Impact of Community Perinatal Mental Health Teams on mental health and birth outcomes (The ESMI-II study).

Perinatal mental health (PMH) problems (i.e., Mental Health problems occurring during pregnancy or the first year after childbirth) can have a severe impact on women and their babies. Severe mental health problems can be associated with significant impairment in social and personal functioning, which might affect the woman's ability to care for herself and her child. Psychiatric causes of maternal death, particularly suicide, continue to be a significant cause of maternal mortality in the UK.

There is a need to support women who may be at risk, such as those with previous mental health problems, so that specialist support can be provided. The provision of specialised perinatal psychiatric care in the UK is minimal and inequitable. Many women with serious illness are not able to access the appropriate type and standard of care as recommended by the National Institute for Health and Care Excellence (NICE) guideline on antenatal and postnatal mental health, 2007.

In 2014, the National Institute for Health and Care Excellence, recommended that women who have or are suspected to have complex or severe mental illness during pregnancy or the postnatal period should be referred to a secondary mental health service, preferably a specialist community perinatal mental health teams (CPMHTs), for assessment and treatment and that inpatient care within 12 months of childbirth should be at a specialist mother-and-baby unit. The Royal College of Psychiatrists published a recommendation on the staff and service composition of specialised perinatal mental health services, which include CPMHTs and mother-and-baby units.

The implementation of specialised CPMHTs in Clinical Commissioning Groups (CCGs) was supported by an investment of £365 million in 2016. It was expected that effective services could reduce the costs related to perinatal MH problems, which are estimated to be £8.1 billion annually (28% due to direct costs and 72% to the effect on the infant).

The London School of Hygiene and Tropical Medicine (LSHTM) is undertaking a study to assess the impact of CPMHTs on the number and duration of mental health admissions in secondary care, on maternity outcomes within 12 months of giving birth, and on the costs of these admissions in women with a who gave birth during the study period. The study will be carried out using existing national electronic health datasets, including Birth Notification data, Hospital Episode Statistics, and the Mental Health Services Dataset (in all its versions), linked at patient-level.

This study for which this data is required, is Work Package (WP) 4 of the NIHR funded project Effectiveness and Cost Effectiveness of Community Perinatal Mental Health Services (ESMI-II). In particular "Investigate the effectiveness of CPMHTs in improving access, outcomes, and preventing relapse, and thus reducing cost, using national NHS datasets".

LSHTM will investigate the impact of CPMHTs on:

- Inpatient admissions to psychiatric hospitals.

- Access to secondary care mental health services, such as contacts with a crisis resolution team (home treatment teams) or secondary care community mental health teams.

- Access to primary care mental health services (e.g., Improving Access to Psychological Therapies Dataset (IAPT).

- Maternity outcomes in women with a history of mental illness, including maternal mortality, morbidity and length of stay, perinatal mortality (stillbirth and death in first week of life),preterm birth, (gestational age at birth < 37 weeks), birth small for gestational age (SGA; birthweight below the 10th percentile for the gestational age), mode of birth (caesarean section and instrumental by type), adjusted for maternal comorbidities as well as pregnancy complications (e.g. hypertension/preeclampsia, diabetes).

-Cost of all secondary care admissions and detentions (economic outcomes).

Survey data will be used to determine if and when CPMHTs were implemented (the exposure) in regions defined according to NHS Clinical Commissioning Group boundaries (CCGs). The survey has been carried out as WP 1 of ESMI-II (Work Package 1). It used a taxonomy of CPMHTs, defining levels of compliance, which capture how they vary in their configurations and the services they provide. LSHTM will use nationally applicable unit costs available from existing sources. The outcomes will be modelled with multiple-baseline interrupted time series analysis using a multilevel logistic regression model with three components (a pre-implementation slope, a level change, and a change in slope).

The lawful basis for processing data will be carried out under Articles 6(1)e and 9(2)j of the General Data Protection Regulations (GDPR). Article 6(1)e states that processing of data is lawful if ‘processing is necessary for the performance of a task carried out in the public interest’. Consequently, the lawful basis for processing is under ‘Public task’. Article 9(2)j states that processing is permitted for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes for the special category health data. The research is in the public interest, will involve processing using sophisticated statistical techniques, and is directly related to health.

The data processing within LSHTM is necessary for the purposes of delivering a national research study funded by National Institute for Health Research. In addition, this project will have no adverse effect on the rights and freedoms of data subjects (i.e. mothers and their babies) the data processing is necessary for improving the quality of mental health and maternity care throughout England, which falls into the categories of “provision of care and treatment” and “management of healthcare services”. LSE are only permitted to process records within a defined period post-birth episode and require destruction of any excess data once they have extracted the relevant records from the full dataset.

The requested data will be used for the purpose of medical research strictly to perform advanced statistical analysis: health economics and outcomes research studies (study of the value of treatments and the health benefits they deliver to enable doctors to make decisions about the best treatment pathways for patients)

The study is funded by the NIHR and it aims to inform improving the quality of mental health and maternity care throughout England.

The data subjects’ interests and fundamental rights are protected through appropriate minimisation of fields; protection of the data in a secure environment and guaranteeing secure destruction at any stage at the request of NHS England or after a defined period on completion of the project.

Globally, mental health problems are stigmatized and may be written about in negative terms. Deciding how to disseminate the findings will include liaising with women from mental health patient groups.

Strict data disclosure control practices will be used to minimise the risk of the data being identified following publication of research results. LSTHM will also work closely with representatives of mental health patient groups advisers to avoid reports and publications stigmatising patients with mental health problems and ensure that mental health problems are not described in a negative way.

The linked national datasets will be used to identify the study population of women with PMH problems who gave birth in the study period, before and after the implementation of a CPMHT in each CCG. These same data will provide information on access to primary and secondary care mental health services, maternal mortality, morbidity and hospital stay of the birth episode, perinatal mortality (stillbirth and death in first week of life), preterm birth, birthweight, mode of birth, and length of (maternity outcomes), as well as on confounding variables including maternal comorbidities as well as pregnancy complications.

The linkage of Birth Notification data to the Mental Health Datasets (all versions) and Hospital Episode Statistics has several advantages for the study:

It will provide information on longitudinal patterns of care, for example, psychiatric hospital admissions before and after birth.

It will allow a detailed study of the women’s comorbidities and their health service use before and during pregnancy which is important for case-mix adjustment.

The date and time of birth field is essential to the study for the following reasons:

Without this information it will not be impossible us to derive outcome measures such as the time to readmission after birth.

Date and time of birth will also assist with determining whether records are duplicates or whether there has been a multiple birth.

As indicated earlier, this study relates to WP 4 of the NIHR funded ESMI-II, which is carried out by a consortium which includes University of Exeter, King’s College London, University of Oxford and University of Liverpool who will act in an advisory capacity only whilst the London School of Hygiene and Tropical Medicine are the Sole Data Controller who also process data for this stage of the study. http://www.esmi2.org.uk.

In summary, the key research question ESMI-II will address if CPMHT is effective in improving mother and infant outcomes and in what contexts.

Perinatal mental health (PMH) disorders (mental disorders occurring in pregnancy or the year after childbirth) are a serious public health issue, associated with distressing symptoms and poor functioning, pregnancy complications, maternal deaths, and long-term negative effects on child cognitive, social and emotional development. These are thought to be mediated prenatally by the effects of cortisol on the developing foetus leading to alterations in infant stress reactivity, and postnatally through deficits in mother-infant interaction. Stigma, lack of specialist services and lack of trained staff have meant access to mental health care has been poor. NHS England recently invested £365 million to improve parent’s access to care and health outcomes.

ESMI-II has four WPs. WP 1 will develop a taxonomy of variations in CPMHTs. WP 2 will develop and validate a measure to assess quality of mother-infant interaction for use by CPMHTs. WP 3 will use a realist evaluation approach and analyse qualitative interviews with women, co-parents and health and social care practitioners. WP 4, as explained earlier, will used linked national data to carry out a before-and-after study to examine how access to care changes over time and whether mothers with a history of serious mental illness are less likely to need acuter care if they live in an area with a CPMHT and their maternity outcomes are better.

The study population is defined as all women who gave birth in England in the study period (from 1 April 2014 to 31 March 2020) with pre-existing severe mental health condition. Women who give birth are identified in Birth Notification data. Women are considered to have had a pre-existing severe mental disorder if they had a contact with secondary care mental health services in the English NHS from 2006 up to the start of pregnancy.

Contact with secondary care mental health services in the English NHS services is defined as at least one community contact (defined as any record in a version of the Mental Health Services dataset that includes a referral to a health professional for a face-to-face consultation such as contact with community mental health teams, crisis resolution and home treatment teams) or a psychiatric hospital admission.

The Birth Notification dataset will be the spine to which all other NHS datasets will be linked.

Datasets requested:

Birth Notification data

Birth Registration data

Civil registration deaths

Hospital Episode Statistics

Mental Health Datasets, including Improving Access to Psychological Therapies Dataset.

Justification for the datasets

Birth Notification data will be used to identify births during the study period in the English NHS. These data will give key demographic details about the mother, including NHS number of mother and baby (needed for record linkage), residential postcode (needed to establish a mother’s CCG and Index of Multiple deprivation), ethnicity and about the birth(s), including number of births and birth order, birthweight, length of gestation, and whether the baby was born alive.

Hospital Episode Statistics (HES) records, including mortality, will give information about maternal characteristics, the birth and inpatient admissions of the mothers and of the babies. It will provide information about maternal outcomes, maternity outcomes, as well as maternal comorbidities and pregnancy complications.

The Mental Health Services Dataset (MHSDS), and its previous version known as Mental Health Minimum Dataset data (MHMDS), and the Mental Health and Learning Disabilities Dataset (MHLD1DS) will provide data about secondary mental health services provided or funded by the NHS, including demographic characteristics, contacts with community service and outpatient contacts, voluntary and involuntary inpatient treatment. Records of mental health contacts in primary care are available from the records of the Improving Access to Psychological Therapies Dataset (IAPT).

Justification for the level of data

LSHTM require pseudonymised data. LSHTM require the month and year of baby birth, because without these items it would be impossible for LSHTM to derive outcome measures such as the time to readmission after birth or to determine whether records are duplicates or belong to a multiple birth, mental health data.

Justification for the number of years requested

LSHTM request Mental Health Data from 1 April 2006 (earliest available) to latest available for all women who gave birth in the study period. LSHTM need to identify women with a history of an episode of secondary mental health care in the 10 years before the start of the study period and LSHTM want to maximise the follow-up duration so LSHTM need to have at least one year follow-up for as many women as possible who gave birth in the study period.

LSHTM request HES data from 1 April 2009 to to latest available for all women who gave birth in the study period. LSHTM need to have HES data from 5 years before the start of the study period to identify comorbid conditions as well as to determine parity, given that this is often missing in the HES record of the birth episode. LSHTM need to have at least one year of follow-up for as many women as possible who gave birth in the study period.

Justification for the geographical spread of the data requested

The CPMHTs were rolled out nationally in all CCGs and therefore national data covering England is needed to evaluate their impact.

Data minimisation

It is impossible to collect this data at national level for all births of mothers that took place during the study period, given the numbers of birth, the number of NHS units that were involved in the care for these women, and the complexity of the data collection structure (e.g., many units that can not be identified in advance can be involved in the care for an individual women).

LSHTM only request data that is necessary:

- to determine the eligible cohort of women and their babies

- to determine their area of residence

- to determine their mental health and other relevant diagnoses;

- to determine the care they received

- to determine the case-mix , including comorbidities, obstetric and mental health history

- to determine their health outcomes and the outcomes of their babies

Only data items which are necessary for the study are requested, with one or more of the following purposes:

- to determine study population, characteristics and case mix, including mental health and obstetric history

- to determine the access to secondary care mental health services during the perinatal period;

- to determine the access to primary care mental health services during the perinatal period (IAPT);

- to determine the health outcomes for the mother and baby up to one year post birth of baby.

For England, a derived cohort of all babies born to women in the scope of this study between 1 April 2014 will be created by NHS England using the birth notification/registrations dataset.

Based on this definition of necessary data, data minimisation was implemented in the following ways:

- LSHTM only request data for women who have given birth during the study period (1 April 2014 and 31 March 2020) and who had an episode of a secondary mental health care episode from 1 April 2004. LSHTM need to consider historical mental health data for at least a 10-year period, because mental health care after a first episode is often delivered in primary care. For example, historical data only for a period of five years (i.e. from 1 April 2009) is unlikely to identify a considerable number of women with severe mental health conditions who have given birth during the study period.

- LSHTM only request HES data for women who have given birth during the study periods, starting from 1 April 2009. LSHTM need a period of 5 years before the start of the study period to identify comorbidities and to determine parity which is one of the key determinants of maternity outcome.

The Project Team at the LSHTM will only receive pseudonymised data from NHS England. No personal information, including name, age, and postcode will be transferred by NHS England to the Project Team at the LSHTM.

Data available from PDS birth notification data set, HES APC, MHSDS, MHLDDS and MHMDS and IAPT for this cohort (including babies’ and mothers’ data) will be returned to the LSHTM.

All historical data from these same individuals is requested. There are three main reasons for requiring the historical data:

1) it enables definition of study population (those with a history of severe mental illness);

2) enables information to be provided on longitudinal patterns of care, for example, maternal hospital readmission following birth.

3) it enables information to be collected on the clinical history of the women before pregnancy, their comorbidities and their health service use during pregnancy which is important for case-mix adjustment.

London School of Hygiene and Tropical Medicine will be processing all data as defined below in the data linkage steps.

Data sources and linkage

Step 1: Birth cohort “spine” (denominator): NHS England to identify all births between 1 April 2014 and 31 March 2020 (mother-baby pairs) in the Birth Notification data.

Step 2: NHS England to link the births identified in step 1 with the mental health data (MHSDS, MHMDS and MHLDDS).

Step 3: NHS England to include in the study cohort: all women identified in step 1 with at least one episode of secondary mental health care since 1 April 2006 according to the linked in mental health service data; and all babies born to women included

Step 4: NHS England to link all women identified in step 3 to IAPT

Step 5: NHS England to link cohort to HES APC records, including linked Civil Registration (Deaths) data records, covering episodes from 1 April 2009 to 31 March 2020 to women in the study cohort

- Link: NHS England to link HES APC records to Cohort defined in step 3 (including maternal and baby records for the birth episode, mother and baby.

- Extract: maternity records from HES for all women included in the cohort : the delivery record and the birth record. (Both types an admitted patient care record with an additional 19 fields, in an appended baby).

- Extract HES APC historic and future records for all women included in the cohort covering episodes from 1 April 2009 to 31 March 2020

- Extract HES APC records for all babies born to mothers included in cohort

- Link: Civil Registration (Deaths) data records to Cohort defined in step 3

- Extract : date and cause of death for mothers and babies born in the cohort

There will be no requirement and no attempt to identify specific individuals from the data.

Step 6: NHS England to transfer the pseudonymised data extract, which includes the birth notification extract (spine- denominator file); linked Birth Notification data; mental health data (including IAPT), and HES data (including linked mortality), to the LSHTM.

LSHTM requests that pseudonymised NHS numbers for mothers and babies, generated from Birth Notification data are added to each dataset provided to the LSHTM.

The LSHTM requests that also the records of the women identified in step 2 (based on linkage of Birth Notification data to the mental health services datasets) that do not link are included in the data extract that is transferred to the LSHTM.

Only members of the Project Team, who are substantive employees of the LSHTM will be involved in the processing and analysis of the data requested from NHS England.

Data will be stored on a secure data server within the LSHTM computer network. Access to data will be limited to the members of the Project Team, who are substantive employees of the LSHTM.

No data can leave the secure environment. The access devices only provide a view into the secure server environment. Access devices may only connect to the secure environment either from within the LSHTM network, or, where accessing the environment from outside LSHTM premises but within the territory of use, using a Virtual Private Network (VPN) connection to the LSHTM network.

The access device is owned by LSHTM and the LSHTM will ensure that access devices comply with security controls specified in the agreement in relation to password complexity; firewall; anti-virus; updates; screen locking; and hard drive encryption.

Data will only be accessed in environments designed to minimise the risk of inadvertent disclosure. This includes not using an access device in public areas, but instead only using the device in private settings such as an office or home environment. Only aggregated data with small number suppression applied as per the HES analysis guide will be disseminated as part of this study.

It is hoped that LSHTM will disseminate the findings through stakeholder events for staff at CPMHTs, maternity mental health leads and policy makers, including places prioritised for sub national site collaborators, regional perinatal network managers and clinical leads. These events will be held in different parts of England each year to facilitate attendance by sub national clinicians and maintain interest and research participation.

The results of this research will also be discussed with NHS England towards the end of the research. LSHTM will also disseminate through other relevant organisational meetings. LSHTM will also disseminate the findings through academic papers and conference presentations.

Expected academic papers will include:

- Maternity outcomes in women with pre-existing mental health conditions.

- Results of the clinical analysis, presenting the estimated impact of CPMHTs maternal and maternity outcomes (June 2022)

- Results of the economic analysis (June 2022)

- Methodological development, describing a classification system for mental health conditions for the purpose of perinatal mental health research.

Update under version 1 the above outputs were published in 2023/2024 as the data was received in March 2022. The team are working on two manuscripts that are within the scope of the programme of work described in Section 5a. Objective for processing. These expected outputs will be focussing on topics that required further methodological development work on mental health conditions + IAPT dataset (primary mental health services).

It is hoped the findings of this study will inform commissioning of perinatal mental health services in England, elsewhere in the United Kingdom and abroad. They will also support the further development of NICE guidance and NHS England policy with respect to the commissioning of perinatal mental health services.

Depending on the results, the study may lead to improvement of services to women with mental health conditions before and during pregnancy and in the first year after birth.

Although there will be no direct benefit to the women or babies included in the cohort for this study, the aim of this study is to evaluate the impact of CPMHTs on psychiatric inpatient admissions during the perinatal period and the health of mothers and babies it is hoped this will help future patients.

It is hoped the findings will inform best practice, the commissioning of perinatal mental health services, the components to commission, future updated NICE and NHSE guidance, and best practice leading to greater staff satisfaction and potential efficiency savings.

Findings could also inform criteria for maternity tariffs for standard, intermediate and high-risk care (currently the high-risk tariff is based on postpartum psychosis history only).

The research should lead to improvement of services; as a result women and their families should benefit through fewer barriers to accessing care, better tailored care across the care pathway, better mental health, functioning and quality of life and better family outcomes.

The results of NIHR ESMI-II: The Effectiveness and cost effectiveness of community perinatal Mental health services study, using this data has been published in high-impact peer reviewed papers, and presented in academic and wider stakeholder meetings including NHS England. The first paper show that women who had more severe and more recent mental illness episodes had substantially higher risk of adverse pregnancy outcomes. The second paper demonstrates that the national roll-out of community perinatal mental health teams reduces risk of acute mental health episodes after birth. The DSA extension is requested for the completion of two other publications.

References:

Langham J, Gurol-Urganci I, Muller P, Webster K, Tassie E, Heslin M, Byford S, Khalil A, Harris T, Sharp H, Pasupathy D, van der Meulen J, Howard LM, O'Mahen HA. Obstetric and neonatal outcomes in pregnant women with and without a history of specialist mental health care: a national population-based cohort study using linked routinely collected data in England. Lancet Psychiatry. 2023 Oct;10(10):748-759

Gurol-Urganci I, Langham J, Tassie E, Heslin M, Byford S, Davey A, Sharp H, Pasupathy D, van der Meulen J, Howard LM, O'Mahen HA. Community perinatal mental health teams and associations with perinatal mental health and obstetric and neonatal outcomes in pregnant women with a history of secondary mental health care in England: a national population-based cohort study. Lancet Psychiatry. 2024 Mar;11(3):174-182.

Tassie E, Langham J, Gurol-Urganci I, van der Meulen J, Howard LM, Pasupathy D, Sharp H, Davey A, O'Mahen H, Heslin M, Byford S. An exploration of service use pattern changes and cost analysis following implementation of community perinatal mental health teams in pregnant women with a history of specialist mental healthcare in England: a national population-based cohort study. BMC Health Serv Res. 2024 Mar 20;24(1):359.