Does acupuncture have a role in the treatment of threatened miscarriage? Findings from a feasibility randomised trial and semi-structured participant interviews

The objectives of this study were twofold – firstly to examine the feasibility and acceptability of offering acupuncture as measured by recruitment and retention. A second objective was to explore clinical outcomes with ongoing pregnancy as a primary outcome and physical and emotional wellbeing as secondary outcomes. This was through a mixed methods study involving a pragmatic randomised controlled trial and semi structured interviews on the women’s experiences on exiting the trial. For women experiencing threatened miscarriage, trial allocation was to a group receiving acupuncture and individualised self-care advice or an active control group receiving a touch intervention receiving medical self-care advice. The randomisation sequence was computer-generated and managed by Sealed Envelope, an Internet randomisation service. Allocation was held centrally and accessed by computer by the researcher (DB).

All women received two visits for the first week of intervention, followed by a weekly visit until 12 completed weeks of gestation. Treatments were delivered in the setting of an acupuncture school clinic or if requested in the women’s home environment. The acupuncture treatment protocol was based on a choice of an eight principle diagnosis from textbook recommendations [4, 24, 26, 42], with additional flexibility to treat underlying patterns of disharmony, historical diagnostic approaches that are still in use today among traditional acupuncture practitioners, according to an established acupuncture text [25]. The needles brand used was ‘Hwato’. Single use disposal needles: 0.22 × 25 mm, 0.25 × 40 mm and 0.30 × 75mm The acupuncture intervention included: needles, moxibustion therapy with smokeless moxa sticks (a therapy that involving providing heat to acupuncture points), and self-care advice specific to the diagnosis made by the researcher (DB), a licensed practitioner with over 20 years clinical experience in delivering pregnancy acupuncture. This self-care advice related to the women’s individual diagnosis and contained explanations that related to traditional Chinese medicine theory (see Additional file 1). Point location and needling depths were according to a recognised acupuncture text [13]. A manual needling technique was used for obtaining deqi (a characteristic needling sensation sought when treating by traditional acupuncturists), with a needle retention time of 20 min, with a maximum of six acupuncture points at any one treatment.

Usual care is recommended as a control for use in pragmatic trials, however, with no usual care other than “watchful waiting’ there were ethical concerns that randomising women to a no-treatment group was unacceptable. Therefore an active control group received touch to non-acupuncture points and medical self-care advice (see Additional file 2). This offered components to control for the non-specific therapeutic effects offered by acupuncture, such as receiving attention, touch, time out to rest and having a health practitioner to consult. This light touch was provided by the researcher (DB). Light touch was used as this has been successfully used in acupressure trials during childbirth to control for these non-specific therapeutic effects [11, 19, 21, 22].

Participants were women with vaginal bleeding and a viable pregnancy of 6–11 completed gestational weeks, living in the Wellington and Hutt Valley areas of New Zealand. Light vaginal bleeding was defined as less than the flow of a menstrual period and lasting longer than a day, with heavy bleeding defined as bleeding heavier than the flow of a menstrual period. Recruitment was undertaken through health professionals providing women with an information pamphlet. Practitioners referred through a fertility unit, a Maternity Assessment Unit (MAU) within a hospital and midwives practicing as lead maternity caregivers in the community.

Data to assess the objective of feasibility was collected for recruitment rates, referral sources, eligibility criteria, and retention rates. Data for ongoing pregnancy rates was collected through telephone follow up and hospital records at 12 and 20 gestational weeks and at birth for pregnancy loss and pregnancy and perinatal outcomes.

Data to assess physical and emotional wellbeing was collected at baseline then weekly until women exited the trial. Wellbeing was assessed using the Measure Yourself Medical Outcome Profile (MYMOP) questionnaire; a validated questionnaire used to assess patient centred outcomes in Complementary and Alternative medicine (CAM) research [30‐32]. Participants were asked to score their general wellbeing and two symptoms they nominated as a concern on a continuum between 0 and 6, with 6 being as bad as they could imagine.

No previous randomised trials have evaluated the potential of acupuncture on pregnancy outcomes for women presenting with threatened miscarriage. In this pilot trial, a modest sample size of 40 women was sought to provide data to answer the study questions. This number of participants allowed for the examination of the study questions relating to feasibility.

Data was entered into an Excel spreadsheet and exported to SPSS version 19.1 for analysis. Analysis was undertaken blind to group allocation. Descriptive statistics were used to examine the demographic and baseline characteristics of trial participants. Analyses of the endpoints used an ‘intention to treat’ approach and compared differences in the outcomes of the groups. Endpoints were analysed between groups using a mean t-test and chi-square tests with P < 0.05 considered statistically significant. Dichotomous data was expressed as Risk Ratio with 95 % confidence intervals.

Women from both groups received an average of four treatment sessions suggesting that women found the acupuncture intervention acceptable (mean of 4.1 (SD 1.6) for the touch group and 4.3 (SD 1.8) for the acupuncture group). From 31 possible TCM diagnoses, 12 were utilised (Table 1). A total of 28 acupuncture points were used (Table 2) with moxibustion also used between visits for three women (16.6 %). No major adverse events occurred.

Most participants were in a relationship and had previously used CAM (Table 3). Mean maternal age was 30.5 years (SD 5.9). The mean gestational age at which vaginal bleeding commenced was 6.2 weeks (SD 1.8) with the mean gestational age for entering the trial 8.3 weeks (SD 1.4). A total of nine women were undergoing fertility treatment at the time of trial entry. There were no significant group differences detected between any baseline variables at trial entry.

There were a total of eight pregnancy losses (Table 4). Although five women experienced pregnancy loss in the touch group compared to three women in the acupuncture group, this was not statistically significant (RR 0.70; 95 % CI, 0.19, 2.13, p = 0.58). There were no differences in pregnancy complications between groups for either stillbirth or neonatal death. While it was not possible to examine miscarriage tissue for chromosomal abnormality, the ultrasound report for one woman in the acupuncture group with early miscarriage did indicate chromosomal abnormality. For late pregnancy losses, one was unexplained, one followed an APH at 27 weeks and one woman had premature labour at 23 weeks. There were no differences in pregnancy complications between groups.

For the MYMOP questionnaire two women did not have any symptoms of concern and not all selected a second symptom of concern. The majority of the 54 symptoms selected related to emotional aspects such as anxiety over bleeding, and feelings of frustration and depression (24, 44.4 %), followed by abdominal cramping (13, 24.0 %) and general symptoms such as back pain (8, 14.8 %).

On exiting the study, women receiving acupuncture indicated less concern for their primary symptom (MD 1.36; 95 % CI, 0.07, 2.66, p = 0.04) compared with the control (Table 4). Whilst we were underpowered to demonstrate differences between groups we have preliminary findings suggesting some benefit to quality of life, however, these findings require caution due to the large confidence in intervals.

Of the 11 women interviewed, seven had received the acupuncture treatment and four were from the control group. An overarching theme emerged from the data of ‘finding something you can do.’ This encompassed the themes: ‘they said there was nothing they could do,’ ‘feeling the benefits,’ and managing while marking time.

This theme captures how women participated in the trial due to a perceived lack of control and how they welcomed the opportunity to participate in self-care in a positive way:

I actually felt like I was actively doing something to try and help my situation, because otherwise you are left in a little bit of a black hole (Shelly).

I think doing nothing feels quite helpless really. It’s all outside of your control anyway but at least trying acupuncture makes you feel like you’re actually doing something (Janice).

For all women, the initial advice they received from their health practitioners was difficult to accept. It was not that practitioners were viewed as neglectful or deliberately unkind—just that, as part of their job, they were treating threatened miscarriage as a common situation and failed to appreciate how distressing it was for women to be told to ‘wait and see; and ‘try and not worry’:

The fact that when you ring up somebody who’s meant to be there—the clinic who’s helped you conceive and they’re meant to be a bit more supportive and all they say is, there’s not much you can do about it. It’s a bit of a shock (Rachel).

I guess what really chimed with me was about being told to go home and try not to worry about it…and it’s impossible. You know it’s utterly impossible to try not to worry about it (Annie).

While women expected, and all received follow up ultrasound appointments, there was also an expectation they would receive advice and support from their practitioners. Women discussed their dissatisfaction in terms of staff not spending time to talk with them, the lack of follow up and a lack of information:

All that happened is every time I had a bleed I’d ring my doctor and I had to go in and have my bloods taken and go for a scan. There wasn’t really anything else. It was just you’ve got to wait and see (Shelly).

The doctors like, well there is nothing you can do. There’s no information given to me. I think I would be more proactive if someone said that to me now, I’d be like, well that’s actually not true. (Lucy).

This theme illustrates why women saw participation in the study as useful to them, including talking to someone who understood. This was seen as support that allowed them to express any guilt and negative feelings and also to reduce the pressure they felt they were placing on their partners and families:

Debra acknowledged how I was feeling …she was not my family. She wasn’t my friend. She was someone that knew about it (Kelly).

While the majority of women described both touch and acupuncture as beneficial in helping with their stress and anxiety, those women receiving acupuncture treatment also described beneficial physical effects relating to changes in bleeding, cramping and back pain that were not discussed by women in the touch group. For two women who had been experiencing heavy bleeding and abdominal cramping for several weeks, acupuncture was perceived as beneficial due to the symptom changes following their first treatment:

The next day my bleeding stopped…yeah, I was shocked and I was really happy. I was a little bit sceptical, and I kind of was thinking oh is it all just —this acupuncture stuff, is it all just in people’s heads kind of thing, but no. I know now that it works (Kelly).

I was going through quite a lot of cramping and I had a treatment and that went away straight away, which was brilliant (Jill).

This theme illustrates how all women saw themselves as having to manage until they achieved the safety of their second trimester—that each week offered more hope and reassurance until they were able to ‘get to that 3 months and everything would be OK’ (Susie). Women sought reassurance through the presence of nausea, which was taken as a positive sign by all women who experienced it. Although ultrasounds were also spoken of as powerful reassurance, women did not necessarily see these as a guarantee that the pregnancy would continue with no further problems. Several women expressed concerns that there would be ongoing problems with their pregnancy:

Even after seeing the scan I kept thinking oh no, something’s going to go wrong. I’m not sure why (Ellen).

The use of a mixed methods approach in this study identified preliminary findings that resulted in a greater understanding of how acupuncture supported women experiencing threatened miscarriage than possible from using a quantitative or qualitative approach in isolation. In addition the use of a pragmatic acupuncture treatment protocol allowed flexibility for diagnosis and treatment, increasing external clinical validity for practitioners by reflecting a treatment approach applicable to real-world acupuncture. Although pragmatic trial designs do not provide information on the effect of specific acupuncture points and needling effects, they are recommended when investigating acupuncture as a complex intervention. Where complex interventions refer to treatment effects not only derived from the use of needles, but also through dietary and lifestyle advice specifically tailored to meet diagnostic criteria that involve the participants [23, 27].

It was also a strength that the same researcher (DB) delivered both acupuncture and light touch intervention, allowing for consistency in the support given to both groups of women. The use of a sole practitioner to deliver the study interventions provided a consistent approach to the acupuncture diagnosis and treatment delivery, however, a sole practitioner delivering treatment also has the potential to introduce a treatment bias that may not be generalisable to other practitioners. Future pragmatic trials involving several acupuncturists would reduce this potential. A limitation was that it was not possible to blind women to treatment allocation and therefore not possible to control for patient expectation. The use of an active control group receiving touch was also a limitation as both groups received non-specific effects relating to supportive care. This may have impacted negatively on the findings for the acupuncture group however; it was not seen as ethical to randomise women to receive no care. To determine the effects of delivering acupuncture as a complex intervention treatment with effects not only derived from the use of needles, but also through dietary and lifestyle advice future research could focus on women receiving usual care rather than touch, for example in a population of women receiving fertility treatment with usual care in the form of ultrasound monitoring. The specific effects of needling such as acupuncture point selection, the depth, degree of stimulation, retention time and number of needles remain to be explored in future efficacy studies using a sham control.