Fen Zhou,Yu-fang Hao, Yan Chen, and Tong Wang

Nursing School, Beijing University of Chinese Medicine, Beijing 100102, China

POLYHYDRAMNIOS is the medical condition in which the amniotic fluid index (AFI) is greater than 20 cm or a single pocket of fluid is greater than 8 cm, which theoretically means an amniotic fluid volume exceeding 2000 mL.1Its incidence has been reported as ranging from 1% to 3%.2The condition has diverse etiology and involves many serious outcomes. The maternal hazards include respiratory compromise, antepartum hemorrhage, abnormal presen- tations, uterine dysfunction, postpartum hemorrhage, and increased operative intervention. The perinatal outcome may be compromised due to prematurity, placental abruption, and umbilical cord prolapse. Moreover, the causes of polyhydramnios such as maternal diabetes mellitus, fetal congenital malformations, and twins are also associated with an adverse perinatal outcome.3-5

For patients with polyhydramnios developing evidence of respiratory embarrassment and excessive uterine activity, one of the therapeutic options is decompression by amniocentesis.6However, it has a risk of preterm labor and infections because of its nature of being an invasive procedure.7Another therapeutic method to decrease the production of amniotic fluid is administration of indome- thacin (an anti-prostaglandin), but it has been reported that main adverse events may occur such as premature closure or constriction of the ductus arteriosus.8Therefore, it remains necessary to seek novel effective drugs for polyhydramnios treatment.

Traditional Chinese medicine (TCM) has developed deep and comprehensive understanding of polyhydramnios very early. This disease was firstly recorded in General Treatise on the Cause and Symptoms of Diseases (Zhu Bing Yuan Hou Lun in Chinese), by the name of “zi man”. In five or six months of pregnancy, it appears as hydramnios, anomalous accretion of belly, fullness in chest and diaphragm, and even inability to lie down because of heavy breath.9Its pathogenesis is mainly due to deficiency of Yang in both the spleen and the kidney, transformation failure of the spleen, and retention of water and dampness. TCM treatment of polyhydramnios involves warming, tonifying kidney-spleen, and activating qi to excrete water.10

This meta-analysis and systematic review aims to assess the efficacy and safety of Chinese herbal medicine in treating polyhydramnios based on randomized controlled trials (RCTs) and controlled clinical trials (CCTs). The trials fell into two groups according to the interventions: Chinese herbal medicine (therapy A) versus indomethacin alone (therapy C), or Chinese Herbal Medicine combined with indomethacin (therapy B) versus indomethacin alone (therapy C). The outcomes include general clinical improvement rate [effective rate (ER)], AFI, maximum vertical pocket (MVP), and rate of constriction of the fetal ductus arteriosus.

MATERIALS AND METHODS

Inclusion criteria

The included RCTs and CCTs must be those in which one group received therapy A or therapy B and the control group received therapy C as the treatment of polyhydramnios. In therapy A and therapy B, any Chinese herbal medicine counts. To allow conclusions regarding the comparison of effects of therapy A or B with that of therapy C, only the trials providing data that enabled outcomes to be tested for significant differences were included.

The trials must be conducted in patients with polyhydramnios (defined as having an amniotic fluid volume exceeding 2 000 mL). The diagnosis of polyhydramnios required one or both of the following conditions: (1) all AFI measurements ≥20 cm； (2) MVP depth＞7 cm. Trials of patients with fetal congenital anomalies detected by ultrasonography were excluded.

Search methods for identification of studies

The published trials were retrieved from the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 4 2012), MEDLINE (1966-April 2012), EMBASE (1980-April 2012), Chinese National Knowledge Infrastructure (CNKI, 1979-April 2012), Chinese Scientific and Technological Periodical Database (VIP, 1989-April 2012), Wangfang Database (1994-April 2012), TCM online (1984-April 2012), and the Chinese Cochrane Center's Controlled Trials Register. Searches were conducted to identify all the relevant studies regardless of language or publication status. The search strategy was constructed by using a combination of subject headings and text words: “Chinese herbal medicine”, “Chinese traditional medicine” or “herb” for people with “polyhydramnios”, “hydramnios” or “excess amniotic fluid”. No terms related to the “comparison” or “outcome” of trials were searched to avoid exclusion of trials in an area where limited research has been conducted. Various combinations of the terms were used, depending on the database searched (full search strategies are available upon request). For example, the search strategies are for the EMBASE are as follows:

#1: “polyhydramnios”/exp OR polyhydramnios

#2: “indomethacin”/exp OR indomethacin

#3: medicine, and (“chineses”/exp OR chinese) AND traditional

#4: drugs, AND (“chineses”/exp OR chinese) AND herbal

#5: “phytotherapy”/exp OR phytotherapy

#6: #3 OR #4 OR #5

#7: #1 AND #2 AND #6

Online journal websites and databases of ongoing trials were also searched. Additional articles were identified by reviewing the reference lists of the retrieved articles and conference proceedings.

Data collection and analysis

Retrieved papers were initially screened based on titles and abstracts by two independent researchers. Full texts of the remaining papers were obtained and assessed independ- ently by both of the two researchers against an inclusion and exclusion checklist. Disagreements were resolved through discussion, or, if failed, consultation with a third researcher.

All data extraction and calculations were performed independently by two researchers using a standardized data extraction form (see Table 1). They were not blinded to study authors, institutions, or journals of publication. The 2 sets of data were then compared for discrepancies, which were resolved through discussion. The data on study characteristics including study design details, participants' characteristics and baseline demographics, interventions, and outcomes were extracted. For dichotomous outcomes, the number of responders and the total number of participants for each study arm were extracted. For continuous outcomes, the mean change and standard deviation (SD) in each group of the trial were extracted along with those of the entirety.

The retrieved articles were assessed by two reviewers independently. To assess the methodological validity of the included studies, the following aspects were evaluated: sequence generation, incorrect performance or absence of sequence allocation concealment, incorrect or absence of blinding, incomplete outcome data, selective outcome reporting, and other potential threats to validity. According to the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions version 5.1.0,11in which the assessments of each aspect was judged to be of low risk, high risk, or uncertain risk (“yes” for a low risk of bias, “no” for a high risk of bias, “unclear” for otherwise). Each study could be divided into three grades as follows: low risk of bias (low risk of bias for all key aspects), unclear risk of bias (unclear risk of bias for one or more key aspects), and high risk of bias (high risk of bias for one or more key aspects). Disagreements were resolved by consultation with a third reviewer.

Statistical analysis

In this study, statistical analyses were conducted using software provided by the Cochrane Collaboration (RevMan 5.1.2). The meta-analysis was performed by pooling similar studies. In comparison in terms of ER of different interventions, the recovery and effectiveness subgroups were first merged into one group to compare with the ineffectiveness subgroup, then the recovery and effectiveness subgroups were compared with ineffectiveness subgroup separately. In comparisons among difference interventions in terms of the rate of fetal ductus arteriosus constriction, the subgroups with different levels (mild, moderate, and severe) of constriction were compared with non-constriction subgroup, first as a whole and then separately. Dicho- tomous data are presented as rate ratio (RR), while continuous data are expressed as weighted mean differ- ence (WMD), all with 95% confidence interval (CI). The presence of statistical heterogeneity was explored by I2test with significance level set at P＜0.1. For I2＜30%, a fixed-effect model was used； for I2≥30%, a random-effect model was applied. All P values were two-sided and P＜0.05 was considered statistically significant.

RESULTS

Description of studies

A total of 168 papers were retrieved, of which 64 were excluded for irrelevance and duplication, and 45 excluded based on titles and abstracts. Full texts were obtained for the remaining 59 papers, of which 54 papers were excluded as they did not meet the inclusion criteria and 5 papers were included in the review (Fig. 1). The number of participants in each of the 5 studies ranged from 30 to 179, with a total of 1017 patients (cases in the investigation groups) in the 5 studies. All the 5 studies were conducted in China, but none was designed as a multicenter study.

Overall, the included studies involved interventions with therapy A, therapy B, and therapy C. The Chinese herbal medicine in therapy A or B was compound Chinese Traditional Medicine (classical prescriptions or self-constituted prescriptions), or modified compound with addition or subtraction. Classical prescriptions with modifications were used in 3 studies (with 3 compound prescriptions, including diuresis and fetus protection, Wupi decoction added with radix astragali, and Wuling powder), and self-constituted prescriptions with modifications used in 2 studies (mainly tonifying the kidneys and the spleen, diuresis and fetus protection).

According to the difference in interventions, the following analyses are divided into therapy A versus therapy C and therapy B versus therapy C. The characteristics of the included studies are listed in Table 1.

Risk of bias of the included studies

The risk of bias of the included is summarized in Figure 2. Three included studies15-17claimed “randomization”, but did not describe the sequence generation method in detail. In addition, none of the included studies described allocation concealment and any information about blinding.

Table 1. Characteristics of the included studies

Figure 1. Flow chart showing the study selection and exclusion process. RCT: randomized controlled trial； CCT: controlled clinical trial.

Figure 2. Risk of bias of the included studies with reviewers' judgments about each risk of bias item presented as percentages across all the included studies.

Therapy A versus therapy C

For therapy A versus therapy C, 3 papers12,15,16were included and fully extracted data were general clinical improvement rate and MVP, whereas only 2 papers15,16provided constriction fetal ductus arteriosus rate.

General clinical improvement rate In the 3 studies, ER was reported as “effective rate ratio” and calculated as the ratio between the proportion of responders in the treatment group and the control group. Although the criteria for efficacy evaluation were self-designed by the authors of the studies, they were all based on clinical manifestations and laboratory examinations. Outcomes were divided into recovery, effectiveness, and ineffectiveness； and the 3 studies12,15,16had the same definitions for these outcomes. Therefore we could conduct meta-analysis to compare the ERs of these studies.

The 3 trials12,15,16involving 740 patients provided data for general clinical improvement rate. A pooled analysis performed using random-effect model (I2=23%-76%, P=0.01) identified no significant differences between therapy A and therapy C (P=0.25-0.48). Compared with therapy C, therapy A did not elevate the general clinical improvement rate (Fig. 3).

MVP A pooled analysis of the 3 trials12,15,16involving 740 patients performed using a random-effect model (I2=73%, P=0.02) indicated that therapy A did not significantly improve MVP compared with therapy C (Fig. 4).

Rate of fetal ductus arteriosus constriction This outcome was reported in 2 trials,15,16which included 640 patients. A pooled analysis using the fixed-effect model (I2=0%, P=0.75) revealed significant differences between therapy A and therapy C (P＜0.01, Fig. 5).

Therapy B versus therapy C

For therapy B versus therapy C, 3 papers12,13,17were included and fully extracted data were general clinical improvement rate and MVP, whereas only 2 papers12,17provided AFI.

General clinical improvement rate The 3 trials12,13,17involving 327 patients provided data for this outcome. A pooled analysis using random-effect model (I2=78%, P=0.01) identified significant differences between therapy B and therapy C ( P＜0.001, Fig. 6).

Figure 3. Forest plot of general clinical improvement in trials comparing therapy A with therapy C in the treatment of polyhydramnios. CHM: Chinese herbal medicine.

Figure 4. Forest plot of normal versus abnormal MVP in trials comparing therapy A with therapy C in the treatment of polyhydramnios.

Figure 5. Forest plot of fetal ductus arteriosus constriction in trials comparing therapy A with therapy C in the treatment of polyhydramnios.

MVP and AFI A pooled analysis of the 3 trials12,13,17using a random-effect model (I2=96%, P＜0.001) did not find any significant difference in MVP between therapy B and therapy C (Fig. 7).

AFI was reported in 2 trials,12,17which included 160 patients. A pooled analysis using the fixed-effect model (I2=80%, P=0.03) detected significant differences between therapy B and therapy C (P＜0.05, Fig. 8).

Figure 6. Forest plot of general clinical improvement in trials comparing therapy B with therapy C in the treatment of polyhydramnios.

Figure 7. Forest plot of normal versus abnormal MVP in trials comparing therapy B with therapy C in the treatment of polyhydramnios.

Figure 8. Forest plot of amniotic fluid index (AFI) in trials comparing therapy B with therapy C in the treatment of polyhydramnios.

Adverse events

Adverse events were not mentioned in the 5 trials in therapy A and therapy B groups, while 2 trials15,16reported the rate of fetal ductus arteriosus constriction in therapy C group.

DISCUSSION

This systematic review included data from 3 studies12,15,16involving 740 patients in which participants received either therapy A or therapy C. The results did not support that either therapy A or therapy C produce better general clinical improvement rate or MVP. But in reducing constriction of the fetal ductus arteriosus, therapy A was found superior to therapy C. On the other hand, for therapy B versus therapy C, data from 3 studies12,13,17involving 327 patients were included. The results showed that compared with therapy C, therapy B could significantly improve general clinical improvement rate and AFI. No significant adverse effects were mentioned in the included studies.

However, concerns about the quality of the included studies are noteworthy in weighing the results of this meta-analysis.

Firstly, the randomized trials included have several methodological flaws. Although 3 trials15-17adopted ‘‘randomization'', none of those mentioned adequate concealment of the allocation sequence, which may produce potential risk of selection bias. None of the trials were blinded, which may lead to performance bias. Another shortcoming about the methodology is that none of the trials provided pre-calculated sample size, which is essential for ensuring adequate sample size and statistical power.

Secondly, only a search strategy in database, however comprehensive, is not adequate. Although we did not limit our search language, we found all articles published in Chinese, geographic biases may be induced. Furthermore, we did not check the references of mannual searching nor contact institutions, societies, specialists, and authors of included studies to identify any additional published or unpublished data.

Thirdly, other flaws of the included trials may produce other bias or decrease the reliability of conclusions. Some selection criteria were common among the studies, as described above； however, selective bias remained due to differences in eligibility for participation in each study. Besides, therapy A in the included trials mainly have modification of classical prescriptions or self-constituted prescriptions, not applying uniform Chinese herbal medicine, but we regarded them as one type of treatment in this systematic review. Furthermore, some trials compare therapy A+C with therapy C, likely to generate false positive results.19And there is a lack of information on patient adherence to interventions, which is a primary determinant of treatment success and clinical benefits.20

High quality trial design (obeying the Consolidated Standards of Reporting Trials statement21), with intention- to-treat analysis, baseline similarity, objective outcome standards, adequate follow-up, and patients' adherence to treatment would advance the quality of study in this field.

In conclusion, therapy A and therapy C appeared to result in comparable outcomes. Compared with therapy C, therapy A could reduce the incidence of fetal ductus arteriosus constriction. Therapy B results in better outcome in terms of general clinical improvement rate and AFI than therapy C does for patients with polyhydramnios. Further large-scale high-quality trials are warranted.

1. Ronald SG, Beth YK, Arthur FH, et al. Danforth's obstertrics and gynecology. Philadelphia: Lippincott Williams & Wilkins； 2008.p.141.

2. Abhyankar S, Salvi VS. Indomethacin therapy in hydramnios. J Postgrad Med 2000； 46: 176.

3. Volante E, Gramellini D, Moretti S, et al. Alteration of the amniotic fluid and neonatal outcome. Acta Biomed 2004； 75 Suppl 1: 71-5.

4. Dashe JS, Mclntire DD, Ramus RM, et al. Hydramnios: anomaly prevalence and sonographic detection. Obstet Gynecol 2000； 100: 134-9.

5. Harman CR. Amniotic fluid abonormalities. Semin Perinatol 2008； 32: 288-94.

6. Cunningham FG, Macdonald PC, Gant NF, et al. Diseases and abnormalitles of the fetal membranes. Williams Obstetrics. 20th ed. Stamford: Appleton & Lange； 1997. p. 657-67.

7. Elliott JP, Sawyer AT, Radin TG, et al. Large-volume therapeutic amniocentesis in the treatment of hydramnios. Obstet Gynecol 1994； 84: 1025-7.

8. Moise KJ Jr, Hunta JC, Sharif DS, et al. Indomethacin in the treatment of premature labor. Effects on the fetal ductus arteriosus. N Engl J Med 1988； 319: 327-31.

9. Ouyang HQ. Obstetrics and gynecology of traditional Chinese medicine. Beijing: People's Medical Publishing House； 2005.p.179-81.

10. You QH, Qiu MJ, Zhao CY. Clinical study of Chinese herbal medicine with the effect of warming and tonifying the kidneys and the spleen in treating syndrome of polyhydramnios. Liaoling J Trad Chi Med 2007； 34:473-4.

11. Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions version 5.1.0. 2011 [cited 2011 March 1]. Available from: http://www. cochrane- handbook.org

12. Xu HS. Modification of promoting urination and fetus protection of prescription in 50 cases with polyhydramnios. Hunan J Trad Chi Med 2003； 19:44.

13. Li LJ, Cai YP, Song SQ. Analysis of combination of Chinese traditional and Western medicine in 124 cases with polyhydramnios. Henan Trad Chin Med 2003； 23:50-1.

14. Le J. Obstetrics and Gynecology. Beijing: People's Medical Publishing House； 1996.p.134-5.

15. Lu YH, Zhou LM, Lv FM, et al. Clinical observation of Chinese herbal medicine in 179 cases with polyhydramnios. Jiangshu J Trad Chin Med 2010； 42: 41-2.

16. Chen GF, Sun JZ, Lan SH, et al. Curative observation of Chinese herbal medicine in 167 cases with polyhydramnios. Guang Ming Zhong Yi 2007； 22: 42-3.

17. Lan XL. Clinical study of wupi and adding radix astragali decoction in treating syndrome of chronic polyhydramnios. Hubei J Trad Chin Med 2011； 33: 22-3.

18. Tang WX. Obstetrics and gynecology of traditional Chinese medicine. Beijing: People's Medical Publishing House； 2002. p.526-31.

19. Ernst E, Lee MS. A trial design that generates only “positive” results. J Postgrad Med 2008； 54: 214-6.

20. World Health Organization. Adherence to long-term therapies: evidence for action. Switzerland: WHO； 2003.

21. Schulz KF, Altman DG, Moher D, et al. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010； 340: c332.