| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
REVIEWS |
From the Department of Obstetrics and Gynecology, Division of Maternal–Fetal Medicine, University of Florida Health Science Center, Jacksonville, Florida.
 |
ABSTRACT |
|---|
TOP
 ABSTRACT SOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
DATA SOURCES: Computerized databases, references in published studies, and textbook chapters in all languages were used to identify randomized controlled trials (RCTs) evaluating the use of progestational agents for the prevention of preterm births in women at elevated risk.
METHODS OF STUDY SELECTION: We identified RCTs that compared progestational agents with placebo for patients at risk for preterm birth and evaluated at least one of the following: delivery before 37 weeks of gestation, birth weight less than 2,500 g, threatened preterm labor, respiratory distress syndrome, and perinatal mortality. The primary outcomes assessed were preterm delivery and perinatal mortality.
TABULATION, INTEGRATION, AND RESULTS: Ten studies met inclusion criteria for this review. For each study with binary outcomes, an odds ratio (OR) with 95% confidence intervals (CIs) was calculated for selected outcomes. Homogeneity was tested across the studies. Compared with women allocated to receive placebo, those who received progestational agents had lower rates of preterm delivery (26.2% versus 35.9%; OR 0.45, 95% CI 0.25–0.80). Similar results were noted when comparing patients who were specifically treated with 17
-hydroxyprogesterone caproate (29.3% versus 40.9%; OR 0.45, 95% CI 0.22–0.93). Additionally, subjects allocated to receive 17-hydroxyprogesterone caproate had lower rates of birth weights less than 2,500 g (OR 0.50, 95% CI 0.36–0.71). No differences in rates of hospital admissions for threatened preterm labor or perinatal mortality were noted for subjects receiving progestational agents in general or for those receiving only 17-hydroxyprogesterone caproate specifically.
CONCLUSION: The use of progestational agents and 17-hydroxyprogesterone caproate reduced the incidence of preterm birth and low birth weight newborns.
Studies of most tocolytic agents have not demonstrated efficacy beyond 48 hours.2 A trial employing betamimetics has shown ritodrine to be effective in delaying delivery by up to a week, although no impact on perinatal mortality was observed.3 A possible reason for the lack of efficacy is the initiation of tocolytic agents only after identification of uterine activity. The prophylactic use of tocolytic therapy in high-risk patients might be associated with more favorable results.4,5
Initial reports assessing progestational agents for the prevention of preterm birth found variable results.6 More recently, 2 randomized controlled trials (RCTs) have rekindled interest in the use of progestational agents for the prevention of preterm birth. A large randomized placebo-controlled trial assessing the efficacy of 17-hydroxyprogesterone caproate to prevent preterm birth in a select, high-risk group of women was stopped early when investigators noted a significant protection against recurrent preterm birth for all races in the group receiving active treatment.7 Similarly, a smaller randomized placebo-controlled trial of vaginal progesterone (100 mg daily) in 142 women at elevated risk for preterm birth revealed a significant reduction in risk among women receiving progesterone.8
Older trials and meta-analyses assessing the efficacy of progestational agents in preventing preterm birth reached contradictory conclusions.6,9,10 In addition, many of the randomized controlled trials had a relatively small number of participants. Since the previous meta-analyses did not adhere to strict methodologic criteria and because of the recent publication of additional trials, we decided to perform an updated systematic review with meta-analysis to further clarify the efficacy of progestational agents for the prevention of preterm births in patients at elevated risk.
|
SOURCES |
|---|
|
TOP
ABSTRACTSOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
We attempted to identify all relevant published and unpublished randomized trials that compared the use of progestational agents with placebo for the prevention of preterm birth in high-risk patients. Without language restrictions, we searched electronic databases (MEDLINE and EMBASE) and the Cochrane Library from January 1966 to January 2004 by using the terms "preterm birth," "progesterone," "17- hydroxyprogesterone," "tocolysis," "randomized trial," "perinatal mortality," and "low birth weight," in combination with generic and trade names of individual preparations of progestational agents. In addition, we manually searched textbook chapters and references of identified journal articles to locate additional studies. We also reviewed abstracts from major national and international meetings and contacted other investigators in the field to learn of possible unpublished studies. Finally, we reviewed the articles analyzed in previous systematic reviews assessing the efficacy of progestational agents.6,9,10
|
STUDY SELECTION |
|---|
|
TOP
ABSTRACTSOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
-hydroxyprogesterone caproate. Published and unpublished RCTs, including full-text articles and abstracts, were included. Nonrandomized (observational) studies were excluded from the analysis, as were RCTs with required outcomes not reported. Disagreements by authors over inclusion and exclusion of studies and interpretation of data were resolved by consensus reached after discussion. Using the study-quality criteria of Schultz and colleagues,12 each study's methodological attributes were assessed for quality by 2 authors (L.S.R. and I.D.). Because of the relatively small number of studies available, quality evaluation was not used for selecting studies. The Materials and Methods and Results sections of each study were photocopied separately after identifiers were removed, masking the investigators to study authors and institutions.
Meta-analysis was performed using STATA 8.0 software (StataCorp, College Station, TX). The primary outcomes assessed were preterm delivery (before 37 weeks of gestation) and perinatal mortality. For each study with binary outcomes, we calculated the odds ratio (OR) and 95% confidence intervals (CIs) of selected outcomes. Estimates of ORs for dichotomous outcomes were calculated using fixed-effects (Mantel-Haenszel) and random-effects (DerSimonian and Laird) models.13,14 Number needed to treat was calculated for outcomes showing significant benefit from the use of progestational agents. To determine the combinability of individual studies, we did a formal test of heterogeneity by using the Mantel-Haenszel method. In addition, homogeneity across studies was assessed by qualitative visual inspection of L'Abbe plots.15 Outcomes with significant heterogeneity (P < .10), were analyzed and reported by using a random-effects model. Additionally, results obtained with a fixed-effects model were compared with those obtained with a random-effects model.
We performed sensitivity analyses to further establish the robustness of our results. To identify any study that may have exerted a disproportionate influence on the summary treatment effect, we deleted studies one at a time and noted the degree to which the size and significance of treatment effect changed. Publication bias was examined by reviewing the Egger test16 and by visual inspection of funnel plots, which plot OR against study sample size. Visual inspection of funnel plots is a technique that may be helpful to estimate whether additional small studies may have been conducted but not published because of unfavorable or negative results.
To explore potential sources of study result heterogeneity, we developed a random-effects regression model that included several variables as covariates. In the meta-regression, the dependent variable was the OR for the outcome of interest (preterm birth rate), and the independent variables were year and country of study and type and dose of progesterone administered.
|
RESULTS |
|---|
|
TOP
ABSTRACTSOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
-hydroxyprogesterone caproate (Papiernik E. Double blind study on an agent to prevent preterm delivery among women at increased risk. Schering ed. Serie IV, 1970:65–8),7,23–29 and 2 assessed other progestational agents.8,31 A total of 1,339 subjects were enrolled in these 10 trials. Four were conducted in the United States, 2 in France, and one each from Brazil, Italy, Israel, and Finland. A summary of these controlled trials is shown in Table 1. The most common indications for progesterone administration were history of preterm birth or multiple spontaneous abortions.
|
Relative to women allocated to receive placebo, those who received progestational agents (17-hydroxyprogesterone caproate and other forms of progesterone) had lower rates of preterm delivery (26.2% versus 35.9%; OR 0.45, 95% CI 0.25–0.80) (Table 2). The number needed to treat to prevent one premature delivery was 10 (95% CI 6–24). Similar results were obtained when we analyzed studies that evaluated only 17-hydroxyprogesterone caproate (29.3% versus 40.9%; OR 0.45, 95% CI 0.22–0.93). The number needed to treat was 8 (95% CI 5–19). Significant heterogeneity was noted for the outcome preterm delivery, and thus a random-effects model was employed. Funnel plots were symmetric, indicating no evidence of publication bias. Sensitivity analyses did not reveal undue influence on the pooled results from any of the individual studies. In the meta-regression analysis, neither year nor country of study nor type nor dose of progestational agent administered was significantly associated with the degree of preterm birth prevention.
|
When compared with those receiving placebo, subjects who received 17-hydroxyprogesterone caproate experienced a lower rate of neonates with birth weights less than 2,500 g (20.3% versus 28.4%; OR 0.50, 95% CI 0.36–0.71). The number needed to treat was 12 (95% CI 7–43) (Table 3). Data for this variable in studies using progestational agents were not available. For this outcome, no significant heterogeneity was noted, and thus a fixed-effects model was employed. Visual inspection of funnel plots and the results of Egger test revealed no evidence of publication bias. Sensitivity analyses did not reveal undue influence on the pooled results from any of the individual studies.
|
When compared with those receiving placebo, subjects who received 17-hydroxyprogesterone caproate or other progestational agents had similar rates of hospital admissions for threatened preterm labor (Table 4). For this outcome, the rate was lower for those receiving 17-hydroxyprogesterone (3.8% versus 5.2%), but this difference did not achieve statistical significance. For the outcome of threatened premature labor, there was evidence of significant heterogeneity. Neither visual inspection of funnel plots nor Egger test revealed evidence of publication bias. For both of these 2 outcomes, sensitivity analyses revealed no undue influence on the pooled results from any of the individual studies.
|
Relative to women allocated to receive placebo, those who received progestational agents (17-hydroxyprogesterone caproate and other forms of progesterone) had lower rates of perinatal mortality (14.8% versus 17.1%; OR 0.69, 95% CI 0.38–1.26) (Table 5). Identical results were obtained when we analyzed studies that evaluated only 17-hydroxyprogesterone caproate (3.8% versus 5.2%; OR 0.69, 95% CI 0.38–1.26). No significant heterogeneity was noted for the outcome preterm delivery, and thus a fixed-effects model was employed. Funnel plots were symmetric, indicating no evidence of publication bias. Sensitivity analyses did not reveal undue influence on the pooled results from any of the individual studies. In the meta-regression analysis, neither year nor country of study nor type nor dose of progestational agent administered was significantly associated with the rate of perinatal mortality.
|
Only 3 studies assessed the impact of progesterone administration on the incidence of respiratory distress syndrome. For this outcome, the rate was lower for those receiving 17-hydroxyprogesterone (9.5% versus 11.6%), but this difference did not achieve statistical significance (OR 0.83, 95% CI 0.25–2.76). There was evidence of statistical heterogeneity, and thus a random-effects model was employed. The lack of statistical significance may reflect a type II error.
|
CONCLUSION |
|---|
|
TOP
ABSTRACTSOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
-hydroxyprogesterone caproate, but also in those that studied other progestational agents. Our review assessed several additional outcomes. Patients who received progesterone, when compared with those receiving placebo, have lower rates of hospital admissions for threatened preterm labor and newborns weighing less than 2,500 g. Overall, only 5 trials assessed the outcome of threatened preterm labor, but the definition of this outcome was not included in most of the studies.
Although the rates of perinatal mortality were lower for women receiving any type of progestational agent, the differences did not achieve statistical significance. Because of the relatively small sample size, this could be the consequence of a type II error. For example, to show a difference in perinatal mortality between 3.8% and 5.2%, for an of 0.05 and 80% power, 3,583 patients would be required in each group.
Although systematic reviews with meta-analyses provide an explicit method for synthesizing evidence and early assessment regarding effectiveness of treatments, they may not be as valuable as a single large RCT. Meis et al,7 in a large, multicenter RCT, noted a significant reduction in preterm births among women who were allocated to receive 17-hydroxyprogesterone caproate. Likewise, the majority of randomized trials selected for this systematic review demonstrated a lower preterm birth rate among women receiving progestational agents. Accordingly, the high agreement of our systematic review with the result of the large multicenter trial justifies aggregating results and confirms our conclusion.
We limited our meta-analysis to RCTs because they are the most reliable method of assessing efficacy of therapeutic interventions. However, studies are more likely to be published if they show statistically significant findings. A review in which only published articles are analyzed might be biased toward larger studies or statistically significant effects or both, and erroneous conclusions might result. It is unlikely that bias of this type was a factor in the findings of our systematic review because individual studies of varying size reached similar conclusions. Unidentified unpublished trials with negative results would have to include a large number of women to negate our findings. Furthermore, funnel plot analyses indicated no obvious publication bias.
Two important issues related to meta-analysis are the quality of the studies and heterogeneity. Based on the results of our blinded and rigorous method of scoring each study, we believe that the studies included were of high quality. In performing meta-analysis, one assumes similar treatment regimens among the various evaluated trials. Excessive heterogeneity in study design might prevent meta-analysis from identifying clinically important benefits of therapy. Although absolute homogeneity among studies might appear desirable, it could in fact preclude generalization to a broad clinical population. Nonetheless, understanding the sources of heterogeneity remains important.
Only one of the studies included in this meta-analysis focused on multiple gestations.26 Although the findings of the Hartikainen-Sorri study failed to show a significant reduction in preterm delivery, its inclusion did not alter the pooled results (OR 0.41, 95% CI 0.29–0.58). Clinicians should be aware, however, that substantially less data support the efficacy of progestational agents in the prevention of preterm births among patients with multiple gestations compared with singletons.
Our systematic review and meta-analysis has several limitations. First, the number of patients and the number of outcome events were modest. Therefore, our meta-analysis lacked statistical power to provide precise estimates of incidence and to detect statistically significant and clinically important differences in some outcome variables (perinatal mortality). Second, the design of the included studies varied. Differences among trials are inevitable, because individual trials involve different samples with different treatment protocols; heterogeneity always exists, even within individual trials. Third, we pooled data from subgroups of patients receiving different types and doses of progestational agents. However, beneficial effects were often noted when these agents were analyzed separately. Finally, meta-analysis remains retrospective research that is subject to the methodologic deficiencies of the included studies.
|
Footnotes |
|---|
Received April 26, 2004. Received in revised form June 25, 2004. Accepted July 8, 2004.
doi:10.1097/01.AOG.0000150559.59531.b2
|
REFERENCES |
|---|
|
TOP
ABSTRACTSOURCESSTUDY SELECTIONRESULTSCONCLUSIONREFERENCES |
|---|
2. King JF, Grant AM, Keirse MJNC, Chalmers I. Beta-mimetics in preterm labour: an overview of randomized controlled trials. Br J Obstet Gynaecol 1988;95:211–22.[Medline]
3. The Canadian Preterm Labor Investigators Group. Treatment of preterm labor with the beta-adrenergic agonist ritodrine. N Engl J Med 1992;327:308–12.[Abstract]
4. O'Leary JA. Prophylactic tocolysis of twins. Am J Obstet Gynecol 1986;154:904–5.[Medline]
5. Allbert JR, Wise CA, Lou CH, Gookin KS, Parmenter MA, Morrison JC. Subcutaneous tocolytic infusion therapy for patients at very high risk for preterm birth. J Perinatol 1992;12:28–31.[Medline]
6. Keirse MJ. Progestogen administration in pregnancy may prevent preterm delivery. Br J Obstet Gynaecol 1990;97:149–54.[Medline]
7. Meis PJ, Klebanoff M, Thom E, Dombrowski MP, Sibai B, Moawad AH, et al. Prevention of recurrent preterm delivery by 17-hydroxyprogesterone caproate. N Engl J Med 2003;348:2379–85.
8. da Fonseca EB, Bittar RE, Carvalho MH, Zugaib M. Prophylactic administration of progesterone by vaginal suppository to reduce the incidence of spontaneous preterm birth in women at increased risk: a randomized placebo-controlled double-blind study. Am J Obstet Gynecol 2003;188:419–24.[Medline]
9. Goldstein P, Berrier J, Rosen S, Sacks HS, Chalmers TC. A meta-analysis of randomized control trials of progestational agents in pregnancy. Br J Obstet Gynaecol 1989;96:265–74.[Medline]
10. Daya S. Efficacy of progesterone support for pregnancy in women with recurrent miscarriage: a meta-analysis of controlled trials. Br J Obstet Gynaecol 1989;96:275–80.[Medline]
11. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomized controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 1999;354:1896–900.[Medline]
12. Schulz KF, Chalmers I, Hayes R, Altman DG. Empirical evidence of bias: dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995;273:408–12.[Abstract]
13. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies. J Natl Cancer Inst 1959;22:719–48.
14. DerSimonian R, Laird N. Meta-analyses in clinical trials. Control Clin Trials 1986;7:177–88.[Medline]
15. L'Abbe KA, Detsky AS, O'Rourke K. Meta-analysis in clinical research. Ann Intern Med 1987;107:224–33.
16. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
17. Noblot G, Andra P, Dargent D, Faguer B, Mellier G. The use of micronized progesterone in the treatment of menace of preterm delivery. Eur J Obstet Gynecol Reprod Biol 1991;40:203–9.[Medline]
18. Kauppila A, Hartikainen-Sorri A, Janne O, Tuimala R, Jarvinen PA. Suppression of threatened premature labor by administration of cortisol and 17-hydroxyprogesterone caproate: a comparison with ritodrine. Am J Obstet Gynecol 1980;138:404–8.[Medline]
19. Breart G, Lanfranchi M, Chavigny C, Rumeau-Rouquette C, Sureau C. A comparative study of the efficiency of hydroxyprogesterone caproate and of chlormadinone acetate in the prevention of premature labor. Int J Gynaecol Obstet 1979;16:381–4.[Medline]
20. Erny R, Pigne A, Prouvost C, Gamerre M, Malet C, Serment H, et al. The effects of oral administration of progesterone for premature labor. Am J Obstet Gynecol 1986;154:525–9.[Medline]
21. Tapil'skaia NI, Petrosian MA, Lesik EA. Therapeutic efficacy of novel 17 -hydroxyprogesterone derivatives in experimental preterm labor [in Russian]. Eksp Klin Farmakol 2002;65:44–5.
22. Corrado F, Dugo C, Cannata ML, Bartolo MD, Scilipoti A, Stella NC. A randomized trial of progesterone prophylaxis after midtrimester amniocentesis. Eur J Obstet Gynecol Reprod Biol 2002;100:196–8.[Medline]
23. Hauth JC, Gilstrap LC, Brekken AL, Hauth JM. The effect of 17 -hydroxyprogesterone caproate on pregnancy outcome in an active-duty military population. Am J Obstet Gynecol 1983;146:187–90.[Medline]
24. Johnson JWC, Austin KL, Jones GS, Davis G, King TM. Efficacy of 17 -hydroxyprogesterone caproate in the prevention of premature labor. N Engl J Med 1975;293:675–80.[Abstract]
25. Yemini M, Borenstein R, Dreazen E, Apelman B, Mogilner M, Kessler I, et al. Prevention of premature labor by 17 -hydroxyprogesterone caproate. Am J Obstet Gynecol 1985;151:574–7.[Medline]
26. Hartikainen-Sorri AL, Kauppila A, Tuimala R. Inefficacy of 17 -hydroxyprogesterone caproate in the prevention of prematurity in twin pregnancy. Obstet Gynecol 1980;56:692–5.
27. Goldzieher JW. Double-blind trial of a progestin in habitual abortion. JAMA 1964;188:651–5.[Medline]
28. Tognoni G, Ferrario L, Inzalaco M. Progestagens in threatened abortion. Lancet 1980;2:1242–3.[Medline]
29. LeVine L. Habitual abortion: a controlled clinical study of progestational therapy. West J Surg 1964;72:30–6.
This article has been cited by other articles:
|
|
 |
|
  S. M. Yellon, C. A. Ebner, and Y. Sugimoto Parturition and Recruitment of Macrophages in Cervix of Mice Lacking the Prostaglandin F Receptor Biol Reprod, March 1, 2008; 78(3): 438 - 444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Rebarber, N. B. Istwan, K. Russo-Stieglitz, J. Cleary-Goldman, D. J. Rhea, G. J. Stanziano, and D. H. Saltzman Increased Incidence of Gestational Diabetes in Women Receiving Prophylactic 17{alpha}-Hydroxyprogesterone Caproate for Prevention of Recurrent Preterm Delivery Diabetes Care, September 1, 2007; 30(9): 2277 - 2280. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chatterjee, J. Gullam, M. Vatish, and S. Thornton The management of preterm labour Arch. Dis. Child. Fetal Neonatal Ed., March 1, 2007; 92(2): F88 - F93. [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. Ng, B. G. Steinetz, S. G. Lasano, and J. T. Zelikoff Hormonal Changes Accompanying Cigarette Smoke-Induced Preterm Births in a Mouse Model Experimental Biology and Medicine, September 1, 2006; 231(8): 1403 - 1409. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. O. Odibo, D. M. Stamilio, G. A. Macones, and D. Polsky 17{alpha}-hydroxyprogesterone Caproate for the Prevention of Preterm Delivery: A Cost-effectiveness Analysis. Obstet. Gynecol., September 1, 2006; 108(3): 492 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Sheehan, G. E. Rice, E. K. Moses, and S. P. Brennecke 5{beta}-Dihydroprogesterone and steroid 5{beta}-reductase decrease in association with human parturition at term Mol. Hum. Reprod., July 1, 2005; 11(7): 495 - 501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Progestins prevent recurrence of preterm birth BMJ, June 4, 2005; 330(7503): . [Full Text] [PDF]
|
|
|
|
|
|
S. Thangaratinam and A. Coomarasamy Progestational Agents to Prevent Preterm Birth: A Meta-Analysis of Randomized Controlled Trials Obstet. Gynecol., June 1, 2005; 105(6): 1483 - 1484. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
