Advertisement

A comparison of two versus four sterile water injections for the relief of back pain in labour: A multicentre randomised equivalence trial

Published:February 10, 2022DOI:https://doi.org/10.1016/j.wombi.2022.02.002

      Abstract

      Background

      Recent trials demonstrated the safety and efficacy of sterile water injections to provide relief from labour back pain. While four injections is the most common approach variations in technique, such as employing two injections, are also used.

      Aim

      To determine if the analgesic effect of two sterile water injections is clinically equivalent to four.

      Methods

      238 women in labour with a Visual Analogue Scale pain score (VAS) of 70 millimetres (mm) (0 = no pain; 100 = worst pain imaginable) were randomised to two or four sterile water injections. The primary outcome was pain measured on a VAS at 30 min post treatment. A priori margin of equivalence was set at ±10 mm. Secondary outcomes included the likelihood of achieving an at least 30% and 50% reduction in pain, birth and neonatal outcomes.

      Results

      At 30 min post-injection the difference in VAS scores between the techniques was −5.97 (95% Confidence Interval [CI] −13.18–1.22). As the lower end of the CI exceeds the margin of −10 mm equivalence was not demonstrated. Both techniques achieved an at least 30% reduction in pain in over 75% of participants though duration of effect was longer in the four injection group. There was no difference in other birth related secondary outcomes.

      Conclusion

      Four injections provided a margin of benefit over two injections in level and duration of analgesia.

      Discussion

      Four injections remains the technique of choice though two injections still provided significant pain relief and would be suitable where it was not possible or desirable to provide four.

      Keywords

      Statement of significance
      Problem
      It is not known if two sterile water injections for managing back pain in labour provides the same degree and duration of pain relief as four injections
      What is already known
      Placebo controlled trials suggest that two sterile water injections provides relief from labour back pain.
      What this paper adds
      While two injections provided significant relief from labour back pain, this was not equivalent to the four injection technique in terms of scale and duration of analgesia.

      1. Introduction

      While the effectiveness of sterile water injections (SWI) to relieve back pain in labour has been demonstrated in our recent large placebo controlled trial [
      • Lee N.
      • Gao Y.
      • Collins S.L.
      • Mårtensson L.B.
      • Randall W.
      • Rowe T.-M.
      • et al.
      Caesarean delivery rates and analgesia effectiveness following injections of sterile water for back pain in labour: a multicentre, randomised placebo controlled trial.
      ] the intensity of the brief pain associated with injections of water has been cited as a deterrent for use by some women [
      • Lee N.
      • Kildea S.
      • Stapleton H.
      “No pain, no gain”: the experience of women using sterile water injections.
      ,
      • Mårtensson L.B.
      • Hutton E.K.
      • Lee N.
      • Kildea S.
      • Gao Y.
      • Bergh I.
      Sterile water injections for childbirth pain: an evidenced based guide to practice.
      ]. One of the key advantages of SWI is the ability to repeat the procedure as needed with no side effects other than the injection pain, however very few of the women participating in recent trials have elected for repeat injections possibly because of the injection pain [
      • Lee N.
      • Gao Y.
      • Collins S.L.
      • Mårtensson L.B.
      • Randall W.
      • Rowe T.-M.
      • et al.
      Caesarean delivery rates and analgesia effectiveness following injections of sterile water for back pain in labour: a multicentre, randomised placebo controlled trial.
      ,
      • Lee N.
      • Webster J.
      • Beckmann M.
      • Gibbons K.
      • Smith T.
      • Stapleton H.
      • et al.
      Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
      ]. The brief but intense painful stimulus provided by the injection of water is thought to be central to the analgesic response through the triggering of ascending and descending neural pain modulation such as gate control and the release of endorphins [
      • Melzack R.
      A new theory. A gate control system modulates sensory input from the skin before it evokes pain perception and response.
      ,
      • Le Bars D.
      • Dickenson A.H.
      • Besson J.-M.
      Diffuse noxious inhibitory controls (DNIC). II. Lack of effect on non-convergent neurones, supraspinal involvement and theoretical implications.
      ]. The procedure normally consists of four injections given at points bordering the Michalis rhomboid of the sacrum [
      • Mårtensson L.B.
      • Hutton E.K.
      • Lee N.
      • Kildea S.
      • Gao Y.
      • Bergh I.
      Sterile water injections for childbirth pain: an evidenced based guide to practice.
      ]. However, the degree of pain triggered by the usual four intradermal injection technique may be in excess of what is required to initiate an acceptable analgesic response.
      A number of studies have sought to address the issue of the injection pain by either varying the depth of the injections from intradermal to subcutaneous or using fewer than the usual four injections to initiate the analgesia. A placebo controlled trial comparing subcutaneous administration to intradermal found no difference in injection pain in laboring women, though a subsequent crossover trial by the same researchers using non-pregnant females did suggest that the injection pain was significantly less in the subcutaneous group [
      • Martensson L.
      • Wallin G.
      Labour pain treated with cutaneous injections of sterile water: a randomised controlled trial.
      ,
      • Martensson L.
      • Nyberg K.
      • Wallin G.
      Subcutaneous versus intracutaneous injections of sterile water for labour analgesia: a comparison of perceived pain during administration.
      ]. The use of a single injection was explored in two studies that also reported significant reductions in back pain compared to a saline placebo [
      • Bahasadri S.
      • Ahmadi-Abhari S.
      • Dehghani-Nik M.
      • Habibi G.R.
      Subcutaneous sterile water injection for labour pain: a randomised controlled trial.
      ,
      • Saxena K.
      • Nischal H.
      • Batra S.
      Intracutaneous injections of sterile water over the sacrum for labour analgesia.
      ]. However, a comparison of a single versus the usual four injection technique found that while a single injection was less painful, the analgesia provided by four injections was significantly greater and lasted for a longer duration than that from a single injection [
      • Lee N.
      • Webster J.
      • Beckmann M.
      • Gibbons K.
      • Smith T.
      • Stapleton H.
      • et al.
      Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
      ]. A recent trial of two injections again demonstrated a significant difference in resulting analgesia compared to placebo [
      • Fouly H.
      • Herdan R.
      • Habib D.
      • Yeh C.
      Effectiveness of injecting lower dose subcutaneous sterile water versus saline to relief labor back pain: randomized controlled trial.
      ]. Though it is not known if a two injection technique would produce a degree and duration of analgesia that is clinically comparable to the standard four injections. Furthermore, if the injection pain from two injections is less than that experienced with four this may increase the woman’s satisfaction and acceptability with SWI treatment during labour, and willingness to use the procedure again during labour, or in future labours.
      The aim of this trial was to determine if the analgesia provide by a two injection SWI technique would be equivalent, within a predetermined clinically relevant margin, to four injections and determine the difference in perceived injection pain.

      2. Design

      We conducted a randomised controlled equivalence trial. As we aimed to investigate if the two treatments were similar in analgesic effect within a predefined clinically relevant margin; the null hypothesis is that there is no difference between the two treatments (equivalence). This should be rejected if the lower end of the 95% confidence interval (CI), of the difference in treatment effect, lies outside of the range of equivalence (−10 mm to +10 mm) [
      • Walker E.
      • Nowacki A.S.
      Understanding equivalence and noninferiority testing.
      ]. The trial took place in six maternity units, five in New South Wales and one in Brisbane, Australia. Annual publically funded births at the participating sites ranged from 500 to 5000. Water injections were available to women as standard care at five of the hospitals and introduced as part of the trial at one hospital.
      Women were eligible to participate if they were 18 years of age or older with a singleton cephalic fetus between 37 and 41 weeks and six days gestation in either spontaneous, induced or augmented labour. Women with back pain in labour equal to or greater than 70 mm on a 100 mmVAS pain scale were offered information about the trial. Previous studies have determined that a pain score of seven or more indicates severe pain [
      • Collins S.
      • Moore R.
      • McQuay H.
      The visual analogue pain intensity scale: what is moderate pain in millimetres.
      ]. Women were excluded if they previously had a significant co-morbidity (e.g. severe hypertension, uncontrolled diabetes), any contraindications to receiving injections (e.g. infection at the injection site, bleeding disorders) had used opioid analgesia less than four hours prior to randomisation or were using of nitrous oxide and oxygen inhalation at time of randomisation.
      Information regarding the trial was provided during routine antenatal visits. Women presenting in labour with back pain and meeting the inclusion criteria were offered further information prior to providing written consent. While obtaining consent during labour for research presents a number of challenges related to the experience of pain and use of medications, evidence suggests that capacity to consent is unaffected, particularly when a staged approach, such as providing trial information in the antenatal period, is used [
      • Dhumale H.
      • Goudar S.
      Ethical issues related to consent for intrapartum trials.
      ,
      • Widmer M.
      • Bonet M.
      • Betrán A.P.
      Would you like to participate in this trial? The practice of informed consent in intrapartum research in the last 30 years.
      ]. If clinicians considered, for reasons such as rapidly advancing labour, that informed decision making may be affected, consent was not sought. Following consent, women were randomised to either a two or four SWI technique. Randomisation schedules were prepared for all sites using a computer generated allocation list in blocks of two to four. Two midwives not directly involved in the woman’s care administered the SWI based on instruction contained in the treatment allocation envelope. The primary midwife, also responsible for VAS data collection, was not present during administration of the SWI. The injections sites were covered with an opaque dressing and the participant was asked not to reveal how many injections she may have experienced.

      2.1 Procedures

      Participants allocated to the intervention group received two injections of sterile water, 0.1–0.3 millilitres (mls) intradermally at the posterior superior iliac spines (PSIS) (Fig. 1). The actual volume injected is determined by a visual assessment of the resulting ‘bleb’ or blister. Participants randomised to the control group were given the standard four injection intradermal technique into the skin surrounding the Michaelis rhomboid over the sacral area: two over the PSIS and two three cm below and one cm medial to the PSIS. Injections in both groups were given by two midwives. Minor discrepancies or changes to the anatomical position or alignment of the four injections have not been shown to impact on any analgesic effect [
      • Mårtensson L.B.
      • Hutton E.K.
      • Lee N.
      • Kildea S.
      • Gao Y.
      • Bergh I.
      Sterile water injections for childbirth pain: an evidenced based guide to practice.
      ]. Following administration of either the intervention or control women received standard care including access to pharmacological (Nitrous Oxide and Oxygen inhalation, opioids and epidurals) or non-pharmacological (e.g. showers, water immersion) which was consistent across all sites. Repeat injections were available to participants as requested. These were not randomised therefore the number of injections used for repeat treatments was based on the participant’s preference.
      Fig. 1
      Fig. 1Sterile water injection points (A Two injections, B Four injections).

      2.2 Outcomes

      The primary outcome was pain at 30 min. post treatment measured on a 100 mm visual analogue scale and the a priori margin for equivalence set at ±10 mm. The VAS is an unbroken line of 100 mm length anchored by the phrases ‘No pain’ (0 mm) and ‘Worst pain imaginable’ (100 mm) and is validated for pain research [
      • Price D.
      • McGrath A.
      • Rafii A.
      • Buckingham B.
      The validation of visual analogue scales as ratio scale measures for chronic and experimental pain.
      ]. VAS outcomes have also been reported as dichotomous outcomes for pain relief of at least 50% or 30% difference between pre and post injection self-reported pain scores at 30, 60, 90 and 120 min as this is the recommended format for clinically relevant outcomes detailed in the Cochrane review [
      • Derry S.
      • Straube S.
      • Moore R.A.
      • Hancock H.
      • Collins S.L.
      Intracutaneous or subcutaneous sterile water injection compared with blinded controls for pain management in labour.
      ]. Other secondary outcomes included at least 30% pain reduction following repeat injections, other analgesia used, mode of birth, neonatal outcomes, satisfaction with SWI, likelihood to use SWI again, and likelihood to recommend it to other women.

      2.3 Statistical analysis

      The measure of equivalence is based on a clinically relevant margin of effect for an existing treatment (four injections) that the new treatment (two injections) would need to achieve to be “close enough” to be considered equally as effective.
      A sample size of 159 in each group was required to achieve 90% power to detect equivalence using a two-sided, two-sample t-test. The significance (alpha) of the test is 0.05. This margin is derived from the minimal difference in VAS scores (±10 mm) that indicates a significant shift in pain perception [
      • Kelly A.-M.
      The minimum clinically significant difference in visual analogue scale pain score does not differ with severity of pain.
      ]. A standard deviation (SD) of ±27 mm on the VAS was estimated from results reported in a recent meta-analysis of SWI studies and our previous work [
      • Hutton E.K.
      • Kasperink M.
      • Rutten M.
      • Reitsma A.
      • Wainman B.
      Sterile water injection for labour pain: a systematic review and meta-analysis of randomised controlled trials.
      ,
      • Lee N.
      • Webster J.
      • Beckmann M.
      • Gibbons K.
      • Smith T.
      • Stapleton H.
      • et al.
      Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
      ]. Given an allowance of 10% for attrition the total number of participants required was 339.
      Data were analysed on an intention to treat basis. Demographic and other baseline characteristics were assessed between groups, however formal statistical testing was not carried out to determine differences based on the recommendations of the CONSORT statement for the reporting of parallel RCTs [
      • Moher D.
      • Hopewell S.
      • Schulz K.F.
      • Montori V.
      • Gøtzsche P.C.
      • Devereaux P.
      • et al.
      CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials.
      ]. The reduction of the mean scores on the VAS between baseline and 30, 60, 90, and 120 min post injection was calculated and analysed using a generalized linear mixed-model repeated measures analysis for the mean difference (95% confidence interval [CI]) between the two treatments. For the secondary analysis of pain outcomes we converted the difference in pre and post treatment VAS scores to a binary outcome (e.g. the proportion of participants that did or did not report an at least 30% and 50% reduction in pain) and expressed as a relative risk (RR). This has been previously cited as the preferred approach to the reporting of pain outcomes [
      • Busse J.W.
      • Bartlett S.J.
      • Dougados M.
      • Johnston B.C.
      • Guyatt G.H.
      • Kirwan J.R.
      • et al.
      Optimal strategies for reporting pain in clinical trials and systematic reviews: recommendations from an OMERACT 12 workshop.
      ].
      For other secondary outcomes an independent t-test was used for variables found to be normally distributed to compare groups, and non-parametric data were analysed using a Mann–Whitney U test. Categorical variables were analysed using chi-squared test or Fisher’s exact test when expected frequency were small (<5). Mean difference and 95% CI are given for the VAS primary outcome and standard deviations (SDs) are use to describe other normally distributed continuous variables. Medians and interquartile ranges (IQR; 25th and 75th centile) are provided for non-parametric data. Relative risk and associated 95% CIs were calculated to compare birth outcomes between the study groups. The level of statistical significance was set at 0.05.

      3. Results

      Between December 2014 and February 2019, 238 participants were recruited to the trial, 119 to each of the two treatment groups. The trial did not achieve the expected sample size due to issues related to limited resources and the COVID-19 pandemic, resulting in a reduction in power for the primary outcome to 70%.There were no post-randomisation exclusions. There were six protocol violations in each group where the inclusion criteria of back pain of VAS of at least 70 mm was not met. Six women in the two injection arm and three in the four injection arm gave birth prior to the primary outcomes. Two women in the four injection group also had epidurals inserted prior to collection of the primary outcome data. Data forms were lost for three participants in both groups and clinical outcome data was incomplete for one participant in each of the trial groups. In keeping with the principles of intention to treat analysis all participants randomized were included in the analysis (Fig. 2). Baseline characteristics were similar between groups. Seventy-eight percent of the participants were nulliparous, and 56% in spontaneous labour. The fetus was assessed as being in the Occipito-Posterior position in 42% of participants and the mean (SD) of the VAS at randomisation was 81.2 (13.1) for the two injection group and 82.3 (13.4) for the four injection (Table 1).
      Table 1Participant baseline characteristics.
      VariableTwo injections (n = 119)Four injections (n = 119)
      Age (years)
       16–2113 (10.9)8 (6.7)
       22–2958 (48.7)65 (54.6)
       30–3532 (26.8)32 (26.8)
       36 or older15 (12.6)14 (11.6)
       Missing data1 (0.0)0
      Parity
       094 (78.9)91 (76.4)
       116 (13.4)17 (14.29
       2 or more9 (7.5)11 (9.1)
      Body Mass Index at booking
       Underweight <18.52 (1.6)3 (2.5)
       Normal 18.5–24.971 (60.7)69 (57.9)
       Overweight 25–3031 (26.2)32 (26.8)
       Obese >3014 (11.86)15 (12.6)
       Missing data01 (0.0)
      Model of care
       Hospital antenatal clinic39 (33.0)34 (28.5)
       Midwives clinic31 (26.2)32 (26.8)
       Team midwifery13 (11.0)11 (9.2)
       GP shared care8 (6.7)6 (5.4)
       Midwifery Group Practice22 (18.6)31 (26.0)
       Private obstetrician5 (4.2)5 (4.2)
       Missing data1 (0.0)0
      Onset of labour
       Spontaneous72 (60.5)61 (51.6)
       Induction of labour47 (39.0)57 (48.3)
       Missing data01 (0.0)
      Fetal position at randomisation
       Occipito-posterior51 (42.8)49 (41.1)
       Occipito- lateral31 (26.0)33 (27.7)
       Occipito-anterior21 (17.6)14 (11.7)
       Not determined8 (6.7)7 (5.8)
       Not recorded8 (6.7)16 (13.4)
      Cervical dilation at randomisation
       0–221 (17.6)13 (10.9)
       3–560 (50.4)68 (57.1)
       6–1020 (16.8)21 (17.6)
       Not recorded18 (15.1)17 (14.2)
      VAS of back pain prior to injections81.2 (13.1)82.3 (13.4)
      Mean (Standard Deviation)Range 1–100Range 1–100
      GP = General Practitioner.
      For the primary outcome the reduction of the mean VAS scores between pre-and 30 min post-injection was −5.97 mm with a 95% CI of −13.18 to 1.22 mm. The lower end of the 95% CI is outside the margin of ±10 mm therefore equivalence was not demonstrated (Fig. 3). While the 95% CI crosses 1.0 the margin of equivalence is independent of statistical significance, which is a measure of superiority, i.e. that one technique is significantly better in effect to the other [
      • Walker E.
      • Nowacki A.S.
      Understanding equivalence and noninferiority testing.
      ]. This result did not differ significantly after a per protocol analysis (mean difference −4.77 mm 95% CI 12.89–3.34). The difference in the proportion of women experiencing at least a 30% reduction in pain was not significant between groups at 30 and 60 min. post injection. This outcome did significantly favour the four injection group at 90 min (p = 0.011), though the difference was again not significantly different at the final timepoint of 120 min. There was no difference in the reported VAS of the injection pain between groups. Sixteen participants (6.7%) requested repeat injections. The mean difference between pre and post injection VAS reported at 30 min was 46.8 mm. This did not differ significantly from the mean difference for the initial administration of the four injection technique of 45.01 mm (p = 0.414).
      Fig. 3
      Fig. 3Equivalence outcome of post injections of sterile water.
      (Note to go under ) To claim equivalence for the intervention group (two injections) at 30 min, the lower end of the 95% confidence interval (−13.18) needed to be within the shaded area of equivalence on the diagram (which equates to the mean VAS score of the control group (α) ±10 mm).
      There was no difference between the trial groups in use of analgesia or birth outcome. Duration of second stage labour was significantly shorter in the two injection group (p = 0.040) though given the small sample size in the context of this outcome the result should be viewed cautiously. There were no differences in neonatal outcomes between groups (Table 2). No adverse events were reported during the trial period.
      Table 2Secondary outcomes.
      VariableTwo injectionsFour injectionsEstimates (95% CI)P value
      N (%)
      ≥30% reduction in pain at:
      30 min post injection0.952 (0.828–1.094)0.492
       Yes84 (76.3)89 (80.18)
       No26 (23.6)22 (19.8)
       Missing98
      60 min post injection0.927 (0.764–1.1240.442
       Yes62 (65.9)69 (71.3)
       No32 (34.0)28 (28.8)
       Missing2522
      90 min post injection0.717 (0.556–0.925)0.011
       Yes42 (52.5)49 (73.1)
       No38 (47.5)18 (26.8)
       Missing3952
      120 min post injection0.912 (0.585–1.420)0.684
       Yes28 (45.9)29 (56.8)
       No33 (54.1)22 (43.1)
       Missing5868
      ≥50% reduction in pain at:
      30 min post injection0.863 (0.706–1.054)0.149
       Yes65 (59.0)76 (68.4)
       No45 (40.9)35 (31.5)
       Missing98
      60 min post injection0.950 (0.721–1.254)0.722
       Yes47 (50.0)51 (52.5)
       No47 (50.0)46 (47.4)
       Missing2522
      90 min post injection0.724 (0.513–1.021)0.066
       Yes32 (40.0)37 (55.2)
       No48 (60.0)30 (44.7)
       Missing3952
      120 min post injection0.912 (0.585–1.420)0.684
       Yes24 (39.4)22 (43.1)
       No37 (60.6)29 (56.8)
       Missing5868
      VAS of injection pain(−6.364 to 11.972)0.547
      Mean (Standard Deviation)56.5 (32.8)59.3 (37.7)
      Reduction in pain following repeat injection (not randomised)
      <30%2 (12.5)
      ≥30%5 (31.2)
      ≥50%9 (56.9)
      Water immersion use0.875 (0.516–1.483)0.620
       Yes21 (17.6)24 (20.1)
       No98 (82.3)95 (79.8)
      Nitrous oxide inhalation0.916 (0.710–1.183)0.508
       Yes69 (57.9)74 (62.1)
       No50 (42.1)45 (37.8)
      IV/IM opioids0.785 (0.537–1.148)0.181
       Yes18 (15.1)26 (21.8)
       No101 (84.8)93 (78.1)
      Epidural1.060 (0.809–1.388)0.673
       Yes38 (31.8)35 (29.4)
       No81 (68.1)84 (70.5)
      Duration of 1st stage labour
      Chi square test.
      0.725
      (minutes) Median (25th–75th)319 (210–465)330 (170–540)
      Duration of 2nd stage labour
      Chi square test.
      0.040
      (minutes) Median (25th–75th)35 (12–88)59 (17–155)
      Birth outcome
       Spontaneous Vaginal69 (58.4)77 (64.7)0.877 (0.679–1.133)0.324
       Instrumental28 (23.7)25 (21.0)1.080 (0.804–1.449)0.615
       Caesarean Section21 (17.8)17 (14.2)1.133 (0.823–1.506)0.461
      Apgar ≥7 at 5 min
       Yes109 (93.1)113 (96.5)0.964 (0.908–1.02)
      Fisher’s exact test.
      0.237
       No8 (6.4)4 (3.4)
       Missing22
      Resuscitation required atbirth1.005 (0.704–1.433)0.977
      Yes18 (15.3)18 (15.2)
      No99 (84.6)100 (84.7)
      Missing21
      Admission to ICN/SCN1.27 (0.930–1.740)0.168
      Yes20 (18.2)13 (11.6)
      No90 (81.8)99 (88.3)
      Missing97
      Reference group for birth outcomes is inclusive of other options.
      IV = Intravenous, IM = Intramuscular, ICN = Intensive Care Nursery, SCN = Special Care Nursery.
      a Fisher’s exact test.
      b Chi square test.
      The postnatal survey was completed by 78 (65.5%) participants from the two injection group and 74 (62.1%) in the four injection arm. The was no difference between groups in response to any of the questions related to satisfaction with the treatment, perceived effectiveness, likelihood to choose the same treatment again or recommend to other women (Table 3). In response to the question “what was the worst aspect of your treatment, over 90% of participants in both groups referred to the pain of the injections. The speed of analgesic effect and the relief from back pain were the most commonly reported “best aspects”.
      Table 3Postnatal survey outcomes.
      VariableTwo injections (n = 78)Four injections (n = 74)P value
      n (%)n (%)
      Effectiveness0.371
       Very effective/effective46 (82.0)64 (86.4)
       Not effective/not very effective,14 (17.9)9 (12.6)
       No response0 (0.0)1 (1.3)
      Satisfied with allocated treatment0.644
       Satisfied/very satisfied68 (87.1)61 (82.4)
       Dissatisfied/very dissatisfied10 (12.8)12 (16.2)
       No response0 (0.0)1 (1.3)
      Choose same treatment again0.425
       Yes56 (71.7)50 (76.5)
       No21 (26.9)22 (29.7)
       No response1 (1.28)2 (2.70)
      Recommend to other women0.555
       Yes68 (87.1)62 (83.7)
       No8 (10.2)11(14.8)
       No response2 (2.56)1 (1.3)

      4. Discussion

      In this randomised trial comparing the use of two sterile water injections to four we did not find that the two injection technique was equivalent to four at the primary outcome time point of 30 min post injection. Neither did the two injections approach achieve equivalence at any of the other time points of 60, 90 and 120 min. The likelihood of achieving an at least 30% or 50% reduction in pain was only statistically significant, in favour of the four injections, at the 90 min time point for the 30% reduction. From a clinical application perspective, the outcomes of our study could then be interpreted as both techniques providing effective pain relief. However the four injections approach has a margin of benefit in terms of degree and duration of analgesia. This is similar to the findings of our previous trial comparing a single to four injections that used a non-inferiority design, similar in concept to equivalence. Though in that trial the single injection was both non-inferior and statistically lesser in analgesic efficacy to four injections [
      • Lee N.
      • Webster J.
      • Beckmann M.
      • Gibbons K.
      • Smith T.
      • Stapleton H.
      • et al.
      Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
      ]. Nevertheless, the majority of women who received the single injection, as did the women in this trial who received two injections, still reported a significant reduction in back pain. A previous superiority RCT also reported significant relief of labour related back pain from two sterile water injections [
      • Fouly H.
      • Herdan R.
      • Habib D.
      • Yeh C.
      Effectiveness of injecting lower dose subcutaneous sterile water versus saline to relief labor back pain: randomized controlled trial.
      ]. This means that, while four injections usually administered by two midwives, appears to be the ideal technique for achieving the best analgesic response, two injections can be used in situations, for example, where only one midwife is present, and still provide reliable analgesia.
      Another important point of no difference between the two trial groups was the pain from the injections reported by the participants. Having only two injections did not result in less injection pain. The mean VAS scores for injection pain were lower than reported in previous trials. The question regarding the pain of the injections was usually put to the participants a couple of minutes after the injections to account for repositioning, impending contractions etc. An issue noted by some midwives with the time delay between injections and VAS assessment was that the mark on the relevant VAS scale made by the participant likely related to any residual injection pain occurring at the time of the assessment, rather than a recall of the pain at the actual time of the injections. Possible resulting in a lower VAS score. Similar issues have been noted in previous trials [
      • Martensson L.
      • Nyberg K.
      • Wallin G.
      Subcutaneous versus intracutaneous injections of sterile water for labour analgesia: a comparison of perceived pain during administration.
      ]. However the distribution of injection VAS scores was similar between groups, so if this was the case then the proportion of no difference may well remain representative.
      A unique feature of water injections as an analgesic is the ability to repeat the injections as often as required. Sixteen participants in the trial opted for a repeat injection. The difference in pre and post injection VAS suggests that repeat injections are effective in re-establishing the analgesia. The low number of women requesting repeat injections may be a factor of the changing nature and origin of pain as labour progresses. In this study, as in previous trials, back pain appears to be more common in the period of transition from latent to active phase labour (cervical dilation of 3–5 cm) [
      • Lee N.
      • Gao Y.
      • Collins S.L.
      • Mårtensson L.B.
      • Randall W.
      • Rowe T.-M.
      • et al.
      Caesarean delivery rates and analgesia effectiveness following injections of sterile water for back pain in labour: a multicentre, randomised placebo controlled trial.
      ,
      • Lee N.
      • Webster J.
      • Beckmann M.
      • Gibbons K.
      • Smith T.
      • Stapleton H.
      • et al.
      Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
      ]. However, it is also likely that the injection pain associated with the procedure acts as a deterrent to repeated use [
      • Lee N.
      • Gao Y.
      • Collins S.L.
      • Mårtensson L.B.
      • Randall W.
      • Rowe T.-M.
      • et al.
      Caesarean delivery rates and analgesia effectiveness following injections of sterile water for back pain in labour: a multicentre, randomised placebo controlled trial.
      ]. In this study the comments regarding the worst aspect of the procedure were dominated by remarks about the severity of the injection pain. A previous case study details the use of water injections by a laboring woman for back and abdominal contraction pain. The woman received several treatments throughout her labour totaling approximately 30 injections [
      • Mårtensson L.
      The patient observer: sterile water injections for labor pain.
      ]. The woman reported that the pain of the injections was ‘insignificant’ in comparison to the resulting pain relief. A theme that was also evident in our qualitative work with women’s experiences of using water injections during labour [
      • Lee N.
      • Kildea S.
      • Stapleton H.
      “No pain, no gain”: the experience of women using sterile water injections.
      ]. Undoubtedly finding a means to effectively reduce the pain of water injections could contribute to facilitating repeated use. Reducing the injection pain may, in turn, affect the depth and duration of pain relief, though the increased acceptability and rate of repeated injections would be an acceptable trade-off. A number of studies are currently exploring the use of topical anaesthetic creams and vapocoolant sprays as a means of mitigating the injection pain.
      This trial had a number of limitations. Notably the full sample size was not achieved. This was due to a number of factors including limited resources to support the trial and recruitment strategies and reliance on midwives at sites to undertake trial activities alongside usual clinical work. The Queensland site was established in 2019 to assist with recruitment, however the advent of the COVID-19 pandemic and a directive from the governing ethics committee to suspend research recruitment prompted the decision to cease enrolling participants to the trial. The reduced sample impacts upon the study power (70%) and as such the primary outcome should be viewed with some caution. Some VAS data were incomplete, mostly due to women progressing to birth or insertion of an epidural. The repeat injections were not randomised and the actual technique used was not recorded. Therefore results consisted of both the two and four injection technique. Our study only reports on the use of water injections for back pain in labour and is not generalizable to other types of labour pain. Well designed trials exploring the use of water injections for abdominal labour pain are needed.

      5. Conclusion

      In this trial we were unable to demonstrate that the two water injections was equivalent to the usual technique of four injections. Our results do suggest that both techniques provide effective relief for back pain in labour with depth and duration of analgesia favouring four injections. The choice between techniques, therefore becomes a matter of clinical applicability, birthing environment and joint decision making between the midwife and laboring woman. As there did not appear to be any difference in injection pain between the two techniques, where two midwives are available to provide the injections simultaneously, four injections remains the technique of choice. Our findings also lend support to the ability of repeat injections to re-establish the analgesia though research into methods to mitigate the injection pain are needed to increase the acceptability and repeatability of water injections.

      Author contributions

      NL. BL YHW and SK designed the study. YHW and BL supervised the recruitment at their respective sites. NL managed the trial, undertook the data analysis and wrote the draft manuscript. SK, YG, LM, BL and YHW reviewed and contributed to the editing of the manuscript.

      Ethical statement

      The study protocol was approved by Northern Sydney Local Health District Human Research Ethics Committee (HREC/14/HAWKE/48) 10th March 2014 for the New South Wales sites and the Mater Misericordiae Ltd Human Research Ethics Committee (HREC/18/MHS/109) 24th July 2018 for the Queensland site.

      Funding

      None declared.

      Conflict of interest

      None declared.

      Acknowledgements

      The authors gratefully acknowledge Alisha Heys, Carina Romero, Catherine Cooper and Dr Bec Jenkinson for their invaluable assistance with undertaking the study and Dr Julie Flynn for assistance with data entry.

      References

        • Lee N.
        • Gao Y.
        • Collins S.L.
        • Mårtensson L.B.
        • Randall W.
        • Rowe T.-M.
        • et al.
        Caesarean delivery rates and analgesia effectiveness following injections of sterile water for back pain in labour: a multicentre, randomised placebo controlled trial.
        EClinicalMedicine. 2020; 25100447
        • Lee N.
        • Kildea S.
        • Stapleton H.
        “No pain, no gain”: the experience of women using sterile water injections.
        Women Birth. 2017; 30: 153-158
        • Mårtensson L.B.
        • Hutton E.K.
        • Lee N.
        • Kildea S.
        • Gao Y.
        • Bergh I.
        Sterile water injections for childbirth pain: an evidenced based guide to practice.
        Women Birth. 2018; 31: 380-385
        • Lee N.
        • Webster J.
        • Beckmann M.
        • Gibbons K.
        • Smith T.
        • Stapleton H.
        • et al.
        Comparison of a single vs. a four intradermal sterile water injection for relief of lower back pain for women in labour: a randomised controlled trial.
        Midwifery. 2013; 29: 585-591
        • Melzack R.
        A new theory. A gate control system modulates sensory input from the skin before it evokes pain perception and response.
        Science. 1965; 150: 971-979
        • Le Bars D.
        • Dickenson A.H.
        • Besson J.-M.
        Diffuse noxious inhibitory controls (DNIC). II. Lack of effect on non-convergent neurones, supraspinal involvement and theoretical implications.
        Pain. 1979; 6: 305-327
        • Martensson L.
        • Wallin G.
        Labour pain treated with cutaneous injections of sterile water: a randomised controlled trial.
        Br. J. Obstet. Gynaecol. 1999; 106: 633-637
        • Martensson L.
        • Nyberg K.
        • Wallin G.
        Subcutaneous versus intracutaneous injections of sterile water for labour analgesia: a comparison of perceived pain during administration.
        BJOG. 2000; 107: 1248-1251
        • Bahasadri S.
        • Ahmadi-Abhari S.
        • Dehghani-Nik M.
        • Habibi G.R.
        Subcutaneous sterile water injection for labour pain: a randomised controlled trial.
        Aust. N. Z. J. Obstet. Gynaecol. 2006; 46: 102-106
        • Saxena K.
        • Nischal H.
        • Batra S.
        Intracutaneous injections of sterile water over the sacrum for labour analgesia.
        Indian J. Anaesth. 2009; 53: 169-173
        • Fouly H.
        • Herdan R.
        • Habib D.
        • Yeh C.
        Effectiveness of injecting lower dose subcutaneous sterile water versus saline to relief labor back pain: randomized controlled trial.
        Eur. J. Midwifery. 2018; 2
        • Walker E.
        • Nowacki A.S.
        Understanding equivalence and noninferiority testing.
        J. Gen. Intern. Med. 2011; 26: 192-196
        • Collins S.
        • Moore R.
        • McQuay H.
        The visual analogue pain intensity scale: what is moderate pain in millimetres.
        Pain. 1997; 72: 95-97
        • Dhumale H.
        • Goudar S.
        Ethical issues related to consent for intrapartum trials.
        Reprod. Health. 2017; 14: 31-33
        • Widmer M.
        • Bonet M.
        • Betrán A.P.
        Would you like to participate in this trial? The practice of informed consent in intrapartum research in the last 30 years.
        PLoS One. 2020; 15e0228063
        • Price D.
        • McGrath A.
        • Rafii A.
        • Buckingham B.
        The validation of visual analogue scales as ratio scale measures for chronic and experimental pain.
        Pain. 1983; 17: 45-56
        • Derry S.
        • Straube S.
        • Moore R.A.
        • Hancock H.
        • Collins S.L.
        Intracutaneous or subcutaneous sterile water injection compared with blinded controls for pain management in labour.
        Cochrane Database Syst. Rev. 2012; 1
        • Kelly A.-M.
        The minimum clinically significant difference in visual analogue scale pain score does not differ with severity of pain.
        Emerg. Med. J. 2001; 18: 205-207
        • Hutton E.K.
        • Kasperink M.
        • Rutten M.
        • Reitsma A.
        • Wainman B.
        Sterile water injection for labour pain: a systematic review and meta-analysis of randomised controlled trials.
        BJOG. 2009;
        • Moher D.
        • Hopewell S.
        • Schulz K.F.
        • Montori V.
        • Gøtzsche P.C.
        • Devereaux P.
        • et al.
        CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials.
        Int. J. Surg. 2012; 10: 28-55
        • Busse J.W.
        • Bartlett S.J.
        • Dougados M.
        • Johnston B.C.
        • Guyatt G.H.
        • Kirwan J.R.
        • et al.
        Optimal strategies for reporting pain in clinical trials and systematic reviews: recommendations from an OMERACT 12 workshop.
        J. Rheumatol. 2015; 42: 1962-1970
        • Mårtensson L.
        The patient observer: sterile water injections for labor pain.
        Birth. 2010; 37: 334-336