You are here

Impact of a randomized trial on maintenance tocolysis on length of hospital admission of women with threatened preterm labor in The Netherlands

European Journal of Obstetrics & Gynecology and Reproductive Biology, pages 8 - 11

Abstract

Introduction

The APOSTEL-II trial was a multicenter randomized placebo-controlled trial, assessing the effectiveness of maintenance tocolysis with nifedipine. The trial showed maintenance tocolysis not to have an effect on perinatal outcome. Objective of the current study is to evaluate the effect of a negative trial on the length of hospital admission of women with threatened preterm labor.

Materials and methods

We evaluated length of hospital admission of all patients admitted with threatened preterm labor with a gestational age <32 weeks in 8 perinatal centers that participated in the APOSTEL-II trial. We studied only the first admission with threatened preterm labor, readmissions were excluded. We distinguished between the period before, the period during and the period after the trial. In a subgroup analysis, we differentiated for the group of women who delivered and for the group of women who did not deliver during the initial admission.

Results

The mean length of hospital admission was 9.3 days before the start of the trial, 8.4 days during the recruitment period and 8.1 days after the trial was completed. The difference in mean length of hospital admission before and during the recruitment period was significantly different (p < 001).

Comments

The length of hospital admission of women with threatened preterm labor is found to be reduced during the recruitment period of the APOSTEL-II trial. This shows that the conduct of a randomized controlled trial itself has the potential to change daily practice.

Keywords: Admission, Threatened preterm labor, Maintenance tocolysis.

Introduction

Preterm birth is the most common cause of neonatal morbidity and mortality worldwide. Annually it affects 12,000 children in The Netherlands, of whom 2000 are born extremely premature (<32 weeks of gestation) [1] . Preterm birth is responsible for approximately 75% of all neonatal deaths [2] and 40% of childhood neurological morbidities. There is an increased risk of lifelong health consequences such as breathing disorders and developmental problems. Prematurity therefore has a major impact on the infant itself as well as the parents. In addition to the health consequences of prematurity, short term and long term costs are substantial [3] . However, no consensus existed in the past in daily obstetrical practice on the length of tocolytic therapy and a strong practice variation existed in The Netherlands for treatment of women with threatened preterm labor. Some obstetricians maintained tocolytic therapy until term – i.e. 37 weeks’ gestational age –, whereas others stopped tocolytics after 48 h irrespective of gestational age while others continued until for example 28 weeks’ gestational age.

This discrepancy regarding the length of tocolytic treatment in women with threatened preterm labor, set the stage for performing the APOSTEL-II (assessment of perinatal outcome with sustained tocolysis in early labor) trial [4] . In this prospective randomized multicenter placebo-controlled trial, women admitted for threatened preterm labor who were treated with 48 h tocolysis and corticosteroids, were allocated to nifedipine (intervention) or placebo (control) for twelve days. The results of the study are presented elsewhere [5] .

Our objective was to evaluate whether participation in a randomized controlled trial on maintenance tocolysis with nifedipine influenced the duration of hospital admission of women with threatened preterm labor, in order to evaluate the effect of a negative trial on this outcome. To do so, we studied duration of hospital stay before, during and after the recruitment period.

Materials and methods

The APOSTEL-II trial was performed between June 2008 and February 2010 in 11 centers in The Netherlands. Data concerning the length of hospital admission were obtained from the Obstetric High Care (OHC) registry from 2006 to 2011, provided by 8 perinatal centers participating in the APOSTEL-II trial. For this retrospective study, we included hospital stays of all patients admitted with threatened preterm labor with a gestational age<32 weeks, whether they participated in the randomized study or not. Only the first admission with threatened preterm labor was used, readmissions were excluded.

We recorded day of admission, day of discharge, gestational age at admission and, when available, whether or not delivery occurred during admission. According to national guidelines, calculation of gestational age was based on the first day of the last menstrual period and verified by a routinely performed first-trimester ultrasound in all patients.

We distinguished the period before the trial, the period during the trial and the period after the trial. We corrected for different start and end dates of the recruitment period between the various hospitals. Since the length of these three periods may vary per center, the mean length of hospital admission was calculated on the basis of a fixed number of days per period for all centers.

Data were presented in mean length of hospital admission (days) with confidence intervals of 95%.

We applied the Kruskal–Wallis test to compare the mean length of hospital admission between the three groups. Then, we applied the Mann–WhitneyU-test to test for differences between two periods. Two sided probability of less than 0.05 was considered statistically significant.

Any decrease in length of hospital admission of women between before and during the APOSTEL-II trial could be caused by a downward trend that started before the APOSTEL-II trial. Therefore, we evaluated also the length of admission of patients admitted 10 to 2 weeks before the start and patients admitted 2 to 10 weeks after the end of the APOSTEL-II trial.

For these data we constructed moving average curves with a regression line. The mean length of hospital admission was plotted against the time of hospital admission, in which the periods before and during and the periods during and after the APOSTEL-II trial, respectively, were compared with each other. Statistical analysis was performed using SPSS software (version 18.0).

In a subgroup analysis we differentiated for the group of women who delivered during hospital admission and for the group of women who did not deliver after the initial admission. Data about whether or not a patient had delivered during hospital admission was available from 4 of the 8 centers.

Results

The fixed number of days per period for all centers was 882 days for the period before the trial, 486 days for the period during the trial and 334 days for the period after the trial.

Table 1 shows the mean length of hospital admission in 3370, 1934 and 1185 women admitted with threatened preterm labor before, during and after the trial, respectively. The mean length of hospital admission was 9.3 days before versus 8.4 days during the trial (decreased 0.86 days,p < 0.001). There was no significant difference in length of hospital admission between the periods during and after the trial (p = 0.548).

Table 1 Length of hospital admission.

  Mean length of hospital admission in days (95% CI) p Value for difference with prior period
All women    
Before the APOSTEL-II trial (N = 3370) 9.30 (8.98–9.62)
During the APOSTEL-II trial (N = 1934) 8.44 (8.05–8.83) p < 0.001
After the APOSTEL-II trial (N = 1185) 8.13 (7.64–8.61) p = 0.548
 
10 to 2 Weeks before start of the APOSTEL-II trial (N = 250) 8.96 (7.87–10.05)  
2 to 10 Weeks after finish of the APOSTEL-II trail (N = 204) 8.50 (7.31–9.69) p = 0.277

When comparing the periods 10 to 2 weeks before recruitment for the trial with 2 to 10 weeks after finish of the trial we also found a trend of reducing mean length of hospital admission (9.0 versus 8.5 days).

Table 2 shows the mean length of hospital admission distinguished for women who delivered and for women who did not deliver during hospital admission. For both subgroups, a reduction in length of hospital admission was seen during the APOSTEL-II trial. For women who delivered during hospital admission the mean length of hospital admission was 10.4 days before versus 9.6 days during the trial (decreased 0.84 days,p = 0.022). Mean length of hospital admission after the trial was 9.6 days too. For women who did not deliver during hospital admission the mean length of hospital admission was 8.2 days before versus 7.3 days during the trial (decreased 0.88 days,p = 0.007) and 6.9 days after the trial (decreased 0.38 days,p = 0.056).

Table 2 Length of hospital admission—subgroups.

  Mean length of hospital admission in days (95% CI) p Value for difference with prior period
Women who delivered during hospital admission    
Before the APOSTEL-II trial 10.44 (9.85–11.04) (N = 1097)
During the APOSTEL-II trial 9.60 (8.85–10.35) (N = 636) p = 0.022
After the APOSTEL-II trial 9.59 (8.59–10.59) (N = 359) p = 0.767
 
Women who did not deliver during hospital admission    
Before the APOSTEL-II trial 8.18 (7.59–8.78) (N = 845)
During the APOSTEL-II trial 7.30 (6.64–7.95) (N = 582) p = 0.007
After the APOSTEL-II trial 6.92 (6.23–7.62) (N = 509) p = 0.056

Fig. 1 shows the moving average curve of length of hospital admission before and during the trial for women who did not deliver during admission. This confirms the coincidence of a reduction trend with the start of the APOSTEL-II trial and this decline stagnated after the end of the APOSTEL-II trial ( Fig. 2 ).

gr1

Fig. 1 Length of hospital admission before and during the APOSTEL-II trial (for women who did not deliver during hospital admission).

gr2

Fig. 2 Length of hospital admission during and after the APOSTEL-II trial (for women who did not deliver during hospital admission).

Comments

The results of this retrospective study show that the APOSTEL-II trial reduced the duration of hospital admission of women with threatened preterm labor. There was a significant reduction of 0.86 days in mean length of hospital admission during the trial compared with the period before the trial. The start of this downward trend in mean length of hospital admission coincides with the moment of start of the APOSTEL-II trial. Possibly, a change in doctors’ behavior causes this reduction. When doctors more often decided to discontinue tocolytics after 48 h during the APOSTEL-II trial, patients will potentially be discharged earlier from the hospital.

Several remarks with respect to the patient can be made in relation to the reduced admission period. First, shorter hospital admissions are associated with less disease caused by hospital admission, such as hospital-acquired infections and venous thromboembolism. Second, discontinuing of tocolytics after a period of initial tocolysis and corticosteroids for 48 h could prevent the rare but severe side-effects of (maintaining) tocolysis on mother and child[6] and [7]. One could argue whether a reduction of 0.86 days is clinically relevant. However, we report on 6489 women who were admitted during the study period, we feel that even a slight mean reduction of hospital stay for this large group is clinically relevant.

Financially, a reduced length of hospital admission is associated with lower costs. Assuming a rate of €420 per admission day for women with threatened preterm labor admitted to the obstetric high care, a reduction of 0.86 days in length of hospital admission results in a cost reduction of €361 per patient. In view of the large group of women admitted with threatened preterm labor, a significant cost reduction can be assumed.

From a methodological point of view, the strength of the present study is that it used data of the OHC registration, which is known for its reliable and consistent data concerning women with threatened preterm labor. We included all patients admitted in a perinatal center with threatened preterm labor, excluding selection bias.

This research shows the impact of a randomized controlled trial (RCT) on doctors’ behavior. Despite that it is normally difficult to affect doctors’ behavior and that implementation of results often fails [8] , the results of this study show that the conduct of a RCT itself has the potential to change daily practice. A trial on the effectiveness of a treatment policy might lead to more awareness by doctors about their own actions and reconsideration of their own clinical policy. However, so far it is not common to monitor the effects of studies on daily practice. To our concern, it would be useful to do so in the future and include the effects of other studies.

One could argue that evolution of diagnosis of threatened preterm labor over the past years, for example by using fetal fibronectin testing (FFN), could lead to a relative higher portion of women who did deliver during the first hospital admission for threatened preterm labor and therefore lead to shorter hospital stay during the course of the APOSTEL-II trial. Indeed, in the same period, we also assessed the value of FFN and cervical length in women with threatened preterm labor [9] . However, the reduction of mean length of hospital admission in the periods before and during the APOSTEL-II trial, was significant for both women who delivered and for women who did not deliver.

Although a shorter admission period is associated with positive consequences, there is a chance that a reduced admission period will have negative consequences for the patient. First, a reduced length of hospital admission may lead to more preterm deliveries. However, this is very unlikely since it never has been shown that bed has an impact on gestational age[10] and [11]. Second, a reduction in length of hospital admission may lead to an increase of readmissions or outpatient visits during pregnancy, since women who were discharged earlier from the hospital could be more anxious. However, this was not a part of this study and we cannot draw any conclusions on these possible negative consequences. For this, further research is required.

The APOSTEL-II trial showed that maintenance tocolysis with nifedipine neither prolongs pregnancy, nor improves perinatal outcome [5] . The effects of publication of the results of the APOSTEL-II trial are not included in this article. The decreasing trend in the mean length of hospital admission stagnated when the trial stopped. Publication of the results may lead to sustainable changes in practice and further reduction in the mean admission period. On the other hand, it is possible that doctors fall back into their old patterns of behavior resulting in an increasing trend in mean length of hospital admission.

In conclusion, we showed that the APOSTEL-II trial reduced the duration of hospital admission of women with threatened preterm labor. This shows that the conduct of a RCT itself has the potential to change daily practice and reduce costs.

Condensation

Length of admission of women with threatened preterm labor is reduced after participation in a randomized trial on maintenance tocolysis with nifedipine.

References

  • [1] P. Stichting. Stichting Perinatale Registratie Nederland. (, 2009)
  • [2] R.L. Goldenberg, J.F. Culhane, J.D. Iams, R. Romero. Epidemiology and causes of preterm birth. Lancet. 2008;371:75-84 Crossref
  • [3] S. Petrou, O. Eddama, L. Mangham. A structured review of the recent literature on the economic consequences of preterm birth. Arch Dis Child Fetal Neonatal Ed. 2011;96:F225-F232 Crossref
  • [4] C. Roos, L.H. Scheepers, K.W. Bloemenkamp, et al. Assessment of perinatal outcome after sustained tocolysis in early labour (APOSTEL-II trial). BMC Pregnancy Childbirth. 2009;9:42 Crossref
  • [5] C. Roos, M.E. Spaanderman, E. Schuit, et al. Effect of maintenance tocolysis with nifedipine in threatened preterm labor on perinatal outcomes: a randomized controlled trial. JAMA. 2013;309:41-47 (the journal of the American Medical Association) Crossref
  • [6] R. de Heus, B.W. Mol, J.J. Erwich, et al. Adverse drug reactions to tocolytic treatment for preterm labour: prospective cohort study. BMJ. 2009;338:b744 Crossref
  • [7] A.J. van Veen, M.J. Pelinck, M.G. van Pampus, J.J. Erwich. Severe hypotension and fetal death due to tocolysis with nifedipine. BJOG. 2005;112:509-510 (an international journal of obstetrics and gynaecology)
  • [8] R. Grol, J. Grimshaw. From best evidence to best practice: effective implementation of change in patients’ care. Lancet. 2003;362:1225-1230 Crossref
  • [9] G.J. van Baaren, J.Y. Vis, F.F. Wilms, et al. Predictive value of cervical length measurement and fibronectin testing in threatened preterm labor. Obstet Gynecol. 2014;123(6):1185-1192 Crossref
  • [10] C.A. Crowther, S. Han. Hospitalisation and bed rest for multiple pregnancy. Cochrane Database Syst Rev. 2010;(7):CD000110
  • [11] C. Sosa, F. Althabe, J. Belizan, E. Bergel. Bed rest in singleton pregnancies for preventing preterm birth. Cochrane Database Syst Rev. 2004;(1):CD003581

Footnotes

a Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam, The Netherlands

b Department of Obstetrics and Gynecology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

c Department of Obstetrics and Gynecology, Leiden University Medical Center, Leiden, The Netherlands

d Department of Obstetrics and Gynecology, Erasmus Medical Center, Rotterdam, The Netherlands

e Department of Obstetrics and Gynecology, University of Groningen, Academic Medical Center, Groningen, The Netherlands

f Department of Obstetrics and Gynecology, University Medical Center, Utrecht, The Netherlands

g Department of Obstetrics and Gynecology, Máxima Medical Center, Veldhoven, The Netherlands

h School of Paediatrics and Reproductive Health, University of Adelaide, Adelaide 5000, SA, Australia

lowast Corresponding author. Tel.: +31 24 3617350/+31 6 11195142; fax: +31 24 3668597.