Randomised controlled trial.
Four Melbourne (state of Victoria, Australia) Local Government Areas (LGAs) of Brimbank, Wyndham, Moonee Valley and Yarra. We selected these LGAs based on a Victorian government request to carry out the study in the north-western region of Melbourne and on annual birth rates in excess of 1000 to maximise recruitment.
All parents of newborn infants seen by their Maternal and Child Health Nurse (MCHN) at their first home visit (day 7-10 postpartum) in the four LGA's. The MCH nursing service is a universal, free service offered to all Victorian families with scheduled visits (covering 93% of all births) post partum, then at 2 weeks of age, and 1, 2, 4, 8, 12, 18, 24 and 42 months.
Parents with insufficient English to complete the questionnaires and take part in the intervention have been excluded. Infants born before 32 weeks gestation or with a serious health concern have been excluded as program material may not be suitable for very premature or ill infants.
Sample Size/Power calculation
In a Victorian survey of 724 mothers, [4
] 34% reported sleep problems at mean infant age of 4.6 months (range 3-6 months). In a community survey of Queensland parents [56
] 27% reported a problem in their 4-6 month old infants (n = 740). Drawing upon this data, a relative reduction in the prevalence of parent report of an infant sleep problem (sleep problem yes/no) by 30% at infant age four months (i.e. from 30% to 20%) is likely to be clinically significant and to have flow on effects for parent mental health. Assuming this, we aimed to detect a reduction in the proportion of infant sleep problems from 30% to 20% at 4-months with 80% power at the 5% significance level. In a trial that randomises individuals, we required 313 subjects in each arm or a total of 780 families, allowing for a 20% loss to follow up. Due to time and budget restraints, we have been able to recruit 750 families only, giving us 78% power to detect a reduction in the proportion of infant sleep problems from 30% to 20% at 4-months, at the 5% level of significance.
Using a computer generated random number sequence, an independent researcher allocated each consenting family to the intervention or control group. The research team and families remained blind to group allocation at the time of recruitment and consent; however, following this, knowledge of group allocation is unavoidable given the intervention type.
Despite being at risk of response and subjective bias, parental perceptions
of either a sleep or cry problem (yes/no) is our most important outcome measure, given its relation to increased parental depression symptoms, early weaning and increased risk of Shaken Baby Syndrome. To counteract the disadvantages of using a subjective outcome measure, we are also using a validated behaviour diary to obtain measures of sleep and cry duration. In a previous study of 446 infants, we found that diary data including total crying time and number of bouts of crying/24 hours were significantly increased in parents who reported problem crying in their infant vs. those who did not, whilst sleep duration and number of bouts of sleep/24 hours were significantly reduced in infant's whose parents reported a sleep problem vs. those who did not [5
All parents allocated to the intervention group will be posted a study-designed booklet and DVD and will be offered an individual telephone consultation (at infant age 6 weeks) and group session (at infant age 12 weeks). Telephone consultations and parent groups will be facilitated by trained health professionals (nurses, psychologists) with a background in infant care. These facilitators will spend a minimum of two hours completing specific training in how to use the Baby Business telephone and group session manuals. This training will be facilitated by either the Chief Investigator (HH) or the project manager (FC), and all facilitators will observe a minimum of two telephone consultations and two groups before independently delivering program content.
The 27-page booklet provides parents with information about normal infant sleep patterns, sleep cycles and therefore, the potential for an infant to wake overnight several times. The benefit of an infant learning to fall asleep independently is discussed. Content highlights the disadvantages of an infant relying on parent dependent cues to fall asleep at the beginning of the night (e.g. by rocking). The dangers of parents sharing their bed with a newborn are also described. Steps to settling an infant are given with emphasis on the importance of putting the infant down for a sleep drowsy but awake, so that the infant can learn to fall asleep independently. Normal infant crying patterns are discussed (including the 'crying curve' showing the natural peak and subsequent decline in infant crying; used with permission from the website: http://www.purplecrying.info
]) together with strategies for managing infant crying in both checklist and pictorial form. Signs and symptoms of uncommon medical causes of crying are outlined. Information on improving parental wellbeing is provided, as well as information on typical sleep and feeding patterns in Australian children after the first three months of life.
The DVD covers very similar content to the booklet but also includes footage of parents discussing the methods they use to settle their infant, the tired signs their infant displays, as well as demonstrations of how they wrap and sooth their infant.
The telephone consultation expands on the content of the booklet and DVD. During the telephone call, a facilitator helps the parent apply the information in a way that is suitable for their family. Parents are encouraged to discuss topics such as: whether they have noticed their infant's sleep cycles; how their infant behaves when over-tired and how to recognise and avoid over-tiredness; the advantages of teaching an infant to fall asleep without hands on help; the risks of co-sleeping; changes they would like to make to their settling routine as well as strategies they will use when their infant is having trouble settling; where their infant currently lies on the 'crying curve' and what they might expect to happen to crying duration in the coming weeks; colic; and, if appropriate, uncommon medical causes of crying.
Parents are invited to attend a 1.5 hour group session at a local venue. The group aims to troubleshoot any problems parents are having with infant sleep and crying. The group facilitator follows a manual that covers daytime feeding and sleeping patterns at 3 to 4 months of age, day time napping and how to encourage longer daytime sleeps, night time feeding, use of dummies, wrapping, myths around infant care, and the importance of parental self-care.
Families allocated to the 'usual care' (control) group receive no intervention from our research team. These families continue to receive the usual assistance and advice provided to all parents of newborns, via usual contact with MCH nurses and other health professionals.
Parent A (the primary caregiver) will complete a baseline questionnaire including infant date of birth, birth weight, birth order, gestation, the type of milk being given, approximate number of feeds during the day and the night, caregiver's date of birth, country of birth, the main language spoken in the home, current marital status and highest level of education completed. Details are also gathered regarding partner date of birth, country of birth and their highest level of education completed. They will also be asked where their infant spends most of the night (parents bed, own cot/bed in other room, own cot/bed in parent's room or 'other') and whether their infant has been diagnosed with or suspected of any major illness.
Parent A will complete the 'doubt' subscale of the Maternal Cognitions about Infant Sleep Questionnaire [25
] (see below). Other subscales are not relevant at such a young age. Feelings of parental doubt may already manifest at 2 to 4 weeks of age and such doubts are associated with increased likelihood of infant sleep problems [25
]. Parent A will also be asked whether they think they are a relaxed or tense person (0 = relaxed to 10 = tense; Sayers, 2004); being 'tense' (scoring 7 or higher) predicts higher infant irritability at nine weeks of age.
Follow up questionnaires
In order to establish the efficacy of this intervention, we are measuring outcomes at infant age four and six months, since most infants 'sleep through the night' by four months of age [58
] and no longer require night time feeding to meet their nutritional needs by six months of age. Measurement at these time points will therefore indicate established sleep patterns of these infants.
Both Parent A and Parent B (secondary caregiver) will complete follow up questionnaires at infant age four and six months. Only Parent A will complete an infant behaviour diary and the feeding questions at these time points.
Primary & secondary outcomes
Our primary outcome is parent report of infant night time sleep problems. Secondary outcomes include parent report of day time sleep problems, infant crying problems and feeding problems. Parents will be asked 'Have any of the following baby behaviours been a problem for you over the last 2 weeks? Daytime sleep (yes/no), Night-time sleep (yes/no), Crying (yes/no)' [59
]. If a parent answers 'yes' to any of these, they will rate the severity of the problem on a seven point Likert scale from 1 = 'hardly any problem' to 7 = 'a severe problem' [43
Parent A will also complete an infant behaviour diary over a 72 hour period, during which they will record whether their infant is sleeping, feeding, awake and content or awake and crying/fussing in 10-minute epochs. This diary has been adapted from the Barr diary which measures such behaviours in 5 minute epochs and has sound psychometric properties [60
Parental sleep quality and quantity
Two items have been adapted from the validated Pittsburgh Sleep Quality Index (PSQI) [61
] to measure parent perception of sleep quantity and quality: (1) 'Over the last two weeks, how would you rate your own sleep quantity?' with responses of 'Not nearly enough', 'Not quite enough', 'Enough' or 'More than enough' and (2) 'Over the last two weeks, how would you rate your own sleep quality?' with responses 'Not nearly good enough', 'Not quite good enough', 'Good enough' or 'More than good enough'.
Parents will also report how many times and for how long on average they attended their infant for night waking over the past week.
The Edinburgh Postnatal Depression Scale (EPDS) [62
] is a validated screen of postnatal depression (PND) [63
] consisting of 10 items. Clinically significant levels of PND are indicated by scores ≥ 10 and ≥ 9 for mothers and fathers in community samples, respectively.
Parental cognitions around infant sleep
The 20-item Maternal Cognitions about Infant Sleep Questionnaire (MCISQ) [25
] has 5 subscales including: limit setting (ability to resist infant demands), anger (anger, regret and helplessness), doubt (uncertainty regarding ability as a parent), feeding (belief about the importance of feeding to settle and concerns about child going hungry) and safety (concerns about cot death). We have not included the 'feeding' subscale as it is not associated with infant sleep problems [25
Parental perception of infant temperament
Parents will rate their infant's temperament on a six point scale: 'Compared to other babies, I think my baby is:' Much easier than average, Easier than average, Average, More difficult than average, Much more difficult than average or cannot say. This single item from the Australian Temperament Project-a longitudinal study of 2000 children-has a moderate correlation with the Australian version of the Revised Infant Temperament Questionnaire [65
Sources of alternate help for infant sleep/crying
Parents will be asked if they have received help or advice (professional or otherwise) from outside the program for either their infant's sleep/crying or for their own stress, and if so, how many appointments they attended for each.
Parent A will indicate the type of food being offered to their infant at present (breast milk, formula or a combination of these) [66
], and if applicable, whether breast feeding has stopped and why, whether their infant's formula has changed at any stage and whether the mother has changed her own diet while breastfeeding. The 6 month questionnaire also asks about timing of introduction and type of solids.
Caregiver support and self-efficacy
At baseline and follow up, parent A will rate the level of support or help they receive from their partner and family and friends living elsewhere (I get enough help, I don't get enough help, I don't get any help, I don't need any help), and how often they feel they need support or help but can't get it from anyone (very often, often, sometimes, never, I don't need it) (LSAC) [68
]. Parent A will rate their efficacy as a parent (from 1 = not very good to 5 = a very good parent) [68
Parental use of intervention materials
Intervention parents will be asked which components of the intervention they received or participated in and usefulness of each, on a study-designed scale ('not at all (useful), a little, quite a bit, a great deal'). Intervention parents will also be asked to rate the helpfulness of specific strategies provided in the intervention materials as either helpful or unhelpful.
Throughout the recruitment phase, MCH nurses approached all families of newborn infants at the first home visit and asked permission to pass on to the research team the contact details of families interested in hearing more about the research. The research team contacted interested families to explain the study further and posted a recruitment pack (information statement, consent form and baseline questionnaire) to parent A and an information pack (information statement and consent form) to parent B if requested. Families were randomised upon receipt of their signed consent form and baseline questionnaire.
Families allocated to the usual care group received a letter explaining that they will continue to see their MCH nurse as usual. Those allocated to the Baby Business intervention group were mailed the booklet and DVD.
A member of the study team completes the telephone consultation with intervention group parents one to two weeks after sending the booklet and DVD. Families are then invited to attend a parenting group session when their infant is around 12 weeks of age. Neither the telephone consultation nor the parent group is compulsory, and parents who do not take part in these components will not be removed from the study given that they will already have received most of the program content via the booklet and DVD. Data will be gathered on participation in each component.
In order to minimize study drop-out, we will call parents if they have not returned their follow up questionnaires two weeks after they were mailed to them. If we do not receive the completed questionnaire in the two weeks following this, we will call the parent again and give the parent the option of completing a shortened version of the questionnaire (covering only the main outcome measures of whether infant sleep, crying and feeding are a problem) over the phone.
We hypothesise that intervention parents compared to control group parents will, at infant age four and six months, report:
• fewer infant night time sleep problems (primary outcome)
• fewer infant day time sleep problems
• fewer crying problems
• decreased mean infant crying duration and increased mean infant sleep duration/24 hours
• improved parent sleep quality and quantity,
• improved mental health with lower mean EPDS scores and lower proportions of mothers scoring > 10 and fathers scoring > 9 on the EPDS
• less difficult infant temperament,
• fewer visits to healthcare professionals for their infant's sleep and crying and their own wellbeing
• higher rates of breastfeeding, fewer formula changes and fewer dietary changes in breastfeeding mothers
Outcome data at 4 and 6 months will be presented using descriptive statistics. Means and standard deviations will be given for continuous outcomes, as well as medians and inter-quartile ranges where continuous data are skewed. Proportions for categorical data will also be given. The primary outcome comparison will be of the proportion of infants with sleep and cry problems at 4 months between the two trial arms. Logistic regression adjusting for potential confounders identified a priori, and measured at baseline (including child gender and family socioeconomic status), will be used to estimate the treatment effect as an odds ratio and 95% confidence interval. Random effects regression models [69
] will be used for further longitudinal analysis examining trends in treatment response, that is, persistence of sleep and cry problems from baseline to 4 and 6 months post intervention. Similar analysis will be carried out for each of the categorical outcomes. We will also compare mean scores for continuous outcomes at the primary endpoint of 4 months (e.g. sleep duration/24 hours) between the two trial arms using t tests, as well as linear regression adjusting for potential confounders. The study sample size is sufficient to enable the use of such techniques when the outcome data are skewed [71
], and empirical bootstrap estimates will be examined to confirm the validity of the inferences made. Trends in treatment response will again be examined from baseline to 4 and 6 months post intervention using random effects regression models. All analyses will be conducted on the basis of intention to treat. The frequency and patterns of missing data will be examined and sensitivity analyses will be performed comparing the results of analyses restricted to families with complete data and analyses where missing data are imputed using a conservative approach [72
A cost-consequences analysis will be conducted from a societal perspective. Costs and outcomes will be taken into account and valued in the analysis regardless of who bears the costs, who benefits or who provides the resources. The incremental costs of the intervention (the difference of costs accrued in the intervention group and costs accrued in the control group) will be compared to a range of the incremental primary and secondary outcomes. Both costs of delivering the intervention and costs of families' use of health and other services outside of the study will be considered in economics costing.
The economic evaluation will draw a comprehensive picture of the costs and consequences of the intervention and assist policy makers to make appropriate decisions on resource allocation to such interventions and determine the cost-effectiveness of nationwide roll-out.
Ethical approval has been obtained from the Royal Children's Hospital Human Research Ethics Committee (HREC 28130) and the Department of Education and Early Childhood Development, Early Childhood Research Committee.