SEARCH

SEARCH BY CITATION

Keywords:

  • Low birthweight;
  • placental complication;
  • preterm birth;
  • small-for-gestational age;
  • termination of pregnancy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

Objective

To compare the effect of medical versus surgical termination of pregnancy (TOP), performed in primigravid women, on subsequent delivery.

Design

Population-based register study.

Setting

Finland 2000–2009.

Population

All primigravid women (n = 8294) who underwent TOP during first trimester of pregnancy by medical (n = 3441) or surgical (n = 4853) method, and whose subsequent pregnancy resulted in singleton delivery.

Methods

The women were identified in the Finnish Register of Induced Abortions, and the data were linked to the Medical Birth and the Hospital Discharge Registries.

Main outcome measures

Risk of preterm birth, low birthweight, small-for-gestational-age (SGA) infant and placental complications (placenta praevia, placental abruption, retained placenta, placenta accreta).

Results

No statistically significant differences in the incidences of preterm birth (4.0% in the medical group versus 4.9% in the surgical group), low birthweight (3.4% versus 4.0%), SGA infants (2.6% versus 2.9%) or placental complications (2.6% versus 2.8%) emerged between the two groups. After adjusting for various background factors, medical TOP was not associated with significantly altered risks of preterm birth (odds ratio [OR] 0.87, 95% confidence interval [95% CI] 0.68–1.13), low birthweight (OR 0.90, 95% CI 0.68–1.19), SGA infant (OR 0.87, 95% CI 0.64–1.20) or placental complications (OR 0.98, 95% CI 0.72–1.34) versus surgical TOP. In a sub-analysis excluding women who underwent surgical evacuation following the index TOP, medical TOP was associated with a reduced risk of preterm birth (< 0.01), but the difference became insignificant after adjusting for gestational age at the time of TOP, inter-pregnancy interval, maternal age, cohabitation status, socio-economic status, residence and smoking during pregnancy.

Conclusions

A history of one medical versus surgical TOP, performed in primigravid women, is associated with similar obstetric risks in the subsequent delivery.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

The method of bringing about termination of pregnancy (TOP) has changed significantly during the past decade. Traditional first-trimester surgical TOP has been increasingly replaced by medical TOP—the antiprogestin mifepristone followed by prostaglandin—in several countries. Mifepristone was first introduced in France in 1988. In 2000 mifepristone was granted market authorisation in several countries of the European Union. In Finland medical TOP has been available since 2000. Since then, the use of medical TOP has expanded within Europe.[1] In 2010, 43% of all TOPs in the UK,[2] 71% in Sweden[3] and 87% in Finland[4] were performed using the medical method.

In the Nordic countries the incidence of TOP is highest among the 20–24-year age group[4] and the same trend can be seen worldwide.[2, 5] Most women seeking termination of pregnancy have no history of previous induced abortions. In 2010 the proportion of women undergoing their first TOP was 63% in Finland,[4] 66% in the UK[2] and 58% in Sweden.[3] Moreover, approximately half of all TOPs are performed in women with no previous childbirths.[2-4]

The results of a recent meta-analysis suggested that a history of one TOP is associated with an increased risk of preterm birth, but not with the risk of having a small-for-gestational-age (SGA) infant.[6] The risk of preterm birth increased as the number of TOPs increased. The studies included in the meta-analysis often lacked information on gestational age at the time of TOP and the method of TOP. In many studies focusing on surgical TOP it has been concluded that a single surgical abortion in the first trimester does not increase the risk of preterm birth, low birthweight, or placenta praevia in a subsequent pregnancy.[7-10] However, long-term safety data concerning the effects of medical TOP on subsequent pregnancy are still scarce.

The purpose of the present study was to compare the effects of medical versus surgical TOP, performed in primigravid women, on subsequent delivery. The study was especially focused on the risks of preterm birth, low birthweight, SGA infants and placental complications (placenta praevia, placental abruption, retained placenta and placenta accreta). To eliminate the possible influence of surgical re-evacuation of the uterus following medical or surgical TOP, a subgroup analysis was performed among women without re-evacuation.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

The study was approved by the Ethics Committee of the Northern Ostrobothnia Hospital District. The National Institute for Health and Welfare gave permission in August 2010 to use confidential personal-level data obtained from the national health registers.

In this register-based national study all primigravid women who underwent TOP during the first trimester of gestation in Finland between 2000 and 2009 were identified in the Finnish Register of Induced Abortions. First trimester was defined as the period of gestation up to 84 days (i.e. ≤12+0 weeks). To collect information on subsequent pregnancies resulting in delivery, the data obtained from the Register of Induced Abortions were linked to the Finnish Medical Birth Register and to the Hospital Discharge Register. All women whose subsequent pregnancy, following the first TOP, ended in singleton delivery, either live or stillborn, were included in the study. Data on subsequent pregnancies up to the end of 2009 were obtained. Women who had miscarriages or underwent another TOP between the index TOP and the first delivery were excluded. Women whose pregnancies continued despite the attempted TOP (ongoing pregnancy) and ended in delivery were also excluded from the study.

The study population was divided into groups of medical and surgical TOP according to the method used at the time of the index TOP. The medical group included women whose TOP was performed with mifepristone alone, or with a combination of mifepristone and misoprostol. The women were assigned to the surgical group if the TOP was performed using either dilatation and curettage, or vacuum aspiration. Furthermore, to focus on the method of TOP per se, women who required surgical re-evacuation after surgical TOP, or evacuation after medically induced TOP up to 42 days after the procedure were identified, and a subgroup analysis was conducted in which these women were excluded. Such women were identified through the Hospital Discharge Register.

According to current Finnish legislation, induced abortion can be performed up to 20 weeks of gestation, or in cases of confirmed fetal anomalies, up to 24 weeks. The legislation requires an indication and approval (by one or two physicians up to 12 weeks, or by a national authority between 12 and 24 weeks) for terminating a pregnancy.[11]

Data collection

The Finnish Register of Induced Abortions has been maintained since 1977 by the National Institute for Health and Welfare.[4] Finnish legislation demands physicians to report TOPs to this register using a specific data collection form. Data are collected from all hospitals and clinics in which TOPs are performed. The data cover information on pregnancy history, occupation, residence, municipality and marital status and information concerning the current pregnancy, such as gestational age at TOP confirmed generally by ultrasonography, contraception before and after TOP, indication and the method of TOP.

All live births and stillbirths of at least 22 weeks of gestation or newborns weighing at least 500 g are registered in the Finnish Medical Birth Register, which is also maintained by the National Institute for Health and Welfare (since 1987).[4] Data on each pregnancy and delivery are reported to the register by using a specific data collection form filled in by the midwife in charge of the delivery. In this study most of the background characteristics (i.e. maternal age, occupation, marital status and residence at the time of birth and smoking during pregnancy) were obtained from the Medical Birth Register. Socio-economic status was defined using the stated occupation or the highest educational level reported to the Medical Birth Register. Coding was based on national standards published by Statistics Finland.[12-14] The groups were further aggregated into five socio-economic status grades: upper white-collar workers, lower white-collar workers, blue-collar workers, students and pupils (level of education not defined), and others (all other groups, such as entrepreneurs, farmers, unemployed women, and housewives with undetermined socio-economic status).

The outcome measures, gestational age at birth, birthweight and some information on placental complications, were obtained from the Medical Birth Register. In this register gestational age at birth is determined according to the last menstruation and in most women also by ultrasonographic examination in early pregnancy. Birth was defined as preterm when gestational age at birth was <37 weeks, very preterm when gestational age was <32 weeks and extremely preterm when gestational age was below 28 weeks.[15] Low birthweight was defined as a birthweight <2500 g. Newborns were defined as SGA if their birthweight was less than the mean – 2SD when compared with the expected birthweight for the same gestational age and sex according to Finnish standards.[16] The SGA status could not be defined among infants who were born at <24 weeks or more than 43 weeks of gestation.

Hospitals in Finland are required to provide information to the Hospital Discharge Register on all inpatient treatment (all hospitals) and outpatient visits (public hospitals). This register contains data on diagnosis (International Classification of Diseases; ICD-10[17]) and procedures (Nordic Classification of Surgical Procedures[18]) of treatment and includes dates of visits. From this register we obtained information on surgical uterine re-evacuations up to 42 days after TOP, and we also complemented the information obtained from the Medical Birth Register concerning placental complications by using specific ICD-10 codes for placenta praevia (O44.0, O44.1), placental abruption (O45.0, O45.8, O45.9), retained placenta (O73.0, O73.1) and placenta accreta (O43.2), and the specific procedure code for removing the placenta manually (MBA30).

Data linkage between the three data sources was performed by using the women's personal identification numbers, which were removed from the data before analysis.

Statistical analysis

Statistical analyses were performed using pasw 18.0 for Windows (SPSS Inc., Chicago, IL, USA). To assess differences between the study groups, the chi-square test or Fisher's exact test were used as appropriate for categorical variables and the Mann–Whitney U-test for continuous variables. The level of statistical significance was set at P < 0.05. Logistic regression analysis was used to identify the risks of preterm birth, low birthweight, SGA infants and placental complications in the medical TOP group in comparison with the surgical TOP group. Both univariate and multivariate analyses were performed. The potential confounders used in the multivariate analyses were gestational age at the time of TOP, the time between TOP and subsequent pregnancy (inter-pregnancy interval), maternal age, cohabitation status, socio-economic status, residence and smoking during pregnancy, because these factors are associated with adverse pregnancy outcomes. The last five confounders were identified at the time of the pregnancy following the index TOP. The estimates are presented as odds ratios (OR) and 95% confidence intervals (95% CI).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

A total of 88 522 women underwent TOP during 2000–2009 (Figure 1). Of these, 39 312 (44%) were primigravid and underwent TOP during the first trimester of gestation. Among these women there were 8307 women whose next pregnancy ended in singleton delivery. Women who had an ongoing pregnancy (n = 13) were excluded. Hence the final population included in the study consisted of 8294 women; 3441 in the medical TOP group and 4853 in the surgical TOP group.

image

Figure 1. Flow chart detailing the study groups.

Download figure to PowerPoint

The background characteristics of the two groups differed in several respects (Table 1). In the medical TOP group the duration of gestation at the time of the index TOP and the inter-pregnancy interval were shorter than in the surgical group. Women in the medical group were more often living with a partner and less often smoked through pregnancy, whereas women in the surgical group were younger and more often were of lower socio-economic status.

Table 1. Demographic characteristics; number of women (%)
CharacteristicsMedical TOP(n = 3441)Surgical TOP(n = 4853)P valuea
  1. IQR, interquartile range.

  2. a

    Mann–Whitney U test for continuous variables and chi-square test for categorical variables.

Maternal age, years
Geometric mean (IQR)24.6 (21–28)24.1 (21–27)<0.001
≤ 19375 (10.9)646 (13.3)<0.001
20–241437 (41.8)2069 (42.6) 
25–29987 (28.7)1397 (28.8) 
30–34486 (14.1)547 (11.3) 
35–39137 (4.0)166 (3.4) 
≥ 4019 (0.6)28 (0.6) 
Cohabitation
Yes2719 (79.0)3569 (73.5)<0.001
No460 (13.4)845 (17.4)
Data missing262 (7.6)439 (9.0)
Socioeconomic status
Upper white-collar worker306 (8.9)346 (7.1)0.002
Lower white-collar worker912 (26.5)1211 (25.0)
Blue-collar worker615 (17.9)887 (18.3)
Students735 (21.4)1070 (22.0)
Other114 (3.3)218 (4.5)
Data missing759 (22.1)1121 (23.1)
Type of residence
Urban2526 (73.4)3554 (73.2)0.828
Densely populated474 (13.8)665 (13.7)
Rural437 (12.7)631 (13.0)
Data missing4 (0.1)3 (0.1)
Gestational age at TOP, weeks
Median (IQR)7 (7–8)9 (8–10)<0.001
≤ 6759 (22.1)220 (4.5)<0.001
71112 (32.3)591 (12.2) 
8919 (26.7)997 (20.5) 
9492 (14.3)1036 (21.3) 
1077 (2.2)956 (19.7) 
1146 (1.3)753 (15.5) 
1236 (1.0)300 (6.2) 
Inter-pregnancy interval, years
Median (IQR)2.1 (1.1–3.5)2.6 (1.3–4.5)<0.001
<1801 (23.3)922 (19.0)<0.001
1–21505 (43.7)1747 (36.0) 
≥31123 (32.6)2156 (44.4) 
Data missing12 (0.3)28 (0.6) 
Smoking during pregnancy
No smoking2183 (63.4)2924 (60.3)<0.001
Stopped smoking in first trimester353 (10.3)436 (9.0)
Smoked through pregnancy828 (24.1)1383 (28.5)
Data missing77 (2.2)110 (2.3)

The incidences of preterm, very preterm and extremely preterm birth did not differ statistically significantly between the medical and surgical TOP groups. Neither were significant differences noted in the incidences of low birthweight, SGA infants, placenta praevia, retained placenta or placental abruption between the study groups (Table 2). There were no cases of placenta accreta in the study population.

Table 2. Risk of adverse events in subsequent delivery according to the method of TOP; surgical method as reference; n (%)
OutcomeMedical TOPSurgical TOP P a OR (95% CI) unadjustedOR (95% CI) adjustedb P b
  1. NA, not applicable.

  2. a

    P value in chi-square test or Fisher's exact test, as appropriate.

  3. b

    Analyses were adjusted for maternal age, cohabitation position, socio-economic status, residence, gestational age at time of TOP, inter-pregnancy interval and smoking during pregnancy.

  4. Information on residence was missing for four women in the medical group and three in the surgical group. Information on gestational age at birth was missing for 12 women in the medical group and 28 in the surgical group and information on birthweight was missing for six women in the medical group and nine in the surgical group. Information on SGA infants was missing for 16 women in the medical group and 33 in the surgical group. In seven women the gestational age at birth was less than 24 weeks. Women with missing information were excluded from multivariable analysis.

Preterm birth
<37 weeks of gestation138/3429 (4.0)236/4825 (4.9)0.0620.82 (0.66–1.01)0.87 (0.68–1.13)0.294
<32 weeks of gestation20/3429 (0.6)36/4825 (0.7)0.3740.78 (0.45–1.35)NA
<28 weeks of gestation6/3429 (0.2)12/4825 (0.2)0.4790.70 (0.26–1.88)NA
Placental complications 91/3441 (2.6)134/4853 (2.8)0.7470.96 (0.73–1.25)0.98 (0.72–1.34)0.905
Placenta praevia5/3441 (0.1)7/4853 (0.1)1.0001.01 (0.32–3.18)NA
Retained placenta75/3441 (2.2)112/4853 (2.3)0.6980.94 (0.70–1.27)NA
Placental abruption11/3441 (0.3)15/4853 (0.3)0.9321.03 (0.48–2.26)NA
Low birthweight <2500 g 116/3435 (3.4)194/4844 (4.0)0.1380.84 (0.66–1.06)0.90 (0.68–1.19)0.450
Small-for-gestational age 89/3425 (2.6)139/4820 (2.9)0.4360.90 (0.69–1.18)0.87 (0.64–1.20)0.401

When compared with surgical TOP, medical TOP was not associated with significantly altered risks of preterm birth, low birthweight, SGA infants or placental complications in univariate analysis (Table 2). The results remained unchanged in multivariate analysis when adjusted for gestational age at the time of induced abortion, inter-pregnancy interval, maternal age, cohabitation status, socio-economic status, residence and smoking during pregnancy. Additional adjustment for year of TOP or for year of birth did not change the result. Gestational age at the time of TOP and the inter-pregnancy interval were not associated with an increased risk of premature birth (data not shown).

There were 208 (6.0%) women who required surgical evacuation after medical TOP and 228 women (4.7%) who underwent re-evacuation after surgical TOP. The difference was statistically significant (P = 0.007). In subgroup analysis excluding these women, medical TOP was associated with a significantly reduced risk of preterm birth in comparison with surgical TOP in univariate analysis (OR 0.74, 95% CI 0.59–0.93). However, the difference became insignificant after adjusting for gestational age at the time of TOP, inter-pregnancy interval, maternal age, cohabitation status, socio-economic status, residence and smoking during pregnancy (OR 0.79, 95% CI 0.61–1.04) (Table 3).

Table 3. Risk of adverse events in subsequent delivery according to the method of TOP among women with no re-procedure; surgical method as reference
OutcomeMedical TOPSurgical TOP P a OR (95% CI) unadjustedOR (95% CI) adjusted P b
  1. NA, not applicable.

  2. a

    P value for chi-square test or Fisher's exact test, as appropriate.

  3. b

    Analyses were adjusted for maternal age, cohabitation position, socio-economic status, residence, gestational age at time of TOP, inter-pregnancy.

  4. interval and smoking during pregnancy. Information on residence was missing for four women in the medical group and three in the surgical group. Information on gestational age at birth was missing for 11 women in the medical group and 26 in the surgical group and information on birthweight was missing for six women in the medical group and nine in the surgical group. Information on SGA infants was missing for 15 women in the medical group and 31 in the surgical group. Women with missing information were excluded from multivariate analysis.

Preterm birth
<37 weeks of gestation119/3222 (3.7)226/4599 (4.9)0.0100.74 (0.59–0.93)0.79 (0.61–1.04)0.090
<32 weeks of gestation17/3222 (0.5)35/4599 (0.8)0.2110.69 (0.39–1.24)NA
<28 weeks of gestation6/3222 (0.2)11/4599 (0.2)0.6210.78 (0.29–2.11)NA
Placental complications 86/3233 (2.7)129/4625 (2.8)0.7300.95 (0.72–1.26)0.98 (0.71–1.36)0.903
Placenta praevia5/3233 (0.2)7/4625 (0.2)1.0001.02 (0.32–3.22)NA
Retained placenta70/3233 (2.2)108/4625 (2.3)0.6180.93 (0.68–1.26)NA
Placental abruption11/3233 (0.3)14/4625 (0.3)0.7711.12 (0.51–2.48)NA
Low birthweight <2500 g 103/3227 (3.2)184/4616 (4.0)0.0650.79 (0.62–1.02)0.86 (0.64–1.15)0.318
Small-for-gestational age 83/3218 (2.6)131/4594 (2.9)0.4680.90 (0.68–1.19)0.87 (0.63–1.21)0.411

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

In the present study a history of one medical TOP, performed in primigravid women, was not associated with altered risks of preterm birth, low birthweight, SGA infants or placental complications in subsequent pregnancy when compared with surgical TOP. There were clinically minor but statistically significant differences in the background characteristics of the two study groups. The risks of adverse effects were calculated after adjustment for these factors. The re-evacuation rate was higher in the medical TOP group than in the surgical group, as also shown in one of our earlier studies.[19] After excluding women who required re-evacuation, medical method was associated with a reduced risk of preterm birth compared with surgical method, but the difference became insignificant after adjusting for background.

To our knowledge, this is the first register-based study, which compares the potential adverse effects of one medical and surgical TOP on subsequent delivery among nulliparous women. The majority of women requesting TOP have no history of previous TOP (for example 63% in Finland and 66% in the UK), and approximately half of TOPs are performed among women with no previous childbirths.[2, 4] Hence, our data are applicable among a large group of women.

Strengths of our study include the large national cohort and high validity of the data. The validity of the information in the Finnish Abortion Register has been assessed in previous studies where information from hospital records was compared with Abortion Register data. A total of 95% of the data in the Abortion Register was identical to that in the hospital records and coverage of the register was 99%.[20, 21] The quality of the Finnish Medical Birth Register has also been shown to be high.[22] Moreover, we controlled for several important confounders that may influence adverse pregnancy outcomes, including smoking, which is a known risk factor of placental abruption, placenta praevia, SGA infants and preterm birth.[23, 24]

One of the limitations of our study is that we could not make a distinction between iatrogenic and spontaneous preterm births. As the two study groups differed slightly but significantly by several background characteristics, we cannot exclude the possibility that iatrogenic preterm births might have been more common in one of the groups. Due to absence of a national register on miscarriages with data on gestational age at the time of the miscarriage, another shortcoming is the lack of data on (late) miscarriages.

The selection of an appropriate control group may be problematic in studies concerning the safety of TOP. Women who choose TOP differ in several ways from those who decide to continue their pregnancies–a range of sociodemographic and lifestyle factors are associated with TOP.[25] Selecting women with no TOP history as a control group might introduce a potential bias. Hence, we chose to compare women undergoing medical versus surgical TOP. In previous studies a single surgical TOP by vacuum aspiration in the first trimester has been found to be safe as regards the risks of premature birth, low birthweight and placenta praevia.[7-10]

The incidence of preterm birth in most developed countries is approximately 5–7% of all live births. The rate of preterm birth occurring before 32 weeks is approximately 1–2% and it is 0.4% before 28 weeks of pregnancy.[26] In Finland the proportion of premature births (< 37 weeks of gestation) was 5.2% of all deliveries in 2010.[4] In our study the incidences of preterm, very preterm and extremely preterm births were somewhat lower in both groups when compared with national figures and those reported in the literature. This suggests that the risk of preterm delivery following a single medical or surgical TOP is not elevated when compared with the overall risk at the population level. The incidences of low birthweight, SGA infants and placental abruption in both study cohorts were also consistent with the national figures, being 4.5%, 2.5% and 0.3%, respectively (unpublished data from the Medical Birth Register).[4]

The incidence of placenta praevia was lower than that reported in the literature (0.3–0.5%)[27] and lower than the incidence seen at a national level (0.4%; unpublished data from the Medical Birth Register and Hospital Discharge Register). The women included in our study were all nulliparous, which is likely to lower the baseline risk, because the risk factors of placenta praevia include a history of caesarean section and increasing parity.[27] In contrast, in both groups the incidence of retained placenta was higher than the overall incidence in the Finnish population (1.4%; unpublished data from the Medical Birth Register and Hospital Discharge Register). However, in the literature the incidence of retained placenta varies in developed countries between 1.3 and 3.3%, which is consistent with our figures.[28] On the other hand, an increased risk of retained placenta after a single TOP was reported in a Danish study (OR 1.17, 95% CI 1.02–1.35), but the association was weak and confounding factors could not be ruled out.[10]

Only a few studies have concerned the long-term effects of medical TOP, and the results are consistent with ours. In a previous Danish cohort study, medical TOP in the first trimester did not increase the risks of preterm birth, low birthweight, spontaneous abortion or ectopic pregnancy in the next pregnancy when compared with surgical TOP.[29] In that study the women may have undergone repeat medical or surgical TOPs before the subsequent pregnancy. Moreover, the method of index TOP was defined according to the TOP closest in time to the subsequent pregnancy. Hence the influence of repeat TOPs by means of various methods could not be fully eliminated. A large study carried out in China concerned evaluation of the risks of preterm delivery and low birthweight in subsequent pregnancy among nulliparous women after a single TOP.[30] Three groups were compared—women with medical TOP, surgical TOP and women with no history of TOP. The risks of preterm delivery and low birthweight were similar in all cohorts. The investigators concluded that a single early medical TOP in nulliparous women is safe as regards the next pregnancy. Another Chinese study concerned evaluation of the effect of first-trimester medical TOP on placental abruption, placenta praevia, placenta accreta and retained placenta in subsequent pregnancy.[31] No increased risk of placental complications was seen in the following pregnancy after a single medical TOP compared with women with no history of TOP. The overall incidence of placental complications was 2.0% in both cohorts, which is also in line with our results. Both of the Chinese studies relied on self-reported TOP history at the time of subsequent pregnancy, which may have caused recall bias.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References

A history of a single medical versus surgical TOP, performed in primigravid women, is associated with similar risks of preterm birth, low birthweight, SGA infants and placental complications in the subsequent pregnancy. This information is of importance when counselling women requesting TOP in early pregnancy.

Disclosure of interests

All authors declare that they have no competing interests.

Contribution to authorship

All authors participated in the design of the study. JM and AB carried out the data analysis. JM wrote the first draft of the manuscript, which was critically revised by all authors. MG is in charge of the Finnish reproductive registries. JM is a guarantor of the study. MN and OH were the supervisors of the study.

Details of ethics approval

The study was approved by the Ethics Committee of the Northern Ostrobothnia Hospital District in April 2010 (No 28/2010). The National Institute for Health and Welfare gave permission in August 2010 to use confidential personal-level data obtained from the national health registers. The data protection ombudsman was notified about the data linkage as required by national data protection legislation.

Funding

This study was funded by the Oulu University Scholarship Foundation (JM), Helsinki University Central Hospital Research Funds (OH) and University Hospital of Oulu Research Funds (JM).

Acknowledgements

We thank Matti Kesti (National Institute for Health and Welfare, Oulu, Finland) for his technical support.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. References