Article Text
Abstract
Objective This study aims: (a) to evaluate patterns of domperidone dispensing to mothers of very preterm (<32 weeks gestation) infants born before and after 2014 when international recommendations were made to limit its use and (b) to examine characteristics associated with domperidone dispensing and impacts on breast milk feeding rates at infant hospital discharge.
Design Retrospective audit using linked electronic medical records and hospital pharmacy records.
Setting Tertiary-referral neonatal intensive care unit at the Women’s and Children’s Hospital in South Australia.
Patients Mothers of preterm infants admitted to neonatal intensive care from January 2004 to December 2018.
Main outcome measures Rate of domperidone dispensing compared pre-2014 and post-2014 recommendations using interrupted time series analyses, and breast milk feeding rates at infant discharge based on domperidone treatment status, adjusted for other factors known to influence breast milk production.
Results Overall, domperidone was dispensed to 691 (41%) of 1688 mothers. Prior to 2014 recommendations, the proportion of women dispensed domperidone was stable. Following the recommendations, there was a significant reduction in trend (−2.55% per half year, 95% CI −4.57% to –0.53%;), reflecting less domperidone dispensing.
Breast milk feeding rates at discharge remained consistently lower in infants of women dispensed domperidone than those who were not (adjusted OR 0.58, 95% CI 0.45 to 0.75).
Conclusion Domperidone dispensing in mothers of hospitalised very preterm infants has declined over time following international regulatory warnings. Breast milk feeding rates remain lower in mothers prescribed domperidone, suggesting further research is needed to optimise lactation support for mothers of very preterm infants.
- Neonatology
- Pharmacology
Data availability statement
Data may be obtained from a third party and are not publicly available. The data for this paper is available only on request to the ethics committee as it contains identifiable details of participants and their children.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Domperidone is commonly used to treat lactation insufficiency following preterm birth, based on evidence of short-term improvements in expressed breast milk volume. In 2014, regulatory warnings were introduced due to potential cardiac side effects of domperidone.
WHAT THIS STUDY ADDS
Domperidone dispensing has decreased significantly from 2014 onwards in mothers of very preterm infants. Despite this reduction, continued breast milk feeding rates at infant discharge remain stable and substantially lower over time, in infants of mothers dispensed domperidone than in those whose mothers were not.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Further research is required to optimise the use of domperidone and other lactation support strategies to improve longer-term breast milk feeding outcomes in mothers of very preterm infants.
Introduction
Breast milk is widely recognised as the best source of nutrition for all infants, particularly for preterm infants.1 However, lactation insufficiency, when a mother cannot produce enough breast milk to meet her infant’s requirements, is frequent after preterm birth and is one of the most common reasons women discontinue breast feeding.2 3 Despite recognition that mothers of preterm infants face additional challenges with breastfeeding,4 lactation support for this high-risk group is often suboptimal.5 When lactation insufficiency persists following non-pharmacological supports, medications such as domperidone are often used to increase breast milk production.6
Domperidone increases serum prolactin, a hormone required for breast milk production.7 8 Meta-analyses of trials involving mothers of preterm infants with diagnosed lactation insufficiency demonstrate that domperidone is effective for increasing short-term breast milk production by up to 88 mL/day at doses of 30 mg/day for 5–14 days.9 In contrast, limited evidence supports using domperidone as a preventative therapy immediately or soon after birth.9 Consequently, use is not typically recommended until after milk letdown has occurred and a trial of non-pharmacological strategies has been completed.2 10
Whether identified short-term benefits of domperidone translate to improved longer-term breast milk feeding outcomes for mothers of preterm infants remains inconclusive. Previous studies report continued breast milk feeding rates at infant discharge from hospital ranging from 58% to 95% among women prescribed domperidone.6 11–15 Clinical trial evidence is inconsistent, with reports of higher breastfeeding rates at discharge or no difference, based on domperidone use.11–13 These studies also have significant limitations, including small sample sizes and differences in the treatment regimens. Consequently, it remains difficult to draw reliable conclusions about the longer-term impacts of domperidone use for lactation insufficiency.
Despite uncertainties surrounding longer-term breastfeeding outcomes, the use of domperidone among mothers of preterm infants has increased in high-income countries such as the UK and Canada, with reports suggesting up to 44% of all mothers of preterm infants are prescribed domperidone.16–19 Concurrently, safety concerns have been raised about domperidone. In 2014, the European Medicines Agency (EMA) reviewed domperidone’s safety and efficacy data and recommended changes in practice due to a small risk of serious cardiac side effects.20
There has been limited exploration of the impact of these revised recommendations on prescribing domperidone among mothers of preterm infants.21 One study undertaken in the UK reported that the publication of the EMA recommendations corresponded to an immediate 15% reduction in domperidone prescribing rates by general practitioners. However, across the subsequent year (2014–2015), the prescribing rate returned to its previous value, suggesting no overall change to practice.19 Additionally, within this study, the publication of EMA guidelines had no impact on domperidone use in mothers of preterm infants, with rates increasing from 29% in 2002–2004 to 44% in 2014–2015.19 In Australia, many neonatal units altered their clinical guidelines to restrict the use of domperidone in line with EMA recommendations.22 Whether this translated to practice change has yet to be investigated.
We aimed to evaluate longitudinal patterns in domperidone use in mothers of very preterm infants (<32 weeks gestation), specifically focusing on changes in use following the 2014 EMA recommendations. Further, we examined characteristics associated with domperidone dispensing and timing and whether dispensing impacts breast milk feeding rates at hospital discharge.
Methods
Setting
The Women’s and Children’s Hospital (WCH) is an Australian tertiary metropolitan hospital and South Australia’s largest maternity, obstetric and neonatal care provider. During the study period, the neonatal unit had no dedicated lactation consultant position, with lactation support instead provided by midwifery staff of the general postnatal wards. Before 2014, the WCH had a standard domperidone prescribing checklist (online supplemental appendix 1), with 30 mg/day of domperidone prescribed for 5 days, followed by a 7-day tapering course (20 mg/day for 3 days, then 10 mg/day for 4 days). Prescriptions were provided by medical officers, following the checklist to ensure that patients had received adequate lactation support prior, were not taking medications or had known cardiac conditions contraindicating domperidone use, and that advice on potential side effects had been provided. From 2014 onwards, the prescribing checklist was amended to recommend a shorter 7-day treatment course at a maximum dose of 30 mg/day (online supplemental appendix 2).
Supplemental material
Supplemental material
Study population and data sources
The inclusion criteria for this study were mothers who initiated breastfeeding an infant (or infants) born at less than 32 completed weeks gestation at the WCH and admitted to the neonatal intensive care unit between January 2004 and December 2018.
Using the maternal unique record number, data from the WCH Perinatal Statistics Collection, the WCH Pharmacy Dispensing Database and the Australian New Zealand Neonatal Network (ANZNN) dataset were linked. Infant data from the WCH Perinatal Statistics Collection and the ANZNN dataset are collected for every baby admitted to the neonatal unit under 32 weeks gestation. Maternal and infant data collected from the Perinatal Statistics Collection included maternal age, body mass index (BMI), socioeconomic status, ethnicity, pregnancy information (public or private care, use of assisted reproductive techniques, parity, plurality), smoking status during pregnancy, any health conditions such as diabetes or hypertension, delivery method and gestation at delivery, and complications such as postpartum haemorrhage. Data from the hospital pharmacy dispensing records included the date of domperidone dispensing. Gestational age groups were stratified into 22+0–27+6, 28+0–29+6 and 30+0–31+6. We calculated the number of days post partum at first domperidone dispensing from the infant’s date of birth. The time to first dispensing was grouped into three postnatal epochs (days 0–7, days 8–28 or greater than 28 days).
Data on whether the mother was still providing breast milk for the infant at hospital discharge were collected from the ANZNN dataset.23 For assessment of breast milk feeding at infant discharge, data for mothers of infants that did not survive until discharge or were first dispensed domperidone within 14 days of infant discharge were excluded.
Data analysis
Differences in categorical variables between groups were compared using Pearson’s χ2 test or Fisher’s exact test if less than five cases. Domperidone dispensing over time was analysed using descriptive statistics. Longitudinal plots were produced with smoothing for a 3-year moving average. We performed an interrupted time series (ITS) analysis with a Newey-West regression to assess changes in dispensing rates after the 2014 EMA recommendations against domperidone using 6 monthly periods from 2004 to 2018, where the interruption occurred from the second half of 2014. The reported coefficients from the ITS analysis were the level change, representing rate of change in frequency of respective outcomes within the first 6 months from 1 July 2014 to 31 December 2014, and change in trend calculated as the difference between the pretrend slope (from 1 January 2004 to 30 June 2014) and post-trend slope (from 1 January 2015 to 31 December 2018) representing difference in rate of change in frequency of respective outcomes per half year, with corresponding 95% CIs.
Breast milk feeding at infant discharge was assessed according to those who were dispensed domperidone or not and for gestational subgroups within those dispensed domperidone. A logistic regression model was fitted for breast milk feeding at discharge according to domperidone dispensing, number of repeat dispensings and timing of dispensing. Each model included adjustment for gestational age at birth, Socio-Economic Indexes for Areas quintile, diabetes, eclampsia, parity, plurality, assisted reproduction, hospital classification, method of delivery, mother’s age, ethnicity, year of delivery, smoking status and BMI in model 2. Missing data on confounders (maternal BMI and smoking status) were addressed with multiple imputations by chained equation method under a missing-at-random assumption. The imputation model contained outcome, exposure and model covariates, with 50 datasets generated. The logistic regression coefficients and standard errors were pooled using Rubin’s Rules.24 ORs with 95% CIs were calculated for each model. Statistically significant differences were defined as p<0.05. Data were analysed using Stata 16 (StataCorp, Texas).
Results
Domperidone prevalence
From 2004 to 2018, 1688 mothers met the eligibility criteria. Of these, 41% (n=691) were dispensed domperidone. Maternal age, BMI and decreased gestational age at birth were significantly associated with an increased likelihood of being dispensed domperidone (all p<0.05; table 1). Prior to 2014, the proportion of women dispensed domperidone was stable. Following the recommendations, there was a non-significant change in level (3.10%; 95% CI −11.10 to 17.29), with a significant change in trend (−2.55% per half year, 95% CI −4.57% to –0.53%; figure 1). Further, similar reductions in dispensing over time were evident across all gestational age groups (online supplemental figure 1).
Supplemental material
Dispensing timing
Of the 691 women dispensed domperidone, 15% (n=107) received their first dispensing in the first 7 days after delivery, while 51% (n=350) received it from postnatal days 8–28 (table 2). Among those receiving domperidone, factors associated with an increased likelihood of domperidone dispensing within the first 7 days post partum included older maternal age, higher socioeconomic status, private patient status and higher gestational age (table 2). The ITS analysis showed a non-significant immediate reduction in the proportion of women first dispensed domperidone within the first 7 days post partum compared with more than 7 days post partum following the 2014 recommendation (level change −4.01%; 95% CI −13.68 to 5.66), with a non-significant change in trend (−0.36% per half year, 95% CI −1.70% to 0.98%; online supplemental figure 2).
Supplemental material
Domperidone courses
Fifty-two per cent (n=362) of women received two or more dispensing of domperidone during the infant’s hospitalisation. Factors associated with an increased likelihood of receiving two or more courses of domperidone included primiparity, gestational age <28 weeks, use of assisted reproductive techniques and higher socioeconomic status (online supplemental table 1). The ITS analysis showed no significant changes over time with respect to the proportion of women who received one compared with two or more treatment courses for domperidone. Following the 2014 recommendations, there was a non-significant change in level (1.58%; 95% CI −14.79% to 17.94%), with a non-significant change in trend (0.18% per half year, 95% CI −2.76% to 3.12%; online supplemental figure 3).
Supplemental material
Supplemental material
Impact on breastfeeding outcomes
Any domperidone dispensing was associated with a reduced likelihood of breast milk feeding at infant discharge (adjusted OR (aOR) 0.59, 95% CI 0.46, 0.77). From 2004 to 2018, breast milk feeding rates among women dispensed domperidone were consistently lower than in those not (figure 2). When stratified according to the timing of first dispensing, compared with those not dispensed domperidone, the likelihood of continued breastmilk feeding at discharge was lowest in women dispensed domperidone 8–28 days post partum (aOR 0.48, 95% CI 0.36 to 0.65; online supplemental table 2). Two or more dispensings were associated with increased odds of continued breast milk feeding at infant discharge (aOR 1.70, 95% CI 1.16 to 2.49) compared with those who received a single domperidone dispensing. Infants born at less than 28 weeks gestation had the lowest rates of breast milk feeding at infant discharge across the study period (56%, n=122, not breast feeding; online supplemental figure 4).
Supplemental material
Discussion
This study demonstrates a significant decrease in the rate of domperidone dispensing in mothers of very preterm infants from 2014 onwards following the EMA recommendations. Throughout the study period, maternal receipt of domperidone was associated with a consistently lower likelihood of continued breast milk feeding at infant discharge. Despite the reduction in domperidone dispensing, continued breast milk feeding rates at infant discharge remained stable and substantially lower over time.
The overall prevalence of domperidone use is similar to that reported in prior Australian and English studies (32%–41%).16 17 19 25 The dispensing of domperidone in this population of mothers of very preterm infants was higher in women who were older and had a higher BMI, consistent with previous studies.11 26 In addition, the highest dispensing rates occurred in women who delivered at less than 28 weeks gestation. This may relate to women having less mammary cell differentiation, causing a lower breast milk volume and hence, a requirement for domperidone, and their longer stay in the neonatal unit, increasing opportunities for identification of lactation difficulties and subsequent domperidone prescribing.2 27
One study has investigated the impact of the EMA recommendations overall but did not focus on the preterm setting. Following the recommendations, an immediate 15% reduction in domperidone prescribing, which then returned to the previous rate over the following year.19 Conversely, the current study reports a sustained reduction in domperidone dispensing from 2014 onwards. While not assessed in the previous study, we also observed a reduction in the proportion of individuals dispensed domperidone within the first week postpartum. Both effects are likely the result of changes to the hospital prescribing guideline for domperidone, which was amended to advise more judicious use of domperidone and discouraged prescribing within the first week postpartum. However, despite changes in overall prevalence and timing of use, we observed no changes in the proportion of women receiving multiple dispensings. While the EMA recommended use should not exceed 7 days, our results indicate that the majority of women undergo treatment for at least 2 weeks. Clearly, balancing the potential benefits of increased duration of domperidone use against potential increased risks remains a challenge, with qualitative studies demonstrating that women’s determination to breastfeed and perceptions of the benefit of breast milk to their infant may outweigh their perceived individual risk of side effects.28 29
We found breastfeeding rates remained lower at infant discharge despite domperidone dispensing compared with women who were not dispensed domperidone. Our real-world evaluation of breast milk feeding rates at infant discharge was 63% and compares to continued breastfeeding rates of 54.6%–95% in previous clinical trials of mothers of preterm infants.12–15 The variation in breastfeeding rates across different clinical trials and our study may stem from variations in the breastfeeding supports available at other sites, and differences in the patient population, such as gestational ages at birth. As noted in the prescribing checklist, domperidone should only be prescribed following the provision and trial of appropriate non-pharmacological lactation support strategies.
Despite a significant reduction in the prevalence of domperidone prescribing following the EMA recommendations, no corresponding changes were identified in the overall prevalence of continued breast milk feeding at infant discharge. This could be interpreted as suggesting that dispensing of domperidone has a limited overall effect on longer-term breast milk feeding outcomes. Within those dispensed domperidone, we found improved breastfeeding outcomes associated with multiple domperidone dispensings compared with a single dispensing of domperidone. These findings may indicate that longer durations of domperidone use (greater than 7 days) may be associated with improved treatment outcomes. Alternatively, differences in outcomes based on treatment duration could reflect those in which domperidone is considered effective, choosing to continue to use it for longer. What remains clear is that determining the optimal treatment duration and identifying individual-level factors influencing treatment response to domperidone remain clear research priorities.
Limitations
The current study has several limitations. We had no information on domperidone dispensing occurring outside of the hospital. However, given that the hospital has a well-established internal process for prescribing and dispensing domperidone from the hospital pharmacy, where the medication is provided free of charge, we anticipate that the number of women choosing to get domperidone dispensed outside the hospital is small. In addition, no information was available about the degree of lactation support women had before their domperidone prescription. Lastly, routinely collected data on continued breast milk feeding at infant discharge did not discriminate between whether infants were exclusively fed their mother’s breast milk or receiving mixed feeds.
Conclusion
The 2014 publication of the EMA guidelines raising safety concerns about domperidone use led to local practice changes and a substantial, sustained reduction in the prevalence of domperidone use in mothers of very preterm infants. While the findings may reflect a more rational use of domperidone, changes in domperidone use had no impact on overall breast milk feeding rates at infant discharge, which were observed to be consistently lower among mothers dispensed domperidone throughout the study period. Further research is needed to improve breastfeeding support and tailor the timing, dose and duration of domperidone treatment to optimise longer-term breastfeeding outcomes in mothers of preterm infants.
Data availability statement
Data may be obtained from a third party and are not publicly available. The data for this paper is available only on request to the ethics committee as it contains identifiable details of participants and their children.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Human Research Ethics Committee at the Women’s and Children’s Hospital (approval number HREC/20/WCHN/79). As this is a retrospective cohort study and individual participation is not required, a waiver of consent was sought and approved by the WCH HREC.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @Nahbucks
Contributors GMM: investigation, writing—original draft preparation, writing—editing and reviewing. ARR: conceptualisation, investigation, writing—editing and reviewing. AKK: writing—editing and reviewing. AK: data curation, formal analysis, writing—editing and reviewing. MB: data curation, writing—editing and reviewing. SJ: writing—editing and reviewing. LS: writing—editing and reviewing. MS: data curation, writing—editing and reviewing. LEG: conceptualisation, data curation, formal analysis, investigation, writing—editing and reviewing. LEG is responsible for the overall content as the guarantor.
Funding GMM was supported by an Australian Government Research Training Program Scholarship. AK was supported by a National Health and Medical Research Council Early Career Fellowship (GNT1161379). LEG receives salary support through a Channel 7 Children’s Research Foundation Fellowship (CRF-210323).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.