The Journal of International Medical Research 2009; 37: 227 239

Is Long-term Tocolysis Effective for Threatened Premature Labour?

K TAKAGI1 AND K SATOH2
1Department

FOR THE

MULTICENTRE PREMATURE LABOUR STUDY GROUP

of Obstetrics and Gynaecology, Saitama Medical University, Saitama Medical Centre, Kawagoe-shi, Saitama, Japan; 2Health Check Centre, Social Insurance Yokohama Central Hospital, Yokohama-shi, Kanagawa, Japan

There is some controversy regarding the optimum duration of tocolysis. This retrospective multicentre study was performed to evaluate whether long-term (> 2 days) tocolysis is effective in treating threatened premature labour. A total of 1147 eligible patients were grouped according to whether or not tocolytics were given, and according to route of administration and whether or not ritodrine (the standard tocolytic regimen in Japan) or other tocolytic was given. They were then further stratified into three subgroups, using the Baumgarten and Gruber tocolysis index (TI), to assess the

efficacy of tocolysis treatment according to the risk of premature labour. Prolongation of gestation was significantly longer in patients treated with tocolytics for > 2 days compared with the unmedicated, bed-rest group. In women receiving tocolysis, the mean duration of prolonged gestation was 2.2 times longer in the intermediate-risk TI group and 3.1 times longer in the high-risk TI group compared with the unmedicated group. In the patients who received tocolysis, IV ritodrine was used in 86% of cases and was considered safe and effective for prolonging gestation in cases of threatened premature labour.
WOMEN;

KEY WORDS: RITODRINE; TOCOLYTICS; PREMATURE LABOUR; PREGNANT GESTATION; NEONATAL OUTCOME

PROLONGATION;

Introduction
The recent global increase in pre-term delivery is believed to be due to a greater use of assisted-reproductive technologies and a higher proportion of pregnant women of advanced age.1,2 Premature labour is one of the most frequently occurring disorders during gestation and is the primary cause of neonatal death: 85% of neonatal deaths due to disorders other than deformity are caused by prematurity at birth.3 There are also longterm issues to consider with prematurity, including a higher incidence of

complications such as cerebral palsy, mental retardation and retinopathy in affected infants. Furthermore, the costs of intensive care that premature neonates require can impose a serious economic burden: Jones et al.4 reported that neonatal intensive care unit (NICU) expenses for newborns delivered at 34 weeks gestation were over seven times higher than those for newborns who did not require specific treatment and that they could be reduced to a statistically significant degree with every additional week of gestation.4

227

K Takagi, K Satoh Efficacy of long-term tocolysis

Although a decrease in premature delivery rates seems highly desirable from medical and socio-economic perspectives, as yet there is no globally or locally established standard treatment for threatened premature labour. In addition, there is some controversy regarding the optimum duration of tocolysis. For example, many obstetricians in Europe and even more so in the USA and Canada affirm the use of short-term (< 2 days) tocolysis:5 7 a Canadian study found that tocolytics such as ritodrine hydrochloride prolonged the duration of pregnancy up to 2 days,8 whereas no largescale studies have assessed the clinical efficacy of long-term (> 2 days) tocolysis. The current multicentre, retrospective study was, therefore, conducted to evaluate whether long-term tocolysis, defined as tocolysis that effectively prolongs gestation > 2 days after treatment initiation, is effective in treating threatened premature labour. complications were recorded on admission. Frequency of uterine contractions, uterine bleeding, membrane rupture and cervical dilatation were recorded. Neonatal outcome including whether alive or dead, neonatal weight, gender and 1-min Apgar score were recorded.

DATA ANALYSIS
The tocolysis regimen for each patient was not identical but was selected by each obstetrician. Thus, patients were stratified into two groups: those treated with tocolytics, and those not treated with tocolytics (unmedicated, bed-rest group). Patients treated with tocolytics were sub-classified into the following four treatment groups: intravenous (IV) ritodrine; IV ritodrine plus oral ritodrine; IV ritodrine plus other drugs; and other drugs. In Japan, ritodrine is the standard tocolytic for treating threatened premature labour and the drugs used for tocolysis in the present study are shown in Table 1 The patients in each group were further stratified into three groups according to the tocolysis index (TI) of Baumgarten and Gruber9, in order to assess the efficacy of treatment according to the risk of premature labour. The TI scores of 0 12 were derived from scores of uterine contraction (none 0, irregular 1 and regular 2), premature rupture of membranes (not ruptured 0, suspected rupture 2 and definite rupture 4), genital bleeding (none 0, spotting 1 and moderate 2) and cervical dilatation (no dilatation 0, 1 cm 1, 2 cm 2, 3 cm 3, 4 cm 4). A TI score 7 indicated a poor prognosis.9 The endpoint of gestational duration was set as day 6 of gestation week 35. Those patients who were admitted to hospital before gestation week 32 for the management of pre-term labour and who had TI > 2 were separately divided from all the other patients into two sub-groups based

Patients and methods

PATIENTS
With the exception of pregnant women participating in other clinical trials, all pregnant women who had been admitted for the treatment of threatened premature labour between 1 July and 31 December 1998, at 37 hospitals in Japan participating in the Multicentre Premature Labour Study Group, were included in the study. As this was a retrospective study of outcomes, ethics approval and informed consent were deemed to be unnecessary.

DATA COLLECTION
Patient characteristics including age, parity, gestational age, multiple pregnancy, presence of uterine deformity, membrane rupture, presence of intra-uterine infection, vaginitis, hydramnios, uterine contraction, uterine bleeding, cervical cerclage and other

228

K Takagi, K Satoh Efficacy of long-term tocolysis

TABLE 1: Standard use of tocolytics in Japan and as used in the present study Tocolytic Ritodrine Ritodrine Terbutaline Magnesium sulphate Indomethacin Urinastatin Dose and administration method IV, 50 200 g/min (maximum 200 g/min) Orally, 15 20 mg/day IV, 5 g/min, increasing to a maximum of 80 g/min IV, 2 g over 20 min then 1 2 g/h drip infusion Loading dose of 50 mg rectally or 50 100 mg orally, then 25 50 mg orally every 6 48 h Vaginal suppository, 500 U/day

on TI scores: sub-group A had a TI of 3 6; sub-group B had a TI 7. Patients whose uterine contractions were properly controlled or whose gestation was prolonged beyond week 35 were evaluated as having a successful outcome, and the success rate for each group was compared. The number of days that gestation was prolonged was retrospectively calculated using the KaplanMeier method in order to estimate the clinical efficacy of tocolysis. Neonatal prognoses and the adverse effects of tocolytic agents were also evaluated.

TOCOLYSIS OF CHOICE
Tocolytics were used in 1054 (91.9%) of the patients eligible for analysis, as shown in Table 3 which also shows the frequencies of single and multiple drug use. Tocolysis with a single drug was the most frequent (54.6% of cases), followed by two-drug tocolysis (32.1%). Administration of IV ritodrine occurred in 909 (79.3%) patients, in singleor multiple-drug regimens as shown in Table 4. Drugs other than IV ritodrine were: oral ritodrine following IV therapy (n = 198), magnesium sulphate (n = 228), terbutaline (n = 39), indomethacin (n = 31) and urinastatin (n = 18). The most frequently used combination therapies were ritodrine plus magnesium sulphate (n = 219) and IV plus oral ritodrine (n = 198).

STATISTICAL ANALYSIS
The KaplanMeier method and a log-rank test were used to determine the efficacy of tocolysis. The 2 tests and MannWhitney Utests were used to assess statistical significance. A P-value 0.05 was considered to indicate statistical significance.

PATIENT CLASSIFICATION FOR ANALYSIS

Since 909 (86.2%) of the patients who underwent tocolysis were treated with IV ritodrine, either as monotherapy, IV plus oral therapy or in combination with other agents (Table 4), analysis according to risk of premature labour was conducted on these patients. Prolongation of gestation in patients given tocolysis was compared with that for the 93 patients who were only given unmedicated bed-rest. As can be seen in

Results
PATIENTS
The total number of case records collected was 1310 of which 1147 were eligible for analysis. Eight cases were excluded due to duplication and 155 were excluded due to insufficient records. Details of the baseline characteristics of the 1147 analysed cases are given in Table 2.

229

K Takagi, K Satoh Efficacy of long-term tocolysis

TABLE 2: Baseline characteristics of the 1147 pregnant women with threatened premature labour Characteristics Age, years Mean (range) 19 years 20 29 years 30 39 years 40 years 35 years Gestational age, weeks Mean (range) Parity Primiparous Multiparous Multiple pregnancies Single Multiple Uterus deformity Ruptured membranes No rupture Rupture Intra-uterine infection Present Absent Vaginitis Present Absent Hydramnios Present Absent Uterine contraction Irregular Regular Absent Bleeding Present Absent Cervical cerclage Present Absent Other obstetric complications Present Absent Other complications Present Absent n (%) 29.5 20 577 526 24 182 (16 46) (1.7) (50.3) (45.9) (2.1) (15.9)

30.0 (11.1 36.9) 659 (57.5) 488 (42.5) 971 (84.7) 176 (15.3) 20 (1.7) 980 (85.4) 167 (14.6) 103 (9.0) 1044 (91.0) 256 (22.3) 891 (77.7) 44 (3.8) 1103 (96.2) 649 (56.6) 435 (37.9) 63 (5.5) 221 (19.3) 926 (80.7) 150 (13.1) 997 (86.9) 411 (35.8) 736 (64.2) 152 (13.3) 995 (86.7)

230

K Takagi, K Satoh Efficacy of long-term tocolysis

TABLE 3: Treatment regimens used to treat the pregnant women with threatened premature labour Regimen Total No. of patients Treated with tocolyticsa Not treated with tocolytics Treated with a single drug Treated with multiple drugs: two drugs three drugs four drugs
a

n (%) 1147 1054 93 626 428 368 55 5 (100) (91.9) (8.1) (54.6) (37.3) (32.1) (14.8) (0.4)

Ritodrine, magnesium sulphate, terbutaline, indomethacin and urinastatin.

TABLE 4: Tocolysis index (TI) score and the treatment received for the pregnant women with threatened premature labour TI score Treatment IV ritodrine alone IV ritodrine + oral ritodrinea IV ritodrine + other drugsb No tocolytic drugs (bed rest) Other treatment Total
aStatistically

n 485 198 226 93 145 1147 198 137 85 24 98 542

36

7 (11.5%) (0.5%) (10.6%) (28.0%) (0%) (9.3%)

(40.8%) 231 (47.6%) 56 (69.2%) 60 (30.3%) 1 (37.6%) 117 (51.8%) 24 (25.8%) 43 (46.2%) 26 (67.6%) 47 (32.4%) 0 (47.3%) 498 (43.4%) 107

significant difference in the distribution of patients according to TI score compared with the IV ritodrine alone and IV ritodrine + other drugs groups (P < 0.01; 2-test), indicating more patients at less severe risk in the IV ritodrine + oral ritodrine treated group. The IV ritodrine and IV ritodrine + other drugs groups were not statistically significantly different from each other. bMagnesium sulphate, terbutaline, indomethacin and urinastatin.

Table 4, a total of 444 (44.3%) women treated with IV ritodrine or unmedicated were at TI grade 2; 451 (45.0%) women were between TI grades 3 and 6; and 107 (10.7%) women had TI grades 7. Patient distribution according to TI grade in the groups treated with IV ritodrine alone or in combination with other tocolytics was comparable. In contrast, differences in the distribution pattern were apparent for patients treated with IV ritodrine and oral ritodrine, with significantly more patients having lower TI scores indicative of a lesssevere risk of pre-term labour (P < 0.01; Table

4). For this reason, patients treated with IV ritodrine and oral ritodrine were excluded from the analyses of clinical efficacy.

CLINICAL OUTCOME
Clinical efficacy by severity and tocolysis The number of patients in sub-group A (admission before gestation week 32 and TI score 3 6) totalled 223, consisting of 212 (ritodrine alone and ritodrine + other drugs) and 11 (unmedicated, bed-rest) (Fig. 1). Gestation was significantly prolonged in patients who underwent tocolysis treatment compared with the unmedicated bed-rest

231

K Takagi, K Satoh Efficacy of long-term tocolysis

1.0

Sub-group A n = 212: Ritodrine alone and ritodrine + other drugs n = 11: Unmedicated, bed-rest

Cumulative prolongation rate

0.8

0.6

0.4 P < 0.05 (log-rank test) 0.2

0.0 0 30 60 90 Prolongation of gestation (days) 120 150

FIGURE 1: Cumulative prolongation of gestation among patients with threatened premature labour who were admitted to hospital before gestation week 32 and had a tocolysis index score of 3 6 (sub-group A) and were treated with IV ritodrine alone and IV ritodrine plus other drugs (magnesium sulphate, terbutaline, indomethacin and urinastatin), compared with unmedicated, bed-rest patients

group (P < 0.05; Fig. 1): ritodrine prolonged gestation significantly longer (5 26 days) compared with no medication (2 11 days). In comparison with the unmedicated group, the mean duration of prolongation was 2.2 times longer (P = 1.2 108; Table 5). Similar results were seen in sub-group B

(admission before gestation week 32 and TI score 7); the number of patients in subgroup B was 45, consisting of 38 (ritodrine alone and ritodrine + other drugs) and seven (unmedicated, bed-rest) (Fig. 2). Compared with the unmedicated bed-rest group, patients who underwent tocolysis had a

TABLE 5: Effect of tocolysis on the prolongation of gestation (mean SE days) in the pregnant women with threatened premature labour Tocolysis index score Treatment IV ritodrine alone and IV ritodrine + other drugsd No tocolytic drugs
aP

n 711 93

2 47.7 1.9a (n = 283) 33.5 6.0 (n = 24)

36 26.8 1.4b (n = 348) 12.3 3.6 (n = 43)

7 5.3 0.8c (n = 80) 1.7 0.3 (n = 26)

dMagnesium

= 0.026, bP = 1.2 x 108, cP = 4.5 x 106 (MannWhitney U-test for IV ritodrine versus no tocolytic drugs). sulphate, terbutaline, indomethacin and urinastatin.

232

K Takagi, K Satoh Efficacy of long-term tocolysis

1.0

Sub-group B n = 38: Ritodrine alone and ritodrine + other drugs n = 7: Unmedicated, bed-rest

Cumulative prolongation rate

0.8

0.6

0.4 P < 0.05 (log-rank test) 0.2

0.0 0 7 14 21 Prolongation of gestation (days) 28 35

FIGURE 2: Cumulative prolongation of gestation among patients with threatened premature labour who were admitted to hospital before gestation week 32 and had tocolysis index scores 7 (sub-group B) and were treated with IV ritodrine alone and IV ritodrine plus other drugs (magnesium sulphate, terbutaline, indomethacin and urinastatin), compared with unmedicated, bed-rest patients mean prolongation of gestation that was 3.1 times longer (P = 4.5 106; Table 5). A 2 analysis was applied in these patients to determine the difference between the number of patients whose gestation was prolonged for > 2 days (from initiation of tocolysis treatment until delivery) and those whose gestation was prolonged for 2 days (Table 6). Analysis showed that the number of patients whose gestation was prolonged for > 2 days in the tocolysis group was significantly greater than in the unmedicated bed-rest group (P = 1.06 104). Even in the TI 2 sub-group, where tocolysis might be considered optional, tocolytics significantly prolonged gestation (P = 0.026) in comparison with no medication (Table 5). Efficacy by tocolysis administered Tocolysis success rates were 65.2% (316/485) in patients who received IV ritodrine alone, 41.6% (94/226) in patients who received IV

TABLE 6: Effect of tocolysis treatment on the prolongation of gestation in severe cases of threatened premature labour (tocolysis index scores 7; n = 106) No. of patients Treatment IV ritodrine alone and IV ritodrine + other drugs No tocolytic drugs n 80 26 > 2 days 44 3 2 days 36 23

P = 1.06 x 104 (2 test for IV ritodrine versus no tocolytic drugs).

233

K Takagi, K Satoh Efficacy of long-term tocolysis

ritodrine and other tocolytic drugs concomitantly, and 89.4% (177/198) in patients who received IV ritodrine plus oral ritodrine. Full-term delivery rates among successfully treated patients were 65.2% (206/316) in patients given IV ritodrine alone, 47.9% (45/94) in patients given IV ritodrine and other drugs, and 68.4% (121/177) in patients given IV and oral ritodrine.

NEONATAL SHORT-TERM OUTCOMES

Data on neonatal outcome (birth weight, mortality and 1-min Apgar score) are shown in Table 7. The total number of neonates produced by the 1147 analysed cases was 1331 and the mortality rate was 3.3%. The survival rate of neonates delivered at 22 24 weeks gestation was 32%, increasing to 75% in neonates delivered at weeks 24 26 and at weeks 30 32 was 94% (Fig. 3A). There was a positive correlation between birth weight and neonatal survival. The survival rate of newborns with birth weights 1500 g exceeded 95% and there were no deaths among newborns with birth weights 2000 g (Fig. 3B). No statistically significant differences in neonatal findings for gestation weeks at birth, birth weight, incidence of respiratory distress syndrome (RDS), 1-min Apgar scores, or survival rates were found between patients treated with tocolytics and those untreated.

ADVERSE REACTIONS TO TOCOLYTICS

Maternal adverse reactions to tocolytics were observed in 113 (12.4%) out of 909 ritodrinetreated women. The most frequent adverse reactions were tachycardia and cardiac arrhythmia (n = 45, 5.0%), hepatic disorder (32 cases, 3.5%), tremors (1.5%) and

234

Outcome

No. of neonates 1331 671 660 968 320 39 4 733 598 Birth weight, mean SD 2295 742 2356 727 2233 751 2369 786 2146 560 1792 473 1328 67 1857 651 2832 428 Neonatal mortality, n (%) 44 (3.3%) 20 (3.0%) 24 (3.6%) 38 (3.9%) 6 (1.9%) 0 (0.0%) 0 (0.0%) 44 (6.0%) 0 (0.0%) 1-min Apgar score, mean SD 7.7 2.0 7.7 2.0 7.6 2.1 7.6 2.1 7.9 1.7 7.5 2.0 7.5 1.1 7.1 2.4 8.4 1.1

Delivery TABLE 7: Neonatal outcome for the 1147 pregnant women who presented with threatened premature labour

All cases

Males

Females

Singletons

Twins

Triplets Quadruplets Pre-term

Term

K Takagi, K Satoh Efficacy of long-term tocolysis

100% 90% 80% 70%

Neonates died Healthy neonates

Survival (%)

60% 50% 40% 30% 20% 10% 0% 22~ (19) 24~ (32) 26~ (39) 28~ (55) 30~ (81) 32~ (125) 34~ (203) 36~ (405) 38~ (247) 40~ (week) (125) (n) Neonates died Healthy neonates

100% 90% 80% 70%

Survival (%)

60% 50% 40% 30% 20% 10% 0% < 500 (13) 500 999 1000 1499 1500 1999 2000 2499 (80) (125) (194) (333) 2500 (586) (weight, g) (n)

FIGURE 3: Survival rate of neonates born to pregnant women who had threatened premature labour grouped according to: (A) week of delivery; and by (B) birth weight vomiting (0.7%). Six patients (0.7%) discontinued medication due to adverse reactions. Fetal and neonatal adverse reactions observed were arrhythmia (one fetus, 0.1%) and difficulty in breathing (one neonate, 0.1%) among 909 cases. birth and delivery practices, lifestyles and medical support systems differ from country to country, making it difficult to identify any single global standard. Pre-term birth rates were reported to be 5.4% in Japan10 and 5 7% in Europe in 2000,11,12 whereas the pre-term birth rate in the US (defined as delivery prior to week 37 of gestation) was reported to be 12.1% in 2002, a 29% increase over the previous two decades.13 Most of this increase related to births between 32 and 36 weeks gestation, whereas the preterm birth rates prior to 32 weeks gestation

Discussion
Threatened premature labour is treated using a variety of tocolytic methods, depending on the diagnostic criteria, the drugs available, and the likely duration and goals of therapy. Furthermore, traditional

235

K Takagi, K Satoh Efficacy of long-term tocolysis

remained stable at about 2%.13 Worldwide, the most commonly used tocolytics were the 2 stimulants, which include ritodrine and magnesium sulphate. In the USA and Canada, however, it appears that ritodrine is considered to be of limited effectiveness: the general view appears to be that ritodrine is only effective in prolonging pregnancy for a period of 24 48 h.8,14 17 Magnesium sulphate has been associated with insufficient inhibition of uterine contractions and safety concerns.18,19 The present study showed that ritodrine therapy maintained pregnancy beyond 35 weeks in women with threatened premature labour with TI 3, prolonging gestation significantly longer (5 26 days) compared with the unmedicated bed-rest group (2 11 days). Moreover, for patients with TI 3 6 and TI 7, prolongation of gestation with ritodrine, was 2.2 and 3.1 times longer, respectively, than with no medication. These results clearly show that treatment of threatened premature labour using ritodrine is effective and useful for long-term tocolysis (defined as tocolysis that prolongs pregnancy > 2 days). Patients registered in the clinical study reported by the Canadian Preterm Labor Investigators Group8 included cases with membrane rupture (25% of cases), cervical length shortened by > 50%, and cervical dilation of > 2 cm of which > 10% were > 4 cm dilated. A clinical study by Garite et al.20 involving cases with membrane rupture showed no significant difference in prolongation of gestation between patients receiving tocolysis with ritodrine and those receiving no medication: 80% of the women gave birth within 24 h of treatment initiation. This indicates that tocolysis using ritodrine may not always be effective in cases with membrane rupture. The reason why the prolongation effects of ritodrine were limited to < 2 days in the Canadian groups clinical trial may be, as suggested by Garite et al.,20 that most of the cases in the Canadian study involved ruptured membranes or a highly dilated cervix, both of which are rather difficult to treat with ritodrine. The present study analysed the effects of tocolysis with ritodrine on the more severe cases (TI 7), which included cases of membrane rupture and cervical dilatation > 4 cm. Whereas the Canadian report8 indicated that tocolytics such as ritodrine extended the duration of pregnancy up to 2 days, the present study showed that, even in the most severe TI 7 cases, the number of cases in which the duration of pregnancy was prolonged by > 2 days was significantly greater in those treated with ritodrine compared with unmedicated cases. The present study, therefore, supports the view that using tocolytics such as ritodrine for > 2 days, or as long as the pregnancy lasts, is effective and useful for the treatment of threatened premature labour. It also confirms the effectiveness of the long-term tocolysis regimens currently used to treat threatened premature labour in Japan. A reason why long-term tocolysis is rarely used in the USA may be the high incidence of adverse effects reported with ritodrine (which include nausea, vomiting, tremors and tachycardia).6,7,21 Lung oedema, which is the most severe side-effect of ritodrine, was reported by Leveno and Cunningham.5 in 3 9% of cases, and by Higby et al.22 in approximately 5% of cases. Its pathogenesis may be secondary overhydration caused by both the anti-diuretic activity of high doses of a 2 stimulant such as ritodrine given concomitantly with a large volume of IV fluids containing 2 stimulants.5 In Japan, the maximum dosage for IV ritodrine is strictly limited to 200 g/min, whereas in the

236

K Takagi, K Satoh Efficacy of long-term tocolysis

USA and Canada 350 g/min is indicated as the maximum dose. This may be another cause of the higher incidence of lung oedema in the treatment of threatened premature labour in the USA5,22 than in the present study (0.3%5 vs 0.15%). The Canadian study reported a high incidence of maternal side-effects with ritodrine including tachycardia (53.4%), nausea (20.5%), tremor (39.2%) and headache (23.0%).8 A much lower incidence of side-effects was recorded in the present study: tachycardia (4.0%), nausea and vomiting (0.7%) and tremor (1.5%). Differences in the incidence of such sideeffects may be attributable to the Japanese practice of gradually increasing the dose of IV ritodrine for long-term tocolysis, whereas in the USA and Canada the generally accepted practice is to increase the dose up to an effective level as quickly as possible within 2 days. In Japan, treatment is expected to prolong pregnancy as much as possible by suppressing uterine contractions enough to delay delivery but, in the US and Canada, it is used to maintain pregnancy by substantial and rapid inhibition of contractions for a period long enough to enable maturation of the fetal lung to be promoted with corticosteroids.23 Recent progress in neonatal medicine has brought remarkable improvements in the primary prognosis for premature neonates in Japan. Although 32% survival rates were reported for neonates delivered at 22 24 weeks gestation in the present study, rates were increased to 75% for neonates delivered at 24 26 weeks and to 94% for neonates delivered at 30 32 weeks. A correlation was also observed between high birth weights and low neonatal death rates: survival was > 95% for neonates weighing 1500 g and no deaths were recorded for neonates weighing 2000 g. Pre-term infants are also prone to physical and mental disorders such as RDS, intraventricular haemorrhage, cerebral palsy, mental retardation and prematurityinduced retinopathy, for which a correlation with gestation weeks and birth weight has been reported.24 Tsuneishi et al.24 reported on extremely low birth-weight infants who had cerebral palsy (5.8%), mental retardation (9.7%) or both (7.7%) at 6 years of age. Hack et al.25 reported that, among infants with birth weights < 750 g, mental retardation was observed in 21% and cerebral palsy in 9% when they reached the lower grades of primary school. Johnson et al.26 also showed that half of all infants born prior to the 29th week of gestation had moderate to severe physical disorders at 4 years of age. Prolonging the duration of pregnancy by long-term tocolysis, as reported in the present study, and the consequential improvement in survival rates for premature newborns may lead to overall reductions in medical costs. A recent survey showed that medical care expenses for infants born at 38 weeks gestation were 10% of those for infants born at 35 weeks (US$441 versus US$4733).27 Rolnick et al.28 reported that medical expenses to treat low birth-weight infants (1500 2499 g) during the first post-natal year were 46% higher than those for normal birth-weight infants. According to a report by Ibara and Ikenoue 29 in Japan, the more that a pregnancy is prolonged, the greater the reduction in neonatal medical expenses from birth to discharge, especially for preterm births. Prolonging pregnancy by a single week can greatly reduce the expenses incurred in treating premature infants who would have been born at earlier stages of gestation.29 If pregnancy is prolonged from 24 to 25 weeks, for

237

K Takagi, K Satoh Efficacy of long-term tocolysis

example, expenses can be reduced by up to US$14 600 (1 600 000) and, if pregnancy is prolonged from 30 to 31 weeks, expenses can be reduced by up to US$4600 (500 000).29 The significance of treating threatened premature labour therefore needs to be considered from an economic, as well as a medical, standpoint. Community Hospital, Saiseikai Yamagata Saisei Hospital, Fukushima Medical College, Medical Centre of Saitama University of Medicine, Nihon University, Nihon University Nerima Hikarigaoka Hospital, Nippon Medical School, Keio University, Perinatal Medical Centre of Tokyo Womens Medical University, Chiba Municipal Kaihin Hospital, Yokohama Municipal University, Kitasato University, Niigata University, Hamamatsu University School of Medicine, Seirei Hamamatsu Hospital, Nagoya Daiichi Red Cross Hospital, Osaka Municipal University, Osaka University, Okayama University, Social Insurance Hiroshima Citizen Hospital, Hiroshima Prefectural Hiroshima Hospital, NTT Matsuyama Hospital, Kochi Medical School, Kyushu University, Saint Marys Hospital, Kumamoto Municipal Kumamoto Citizen Hospital, Miyazaki Medical University, Miyazaki Prefectural Miyazaki Hospital, and Kagoshima Municipal Hospital.

Acknowledgements
The authors express deep gratitude to the following investigators (all Japan): Toshihiko Terao (Hamamatsu University School of Medicine), Seiichiro Fujimoto (Hokkaido University), Yuji Murata (Osaka University), Tsuyomu Ikenoue (Miyazaki University), Naohiro Kanayama (Hamamatsu University School of Medicine), Toru Kanzaki (Kanzaki Ladies Clinic), Tadashi Sagawa (Hokkaido University), and Yoshio Matsuda (Tokyo Womens Medical University). The following institutions participated in the study (all Japan): Sapporo Municipal Hospital, National Hirosaki Hospital, Akita University, Akita Red Cross Hospital, Tohoku University, Sendai Municipal Hospital, National Sendai Hospital, Shirataka

Conflicts of interest
The authors had no conflicts of interest to declare in relation to this article.

Received for publication 20 July 2008 Accepted subject to revision 1 August 2008 Revised accepted 2 December 2008 Copyright 2009 Field House Publishing LLP
References 1 Tough SC, Greene CA, Svenson LW, et al: Effects of in vitro fertilization on low birth weight, preterm delivery, and multiple birth. J Pediatr 2000; 136: 618 622. 2 Tough SC, Newburn-Cook C, Johnston DW, et al: Delayed childbearing and its impact on population rate changes in lower birth weight, multiple birth, and preterm delivery. Pedatrics 2002; 109: 399 403. 3 Rush RW, Keirse MJNC, Howat P, et al: Contribution of preterm delivery to perinatal mortality. Br Med J 1976; 2: 965 968. 4 Jones JS, Istwan NB, Jacques D, et al: Is 34 weeks an acceptable goal for a complicated singleton pregnancy? Manag Care 2002; 11: 42 47. 5 Leveno KJ, Cunningham FG: -Adrenergic agonists for preterm labor. N Engl J Med 1992; 327: 349 351. 6 Leveno KJ, Klein VR, Guzick DS, et al: Singlecentre randomised trial of ritodrine hydrochloride for preterm labour. Lancet 1986; i: 1293 1296. 7 Lamont RF, International Preterm Labour Council: Evidence-based labour ward guidelines for diagnosis, management and treatment of spontaneous preterm labour. J Obstet Gynaecol 2003; 23: 469 478. 8 The Canadian Preterm Labor Investigators Group: Treatment of preterm labor with the beta-adrenergic agonist ritodrine. N Engl J Med 1992; 327: 308 312. 9 Baumgarten K, Gruber W. Tocolyseindex: In: Perinatale Medizin (Dudenhausen JW, Saling E, eds). Stuttgart, Germany: Georg Thieme, 1974; pp 197 199.

238

K Takagi, K Satoh Efficacy of long-term tocolysis

10 Mothers and Childrens Health and Welfare Association: Maternal and Child Health Statistics of Japan. Tokyo: Mothers and Childrens Health and Welfare Association, 2008. 11 Morken NH, Vogel I, Kallen K, et al: Reference population for international comparisons and time trend surveillance of preterm delivery proportions in three countries. BMC Womens Health 2008; 8: 16. 12 MacFarlance A, Mugford M: British counts. In: Statistics of Pregnancy and Childbirth, 2nd edn. London: The Stationary Office, 2000. 13 Martin JA, Hamilton BE, Sutton PD, et al: Births: final data for 2002. Natl Vital Stat Rep 2003; 52: 1 113. 14 Simhan HN, Caritis SN: Prevention of preterm delivery. N Engl J Med 2007; 357: 477 487. 15 ACOG Committee on Practice Bulletins, American College of Obstetricians and Gynecologists: ACOG Practice Bulletin. Clinical management guidelines for obstetriciangynecologist. Management of preterm labor. Obstet Gynecol 2003; 101: 1039 1047. 16 Anotayanonth S, Subhedar NV, Garner P, et al: Betamimetics for inhibiting preterm labour. Cochrane Database Syst Rev 2004; 4: CD004352. 17 Royal College of Obstetricians and Gynaecologists: Tocolytic drugs for women in preterm labour. Clinical Guideline No. 1(B), 2002. 18 Cox SM, Sherman ML, Leveno KJ: Randomized investigation of magnesium sulfate for prevention of preterm birth. Am J Obstet Gynecol 1990; 163: 767 772. 19 Crowther CA, Hiller JE, Doyle LW: Magnesium sulphate for preventing preterm birth in threatened preterm birth. Cochrane Database Syst Rev 2002; 4: CD001060. 20 Garite TJ, Keegan KA, Freeman RK, et al: A randomized trial of ritodrine tocolysis versus

expectant management in patients with premature rupture of membranes at 25 to 30 weeks of gestation. Am J Obstet Gynecol 1987; 157: 388 393. 21 Gyetvai K, Hannah M, Hodnett ED, et al: Tocolytics for preterm labor: a systematic review. Obstet Gynecol 1999; 94: 869 877. 22 Higby K, Xenakis EMJ, Pauerstein CJ: Do tocolytic agents stop preterm labor? A critical and comprehensive review of efficacy and safety. Am J Obstet Gynecol 1993; 168: 1247 1259. 23 Goldenberg RL, Cliver SP, Bronstein J, et al: Bed rest in pregnancy. Obstet Gynecol 1994; 84: 131 136. 24 Tsuneishi S, Uetani Y, Nakamura H: Preschool growth of extremely low birth weight infants. World Obstet Gynecol 2003; 55: 49 57 [in Japanese]. 25 Hack M, Taylor HG, Klein N, et al: School-age outcomes in children with birth weight under 75 g. N Engl J Med 1994; 331: 753 759. 26 Johnson A, Townshend P, Yudkin P, et al: Functional abilities at age 4 years of children born before 29 weeks of gestation. BMJ 1993; 306: 1715 1718. 27 St John EB, Nelson KG, Cliver SP, et al: Cost of neonatal care according to gestational age at birth and survival status. Am J Obstet Gynecol 2000; 182: 170 175. 28 Rolnick SJ, Jackson JM, OConner P, et al: Impact of birthweight on healthcare charges within a managed care organization. Am J Manag Care 2000; 6: 1289 1296. 29 Ibara H, Ikenoue T: Survey on medical expenses among neonates with low birth weight by the gestational week of delivery; comparison to those among neonates delivered from women with pregnancy induced hypertension. World Obstet Gynecol 1999; 51: 69 73.

Authors address for correspondence Associate Professor Kenjiro Takagi Department of Obstetrics and Gynaecology, Saitama Medical University, Saitama Medical Centre, 1981 Tsujido-machi, Kamoda, Kawagoe-shi, Saitama 350-8550, Japan. E-mail: 0384504001@jcom.home.ne.jp

239