Randomised trial of the effectiveness of Group Family Nurse Partnership (gFNP) programme in improving outcomes for high-risk mothers and preventing abuse

Study design: The study will comprise a randomised controlled trial in which eligible families will be randomly allocated to one of two arms i) gFNP delivered over 76 weeks; ii) standard care. Setting: Family Nurses employed by the local NHS trust in conjunction with local councils as part of the Family Nurse Partnership team will provide the intervention in local Sure Start Children's Centres. Inclusion criteria: Expectant mothers prior to 16 weeks gestation and with expected delivery dates within 4 to 6 weeks of each other (per site). Eligibility criteria are either: aged under 20, expecting second or subsequent child, no qualifications or maximum 4 GCSEs at Grade C; or aged 20 to 23, expecting first or subsequent child, no educational qualifications or maximum 4 GCSEs at C, and one of: current or history of mental health problems, ever in care of social services, lives in workless household, or never employed. Intervention: Group Family Nurse Partnership (gFNP) - groups of on average 10 women with similar due dates and partners if applicable starting at 16 weeks gestation and extending until infants are one year; comprises 14 2-hour pregnancy sessions and 30 in infancy, all based on materials originally used for the Nurse Family Partnership(NFP) programme. One of the two Family Nurses is also a qualified midwife who provides routine antenatal midwifery care; both FNs provide postnatal routine infant checks. Recruitment (see Figure 1) Parents will, with the support of the midwifery team, be recruited by researchers at or soon after their the booking-in visit. The midwife will identify all eligible women and, with oral agreement, provide their details to the researcher who will then explain the study in more detail including the randomization. They will be given 24 hours to make a decision after which time the researcher will telephone. If verbal consent is given a home visit will be made immediately or as soon as possible to obtain written consent and baseline details. Researchers will telephone the clinical trials unit so that randomization can occur. Consenting participants will be informed about the group to which they have been allocated by the researcher and those receiving the intervention will be contacted by the FNP team, who will be given the details of intervention families by the Trial Manager. Sample The standard deviation (SD) of the short form Child Abuse Potential Inventory in the target population is approximately 6.3 (Ondersma et al., 2005). We propose to recruit sufficient mothers and babies (families) to allow the trial to detect a difference between groups of 0.5 standard deviations, with 90% power at a significance level of 0.05 (2-tailed); this is considered to represent a moderate size of effect . Very conservatively assuming a correlation of 0.4 between pre and post intervention scores we would need 71 families in each arm of the trial to detect this difference. Conservatively, allowing for an expected 30% drop out rate (based on pilot sites) we would therefore need to recruit a minimum of 84 families per arm of the trial. We therefore propose to recruit 100 families per arm. In the intervention arm this will consist of 10 groups of (generally) 10 families, provided in 5 FNP sites (N=100). Allocation to groups: Consenting parents will be randomly allocated to one of two arms. Randomisation will be conducted by the LSHTM Clinical Trials Unit, who will provide a 24 hour telephone contact system. Data collection: The data will comprise a range of self-report questionnaires, and (at 12-13 months) recorded observations of the parent and infant and hospital attendance data from HES. Data will be collected at baseline (before randomisation) for demographic data and all questionnaire outcomes, at 2-3 months (CAP, EPDS; PSI); and immediately post-intervention when the infants are 12 months (all outcomes including the CARE index). Maternal quality of life information (EuroQuol EQ-5D, SF-36) wi