Overview data collection is presented in Figure . Baseline characteristics of the populations are described in Table . The respondents in the cross-sectional studies were representative for the pregnant population in their area during the study period in regard to age, parity and smoking habits (data from the Medical Birth Registry Norway, not shown).
Figure 1 Trial profile. Trial profile of total births, reports of decreased fetal movements and the cross-sectional population pre- and post intervention. *Counting group: women reporting using a kick chart more than once per week. Yellow boxes: The cross-sectional (more ...)
Descriptive characteristics: DFM and Cross-sectional populations
Information and maternal awareness of fetal activity
Data from the cross-sectional studies showed that one in four women did not recall receiving information about normal expected fetal activity by their health provider, both pre- and post-intervention. Recall of receiving information was associated with higher awareness of fetal activity, both pre-intervention (OR 2.0, 95% CI 1.2-3.3) and post-intervention (OR 1.8, 95% CI 1.0-3.1, p = 0.043). Pre-intervention, recall of receiving information was associated with more frequent maternal concern (OR 1.7, 95% CI 1.2-2.4); while this association was not longer present post-intervention (OR 1.3, 95% CI 0.9-1.9).
Maternal recall of information about limits for normality was more homogeneous in the intervention period, e.g. 22% recalled having seen the thumb rule (10 kicks in two hours) at baseline measurement, versus 42% in the intervention period (p = 0.022). Pre-intervention, low maternal awareness to fetal activity was associated with an increased risk of having an SGA baby; [10
] this association was not observed in the post-intervention period (OR 1.3, 95% CI 0.6-2.9).
Maternal behavior and pregnancy outcomes
Among women with DFM, the stillbirth rate was lower in post-intervention period; 4.2% versus 2.4% (Tveit et al, submitted 2009). The reduction in stillbirth was isolated to primiparous women only. Primiparous women also reported DFM earlier than all other women included (Table ). In the total population, the mean gestational age at the time of reporting DFM was two days lower during the post-intervention period; 366 versus 364 weeks, p = 0.006.
DFM population: Effects of intervention on maternal behavior and stillbirth rates, stratified by subgroups
In the post-intervention group, overweight women in the cross-sectional populations described higher awareness of fetal activity (Table ). No behavior changes were observed among overweight women if they perceived DFM (Table ).
Cross-sectional population: Low maternal awareness of fetal activity and maternal characteristics (N = 1431)*
Pre-intervention, smoking mothers in the cross-sectional population recalled less receipt of information about fetal activity than non-smokers, OR 0.5 (95% CI 0.3-0.9). This association was not present in post-intervention, OR 0.6 (95% CI 0.3-1.2). No changes in maternal behavior were observed among smoking women perceiving DFM (Table ).
Non-Western women in the cross-sectional study post-intervention, remained the only risk group reporting both less receipt of information (adjusted OR 0.4, 95% CI 0.2-0.8) and low awareness of fetal activity (Table ). Among the non-Western women who perceived DFM, the intervention showed no changes in maternal behavior, frequency of concerns or outcomes (Table ).
The hospital-specific percentage of women reporting having received written information (proxy for distribution) was negatively associated with mortality rates - the more information, the lower mortality (β = 0.974, p = 0.031). This was done to assess the effect of the distribution of information and maternal internalization of it on the number of stillbirths.
Maternal concerns - as reported by women in the cross-sectional studies
Mothers in the post-intervention period did not report concerns or have a DFM consultation more frequently (Table ). Overweight women were the only subgroup reporting increased concerns; however, this was not significant after Bonferroni correction (Table ). When concerned, the mothers more often related their concern to the fetal activity level earlier in the actual pregnancy (44% vs. 51%, p = 0.011). More often, the concerned mothers assessed their perception of DFM not being normal for their baby and that their concern was a true reason for being concerned (28% vs. 33%, p = 0.022). Being concerned was associated with being examined at hospital both pre-intervention (OR 4.9, 95% CI 3.0-7.8) and post-intervention (OR 5.8, 95% CI 3.7-9.2).
Cross-sectional population: Effects of intervention on maternal awareness, concern and maternal behavior (N = 1431)
Fetal movement counting in the intervention group
In the post-intervention group, 235 (32%) reported using a kick chart, as opposed to 8 (1%) pre-intervention. Post-intervention, 64 (9%) of women used a kick chart more than once per week (counting group); versus 8 (1%) pre-intervention. Primiparous women were more likely than multiparous women to use a kick chart more than once per week (OR 2.3, 95% CI 1.3-4.2). No non-Western mothers used a kick chart.
Maternal experiences with use of a kick chart in the intervention period are presented in Table , illustrating the benefits of maternal receipt of receiving information on how and why to use the kick chart. The use of a kick chart was not associated with increased maternal concerns about DFM (32% in the non-counting group vs. 42% in the counting group, p = 0.090). Use of a kick chart was associated with a reduced risk of having a DFM consultation, 18% vs. 9% (p = 0.045). One of ten babies was SGA in both groups. Eleven (7%) of the non-counting group had an emergency caesarean section, as opposed to one (2%) in the counting group (p = 0.047).
Cross-sectional population post-intervention: Experiences with use of a kick chart (N = 235)*
The hospital-specific percentage of women reporting having used the kick chart more than once per week or more (proxy for internalization) was negatively associated with mortality (β = 0.922, p = 0.005). This does not reflect the effect of kick counting on an individual level, as there are no data to support this, only the benefit of effective information.