Background
Recent studies suggest that the basis for adverse pregnancy outcomes is often established early in pregnancy, during organogenesis. It is therefore important to take preventive action as early as possible, preferably before pregnancy. Because most adverse pregnancy outcomes occur in women who are unaware of being at risk, we conducted a randomized controlled trial, “Parents to Be.” With this study, we sought to assess the extent to which women who have participated in preconception counseling (PCC) increase their knowledge on pregnancy-related risk factors and preventive measures and change their behavior before and during pregnancy and to provide an overview of adverse pregnancy outcomes among such women.
Methods
Knowledge: Women aged 18–40 who attended PCC and women who received standard care were matched on previous pregnancy, time since last pregnancy, age, country of birth, and educational achievement. They were sent a questionnaire on knowledge about pregnancy-related risk factors and preventive measures. Behavior: Data on pregnancies and outcomes were collected. Two months after pregnancy, a questionnaire was sent regarding behavior before and during pregnancy.
Results
Knowledge of women who received PCC (81.5%; n=211) exceeded that of women who did not (76.9%; n=422). Levels of knowledge in women who were not yet pregnant after PCC were comparable to those in women who became pregnant after PCC, indicating that, even before pregnancy, PCC increased knowledge in women contemplating pregnancy. After PCC, significantly more women started using folic acid before pregnancy (adjusted odds ratio [OR], 4.93; 95% confidence interval [CI], 2.81–8.66) and reduced alcohol use during the first 3 months of pregnancy (adjusted OR, 1.79; 95% CI, 1.08–2.97). Among the group receiving standard care, about 20% of all pregnancies ended in an adverse outcome; in the group with PCC this was 16% (OR, 0.77; 95% CI, 0.48–1.22).
Conclusion
After PCC, women have more knowledge about essential items. Importantly, they gained this greater knowledge before pregnancy and more women changed their behavior to reduce adverse pregnancy outcomes.
Introduction and Background
There has been much debate on the prevention of adverse pregnancy outcomes (e.g., miscarriage, preterm birth, or congenital malformations), especially those related to maternal and infant morbidity and mortality (
Frishman, 2003
, Gottesman, 2004
, Konchak, 2001
, Kuller and Laifer, 1994
, Moos, 2004
, Muchowski and Paladine, 2004
, Taysi, 1988
). Over time, the optimization of care during pregnancy and labor has reduced morbidity and mortality rates, especially in women known to have a high risk of adverse pregnancy outcomes (e.g., those with diabetes; Grubbs& Brundage, 2002
, Herman et al., 1999
, McElvy et al., 2000
). But, despite the continued improvements in care, these rates have stabilized during the past 3 decades (Bennebroek Gravenhorst et al., 2001
, Garssen and van der Meulen, 2004
, Schuitemaker, 1998
).Recent studies suggest that the basis for adverse pregnancy outcomes is often established early in pregnancy, during organogenesis. It is therefore important to take preventive action as early as possible, preferably before pregnancy, as organogenesis takes place from days 17 to 56 after conception, a period during which women are often unaware of their pregnancy (
Grubbs and Brundage, 2002
, Moos, 2002
, Moos, 2003
).Women contemplating pregnancy have limited knowledge about risk factors and preventive measures regarding adverse pregnancy outcomes (
de Jong-Potjer et al., 2008
). To minimize such risks, they and their partners can attend preconception counseling (PCC), which provides information on general and personal risk factors and preventive measures (Cefalo and Moos, 1995
). Pregnancy can be influenced by a number of risk factors and preventive measures. A well-known example of the latter is folic acid use, which reduces the risk of neural tube defects. To obtain maximum benefit, folic acid use should start as early as 4 weeks before conception—an example that stresses the need for early action (Centers for Disease Control and Prevention, 1993
).Because most adverse pregnancy outcomes occur in women who are unaware of being at risk, we conducted a randomized controlled trial, “Parents to Be,” which was intended to study the effects of general practitioner (GP)-initiated PCC in the general population. During this project, PCC provided women with information on a healthy lifestyle before and during pregnancy. It also provided information on risk factors specific to their own medical, reproductive, and family histories (
Elsinga et al., 2006
). The results showed that the prevalence of risk factors among couples contemplating pregnancy was high, even in couples assumed to be at low risk, suggesting that PCC is beneficial for the general population (van der Pal-de Bruin et al., 2005
).On the basis of earlier research showing that personal counseling increased risk-reducing behavior (e.g., stopping smoking) and the increased awareness and use of folic acid (
Lumley and Waters, 1999
, Pastuszak et al., 1999
, Sayers et al., 1997
), we hypothesized that personal counseling causes women to increase their knowledge on risk factors and preventive measures, thus enabling them to adapt their risk behavior toward a favorable pregnancy outcome.We therefore studied the possible effects of PCC on women's knowledge as well as behavior before and during pregnancy. To determine whether attendance in PCC would increase women's knowledge before pregnancy, we compared knowledge among women who attended PCC with that of women who received standard care, matching the 2 groups on previous pregnancy, time of last pregnancy, age, country of birth, and educational level. To determine whether attendance in PCC would change women's behavior, we compared the behaviors before and during pregnancy of women who attended PCC with the behaviors of women who had received standard care. Furthermore, we present the pregnancy outcomes of the 2 latter groups.
Methods
Design
This project is part of “Parents to Be,” a randomized controlled trial in which randomization occurred at the level of general practices. The study design was published elsewhere (
Elsinga et al., 2006
).The intervention consisted of an annual invitation for PCC to women aged 18–40. A risk-assessment questionnaire was sent to women who were interested in PCC and who were also contemplating pregnancy within 1 year. GPs then invited these women and their partners for PCC, where they provided the couple with information on general risk factors and on their personal risk factors identified in the risk-assessment questionnaire. This study was approved by the Medical Ethics Committee at Leiden University Medical Center.
Knowledge assessment
In 2000, knowledge of pregnancy-related risk factors and preventive measures was assessed among a random selection of half the women aged 18–40 registered at the general practices prior to their offer of PCC (
de Jong-Potjer et al., 2008
). In 2003, knowledge was assessed among all women who attended PCC during the trial and among a random selection of half of the women who received standard care.Knowledge assessment procedure
GPs excluded women with adverse social circumstances. Knowledge levels were assessed in a questionnaire consisting of 94 questions on pregnancy-related knowledge, to which 12 questions on socioeconomic factors and family planning had been added.
A distinction was made between essential items, that is, subjects that should always be addressed during PCC, and items indicating the extent to which a woman was aware of specific risk factors that needed to be discussed because they were relevant to her lifestyle, medical history, or family history.
Twenty items divided over 4 categories were defined as essential. The first category was composed of items related to different aspects of timing of conception. The 3 other essential categories were composed of items about infectious diseases, folic acid need, and exposure to harmful substances, covering risks that were applicable to all women, or risks that a woman might easily run (
Mullen et al., 1994
).Comparisons between groups
Of the 353 women who received PCC, GPs excluded 59 women because they registered with another general practice or adverse social circumstances. Of the remaining 294 women, 72% returned the questionnaire. A total of 211 questionnaires could be used for analysis.
To assess whether knowledge was influenced by PCC, each woman who had attended PCC and completed the knowledge questionnaire after the intervention was matched with 2 women from general practices offering standard care. Each of these 3 women was selected for the similarity of her demographic characteristics (previous pregnancy (yes/no), year of last pregnancy (2000, 2001, 2002, 2003, or before entering trial), age (in years), country of birth (Netherlands, Surinam/The Antilles, Turkey/Morocco, or other) and educational level (basic, intermediate, or high).
Some of the women who attended PCC completed the questionnaire both in 2000 and 2003 (n=74), so their knowledge over these 3 years could be compared. Forty-six of these women had become pregnant since attending PCC. All women who attended PCC also completed a risk assessment questionnaire beforehand. To determine whether PCC had increased these women's knowledge about personal risk factors for adverse pregnancy outcomes, the reported presence or absence of these personal risk factors was linked with their knowledge about these risk factors in 2000 and 2003.
Assessment of pregnancy outcomes and behavioral changes
Data were collected from all participating practices on pregnancies in women whose first day of the last menstrual cycle had occurred between April 2000 and April 2003. Pregnancy was defined as any new entry of W78, namely, the code for pregnancy under the International Classification of Primary Care, in the electronic patient file (
Rodgers et al., 2004
). Parity, first day of last menstruation, term date, date of the end of the pregnancy, and outcome of pregnancy were recorded. Birth announcement cards sent to the general practice were checked for missing pregnancies or data about pregnancy. Each practice was visited every 2 months to collect these data. We ensured that newly registered female patients had an opportunity to receive PCC before they were included in the trial.Provided the GP gave his or her approval, we sent a questionnaire1 to all women within 2 months after delivery, enquiring about their pregnancy outcome and about their behavior before and during pregnancy. The GP could exclude women for social reasons, such as a recent divorce. A postage-free envelope addressed to the researchers was included with the questionnaire. A reminder was sent after 2 months.
The questionnaire contained 27 questions. As well as questions on lifestyle factors before and during the whole pregnancy, there were questions on pregnancy complications, pregnancy outcome, and 7 questions on socioeconomic factors, anxiety, and family planning. The variable “folic acid use” also elaborated into a question on multivitamin supplements specifically meant for pregnant women.
A total of adverse pregnancy outcomes was calculated on the basis of the following definitions: miscarriage, extrauterine pregnancy, still birth, premature birth, low birth weight (<2,500 g), small for gestational age (<p2.3), and congenital anomalies. Abortions for either social or medical reasons have not been taken into account. When a pregnancy had multiple adverse outcomes (e.g., both preterm birth and low birth weight), it was counted only once. Live births where data were lacking about duration, low birth weight, and weight related to gestational age were assumed to be in the normal range, because abnormal outcomes are almost always accompanied by a letter from a specialist or a remark from the GP in the file. An odds ratio (OR) was calculated for adverse pregnancy outcomes among women with PCC versus women with standard care.
Statistics
Analyses were performed using SPSS 11.0 for Windows.
Knowledge assessment
For comparisons between the knowledge level of women with PCC and their matched controls analysis of variance was used, with a fixed group effect and a random factor indicating the matching group.
Paired Student's t-tests were used to test for changes in the knowledge of women who completed the questionnaire both in 2000 and 2003. Changes in knowledge between 2000 and 2003 were compared in a multiple linear regression, adjusting for country of birth, educational level, and parity in a multivariate regression analysis. p-values < .05 were considered significant.
Assessment of pregnancy outcomes and behavioral changes
Actual PCC attendance was lower than expected (
Elsinga et al., 2006
). Because of the low numbers, an intention-to-treat analysis, which is the recommended method of analysis of randomized controlled trials, could not be performed. To gain insight into the efficacy that PCC might have had on risk-reducing behavior, we analyzed the results as we would have analyzed those of an observational study in which women with PCC (the intervention group) were compared with women receiving standard care (the control group), adjusting for confounders. ORs were calculated for behavioral changes in women with PCC versus women with standard care, with 95% confidence intervals (CI). An OR>1 indicates that women who received PCC more often showed risk-reducing behavior. Multiple logistic regression was used to calculate ORs, which were adjusted for the possible confounders, age (continuous), country of birth (Dutch/non-Dutch), and educational level (low, intermediate, high). ORs with 95% CIs not containing 1 were considered statistically significant.Results
Knowledge assessment
Table 1 compares the knowledge of women after PCC with that of the matched women who received standard care. Women who had received PCC had significantly more knowledge of the 20 essential items in total and of the subgroups infection prevention and folic acid. Compared with the matched women receiving standard care, women of lower educational level who had attended PCC also had greater knowledge of hazardous substances (+10.4%; p=.04), folic acid (+17.6%; p=.01), and the 20 essential items (+6.5%; of p=.03). The knowledge of women who had attended PCC but had never been pregnant (73.5%) was substantially higher than that of matched control women who had never been pregnant (62.2%; p < .01). This difference was smaller, but remained significant among women who had been pregnant before (p=.01).
Table 1Percentage of Items Answered Correctly by Women Who Attended Preconception Counseling (PCC) and by Those Who Received Standard Care Matched by Age, Country of Birth, Educational Level, Previous Pregnancy, and Time of Last Pregnancy (with 2 matched standard care women per PCC woman)
| PCC | Standard care | Difference in knowledge level(95% CI) | |
|---|---|---|---|
| Total, n | 211 | 422 | |
| Hazardous substances | 73.0 | 69.2 | 3.8 (−0.1 to 7.6) |
| Prevention of infection | 94.2 | 89.8 | 4.4 (2.0–6.8) |
| Folic acid | 91.9 | 80.6 | 11.3 (7.3–15.3) |
| Timing of conception | 77.0 | 74.4 | 2.6 (−0.2 to 5.3) |
| Total for 20 essential items | 81.5 | 76.9 | 4.6 (2.6–6.6) |
| Educational level, n | |||
| Low | 30 | 59 | |
| Hazardous substances | 72.2 | 61.9 | 10.4 (−0.5 to 20.2) |
| Prevention of infection | 86.7 | 83.1 | 3.7 (−3.5 to 10.8) |
| Folic acid | 86.7 | 68.9 | 17.6 (4.1–31.1) |
| Timing | 70.0 | 70.0 | 0.0 (−8.9 to 8.9) |
| Total for 20 essential items | 76.5 | 70.0 | 6.5 (0.7–12.2) |
| Intermediate, n | 86 | 179 | |
| Hazardous substances | 70.7 | 71.1 | −0.4 (−6.8 to 6.0) |
| Prevention of infection | 93.9 | 89.7 | 4.3 (0.5–8.0) |
| Folic acid | 91.5 | 81.6 | 9.9 (3.6–16.1) |
| Timing | 76.7 | 72.9 | 3.8 (−0.3 to 7.9) |
| Total for 20 essential items | 80.6 | 77.0 | 3.5 (0.5–6.6) |
| High, n | 94 | 182 | |
| Hazardous substances | 75.9 | 70.2 | 5.6 (0.2–11.1) |
| Prevention of infection | 97.1 | 92.7 | 4.4 (1.0–7.7) |
| Folic acid | 94.3 | 84.4 | 9.9 (4.3–15.5) |
| Timing of conception | 79.8 | 77.7 | 2.1 (−1.8 to 6.0) |
| Total for 20 essential items | 84.3 | 79.5 | 4.8 (1.9–7.8) |
| Pregnant | |||
| Never, n | 40 | 80 | |
| Hazardous substances | 63.3 | 52.9 | 10.4 (0.2–20.6) |
| Prevention of infection | 87.5 | 76.6 | 11.9 (3.5–20.2) |
| Folic acid | 85.8 | 49.2 | 36.7 (22.5–50.9) |
| Timing of conception | 68.9 | 68.0 | 0.9 (−6.7 to 8.5) |
| Total for 20 essential items | 73.5 | 62.2 | 11.3 (4.6–18.0) |
| More than once, n | 171 | 342 | |
| Hazardous substances | 75.2 | 73.1 | 2.2 (−1.8 to 6.2) |
| Prevention of infection | 95.8 | 93.1 | 2.6 (0.4–4.8) |
| Folic acid | 93.4 | 88.0 | 5.4 (2.0–8.7) |
| Timing of conception | 78.9 | 75.9 | 3.0 (0.1–5.8) |
| Total for 20 essential items | 83.3 | 80.3 | 3.0 (1.2–4.84) |
∗ Statistically significant (p < .05).
Women who attended PCC and completed both knowledge questionnaires (2000/2003) were divided into 2 groups: women who became pregnant after PCC and women who were not pregnant when they completed the second knowledge assessment (Table 2). Regarding the 20 essential items, the knowledge of both groups increased substantially after they had attended PCC (23.4% [95% CI, 17.5–29.2] and 12.0% [95% CI, 2.9–21.0], respectively). Among women who had not become pregnant after PCC, there was a substantial increase in knowledge regarding 2 subgroups: the prevention of infection (17.9% [95% CI, 5.9–29.8]) and folic acid (28.6% [95% CI, 13.0–44.1]).
Table 2Percentage Increase in Items Answered Correctly by Women Who Attended Preconception Counseling (PCC)
| Total PCC (n=74) | Not Pregnant After PCC (n=28) | Pregnant After PCC (n=46) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 2000 | 2003 | Change in Knowledge Level (95% CI) | 2000 | 2003 | Change in Knowledge Level (95% CI) | 2000 | 2003 | Change in Knowledge Level (95% CI) | |
| Hazardous substances | 55.4 | 74.5 | 19.1 (11.5–26.8) | 63.1 | 71.4 | 8.3 (−5.2 to 21.9) | 50.7 | 76.4 | 25.7 (16.6–34.8) |
| Prevention of infection | 72.6 | 94.9 | 22.3 (15.4–29.2) | 74.1 | 92.0 | 17.9 (5.9–29.8) | 71.7 | 96.7 | 25.0 (16.5–33.5) |
| Folic acid | 57.7 | 94.1 | 36.5 (26.3–46.7) | 65.5 | 94.0 | 28.6 (13.0–44.1) | 52.9 | 94.2 | 41.3 (27.8–54.8) |
| Timing of conception | 68.7 | 78.4 | 9.7 (4.2–15.1) | 70.9 | 75.5 | 4.6 (−4.9 to 14.1) | 67.4 | 80.1 | 12.7 (5.9–19.5) |
| Total for 20 essential items | 63.9 | 82.9 | 19.1 (14.1–24.1) | 68.4 | 80.4 | 12.0 (2.9–21.0) | 61.1 | 84.5 | 23.4 (17.5–29.2) |
∗ Statistically significant (p<0.05).
There is also a relation between risky behavior and knowledge as an increase in the proportion of women responding correctly to risk-related knowledge statements was found (data not shown). For example, before attending PCC, only 50% of women who smoked could correctly indicate that passive smoking by a pregnant woman for 2 hours a day can be harmful to the baby in her womb. After PCC, 85.5% of smokers answered this item correctly. Similarly, more women could correctly answer statements regarding personal risk factors (age and non-use of folic acid), indicating that PCC increases not only essential knowledge, but also knowledge regarding personal risk factors.
Assessment of pregnancies and behavioral changes
There were 150 pregnancies among these women after they received PCC. Among women who received standard care, 1,914 pregnancies were registered. Questionnaires about behavior and pregnancy outcome were sent for 139 (92%) pregnancies after PCC and for 1,703 (82%) pregnancies with standard care. Of the questionnaires returned, 114 (82%) were usable for pregnancies after PCC and 1,158 (68%) for pregnancies with standard care. A higher number of women with PCC had been born in the Netherlands and had a high educational level compared with women with standard care (Table 3).
Table 3Distribution of Demographic Characteristics
| PCC (n=114) | Standard care (n=1,158) | |||
|---|---|---|---|---|
| n | % | n | % | |
| Age (yrs) | ||||
| <20 | 0 | 0 | 2 | 0.2 |
| 20–24 | 1 | 0.9 | 60 | 5.3 |
| 25–29 | 29 | 25.9 | 260 | 23.0 |
| 30–34 | 53 | 47.3 | 479 | 42.5 |
| 35–39 | 22 | 19.6 | 259 | 23.0 |
| ≥40 | 7 | 6.3 | 68 | 6.0 |
| Country of birth | ||||
| The Netherlands | 108 | 94.7 | 1,017 | 87.8 |
| Other | 6 | 5.3 | 141 | 12.2 |
| Educational level | ||||
| Basic | 11 | 9.8 | 259 | 23.2 |
| Intermediate | 51 | 45.1 | 438 | 39.2 |
| High | 51 | 45.1 | 420 | 37.6 |
| Marital status | ||||
| Married/cohabiting | 111 | 99.1 | 1,081 | 94.3 |
| Permanent relationship | 1 | 0.9 | 30 | 2.6 |
| Single | 0 | 0 | 35 | 3.1 |
| Medical insurance | ||||
| Public | 69 | 61.6 | 787 | 68.3 |
| Private | 6 | 5.4 | 35 | 3.0 |
| Employment based | 37 | 33.0 | 330 | 28.6 |
| None | 0 | 0 | 1 | 0.1 |
a Numbers may vary owing to missing data.
Table 4 shows that tobacco use before pregnancy did not differ between women who received PCC and women who received standard care (24.8% vs. 26.8%). Women who received PCC used alcohol slightly more often before pregnancy (71.7% vs. 61.7%), although this difference was not significant. There were no differences with regard to reported hereditary diseases with possible consequences for future pregnancies.
Table 4Reported Risk Factors for Adverse Pregnancy Outcome Before Pregnancy
| PCC (n=114) | Standard Care (n=1,158) | |||
|---|---|---|---|---|
| n | % | n | % | |
| Smoking before pregnancy | ||||
| Yes | 28 | 24.8 | 309 | 26.8 |
| No | 85 | 75.2 | 842 | 73.2 |
| Alcohol use before pregnancy | ||||
| Yes | 81 | 71.7 | 711 | 61.7 |
| No | 32 | 28.3 | 442 | 38.3 |
| Reported hereditary diseases with possible consequences for future pregnancies | ||||
| (Future) mother | 2 | 1.8 | 20 | 1.7 |
| (Future) father | 2 | 1.9 | 16 | 1.4 |
| Offspring | 0 | 0 | 9 | 0.8 |
a Numbers may vary owing to missing data.
Table 5 shows behavioral changes before and during pregnancy. Compared with women receiving standard care, more women with PCC quit smoking before pregnancy (10% vs. 18%; unadjusted OR, 3.04 [95% CI, 0.95–9.69]), and used folic acid in the recommended period (53% vs. 86%; unadjusted OR, 5.40 [95% CI, 3.15–9.28]). Fewer of these women used alcohol in the first 3 months of pregnancy (32% vs. 45%; unadjusted OR, 1.68 [95% CI, 1.03–2.75]). After adjustment for possible confounders, changes in behavior remained statistically significant for folic acid use and for not drinking alcohol in the first 3 months of pregnancy (OR, 4.93; 95% CI, 2.81–8.66, respectively, OR, 1.79 [95% CI, 1.08–2.97]). For other behavior during pregnancy, it was also found that more women who had had PCC reported a healthy lifestyle than women with standard care, although none of these differences were significant.
Table 5Behavioral Factors in Relation to Pregnancy
| PCC | Standard Care | |||||
|---|---|---|---|---|---|---|
| n | % | n | % | OR(95% CI) | Adjusted OR(95% CI) | |
| Smoking during pregnancy (given that tobacco was used before pregnancy) | ||||||
| Quit before pregnancy | 5 | 17.9 | 30 | 9.7 | 3.04 (0.95–9.69) | 2.94 (0.70–8.84) |
| Quit when pregnancy known | 14 | 50.0 | 115 | 37.2 | 2.22 (0.93–5.30) | 1.85 (0.74–4.60) |
| Smoked until few months, then quit | 0 | 0 | 18 | 5.8 | ||
| Smoked entire pregnancy | 7 | 25.0 | 115 | 37.2 | 1.0 | |
| Other | 2 | 7.1 | 31 | 10.0 | ||
| Binge drinking on ≥1 occasion just before or during pregnancy (given that alcohol was used before pregnancy) | ||||||
| Yes | 6 | 7.4 | 77 | 10.8 | 1.0 | |
| No | 75 | 92.6 | 634 | 89.2 | 1.52 (0.64–3.60) | 1.51 (0.63–3.63) |
| Alcohol use during first 3 months of pregnancy (given that alcohol was used before pregnancy) | ||||||
| Yes | 26 | 32.1 | 315 | 44.3 | 1.0 | |
| No | 55 | 67.9 | 396 | 55.7 | 1.68 (1.03–2.75) | 1.79 (1.08–2.97) |
| Alcohol use during rest of pregnancy (given that alcohol was used before pregnancy) | ||||||
| Yes | 37 | 45.7 | 295 | 41.5 | 1.0 | |
| No | 44 | 54.3 | 395 | 58.5 | 0.84 (0.53–1.34) | 0.95 (0.59–1.54) |
| Drug use during pregnancy | ||||||
| Yes | 0 | 7 | 0.6 | NA | ||
| No | 114 | 1,151 | 99.4 | |||
| Folic acid use | ||||||
| Yes, started before pregnancy | 98 | 86.0 | 612 | 53.1 | 5.40 (3.15–9.28) | 4.93 (2.81–8.66) |
| No, | 16 | 14.0 | 540 | 46.9 | 1.0 | |
| Medication during pregnancy | ||||||
| Yes, safe medication | 50 | 47.6 | 470 | 45.3 | 1.35 (0.64–2.82) | 1.31 (0.62–2.77) |
| Yes, (possibly) harmful medication | 9 | 8.6 | 114 | 11.0 | 1.0 | |
| No | 46 | 43.8 | 454 | 45.7 | 1.28 (0.61–2.70) | 1.40 (0.66–2.99) |
a Numbers may vary owing to missing data.
b Each odds ratio indicates behavioral changes that are in favor of pregnancy outcome.
c Adjusted for age, education level and country of birth.
d Reference category for the calculation of the ORs. When multiple categories are marked, these have been combined into 1 reference category.
e The “No” category includes women who never used folic acid during pregnancy and women who started using folic acid from the moment they knew they were pregnant.
f According to the Swedish classification system for the safety of drugs during pregnancy, category A was considered to be safe; the other categories have been combined into the category “(possibly) harmful.” Medication that could not be classified has not been taken into account.
g p < .05.
Compared with women receiving standard care, folic acid use in the recommended period increased strongly among women with PCC (53.1% vs. 86.0%; p < .01). This increase is present not only among women with a high educational level (90.2% vs. 56.4%; p < .01), but also among women with intermediate (84.3% vs. 55.1%; p < .01) and low educational (81.8% vs. 46.5%; p=.02) levels (data not shown).
Table 6 describes pregnancy outcomes of women receiving PCC versus standard care. Among women who received standard care, 20.2% of pregnancies ended in an adverse outcome. Among women who attended PCC, this was 16.2%. This reduction in total adverse pregnancy outcome was in the expected range, but was not significant (OR, 0.77; 95% CI, 0.48–1.22). Also, in the specific adverse pregnancy outcomes, a trend toward improved pregnancy outcome in the PCC group was found (miscarriage, 8.3% vs. 9.3%; preterm birth, 6.2% vs. 8.2%; low birth weight, 3.1% vs. 5.3%; congenital anomalies, 3.9% vs. 4.5%).
Table 6Pregnancy Outcomes of All Included Pregnancies
| PCC | Standard care | |||
|---|---|---|---|---|
| n | % | n | % | |
| No. of pregnancies with known outcome | 145 | 1,740 | ||
| Pregnancy outcomes | ||||
| Miscarriage | 12 | 8.3 | 162 | 9.3 |
| Termination of pregnancy for social reasons | 2 | 1.4 | 60 | 3.5 |
| Termination of pregnancy for medical reason | 3 | 2.1 | 2 | 0.1 |
| Extrauterine pregnancy | 0 | 0 | 6 | 0.3 |
| Perinatal death | 1 | 0.7 | 14 | 0.8 |
| Live birth | 127 | 87.6 | 1,496 | 86.0 |
| Live born children | 129 | 1,520 | ||
| Gestational age | ||||
| Preterm (<37 weeks) | 8 | 6.2 | 124 | 8.2 |
| Term (≥37 weeks) | 119 | 92.2 | 1,356 | 89.2 |
| Missing data | 2 | 1.6 | 40 | 2.6 |
| Birth weight | ||||
| Low (<2,500 g) | 4 | 3.1 | 81 | 5.3 |
| Normal (≥2,500 g) | 102 | 79.1 | 1,124 | 73.9 |
| Missing data | 23 | 17.8 | 315 | 20.7 |
| Growth <p2.3 (small for gestational age) | ||||
| Yes | 1 | 0.8 | 25 | 1.6 |
| No | 105 | 81.4 | 1,160 | 76.3 |
| Missing data | 23 | 17.8 | 335 | 22.0 |
| Congenital anomalies | ||||
| Yes | 5 | 3.9 | 68 | 4.5 |
| No | 124 | 96.1 | 1,452 | 95.5 |
| Total adverse pregnancy outcomes d Adverse pregnancy outcomes include miscarriage, extrauterine pregnancy, perinatal death, preterm birth, birth weight <2,500g, small for gestational age (<p2.3), and congenital anomalies. Normal outcomes (e.g., term birth, normal birth weight, and not small for gestational age) were assumed for children born alive but on whom data were missing regarding duration of pregnancy, birth weight, and weight for gestational age. When a pregnancy had multiple adverse outcomes (e.g., preterm birth and low birth weight), it was counted only once. | 23 | 16.2 | 343 | 20.2 |
a Perinatal deaths are all pregnancies with stillbirth or infant death ≤1 month after birth.
b The number of live-born children exceeded the number of pregnancies ending in live birth owing to twins.
c Only for children who were born alive.
d Adverse pregnancy outcomes include miscarriage, extrauterine pregnancy, perinatal death, preterm birth, birth weight <2,500 g, small for gestational age (<p2.3), and congenital anomalies. Normal outcomes (e.g., term birth, normal birth weight, and not small for gestational age) were assumed for children born alive but on whom data were missing regarding duration of pregnancy, birth weight, and weight for gestational age. When a pregnancy had multiple adverse outcomes (e.g., preterm birth and low birth weight), it was counted only once.
Discussion
Compared with matched women who received standard care, women who had attended PCC had greater knowledge about hazardous substances (73.0% vs. 69.2%), infection prevention (94.2% vs. 89.8%), folic acid intake (91.9% vs. 80.6%), timing of conception (77.0% vs. 74.4%), and the total of 20 essential items (81.5% vs. 76.9%; Table 1). These differences are all significant, with the exception of the difference in knowledge for infection prevention. Furthermore, women who attended PCC gained this knowledge before pregnancy (Table 2). Not only essential knowledge increased, knowledge of personal risk factors increased as well. For instance, women who smoked answered more items correctly regarding the possible hazards of smoking for the unborn child (data not shown).
More important, PCC resulted in behavioral changes both before and during pregnancy. Compared with the control group, more women who received PCC quit smoking before pregnancy and used folic acid in the recommended period; and fewer of these women drank alcohol in the first 3 months of pregnancy (Table 5). A somewhat lower percentage of adverse pregnancy outcomes was found when pregnancies had been preceded by PCC (Table 6).
Historically, there have been very few initiatives offering a comprehensive PCC program, covering multiple risk factors and a subsequent number of preventive measures. Prior initiatives have not measured women's knowledge of pregnancy-related risk factors and preventive measures or behavior both before and after PCC was provided. Effects on knowledge and behavior have only been measured for separate risk factors. For instance, several studies have described increased awareness and use of folic acid after personal counseling (
Pastuszak et al., 1999
, Sayers et al., 1997
). This is in agreement with the finding that a higher number of items on folic acid were answered correctly after PCC as well as the increase in folic acid use before pregnancy.The main limitation of this study is the small number of women attending PCC. This is partially explained by the GPs' large-scale exclusion of eligible women (
Elsinga et al., 2006
). Consequently, it was not possible to conduct an intention-to-treat analysis, but we analyzed this study as if it had been an observational study. Of the women who attended PCC, a higher proportion had an intermediate or high educational level. For the analysis of the knowledge data, women with PCC were therefore matched with women of the control group on the basis of educational level, besides other demographic characteristics. Even after adjustment, women who had attended PCC revealed a higher level of knowledge, irrespective of their educational level. For the analysis of the data on behavior before and during pregnancy and pregnancy outcomes, women after PCC were compared with those who had received standard care in the control group. To compare the results in these groups, the analysis accounted for differences in age, educational level, and country of birth.The knowledge assessment questionnaire consisted of 94 items on a broad range of subjects. To be able to compare the knowledge applicable to all women, we selected 20 essential items comprising different aspects of timing of conception, and risks that either apply to all women, or risks woman can easily encounter. This allowed for accurate comparisons between knowledge in different groups of women.
Some women who attended PCC completed the knowledge questionnaire both in 2000 and 2003. Completing the questionnaire in 2000 may have induced a learning effect. However, comparison with the level of knowledge in 2003 did not show any differences between women who did and did not complete the 2000 questionnaire, suggesting that completing the knowledge assessment at the start of the project did not influence the level of knowledge at the end of the project.
Time in itself may have been a factor responsible for an increase in knowledge. In 2000, we assessed baseline knowledge in half the women at the general practices who were going to be offered PCC; in 2003, knowledge assessment was repeated in a similar random selection of the women registered at the general practices offering standard care. In this way, we could detect any differences in knowledge that occurred in the general population over time. After adjusting for age, country of birth, and educational level, the results show that, over time, knowledge of the 20 essential items had increased in the general population by 3.2%. Women who attended PCC and completed both questionnaires displayed a significant higher increase in knowledge than time in itself had caused.
We cannot rule out the possibility that knowledge increased and behavior changed among women who chose not to attend PCC—after all, the invitation may have induced greater awareness of the PCC issues. Whatever the case, we found that women who attended PCC attained a high level of knowledge before pregnancy, and were thus known in time to minimize risks. Furthermore, behavior during pregnancy among women who chose not to attend PCC was similar to or less favorable than that of women who received standard care (data not shown). We therefore conclude that risk reducing behavior was brought about by actual PCC attendance, and not by any increased awareness that may have been induced by the invitation to attend PCC.
Because PCC is a new concept, it is reasonable to assume that its very newness also influenced participation. First, we believe women to be hesitant toward such a new concept. Second, those women who did participate were relatively well educated and thus probably more motivated. This stronger motivation may also have contributed to a higher tendency to learn and change behavior. This suggests that the results we describe may overestimate the real impact. However, several studies have shown that, regardless of their motivational status, more people changed their risk behavior after personal counseling than people with other forms of counseling or with no counseling at all (
Floyd et al., 1993
, Mullen et al., 1994
, Secker-Walker et al., 1998
, Walsh and Redman, 1993
). This suggests that behavioral changes would also occur if PCC were more common.It should also be noted that in the calculations for the percentage of total adverse pregnancy outcomes, data missing on duration of pregnancy, birth weight, and weight for gestational age were recorded as “normal” because abnormal outcomes are almost always accompanied by a letter from a specialist or a remark from the GP in the file. The percentage of adverse pregnancy outcomes is, therefore, a conservative estimate.
Although this study indicates that PCC can have positive effects on knowledge and risk behavior favoring pregnancy outcomes, the target group is not easy to reach for PCC. This was not only the case in our study, but seems to be a general problem, encountered by other initiatives in the Netherlands as well. In Rotterdam, only a limited amount of women attended a preconception consultation after a door-to-door distribution of 15,000 information leaflets; various initiatives of Dutch midwives have resulted in only a few consultations as well (
Aalhuizen and van der Stouwe, 2008
).It may be useful to educate students in secondary school about the aims and relevance of PCC. This can be combined with sex education and may contribute to reaching all future women of childbearing age. In addition, new national initiatives are needed to raise awareness about the value of PCC. The costs of such initiatives are substantial. However, PCC may save health care costs in the long term, for example, by reducing the number of preterm births. In our study, we also asked GPs about the impact of PCC on their time. Although GPs indicated that a preconception consultation was time consuming, the time spent on the first pregnancy consultation was reduced. Therefore, it is an option that health insurance companies will fund preconception consultations, because they will also profit from long-term cost reductions. Furthermore, the behavioral changes initiated by PCC also improve the health of the future parents themselves, another reason why funding by health insurance companies seems a reasonable option.
Because only a small number of pregnancies were preceded by PCC, we could only give a first impression of pregnancy outcomes. These seemed to be improved in the group of women who attended PCC compared with the control group, but the differences were not statistically significant (OR, 0.77; 95% CI, 0.48–1.22). Furthermore, adjustments for demographic differences were not made. Therefore, the results should be interpreted with caution. But if there are real differences and PCC attendance increases over time, PCC may prevent a considerable number of miscarriages, preterm births, babies born at a low birth weight, and congenital anomalies.
Acknowledgments
The authors thank all the general practitioners and their patients for their participation in this study.
References
- [The first experiences with preconception care].Tijdschr v Verloskundigen. 2008; : 33
- Toename van de moedersterfte een reden tot ongerustheid?.Ned T Obstet Gyn. 2001; 114: 75-76
- Preconception health care.American Family Physician. 2002; 65: 2507-2514
- Preconceptional health promotion: A practical guide.2nd ed. Mosby-Yearbook, St. Louis1995
- Recommendations for use of folic acid to reduce number of spina bifida cases and other neural tube defects.Journal of the American Medical Association. 1993; 269: 1233-1238
de Jong-Potjer, L.C., Elsinga, J., le Cessie, S., Pal-de Bruin, K.M., Schoorl, E., Sneeuw, K.C.A., et al. (2008). Future mother's knowledge of pregnancy-related risk factors: The need for preconception care. (in press).
- Preconception counselling initiated by general practitioners in the Netherlands: Reaching couples contemplating pregnancy.BMC Family Practice. 2006; 7: 41-49
- A review of smoking in pregnancy: Effects on pregnancy outcomes and cessation efforts.Annual Review of Public Health. 1993; 14: 379-411
- Preconceptional counseling and care: A unique window of opportunity. Medicine and Health.Rhode Island. 2003; 86: 16-18
- Perinatal mortality in the Netherlands. Backgrounds of a worsening international ranking.Demographic Research. 2004; 11: 357-394
- Preconception education: Caring for the future.Journal of Pediatric Health Care. 2004; 18: 40-44
- Preconception management of chronic diseases.Journal of the South Carolina Medical Association. 2002; 98: 270-276
- Diabetes and pregnancy. Preconception care, pregnancy outcomes, resource utilization and costs.Journal of Reproductive Medicine. 1999; 44: 33-38
- Preconception care: “VITAL MOM”—A guide for the primary care provider.Journal of the American Osteopathic Association. 2001; 101: S1-S9
- Preconceptional counseling and intervention.Archives of Internal Medicine. 1994; 154: 2273-2280
- Interventions for promoting smoking cessation during pregnancy.The Cochrane Library. 1999; 2: 1-43
- A focused preconceptional and early pregnancy program in women with type 1 diabetes reduces perinatal mortality and malformation rates to general population levels.Journal of Maternal and Fetal Medicine. 2000; 9: 14-20
- Preconceptional health promotion: opportunities abound.Maternal and Child Health Journal. 2002; 6: 71-73
- Preconceptional wellness as a routine objective for women's health care: An integrative strategy.Journal of Obstetric, Gynecologic, and Neonatal Nursing. 2003; 32: 550-556
- Preconceptional health promotion: Progress in changing a prevention paradigm.Journal of Perinatal and Neonatal Nursing. 2004; 18: 2-13
- An ounce of prevention: the evidence supporting periconception health care.Journal of Family Practice. 2004; 53: 126-133
- A meta-analysis of randomized trials of prenatal smoking cessation interventions.American Journal of Obstetrics and Gynecology. 1994; 171: 1328-1334
- Preconception counseling and women's compliance with folic acid supplementation.Canadian Family Physician. 1999; 45: 2053-2057
- ICPC multilingual collaboratory: A Web- and Unicode-based system for distributed editing/translating/viewing of the multilingual International Classification of Primary Care.Medinfo. 2004; 11: 425-429
- A survey of knowledge and use of folic acid among women of child-bearing age in Dublin.Journal of Public Health Medicine. 1997; 19: 328-332
- Confidential enquiries into maternal death in the Netherlands 1983–1992 Thesis.Rijksuniversiteit, Leiden1998
- Reducing smoking during pregnancy and postpartum: Physician's advice supported by individual counseling.Preventive Medicine. 1998; 27: 422-430
- Preconceptional counseling.Obstetrical and Gynecologic Clinics of North America. 1988; 15: 167-178
- Preconception counselling in primary care: Prevalence of risk factors among couples with pregnancy wish.Paediatric and Perinatal Epidemiology. 2008; 22: 280-287
- Smoking cessation in pregnancy: Do effective programmes exist?.Health Promotion International. 1993; 8: 111-127
Biography
Joyce Elsinga, PhD, is a researcher at the Department of Public Health and Primary Care at Leiden University Medical Center. Her research interests focus on preconception counseling.
Lieke C. de Jong-Potjer, PhD, MD, is researcher at the Department of Public Health and Primary Care at Leiden University Medical Center and general practitioner in Zoetermeer. Her interests focus on preconception counseling.
Karin M. van der Pal-de Bruin, PhD, is epidemiologist at TNO Quality of life. Her research focuses on reproduction and perinatology.
Saskia le Cessie, PhD, is an Associate professor of Medical Statistics at the Leiden University Medical Center. Her interests focus on logistic regression, survival analysis, goodness of fit, prediction and reliability.
Willem J.J. Assendelft, PhD, MD, is professor and head of the department of Public Health and Primary Care at the Leiden University Medical Center. He is involved in connecting research to daily general practice. His interests focus on public health, primary care, systematic reviews and evidence based medicine.
Simone E. Buitendijk, PhD, MD, is programmanager Youth at TNO Quality of Life and professor at the Leiden University Medical Center. She is involved in preventive health care for children.
Article info
Publication history
Accepted:
September 9,
2008
Received in revised form:
September 9,
2008
Received:
February 20,
2008
Footnotes
The authors have no direct financial interests that might pose a conflict of interest in connection with the submitted manuscript.
The Netherlands Organisation for Health Research and Development (ZonMw) provided funding for this study. Professor Assendelft is member of a scientific committee of ZonMw, from which he derives no financial gain. The other authors declare that they are independent of the Netherlands Organisation for Health Research and Development (ZonMw).
Identification
Copyright
© 2008 Jacobs Institute of Women's Health. Published by Elsevier Inc. All rights reserved.
