Identification of substance use in women and pregnancy
: national and international review of prevalence and access to evidence-based treatments with an examination of the maternal and perinatal outcomes associated with methadone opioid substitution treatment in pregnancy

Student thesis: Doctoral ThesisDoctor of Philosophy


To contextualise this series of experimental studies introductory chapters examined the understanding of the aetiology and prevalence of substance use in women and access to WHO recommended evidence-based opioid substitution treatment (OST) nationally and internationally. This understanding helps inform estimations of the prevalence rate of substance use in pregnancy which has historically been challenging. Perinatal outcomes associated with maternal methadone treatment in pregnancy – primarily Neonatal Abstinence Syndrome (NAS) – may impact negatively on women’s access to evidence-based opioid substitution treatment. NAS as a syndrome is poorly defined whilst the relationship between OST and NAS remains uncertain which the experimental sections of this study aimed to examine.
Introductory chapters - A literature review examined current knowledge of female substance use in pregnant and non-pregnant women nationally and internationally and considered pregnant women’s current access to evidence-based treatments for opioid dependence. Access to OST treatment during pregnancy was examined and a review of the current evidence describing the relationship between methadone treatment in pregnancy and perinatal outcomes undertaken. A detailed examination of the pharmacological properties of methadone were undertaken to assist understanding of the perinatal outcomes of intrauterine methadone exposure.
Experimental chapters - A national self-report Survey of Neonatal Units in England and Wales was undertaken to examine the objectivity and standardisation of current diagnostic processes for NAS and estimate prevalence rates for maternal substance use affected pregnancies. The National Survey was a repeat of one undertaken 8 years previous. A prospective study recruited pregnant women receiving methadone for the treatment of opioid dependence from an NHS antenatal clinic in West Yorkshire. Blood and urine samples were collected from subjects and their neonates to investigate maternal and neonatal methadone pharmacokinetics in pregnancy and early neonatal life. A self-report Critical Staff Survey was undertaken at a secondary care hospital in the North of England. The aims of the survey were to investigate the knowledge, skills, and attitudes of front-line clinical staff providing care to methadone exposed neonates and their mothers. Ethical approval was obtained collectively for all three studies. The NRES approval committee was Camberwell St Giles, London (Skipton House, SE1 6LH): REC number 13/LO/1370. Health Research Authority approval was also awarded (IRA Project ID: 124847). Local Research and Development approvals were obtained from appropriate local NHS Trusts.
Literature Review - The aetiology of female substance use is complex but often associated with a history of personal trauma. Substance use behaviours in women are changing in the UK with alcohol and prescribed medication use increasingly prevalent. The true prevalence of pregnancies affected by maternal substance use is not known. It is likely official prevalence rates are underestimates due to the associated stigma preventing women from disclosing. Women’s access to evidence-based treatments for opioid dependence internationally is inconsistent. Perinatal outcomes are variable and NAS remains poorly defined with its relationship to intrauterine methadone exposure complex and subject to numerous confounders.

National Survey - There was an 80% response rate from all neonatal units in England and Wales. The prevalence rate for drug affected pregnancies had reduced from 1.30% in 2008 to 1.09% in 2016. The highest prevalence rates for drug affected pregnancies were identified in the Midlands, North West regions of England, and East of England. Ninety-six percent of responding units had a written protocol for the management of NAS: an increase from 90% in 2008. Most units (81%) commenced monitoring for NAS within 12 hours of birth, 12% of responding units only commenced monitoring if an infant became symptomatic for NAS. The use of clinical score charts to assess NAS had reduced from 94% in 2008 to 89% in 2016. Over the same period use of Finnegan had increased from 56% in 2008 to 68% in 2016. Respondents identified inter-rater variability as a negative consequence of clinical score charts. Oral morphine sulphate remained the most used first line pharmacological agent for the treatment of NAS in both surveys. In 2016 2% of units used methadone. Phenobarbital was the most used second line pharmacological agent (43%). Since 2008 clonidine has been introduced by 4% of units as a second line agent. Ninety-three percent of responding units supported breastfeeding in ‘at risk’ neonates. Staff reported concerns about maternal drug use and methadone dose levels as reasons not to encourage breastfeeding.
Prospective studies of methadone kinetics and neonatal outcomes - In the prospective study 11 women were recruited who had 13 live births. Only 64% of women presenting to antenatal services were engaged with substance misuse services at the time of conception. Mean birth weight was 2.17 ± 0.83kg (median 2.13kg), range 1.08 to 3.72 kg. Maternal mean age was 33 ± 5.01 years (median 35 years), range from 24 to 39 years. Mean gestation at booking was 15 ± 6.58 weeks (median 15 weeks), range 6 to 27 weeks. The mean number of previous pregnancies was 2.3. All subjects smoked cigarettes and used illicit drugs during pregnancy. Mean gestation at delivery was 36.38 ± 3.18 weeks (median 37 weeks), range 30 to 40 weeks. There were children’s safeguarding concerns in over 90% of subjects (10/11). At delivery, mean daily methadone was 60 ± 31.25 (median 60 mg/day), range 5 to 95mg/day. Maternal methadone dose at delivery and maternal methadone plasma concentration at delivery were associated (p = <0.01). Mean methadone concentration of neonatal urine at Day one after birth was 4, 803ng/mL dropping by Day 5 to 774 ng/mL. The longest methadone was detected in neonatal urine was twenty-eight days. Neonates not requiring admission to SCBU reached peak NAS score within 100 hours of birth. Mean time to peak NAS score in breastfed neonates was 194hr. The prevalence of pharmacologically treated NAS within the study population was 17%. Morphine sulphate was utilised as the first line pharmacological intervention – one neonate also required phenobarbital. Neither maternal methadone dose nor maternal methadone plasma concentrations are directly related to the development of NAS. Neonatal cord methadone concentrations are not directly related to the development of NAS.
Critical Staff survey - There were 72 respondents representing medicine, midwifery, nursing, and healthcare assistants. The largest staff group participating in responses was midwives (35%). Neonatal nurses have the highest median clinical experience duration of 12.5 years, doctors the lowest at 3 years. 47% of respondents advocated pregnant women to be stabilised on methadone and maintained at the lowest clinical dose’. Twenty-two percent of doctors advocated detoxification from heroin in pregnancy. Almost 70% of respondents expected a neonate to become symptomatic for NAS within the first few hours of life. Most respondents described NAS as a ‘prolonged condition’ lasting weeks (32%) or months (48%). Fifty-four percent of respondents supported breastfeeding in mothers treated with methadone. Midwives demonstrated the greatest resistance to breastfeeding as a group. Eighty-five percent of respondents identified a need for formal clinical training in the management of NAS.

The prospective study identified a cohort of women who were not known to local drug services prior to presentation to specialist antenatal services. Pregnancies were marked by late presentation for antenatal care and high-risk due to maternal smoking and active substance use. Neonates demonstrated physiological compromise due to low birth weights. The prevalence of safeguarding concerns within the cohort evidenced vulnerable family units and potential inter-generational perpetuation of social exclusion and trauma. Pregnancy was not identified as a period of positively sustained maternal behaviour change. A key finding of the Prospective study was a low prevalence rate of pharmacologically treated NAS (17%) in contrast to many published reports. Maternal methadone plasma concentrations were associated with methadone dose but were not associated with NAS. Case studies highlighted the inter-individual variability in unwell neonates and the complexity of defining and describing NAS as a syndrome. Urine analysis in neonates revealed that methadone was long acting. This was a novel finding of the study. A total of 138/173 (80%) neonatal units in England and Wales responded to an invitation to participate in the national survey. Prevalence rates for substance use affected pregnancies were reduced from 2008. The largest reduction was observed in the North of England. The majority of responding units (68%) reported the use of Finnegan clinical assessment charts (89/138), morphine as the first line pharmacological intervention (96%) (132/173), and four-hourly clinical assessments of ‘at-risk’ neonates (46%) (63/138). The staff survey identified at a local level the direct impact of care and policy delivery. The most significant finding was a lack of staff training and education regarding the care of women who use substances during pregnancy and their neonates. 85% of respondents identified a training need in this area. Although nationally there is a consensus for the use of clinical score charts and breastfeeding locally this was not demonstrated. These findings highlight the challenges of standardising NAS as a syndrome and applying a consistent and unified care approach optimising outcomes for both mother and child.

Female substance users are a vulnerable and often hidden population impacting in some circumstances on their accessing critical care during pregnancy. There is no evidence from the prospective study that maternal methadone doses in pregnancy impact negatively on perinatal outcomes. National Survey data suggests the prevalence of drug affected pregnancies has reduced since 2008 with the greatest reduction in the North of England. Key findings from the national survey highlight a consensus amongst neonatal units in the assessment and treatment of NAS although when examined at the local level lack of staff training may impact on the consistency and standardisation of NAS diagnosis and care translating into inaccurate prevalence reporting. Anecdotally staff highlight increased presentations of mothers who use prescribed medications during pregnancy which triangulates with national data regarding female substance use trends.
Date of Award1 Nov 2021
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorKim Wolff (Supervisor), Jane Marshall (Supervisor) & Raul Perez-Montejano (Supervisor)

Cite this