Characteristics of the included studies – design, setting and demographics
A total of nine primary research studies met the inclusion criteria: one reporting antenatal intervention only , six reporting postnatal interventions only [28, 31, 34-37] and two reporting combined antenatal and postnatal interventions [27, 38] (Additional File 1). Our review included five randomized controlled trials (RCTs) [27, 28, 34-36] (two of them being cluster RCTs) [27, 28], three quasi-experimental studies with a control or comparison group [31, 37, 38] and one retrospective chart review . Studies were conducted in five different countries: two in Canada [31, 36], three in the United States [29, 35, 38], two in Brazil [27, 34], one in Ireland , and one in Iran .
Overall, demographics of the participants (both mothers and children) were poorly described in the studies. The mean age of the mothers was only mentioned in two studies and ranged from 25.7 years  to 26.4 years  at delivery. Five of the studies provided information about race/ethnicity, which included Latina , white , Punjabi-speaking South Asian [36, 39] and Vietnamese women . Only four provided information on education level [27, 28, 37, 38] and five on socioeconomic status of participants [27-29, 37, 38]. Mostly, studies included participants residing in areas of high social deprivation, or from high-risk, impoverished, and socioeconomic-challenged communities. The mean age of participating children were only described in four studies and ranged from 11 months  to 28 months , although the follow-up period ranged from 0 to 7 years. The studies included diverse samples of participants and none of the samples had similar characteristics.
Methodological quality of the papers
Overall the aims, design, population and settings, intervention and data collection methods were poorly described (Additional Files 1 and 3) with a total of six studies being classified as weak.
Type of non-dental health professionals who provided the intervention
Interventions were provided by different groups of non-dental professionals including health counsellors – local South Asian lay women [36, 39] and a lay Vietnamese woman  – community based nurses , midwives , healthcare workers – physicians, nurses and administrative staff  – field workers , general vaccination health staff , outreach coordinator – a health department employee  – and multidisciplinary team formed by nurses, obstetricians, social workers, nutritionists, oral and maxillofacial surgeons and support staff . The majority of studies described these professionals receiving trainings/introductory workshops; nonetheless, only five provided details of the trainer’s background, which included nutritionists [27, 34] and dentists [28, 36, 39].
The intervention methods varied across studies and included (i) oral health education, (ii) oral health assessment/screening and (iii) referrals of participants to dental services. Provision of oral health education by a non-dental health professional was the focus of all the studies. Three also included referrals for dental care by non-dental health professionals – obstetricians , community-based nurses  and outreach coordinator (OHSC) who was a health department employee  – and one involving dental screening initiated by a multidisciplinary team .
Oral health education encompassed verbal oral health advice and information such as discussing that dental care is safe during pregnancy [28, 29], one-to-one counselling sessions [27, 31, 38], motivational interviewing [36, 39], home visits for one-to-one preventive advice [34, 37, 38], as well as follow-up phone calls to provide support and to coach mothers , to reinforce and maintain behavioural changes [36, 39], and as reminders of oral health instructions  and child’s dental appointment . It also involved written oral health promotion materials such as postcards [36, 39], pamphlets [27, 28, 36, 39], brochures , leaflets , posters displayed in clinics , letters to remind parents about a child’s dental appointment  and a toolkit containing educational material . Some visual tools were also used in interventions and consisted of educational videos [36, 39] and DVDs . Finally, in some interventions, community-wide initiatives (video, written information, window displays and brochures) were also employed  as well as the distribution of ‘goody bags’ that included items such as an adult or infant toothbrush, toothpaste, training cups, finger cots and table mats [31, 35, 37].
The screening/assessment component of the intervention described by Larsen et al.  included (but was not limited to) prompting pregnant women regarding ‘current oral health problems, previous dental problems, and the availability of a dental provider’. Those identified as having a ‘current oral health problem’ or ‘not having a dental visit in the past six months’ were referred to a dentist, preferably before 20 weeks gestation. The other two studies that also had a referral component involved distribution of dental registration vouchers by nurses and referral by the OHSC during home visits.
The exact point in time during the antenatal period, in which the intervention was provided, was not specified by the authors. The postnatal interventions occurred sometime between immediate (0-5 days) and 24 months post-partum[36, 39]. Interventions that took place in both antenatal and postnatal periods occurred from second trimester of pregnancy to within 2 months of delivery.
Eight out of the nine studies measured clinical outcomes in children (i.e. dental health status) using proxies such as presence of dental caries/decayed teeth/cavitation[27, 29, 35], extractions, decayed surfaces [28, 31, 34, 36, 38], enamel caries , and child’s caries risk. In addition to clinical outcomes, six assessed mother’s behaviours including service uptake, beliefs, and dietary, hygiene and parenting practices [27, 29, 31, 34-36]; one children’s dietary behaviour ; one mother’s knowledge/awareness of oral health; one mother’s experiences with the intervention and one assessed mother’s perceptions of the effectiveness of the intervention. One study assessed only behavioural outcome through service uptake by preschool children.
Using a retrospective chart review, Larsen et al  investigated the efficacy of an antenatal intervention involving oral health education, referral and screening delivered by a multidisciplinary team (involving obstetricians, nurses, social workers, a nutritionist, oral and maxillofacial surgeons, dentists and support staff), in addition to dental evaluation and consultation by dental professionals to pregnant women. The authors reported that children of mothers who received the intervention had significant clinical outcomes such as less dental caries (p=0.019), fewer extractions (p<0.021) and number of teeth with caries at 2-3 years of age (p<0.001) compared with children of mothers who did not participate in the intervention. Oral health service uptake was also increased overtime following the intervention suggesting its effectiveness in improving oral health of young children.
Interventions conducted in the postnatal period also showed meaningful improvements in children’s clinical and mother’s behavioural outcomes. Clinical improvements went from fewer decayed surfaces (p=0.03) and lower enamel caries (de) increment (p<0.05) in the short-term (up to 1 year after the intervention) to fewer decayed surfaces measured as defs (p<0.005) and fewer new carious lesions (p<0.01;  and p<0.02) ; sustained over the 1-year  and follow-up periods , confirming the protective effect of oral health education interventions. Due to the high rates of no-show in follow-up assessments, clinical evaluation of the oral health education (using DVDs) described by Hallas et al.  was compromised. Baseline data however, highlighted mother’s lack of oral health knowledge, particularly awareness of vertical transmission of S. mutans during immediate postpartum.
Behaviourally, changes in mother’s parenting practices such as less use of sleep-time and daytime bottles (p<0.005), dietary practices including duration of exclusive breastfeeding (p=0.000) and introduction of sugar (sugar cane and honey in fruits, milk and porridge etc) (p=0.005) were also significant and contributed to improved oral health outcomes in children. Despite these positive outcomes, the results reported by Feldens et al. were short-term only. Service uptake was also significantly increased for 0-2 year old children, 5 months after a combination of oral health education and referral to dental services intervention but showed no equivalent effect for the 3-5 year old group. Conversely, the oral health education only intervention conducted by Weinstein et al. [36, 39] showed no difference in service uptake between intervention and control groups after 1-year follow-up (around 1-2 years of age).
Both antenatal and postnatal periods
Interventions offered in both pregnancy periods and postnatal had mixed results. The combined oral health education and referral intervention described by Milgrom et al. showed significant clinical outcomes with reduction in the mean number of teeth with decay (p=0.04) in children up to 2 years living in rural areas. Nonetheless, this finding was primarily attributed to the dental care component of the intervention with the authors acknowledging that non-dental health professionals played a minor role in referring and providing education at home visits. In addition, the study by Chaffee et al.  provided oral health education as an intervention during both periods, showed no significant reduction when compared with the control groups. Feldens et al.’s  study was similar to the intervention by Chaffee et al.  which focussed on nutrition. Physicians and nurses were trained in infant complementary feeding by a nutritionist to incorporate into maternal consultations . However, the weakness of Chaffee et al.’s study was that the number of times women received counselling was not monitored and the accuracy and consistency of the messages relayed to mothers were unknown .