Current Resources for Evidence-Based Practice, January/February 2020.

Abstract

Does breast massage prevent engorgement and/or aid in the treatment of engorgement, breast pain, or mastitis? After perceived inadequate milk supply, nipple or breast pain from engorgement or plugged ducts is the second most common reason for early breastfeeding cessation.1 Early breastfeeding cessation has significant consequences on both infants’ and women's health. Suboptimal breastfeeding rate and duration in the United States are estimated to cause 3340 excess maternal and infant deaths and cost $3.9 billion a year; 79% of this cost is attributed to potentially avoidable maternal morbidity.2 Breast massage is relatively easy for health care providers to learn and teach to their patients. The authors of this systematic review of the effectiveness of breast massage for breastfeeding problem used a search strategy that included both English and Japanese language and identified only 9 studies meeting their criteria, of which one-third was excluded for significant study design flaws.3 The included 6 randomized or quasiexperimental trials included 391 women with study sizes ranging from 36 to 152. Because of the variety of massage techniques used and different comparison control conditions, it was not possible to combine results. The massage protocols included 1) Gua Sha, a Chinese scraping technique with a soft instrument along certain medians of the breast (one study); 2) Oketani massage, popular in Japan and South Korea, which stimulates the pectoral muscles and is combined with a rolling massage of the breast (one study); 3) therapeutic breast massage, described by Bolan and colleagues,4 which was learned from several trips to Russia (one study); and 4) other kinds of manual massage (3 studies). The study using Gua Sha randomized 54 women who had uncomplicated births and were diagnosed with engorgement.5 The intervention group received 7 light unidirectional strokes starting at 1 of 3 meridians from the base of the breast toward the nipple for 2 cycles, which took about 2 minutes. The control group had warm compresses applied to the breasts followed by manual massage taking about 20 minutes. Validated scales measured subjective engorgement, pain, and discomfort before interventions and at 5 and 30 minutes after interventions. By 30 minutes, both groups had improvement in all 3 measures. At 5 minutes, the Gua Sha treatment yielded significantly lower mean differences of at least 2 points on the scales. Oketani massage involves gentle massage of the pectoral muscles to increase circulation and drainage from the breast and separate any adhesions of breast connective tissue so the blood vessels can freely expand. In Japan, women may undergo this massage once in early lactation to prevent any breast problems, weekly if they need to increase their milk supply, or only when experiencing a breastfeeding problem like a plugged duct or engorgement.6 Oketani massage was employed in a randomized controlled trial (RCT) in Korea that enrolled 47 hospitalized postpartum women with diagnosed breast engorgement.7 The Oketani massage was done over 30 minutes, and the control massage method, described only as “routine massage,” was for the same length of time. It was unclear if women were truly randomized, although the 2 groups had similar baseline characteristics, except that the neonates in the Oketani group were 2 days older than those in the control group (12 vs 9.4 days). The Oketani massage group had a very large decrease in breast pain assessed on a visual analog scale from 7.10 (SD 1.92) to 1.95 compared with the control group from 7.16 (SD 1.58) to 6.14. There was also a statistically significant increase in breast milk pH and sucking speed, although this measure has not been validated and no measure of infant swallowing was assessed. Presumably, these last 2 measures were aimed at assessing a change in the quality and taste of the milk that Oketani claimed occurred following the use of the technique. There is limited evidence about the meaning of pH in breast milk, although it is documented that colostrum has a higher pH than milk at 2 weeks.8 Additionally, foremilk has less fat content with a lower pH than hindmilk with its higher fat content and higher pH.9 Therapeutic breast massage (TBM) can be taught to a breastfeeding woman for her to use as needed. It starts with gently softening the areola area by gently stroking from the areola toward the axilla, a type of reverse pressure softening.10 This technique can also be done for plugged ducts or mastitis. This study was as a nested case-control study with the cases being 42 women seen with either breast engorgement (36%) or plugged duct and/or mastitis (64%) and taught TBM.10 Women in the control group (n = 70) were seen by a lactation consultant during a pediatric examination and were excluded if they used TBM. Women in the TBM group were assessed before massage and after treatment. Their results after a 30-minute massage session showed significant improvement in pain and engorgement, and over 90% of women had their plugged duct resolve or reduced to less than 3 cm. All participants were asked to complete electronic surveys after treatment or regular visit at 2 days and at 2 and 12 weeks for breast engorgement, breast pain, and nipple pain, and at 12 weeks for exclusive breastfeeding and occurrence of breastfeeding problems. Although the text reports that 38 women in the TBM group completed a 2-day survey, the table comparing the case and control groups only shows the baseline characteristics and 2-day results for 15 women in the case group with 13 in the case group and 64 in the control group completing 12-week surveys. Given the fact that only 36% of case group participants were in the initial 2-day comparison, it is difficult to draw any conclusions from comparisons between the 2 groups because of a large potential bias between those in the TBM who responded and did not respond to follow-up surveys. Among the 3 studies with a variety of massage methods was an RCT from China among 200 women with plugged ducts who used either breast massage, aloe and cacti cold compresses for 15 minutes, or both interventions daily for 7 days.11 The massage technique started at the base of the breast and gently stroked toward the nipple followed by using the forearm to massage over the plugged duct in a circular manner and finally massaging the nipple to release milk. It was unclear if women returned daily to have treatments by the study investigator or were instructed to do the interventions at home, where presumably someone else would have had to be instructed in the technique. There was no information about adherence to the study protocol by participants. The one-week loss to follow-up rate was somewhat high at 24% (n = 48), but the women in the 3 groups with outcome measures were similar on baseline characteristics. Breast hardness and tenderness were measured using an unvalidated scale. Breast pain was measured with a visual analog scale and was categorized into 3 categories, instead of being examined as a continuous measure. After one week, the women using both massage and compresses had the best results with least pain, breast hardness, and a higher exclusive breastfeeding rate (72%) compared with the massage-only (20%) and compress-only (54%) groups. Performing massage only resulted in twice as many women having the most severe category of breast hardness and the most pain. In this study, the massage technique was most effective in conjunction with the aloe and cacti cool compresses. Another study used either massage, massage and cool cabbage leaves, or a control condition, which was not well specified, in women who were postcesarean and primiparous.12 Assignment of 20 women to each of the 3 groups was nonrandom, although the groups were similar on key characteristics. The massage technique was not described well enough to be replicable. By day 2 postpartum, the massage and cabbage leaves group had the lowest scores on breast hardness. On days 3 or 4 postpartum, both intervention groups had lower breast pain and hardness, with the combined massage and cabbage leaves group having the best outcomes. Lastly, among the other massage types was a small US study among 36 women who were pumping because their newborns were in the neonatal intensive care unit.13 This study's primary objective was to assess sequential versus simultaneous breast pumping, and additionally each woman was randomly assigned either to massage the breast prior to pumping or not. The authors report stratified randomization on parity and gestational age, but they do not report any inclusion or exclusion criteria or the gestational age and parity of the groups, so external validity of this study is unclear. The outcome measures were milk volume and fat content. There was no difference in fat content by pumping method or massage. However, simultaneous pumping was clearly superior for producing the most volume, and massage had an additive effect on average of 9 g for the sequential group and 37 g for the simultaneous one. These results were assessed on a single pumping session for each condition. What kind of massage, what frequency, and does the technique differ for prevention or the type of presenting breast problem? These are important questions. Unfortunately, this systematic review cannot really provide strong evidence to answer them. Overall, some kind of massage seems to improve engorgement and breast pain and plugged ducts; however, it is not clear if type of massage matters. The massage techniques were quite varied with some stroking from the base of the breast toward the areola and others stroking in the reverse direction. Additionally, it seems that in East Asia (China, Korea, and Japan), postpartum nurses routinely provide some kind of breast massage daily to all hospitalized postpartum women. This area is one requiring better research and an improved understanding of the physiologic functioning of fluid dynamics both in the blood vessels and the lymphatics of the breast, which would help in understanding the rationale for different kinds of techniques. Breastfeeding researchers are urged to identify reliable and valid methods to assess outcomes and to agree on important parameters to be measured. These efforts would make it possible to pool the results from smaller studies that tend to be done by nurses and midwives who have the interest but not the funding for large trials to compare among interventions. Although two-thirds of abortions are done at 8 weeks’ gestation or earlier, only an estimated 39% of women obtain a medical abortion.14 The cost from time off work or the provision of childcare for follow-up visits for medical abortion can be a barrier to women who perceive scheduling a one-time vacuum aspiration to be more feasible. The privacy of the ability to self-administer and self-assess outcomes as well as one-third the cost of vacuum aspiration may make medical abortion more readily acceptable to women. Therefore, this meta-analysis of the safety of self-assessment may encourage health care providers to consider this approach to medical abortion. The review's objective was to evaluate effectiveness and safety of self-assessment of the outcomes from self-administered medical abortion at home to routine clinic follow-up of outcomes. Self-assessment by the women at home was performed with a urine pregnancy test, which was low sensitivity, semiquantitative, or high sensitivity. These were combined with a pictorial instruction sheet and sometimes included a symptom checklist. The measured primary outcome was a successful complete abortion without the need for surgical or repeat medical intervention within 3 months of the abortion, and secondary aims included side effects, number of and types of contact, and satisfaction. The review identified 4 RCTs that took place in Western Europe,15 Central Asia,16 India,17, 18 and Vietnam19 and included 5493 women. All studies were rated as high quality on the majority of study characteristics with the exception of unclear risk for blinding of participants and assessors. The overall loss to follow-up among all the studies was low, 3% to 6%. On the primary outcome of effectiveness of the medical abortion, there was no difference between groups (relative risk [RR], 1.00; 95% CI, 0.99-1.01) with no heterogeneity among studies. Therefore, this was given a rating of high certainty around the result. Similarly, the number of ongoing pregnancies was a homogenous measure and showed no difference between groups (RR, 0.90; 95% CI, 0.50-1.62). Safety of self-assessment was measured on 4 aspects. Need for surgery showed no difference between approaches with moderate certainty of the estimate of RR of 0.92 (95% CI, 0.61-1.68). There was more heterogeneity of study results on the occurrence of excessive bleeding with an RR of 1.48, but the 95% CI (0.94-2.60) crossed 1, indicating no statistically significant difference. Only 2 studies measured occurrence of fever and infection and need for a drug to treat hemorrhage, with the latter having a low number of events (n = 14). Therefore, the CIs were very wide, but the results showed no statistically significant difference between groups. The certainty around these estimates was rated as low for excessive bleeding and infection and moderate for need for hemorrhage drugs. Other outcomes measured only in single studies, such as need for hospitalizations or additional contacts (eg, phone call or clinic visits), showed no differences between groups. Acceptability was measured as preference for type of follow-up for a future medical abortion. Among those assigned to self-assessment groups with phone or text follow-up, 76% to 88% endorsed this method among the 4 studies. For those assigned to clinic follow-up, when asked to choose between phone contact or clinic follow-up, more women in India (70% vs 30%)18 and Vietnam (60% vs 40%)19 chose clinic follow-up; this was not the case in the other 2 studies. This meta-analysis provides good evidence that self-assessment of self-administered medical abortion by women is equivalent to a scheduled clinic follow-up. In the United States, where the presence of ovulation and pregnancy self-assessment kits are ubiquitous in drug and grocery stores, there should be low concern about the ability of women to use and correctly interpret results from self-assessment point of care tests. There is good evidence of the ability of nurse-midwives to provide early abortion-related services to women, both nationally20 and internationally.21 This ability of nurse-midwives to provide both medical abortion and vacuum aspiration is endorsed by both the American College of Nurse-Midwives22 and the American College of Obstetricians and Gynecologists.23, 24 With only a minority of states having no restriction on insurance coverage for abortion services, a cost-effective and acceptable approach to abortion services is highly needed.25 Hopefully these results will encourage midwives to offer this lower-cost alternative approach to women, especially in those geographical areas where abortion care services are severely limited. The evidence is strong that harassment and discrimination in multiple spheres are pervasive in the lives of transgender people.26 As a result, this population suffers from a disproportion of mental health issues, such as depression and anxiety, and more importantly, more than 40% have attempted suicide.27 Because psychological well-being is a key factor to overall health, the presence of adverse mental health has a significant effect on the overall health of transgender people. The prevalence of transgender persons in the United States is not well documented. A recent survey of young adults estimates the prevalence at 0.5% or 5 in 1000,28 although epidemiologic studies with a wider range of ages estimates it at around 5 in 100,000.29 Not all of these individuals will seek gender-affirming treatment, but many will. The use of hormonal treatment to assist an individual to transition to their desired gender can improve their quality of life. This systematic review was aimed at summarizing the effect of transgender treatment on quality of life, depression, and anxiety. The authors specifically wanted to find information useful for primary care providers. Unlike many reviews that are limited to experimental studies, studying this question with an experimental design would be difficult to carry out ethically. Therefore, the authors cast their search strategy quite broadly to include not only experimental studies but also cohort and case-control studies, cross-sectional studies, and case series. The inclusion criteria were studies of individuals undergoing hormonal therapy under medical supervision who had not experienced gender-affirming surgery, and the study measured at least 1 of the 3 study outcomes using validated instruments. Excluded studies included those that used the internet for recruitment in which clinical transgender diagnosis or the proper use of hormones under medical supervision could not be verified. The authors identified 7 studies: 1 case-control study, 2 cross-sectional studies, and 4 case series. Three studies were rated as high quality and 4 as moderate quality. These studies included 552 people, but most were quite small with single-study sample sizes ranging from 14 to 163. An important issue limiting synthesis of the results was the varying duration of hormone use, 3 to 12 months, with one study not specifying the duration. Additionally, some studies provided results for trans men and trans women together but not separately, and vice versa. The 3 quality of life studies all used different validated instruments, so their results could not be combined. All 3 studies showed significant improvements in quality of life, including better emotional functioning among those taking hormones compared with those not taking hormones. A study of 87 trans women and trans men taking hormones from the start of therapy to one year later showed improvements over their baseline scores in quality of life, sexual quality of life, body image, and social relationships.30 However, trans men reached statistical significance only on body image and social relationships, which may have been due to the study having only 27 trans men. One study found that transgender persons not taking hormones had significantly poorer scores compared with the general population in all spheres except for general health.31 However, those taking hormones had better than the population scores on mental health functioning and general health. Five studies examined depression, and 2 studied anxiety. For both outcomes, the studies uniformly found improvements in those taking hormones between 3 months and 2.5 years compared with those not taking hormones. The authors agree that the evidence showed that those transgender persons taking hormones had improvements in quality of life, depression, and anxiety, but because the quality of the evidence was low because of the observational type of study designs, the use of hormones to improve these aspects of a trans person's life could not be endorsed. A limitation noted by the authors was the exclusion of nonbinary or genderqueer people. These individuals are a significant proportion of the population of the transgender population,27 and there is no knowledge if there is a desire to seek any kind of hormone treatment. A recent systematic review of nonbinary and genderqueer people identified one intervention study in which individuals who underwent chest surgery improved their quality of life and comfort with their physical appearance, exercise, and sex life.32 Nurse-midwives may be caring for transgender people in their practices, and with the appropriate education, the American College of Nurse-Midwives endorses transgender people's care and hormone treatment as part of the midwifery scope of practice.33 The Journal of Midwifery and Women's Health has had several excellent recent review articles on the care of transgender persons with education and practice guideline resources clearly identified.34-36 Many transgender individuals may be seeking hormone treatment, and the midwifery model of care of respectful practice and shared decision making makes midwives poised to be important health care providers to this population.