undergoing medical abortion were invited and enrolled after they took the mifepristone, and after they had received all treatment instructions and plans.

We collected quality of life data during structured interviews using a validated instrument, the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ C30) [15], [16]. Subjects underwent the EORTC quality of life assessments three times: in person at the initial visit and again at the routine 1-week follow-up visit, and finally during a telephone interview 1 month later. A few subjects who could not complete the routine follow-up visit as scheduled completed their 1-week questionnaire by telephone. Subjects answered the questions with regard to their symptoms during the previous 7 days. The EORTC instrument was relevant for women undergoing termination of pregnancy because it contains items for physical, emotional, cognitive and social function as well as a global function scale. Specific symptom questions regarding fatigue, nausea, vomiting, and other gastrointestinal upset are highly relevant for pregnant women. The EORTC consists of 30 items; the responses combine to yield the functional scale scores and the specific symptom scores. None of the items contribute to more than one of the scales.

Clinical information and demographic data were abstracted from clinical records and the quality of life scores were calculated from the EORTC. Statistical analyses were performed using the SPSS statistical package (version 9.0). Analyses of variance (ANOVA) and t-tests were performed to assess differences in EORTC scores during follow-up for the entire study population as well as to compare baseline characteristics and changes in EORTC scores between the medical and surgical patients. Most of the EORTC scores were not normally distributed; therefore, we also evaluated these using nonparametric tests with similar results (data not shown). All statistical analyses were performed and results are presented using EORTC raw scores; these were not converted to percentages. No adjustments were taken for multiple comparisons. Because this study provides a preliminary description of quality-of-life changes following early abortion, no sample size was predetermined for the purpose of hypothesis testing.

3. Results 

Ninety-seven women enrolled in the quality of life study. Their baseline characteristics are summarized in Table 1. Women choosing medical and surgical abortion were similar in age, parity, race, marital status and gestational age. Women choosing medical abortion in this practice had somewhat less educational attainment and were less likely to have commercial insurance. Most subjects in both groups reported at the time of making the appointment that their partners were aware of the pregnancy. All of the subjects in this study experienced a successful abortion, and did not require any repeat treatments.

Table 1. Baseline characteristics of surgical and medical termination of pregnancy patients

	Characteristic	Surgicala (n=55)	Medicala (n=42)	p-valueb
	Age (years; mean ± SD)	29.2 ± 6.2	28.4 ± 6.2	0.53
	Gravidity (mean ± SD)	2.6 ± 2.0	2.5 ± 1.4	0.74
	Parity (mean ± SD)	0.7 ± 1.0	0.5 ± 0.8	0.34
	Gestational age (days; mean ± SD)	49.9 ± 7.1	49.1 ± 5.8	0.56
	Race			0.63
	White/Caucasian	25 (45)	19 (45)
	African American/Black	8 (15)	7 (17)
	Asian	4 (7)	5 (12)
	Hispanic/Latin	18 (33)	10 (24)
	Other	0 (0)	1 (2)
	Education			0.02
	High school graduate or GED	7 (13)	4 (10)
	Some college	11 (20)	21 (50)
	Bachelor’s degree completed	17 (31)	9 (21)
	Graduate or professional school	20 (36)	8 (19)
	Employment status			0.20
	Student	14 (25)	12 (29)
	Employed part-time	3 (5)	2 (5)
	Employed full-time	37 (67)	23 (55)
	Unemployed	1 (2)	5 (12)
	Income (annual)			0.28
	<$25,000	10 (19)	9 (21)
	$25,000–$75,000	28 (54)	17 (40)
	>$75,000	13 (25)	16 (38)
	Insurance status			0.05
	Self-pay	8 (15)	14 (33)
	Private insurance	43 (78)	23 (55)
	Medicaid	4 (7)	5 (12)
	Marital status			0.40
	Single	30 (55)	18 (43)
	Married/living with partner	20 (36)	21 (50)
	Divorced/separated	5 (9)	3 (7)
	Household			0.3
	Lives with partner	13 (23)	13 (32)
	Does not live with partner	44 (77)	27 (66)
	Cohabitant knowledge of pregnancy			0.1
	Yes	16 (28)	19 (46)
	No (or NA)	40 (70)	22 (54)
	Partner knowledge of pregnancy			0.8
	Yes	46 (81)	34 (83)
	No	11 (20)	7 (17)
	Partner support			1.0
	Yes	32 (56)	22 (54)
	No	12 (21)	8 (20)
	Not asked	13 (23)	11 (27)

a Values in parentheses are percentages. b Pearson chi-square for categorical variables. T-test for continuous variables. Not all subcategories sum to 100% due to missing data or rounding.

Thirty-three patients undergoing elective surgical abortion and 22 patients undergoing medical abortion during the period of study enrollment did not participate in the study. Nonparticipation was usually due to having no staff present to enroll the patient or due to lack of fluency in English that prevented completion of the EORTC; eligible patients rarely declined to participate. The main reason for refusal was the patient’s need for secrecy regarding the abortion. Participants and nonparticipants were otherwise similar with regard to the demographic and pregnancy characteristics (data not shown).

The three quality-of-life scores from the EORTC are shown in Table 2 for all subjects across the study month. There were 97 participants who completed the baseline assessment, 87 who completed the 1-week and 85 who completed the 1-month assessment. The scores in the table include only the 85 subjects who completed study participation. The 12 subjects who did not complete all three questionnaires were similar at baseline to those who completed the study (data not shown). Baseline EORTC scores were unrelated to subjects’ age, parity or gestational age. Similarly, race, insurance status, educational status and measures of social support for the abortion were unrelated to the quality of life scores. Over 1 month, the study subjects experienced a clinically and statistically significant improvement in their scores for all quality of life variables. These substantial improvements over 1 month were experienced to a similar degree in both the surgical abortion and medical abortion subgroups. Additional symptom scales from the EORTC included dyspnea, constipation and diarrhea; subjects also reported on financial difficulties. Scores for all of these scales were low at baseline, indicating minimal symptoms; nonetheless, they also improved during follow-up, similar to the other scales (data not shown).

Table 2. EORTC quality-of-life measures following abortion

	Variable (score range and direction)	Mean score (n=85)
		Baseline	1 Week	1 Month	p-valuea
	Global health status (2–14, high score good)	8.8	9.7	12.0	<0.01
	Function scales (low score good)
	Emotional (0–12)	5.2	3.2	2.1	<0.01
	Cognitive (0–6)	1.2	0.6	0.3	<0.01
	Social (0–6)	1.2	1.1	0.3	<0.01
	Physical (0–5)	0.5	0.7	0.1	<0.01
	Work (0–1)	0.2	0.2	0.1	<0.01
	Symptom scales (low score good)
	Pain (0–6)	1.2	2.0	0.6	<0.01
	Fatigue (0–9)	4.2	3.0	1.3	<0.01
	Nausea/vomiting (0–6)	1.5	0.7	0.3	<0.01
	Appetite loss (0–3)	0.9	0.5	0.1	<0.01
	Insomnia (0–3)	1.0	0.7	0.4	<0.01

a p-values are from repeated measures ANOVA to test for trend in mean EORTC scores across all three measurements. For global health status, a high score is desirable. For all other measures, a low score is better.

Improvements in the quality of life following abortion were already seen in most of the scales by 1 week. There were, however, two measures with worse scores at 1 week—pain and physical function. The worsening in the physical function score from 0.48 to 0.65 was entirely due to an increase in the scores among subjects who underwent surgical abortion; however, these small differences could easily be explained by chance (p = 0.3). The pain score increased from 1.2 to 2.0 at week 1 for women in both the medical and surgical abortion subgroups (p < 0.01 overall and for the medical subgroup, p = 0.02 for the surgical subgroup). It is not surprising that women would experience pain due to uterine cramps during the week after an abortion.

There were some differences between the subjects undergoing medical abortion and surgical abortion (Table 3). At baseline, women who had selected and undergone a surgical abortion had higher, i.e., worse, scores on the emotional (5.9 vs. 4.2, p = 0.01), cognitive (1.6 vs. 0.7, p < 0.01), and social (1.5 vs. 0.8, p = 0.01) function scales compared to those who chose a medical abortion. At 1 week, these baseline differences between the medical and surgical group disappeared. At 1 week the surgical group had better, i.e., lower, symptom scores than the medical abortion subjects for nausea/vomiting (0.4 vs. 1.0, p = 0.02) and a similar improvement in loss of appetite (0.3 vs. 0.6, p = 0.06). There were no other important differences between the medical and surgical abortion subgroups in quality of life during the first week. At the 1-month follow-up, there were only minimal differences, none statistically or clinically significant, between the women in the medical and surgical abortion groups.

Table 3. Procedure type and mean function and symptom scores

	EORTC scale (score range and direction)	Mean score
		Baseline		p-valuea	1 Week		p-valuea	1 Month		p-valuea
		Medical	Surgical		Medical	Surgical		Medical	Surgical
	Number of subjects	42	55		40	47		38	47
	Global health (2–14, high score good)	9.4	8.4	0.08	9.6	9.8	0.7	11.7	11.3	0.4
	Function scales (low score good)
	Emotional (0–12)	4.2	5.9	0.01	3.2	3.2	1.0	1.8	2.4	0.3
	Cognitive (0–6)	0.7	1.6	<0.01	0.7	0.6	0.5	0.3	0.4	0.4
	Social (0–5)	0.8	1.5	0.01	1.1	1.0	0.8	0.3	0.2	0.7
	Physical (0–1)	0.5	0.4	0.7	0.6	0.7	0.7	0.1	0.1	0.9
	Work (0–1)	0.2	0.3	0.2	0.2	0.2	1.0	0.03	0.06	0.5
	Symptom scales (low score good)
	Pain (0–6)	1.0	1.4	0.2	2.3	1.9	0.3	0.5	0.7	0.4
	Fatigue (0–9)	3.5	4.8	0.01	3.2	2.8	0.4	1.2	1.3	0.9
	Nausea/vomiting (0–6)	1.8	2.0	0.4	1.0	0.4	0.02	0.3	0.4	0.7
	Appetite loss (0–3)	1.0	0.9	0.9	0.6	0.3	0.06	0.1	0.1	0.9
	Insomnia (0–3)	0.7	1.2	0.02	0.7	0.7	1.0	0.3	0.5	0.5

a p-values assess differences between medical and surgical abortion subgroups based on independent t-tests. For global health status a high score is desirable. For all other measures a low score is better.

The only baseline characteristic that was associated with the EORTC scores was partner knowledge of the pregnancy. We asked women at the time of making the abortion appointment if their partner or anyone else was aware of the pregnancy. Most women who chose to participate in this study had reported that their partner was aware at the time of scheduling the appointment. Some women may have subsequently informed their partners about the pregnancy; however, we did not collect that information. As shown in Table 4, partner knowledge of the pregnancy was associated with worse scores at baseline compared to subject’s whose partner was not aware of the pregnancy for all scales except pain. The magnitude of the differences diminished during follow-up; although the scores tended to remain less favorable for the subjects who reported partner knowledge, these differences were no longer statistically significant at follow-up. These associations were similar whether the patient chose medical or surgical abortion. Among those whose partners were aware of the pregnancy, we assessed whether partner support of the abortion decision was related to the EORTC scores. We found no association at all between the EORTC scores and partner support for the abortion decision. Subjects in the study also reported that other family members, roommates and friends were aware of and supported the abortion. We did not find any associations between the quality-of-life scores and knowledge of the abortion by these other individuals; however, there were only a few subjects in each of these many possible subgroups, so associations could have easily been missed.

Table 4. Partner knowledge of pregnancy and mean function and symptom scores

	EORTC Scale	Mean score
		Baseline			1 Week			1 Month
		Partner Knowledge
		Yes n=80	No n=18	p-valuea	Yes n=70	No n=15	p-valuea	Yes n=71	No n=16	p-valuea
	Global health status (2–14, high score good)	8.4	10.4	<0.01	9.5	10.6	0.07	11.4	11.9	0.4
	Function scales (low score good)
	Emotional (0–12)	5.6	3.2	<0.01	3.4	2.1	0.1	2.3	1.6	0.2
	Cognitive (0–6)	1.4	0.7	0.01	0.7	0.3	0.05	0.4	0.1	0.04
	Social (0–6)	1.4	0.4	<0.01	1.1	0.7	0.3	0.3	0.0	<0.01
	Physical (0–5)	0.5	0.4	0.5	0.6	0.7	0.9	0.1	0.3	0.5
	Work (0–1)	0.2	0.1	0.02	0.2	0.2	0.6	0.03	0.1	0.3
	Symptom scales (low score good)
	Pain (0–6)	1.3	0.9	0.2	2.1	2.0	0.9	0.5	0.9	0.2
	Fatigue (0–9)	4.4	3.2	0.03	3.2	1.9	0.04	1.3	0.9	0.4
	Nausea/vomiting (0–6)	2.2	0.8	<0.01	0.7	0.2	<0.01	0.4	0.1	0.06
	Appetite loss (0–3)	1.1	0.4	<0.01	0.5	0.1	<0.01	0.1	0.0	0.02
	Insomnia (0–3)	1.1	0.3	<0.01	0.8	0.3	<0.01	0.4	0.6	0.4

a p-values assess EORTC mean score differences between social support groups based on independent t-tests.

4. Discussion 

This study found that during the 4 weeks after an early abortion, quality of life improved dramatically across all of the outcomes that are included in the EORTC questionnaire. The 15 outcome scales included measures of global health status, function and symptoms. No other studies of abortion have obtained repeated, standard assessments of this large range of physical and psychological outcomes. Although the study was small with only 97 participants, high rates of follow-up were achieved (87%), increasing the credibility of the results. The reported improvements in quality-of-life measures were similar for women undergoing medical abortion with mifepristone and women undergoing a surgical abortion with local anesthesia.

In previous studies comparing medical and surgical abortion outcomes, participants had incomplete control over which treatment was used. Choices of the type of abortion were sometimes decided by the clinician on medical grounds [9], [10], [11], sometimes decided by the investigator to yield equal size treatment groups [11], [12] and sometimes determined entirely by the availability of mifepristone [17]. In two studies, the few patients without a strong treatment preference were randomized to medical or surgical abortion [12], [18]. We chose not to randomize subjects, so that we could assess quality of life among patients who received their preferred treatment. Most women seeking abortion do have a strong preference [12], [18]. All subjects in the present study were drawn from the same referral base, and they chose between medical and surgical abortion techniques after complete counseling regarding both of the methods. At the time this study was carried out, all abortion patients had the same schedule for return visits and were charged the same fees. This minimized treatment choices being made for reasons of cost or convenience. Although mifepristone was not USFDA-approved until just after the present study, it had been used in clinical studies in this faculty practice for 4 years; therefore, the referring physicians and the patient population already had substantial familiarity with this option. It is nonetheless possible that the preapproval status of mifepristone may have skewed patient treatment choice or affected the quality-of-life responses.

The EORTC questionnaire is mainly used in studies of cancer patients [16]. We chose it for this study because it is easy to understand, quick to complete and relevant to the symptoms of pregnant women. Study enrollment and thus the baseline EORTC questionnaire occurred at the very end of the first treatment visit. We specifically waited to introduce the study until late in the first visit to minimize coercion to participate. We recognize this may have influenced the baseline scores. Women choosing surgical abortions completed the baseline questionnaire very shortly after their procedure, and several of their baseline scores were significantly worse than those of the medical abortion group. This difference may have been due to the timing of the baseline questionnaire; alternatively, women who had more symptoms or distress during the week before the appointment may have selected a surgical abortion in order to complete the process more quickly. In future studies of quality of life following abortion, we would recommend earlier administration of the baseline questionnaire.

Subjects in both the medical and surgical abortion groups improved substantially and reported similar functioning and symptoms at the 1-week and 1-month assessments when compared to each other. The only important differences were more nausea and vomiting and loss of appetite among the medical abortion patients during the week after enrollment compared to the surgical abortion patients. Previous studies consistently show that misoprostol used during medical abortion is associated with nausea and vomiting in many women [6]. These symptoms improved in both of our groups during the first week, but the surgical patients improved more than the medical abortion patients. Pain, usually uterine cramps, increased in both groups during the first week and, as expected, increased more in the medical abortion patients who used misoprostol during the first week. Clinicians should not forget that surgical abortion patients often continue to cramp after the procedure is completed. Two other studies comparing medical and surgical abortion also reported similar pain 2 weeks after medical or surgical abortion [12], [17]. In contrast, when completing symptom diaries, medical-abortion subjects reported far more symptoms than surgical-abortion patients [11]. Interval interviews or questionnaires may be less sensitive than diaries for identifying short-term symptoms.

Our standard telephone triage for women seeking abortion includes questions about social support in order to identify confidentiality concerns and needs for social services. In our analysis, we found that partner knowledge of the pregnancy at triage was strongly associated with adverse symptoms and functioning at the baseline EORTC; this is the opposite of what we expected. Most women who participated in this study stated their partners knew about the pregnancy and most of these partners supported the decision to have an abortion. We attempted to assess the effect of partner support separate from partner knowledge, and found no difference in the EORTC scores of the women; however, our ability to disentangle this was limited by the small study size. Partner knowledge of the pregnancy had a much stronger association with the EORTC scores than any other baseline variable. In our study, this was a secondary analysis, and the surprising association deserves a more systematic evaluation.

Most abortion care in the US is provided in free-standing clinics; few abortions are provided in a private office setting like that used in this study [19]. The setting of this study may limit the generalizability of our results. A woman’s baseline level of distress and her subsequent ability to recover physically and emotionally from a crisis pregnancy may vary by her access to care or by features of the abortion setting. In our analyses, however, results did not vary in any important fashion according to education, income, or insurance status; this suggests our results may apply to a broad range of women. Overall, abortion care in the US is extremely safe, and the clinics in which most care is provided have developed for medical, political and cost-efficiency reasons. Medical abortion, still novel, is being incorporated into existing clinics and also is diffusing into other practice settings. It may be valuable to explore more formally how the setting in which abortion is provided affects the woman’s quality of life.

This study has shown clearly and unambiguously that a woman’s quality of life improves substantially during the month after an abortion. We assessed 15 symptoms and functional domains using a standardized instrument, and every domain showed clinically and statistically significant improvement. These dramatic improvements were similar whether the woman chose a medical abortion with mifepristone or a surgical abortion. Many reproductive age women have serious chronic medical and psychiatric problems; our findings, of course, do not negate the need to identify and care for such ongoing problems in women seeking abortion care. Our findings are reassuring for the general population regarding the quality of life after early medical or surgical abortion. Physician reluctance to provide medical abortion due to worry about side effects is not justified. These findings should support improved access to early medical and surgical abortion in physician offices and other settings.