Abstract
Usual referral pathways to psychiatric services can miss opportunities for timely intervention in maternal perinatal psychiatric ill health. Psychiatric illness leading to suicide is a significant factor in at least 10% of maternal deaths. Despite Royal College of Psychiatry and National Institute for Health and Clinical Excellence recommendations for specialist provision of perinatal mental health services, this remains sporadic and insufficient. We set out to develop a new integrated antenatal–psychiatric direct referral pathway and present a year of experience using this service model. The psychiatric service was delivered from within the antenatal clinic setting with a direct health-care professional (HCP) led referral pathway between 2003 and 2004. The service comprised one session per week of a senior psychiatric specialist registrar and provided three new patients and two follow-up appointments per week. During this period, a total of 75 referrals to the service were made with 57 individuals attending for an appointment. There was a range of diagnoses among the women who attended, with only 24% meeting eligibility criteria for referral to secondary psychiatric services. The majority diagnosis was depression. More severely ill women were not referred to this clinic by obstetric HCPs. In conclusion, this model for developing and delivering a specialist perinatal psychiatric service using direct links to antenatal medical care was not successful despite requiring minimal funding. Nevertheless, it has been used to inform development of a new perinatal service in keeping with the Royal College of Psychiatrists' recommendations and incorporating enhanced training of HCPs responsible for the referral pathway.
INTRODUCTION
The perinatal period can have a profound impact on women's mental health. Patients with severe affective disorders are at high risk of relapse following delivery, treatments for pre-existing disorders may need to be changed because of pregnancy and breast feeding, and women are at elevated risk of developing their first episode of a range of disorders. Psychiatric illness leading to suicide is a significant factor in at least 10% of maternal deaths, and the National Institute for Health and Clinical Excellence (NICE) guidelines on antenatal and postnatal mental health highlight these risks and recommend opportunistic screening for mental illness throughout the perinatal period. 1
Much is known about the high risks of postnatal relapse in women with severe affective disorders, particularly bipolar affective disorder 2 ; and yet the care provided for pregnant mental health service users in generic community mental health teams has been criticized. The enquiry into the ‘extended suicide’ of Daksha Emson and her baby daughter Freya revealed poor communication between mental health teams, primary care and obstetric services, and highlighted a general failure to recognize the risks of relapse following childbirth. 3
The report Why Mothers Die: Confidential Inquiries into Maternal Death recommends the provision of specialist perinatal mental health services to every woman at risk. 4 The Royal College of Psychiatrists (2001) recommends that every health authority has a perinatal mental health strategy and that all women with perinatal mental health problems should have access to a consultant psychiatrist with a special interest in perinatal psychiatry, supported by professionals with experience and skills in this area. 5 Following these recommendations, the London Development Centre for Mental Health created a database of information about protocols and services and published their own recommendations arising from an audit of current perinatal psychiatric services. 6 These recommendations included the provision of four sessions of a community consultant perinatal psychiatrist with junior doctor and multidisciplinary team support per 4000 births or per primary care trust.
Yet services remain inadequate nationwide: a recent survey of perinatal services in England found a reduction from 35 dedicated inpatient mother and baby units in 1991 to 21 in 2005. 7 Liaison psychiatrists are more involved in the provision of these services over the last 10 years. Since trust restructuring in 1999 it seems that some trusts have reduced or stopped providing perinatal services, indicating that primary care trusts do not consider this a priority.
Working in partnership with the local obstetric service, we established a perinatal mental health clinic. We aimed to identify pregnant women at risk of moderate to severe illness to plan perinatal psychiatric care and link them to local community mental health services. The service was set up in response to the national guidelines mentioned above and also following two local maternal deaths secondary to mental illness within the trust, highlighting the need for a specialist service. This article explores issues raised by the project, which was ultimately unsuccessful in identifying those patients most in need of specialist psychiatric input.
DESCRIPTION OF SERVICE
The Whittington Hospital has an annual delivery rate of 3264 per year (mean rate 2002–2005). Based on published prevalence rates, it was assumed that there would be around 70 women with pre-existing enduring and disabling mental illness annually within this population. 6,8 Data from another perinatal mental health service in London indicated that around 150–200 referrals per year could be expected in total, comprising women with pre-existing illness, those with new episodes of illness and a number of women referred inappropriately. 9 Constrained by available resources, a liaison model was adopted, working closely with the antenatal clinic and accepting referrals from no other sources. The service was established as an unfunded pilot, with the hope that a permanent and more comprehensive service would later be commissioned.
Women were seen in an outpatient clinic held weekly by a psychiatry specialist registrar (SpR) attached to the hospital psychiatric liaison service. Complex cases were presented at a weekly multidisciplinary meeting along with relevant audit and research data. The SpR also attended the monthly social services meeting at which the most high-risk cases were discussed.
METHODS
The psychiatric clinic was located in the antenatal department in an attempt to reduce the non-attendance rate associated with the stigma of mental illness. One session of senior-level SpR time was dedicated to this service, with supervision from a consultant liaison psychiatrist. During this session, there were appointment slots for two new patients and three follow-ups, although some flexibility was exercised. This was in addition to a separate service already providing two sessions of psychology input to the antenatal department; while they managed women with milder conditions, our goal was to identify and treat women with more severe illness. Diagnoses were made in accordance with ICD-10 criteria on the basis of a full psychiatric history and mental state examination.
A referral form was made available in the antenatal clinic (available on request from the authors) and most referrals were generated by midwives and obstetricians. All women receiving antenatal care at the Whittington Hospital were eligible for referral to the service resulting in referrals from a wide geographical area. This is in contrast to the way that psychiatric services are provided in England. Mental health trusts, in common with primary care trusts, function within strict geographical boundaries and therefore this created difficulties in organizing psychiatric follow-up after delivery as our own trust only provided care to women living within the borough.
Both formal and informal training was offered to midwives to inform them of referral criteria and teach the basics of perinatal mental illness, as recommended in the NICE guidelines. Due to limitations of time, only one session of non-mandatory formal training was arranged at which there was poor attendance. Informal training was conducted opportunistically, through conversations with maternity colleagues in clinics and at meetings.
No specific screening tools were employed; referrals were derived from routine booking data regarding past or current mental health difficulties and any family history of psychiatric problems. While the intention was to provide a service to women with more severe illness, it was recognized that referrers may not have been able to determine severity; referrals of all women with mental health problems were therefore encouraged. The service was launched after presentation of its remit to the weekly perinatal meeting at the Whittington Hospital, which was attended by obstetricians, midwives and neonatologists. Active liaison was also attempted with hospital social workers, psychology services and any other relevant groups who might generate referrals.
RESULTS
Demographic and clinical data (Tables 1 and 2)
In the first year of the study period (Table 1), we received 75 new patient referrals, of which 41 (55%) were of women living in the local health authority. Of these, 18 did not attend an appointment (24%) which is in keeping with current expected DNA rates in general psychiatric outpatients. No data are available on the individuals who did not attend their appointment. Ethnicity data were recorded for 43 women with only 5/43 recorded as being of White British ethnicity. The ethnicity of the remaining 38 women was mixed but consistent with the ethnic minority population in this part of inner London. Eleven out of 75 women did not have English as their first language and required interpreters. The mean age was 29.7 years (range 17–43 years).
Data from the psychiatric clinic for initial 12 months (n = 75)
DNA, did not attend; MW, midwife; O, obstetrician; CMHT, community mental health team
Primary diagnoses (ICD-10)
Using this referral system half the women were referred by obstetricians across all grades, while the remainder were referred by midwives. Eleven out of 57 were referred via this specialist service to the existing psychology clinic, despite the provision of a direct referral pathway from obstetric health-care professionals (HCPs) already in place. Only 14 of 57 patients referred met eligibility criteria for onward referral to secondary mental health care. Twenty-two out of 57 women referred to this service by obstetric HCPs did not have any mental illness. The remaining 10 women had diagnoses that could be effectively managed in a primary care setting. Table 2 summarizes the range of psychiatric diagnoses among the women seen in this clinic. The most frequent diagnosis (17/57, 29%) was of depression, although 33/57 (58%) referred did not require any further psychiatric input.
DISCUSSION
The requirement for obstetrically trained HCPs to initiate referrals without sufficient training in mental illness and without the use of screening questions such as those recommended in NICE antenatal and postnatal mental health guidelines led to inappropriate patient identification. The service fostered a close working partnership between psychiatric and obstetric HCPs and with time could have been optimized to enhance training of obstetric teams to identify appropriate referrals. It is essential to incorporate regular maternity staff training as part of any such ongoing service. Obstetricians and midwives expressed a high degree of satisfaction with the service provided (personal correspondence available from author) and we felt that this was in part due to the face-to-face contact that was ensured by the location of the clinic and to the meetings attended by the SpR outside routine clinic times, though these inevitably increased further the time commitment necessary to run the service.
The rate of referrals into this service was at an appropriate level for the clinical time available, although the number of severely ill women identified was low: some of the women with more severe illness were missed by this process. The population of patients seen in this clinic differed from other psychiatric populations in general psychiatry outpatient clinics in terms of the range and prevalence of various types of mental illness. 10 Of particular note was the lack of women with severe and enduring illness, who would be eligible for service provision by the local psychiatric services. The reasons for this are likely to be threefold: firstly, the DNA subgroup may have comprised the most seriously unwell. 11 Secondly, without a successful training programme for the obstetric team or use of a validated screening tool, it is possible that most such patients were missed.
Obstetric HCP stigmatization of psychiatric illness may also play a part in this, although locating the clinic within the antenatal department was an attempt to tackle this. Finally, women with diagnoses such as schizophrenia, borderline personality disorder and affective psychosis were more likely to be already receiving psychiatric support from community mental health teams, and may possibly have been less likely to consent to referral to the liaison clinic. With retrospect, referrals should have been accepted from primary care and community mental health teams to enable those most in need of specialist input to be identified and referred.
Our clinic model more closely resembled a psychiatric liaison clinic than a dedicated perinatal service, and highlights a controversy that currently exists about whether perinatal service provision should be met by a specialist perinatal service or whether it is more pragmatic at least in the short term to start services up within existing psychiatry liaison services (where these exist). Liaison services assess and triage a high volume of patients to locality mental health services and manage emergencies. Dedicated perinatal services on the other hand aim to provide more comprehensive care for the most severely unwell by managing women throughout pregnancy and for the first postpartum year. In addition, perinatal services provide advice and consultation to general practitioners, accepting referral from primary care, and are available for advice regarding prescribing psychotropic medication to all women of childbearing age. This specialist expertise cannot be assumed from a more general liaison service, which takes referrals from all other non-obstetric medical and surgical departments.
The most pertinent issue is of course related to funding. Funding this clinic was a complex matter due to lack of geographical alignment between obstetric and psychiatric services. Local psychiatric services restricted their input to North Islington residents, whereas women living in a much more widespread area including Islington, Haringey and Camden were all eligible to use the obstetric services. This reflects in part the consumer-led nature of an obstetric service. Perhaps the consumers could be an effective lobby for ongoing funding. Smith 12 made it clear that her two special interest sessions per week were inadequate to set up formal referral procedures and protocols, and her service was soon inundated with referrals that largely came from local psychiatrists. Our clinic was significantly different from that reported by Smith in that we saw 57 referrals in the year described (Smith saw 26), they were seen in the antenatal department (Smith saw women within the psychiatric department or at home), and we did not accept postnatal referrals that were picked up by the liaison service as before. Our clinic operated from the liaison service base: secretarial services and supervision were both provided by the liaison consultant. After two years, despite submission of a business plan to the relevant authorities, no funding was allocated for this clinic and the service was discontinued.
More recently however attempts to fund a comprehensive perinatal mental health service at the Whittington Hospital have been successful, and the first author has been appointed to provide five sessions per week of Consultant Perinatal Psychiatrist input, backed by a part-time Specialty Grade doctor, and a WTE nurse specialist. The service will accept referrals from primary care and community mental health teams, and will provide ongoing psychiatric care to women throughout the perinatal period. Lessons will be learned from this ultimately unsuccessful pilot, and we hope to report on more successful outcomes in the future.