Satisfaction with caregivers during labour among low risk women in the Netherlands: the association with planned place of birth and transfer of care during labour

The Deliver study is a multicenter prospective cohort study into the quality, organisation and accessibility of midwifery care in the Netherlands, which was described extensively elsewhere [15].

Briefly, the means of recruitment of clients was through midwifery practices. Purposive sampling was used to select practices, using three stratification criteria: region (north, centre, south), level of urbanisation (urban or rural area), and practice type (dual or group practice). Twenty of the 519 midwifery practices across the Netherlands participated in this study. Between September 2009 and December 2010 client data were collected using questionnaires. Clients who received antenatal care and who gave informed consent, were given up to three questionnaires: one before 34 weeks gestation (the 1st questionnaire), one between 34 weeks gestation and birth (the 2nd questionnaire), one approximately 6 weeks postpartum (the 3rd questionnaire). The response rate of the DELIVER study was 62%.

The Deliver client data were linked to primary midwife-led care data from the Netherlands Perinatal Register (PRN, “Landelijke Verloskundige Registratie”, LVR1). Linkage was successful in 76.1% of the women included in this study.

For the women with successful linkage, agreement between LVR1 and Deliver data for women who started labour in midwife-led care was 99.1% for vacuum or forceps extraction, 99.9% for caesarean section and 99.4 (hospital) to 94.7% (home) for actual place of birth. In case of disagreement, we used data from the DELIVER study.

For this study, participants with singleton term pregnancies that were in midwifery care at the onset of labour were selected. The definition for onset of labour in primary care is based on information from the LVR1 database. This information showed internal inconsistencies in 3.7% of the cases. Women who had care transferred for prolonged rupture of membranes (>24 h without contractions) were excluded. Among these women, transfer to secondary care occurred before start of the dilation (first) stage, and thus planned place of birth is unlikely to have affected satisfaction with the caregiver. Women who were transferred to secondary care during pregnancy and women who were advised to give birth in hospital in midwife led care because of a condition that would increase the risk of complications for the woman or baby were also excluded.

Planned place of birth (home or hospital under midwife-led care) is recorded on the LVR-1 form at some point during pregnancy. This information is missing for some women; midwives may forget to record the details or the women may not have made a decision on where to give birth until the onset of labour. The number of women with unknown planned place of birth in this study will be reported in the Results section.

When complications arise such as listed in the VIL [13], care is transferred from midwife-led to obstetrician-led care. When a woman is at home, this requires transport to a hospital facility prior to transfer of care, either by car, or if transport by car is too inconvenient or in case of an emergency, by ambulance. In this study, both transfer during labour or immediately postpartum, were defined as transfer of care.

From information on planned and actual place of birth two groups were formed of women who were transferred, according to their planned place of birth. Three groups were formed for women who gave birth in primary care (and thus were not transferred), see Table 1:

In the analyses for the primary research question planned hospital birth (including hosp-transferred and hosp-hosp) was the comparison group, because the hypothesis was that quality of treatment and general satisfaction with the caregiver is negatively influenced by transportation from home to hospital when transfer of care is necessary.

Satisfaction with the caregiver during labour was assessed with the 3rd postpartum DELIVER client questionnaire, which was filled in 6 weeks postpartum on average (see Additional file 1).

Satisfaction of care by caregiver was assessed using 1: general and 2: evaluative questions which both were adapted from the Consumer Quality Index (CQ-index) for maternity care. This instrument was developed to measure client experiences. The questions were identified and validated in item- and factor analyses [8, 16]. The cut-off values for both outcomes were chosen based on the fact that self-reported satisfaction with care is usually very high [17], particularly in relation to maternity care [18]. We aimed to identify whether a planned home or hospital birth is associated to rates that indicate anything less than very good care.

The cut-off of 9 was chosen, based on a previous study showing that ‘general satisfaction with caregiver’ during labour and birth was rated 9 or 10 in more than half of the women. The cut-off for ‘quality of treatment’ was set on 4 vs lower than 4 (1, 2, 3), based on the high mean score of this variable in the previous study (e.g. 3.61 to 3.92) [8].

Maternal age, ethnic background (Dutch, western background or non-western background) and social status (score 1–4) were taken into account, because of their association with planned place of birth [19, 20] and satisfaction with childbirth [21]. For social status we used the score developed by the Netherlands Institute for Social Research (SCP), using postal codes, based on education, income and employment rates. A low score equals high social status (www.​scp.​nl/​Onderzoek/​Lopend_​onderzoek/​A_​Z_​alle_​lopende_​onderzoeken/​Statusscores).

Results were presented separately by parity, because the association between planned place of birth and satisfaction with the caregiver might differ for primi- and multiparous women. Once a woman has given birth, her choice about place of birth is influenced by her previous experience [22], and multiparous women are more satisfied with the birth experience and thus might rate satisfaction with care differently [21].

Some factors were assessed to explore if they could explain associations between planned place of birth and satisfaction with the caregiver. Some of these factors are only applicable to women in obstetrician-led care. However, both among the women who planned home birth as well as among the women who planned hospital birth in midwife-led care, transfer during labour to obstetrician-led care occurred in some cases, and consequently some women who planned home or hospital birth in midwife-led care received medical pain relief during labour for example.

Medical pain relief (yes-no) and experience of labour pain (rating scale from 1 to 10) were taken into account, because they have been reported to influence satisfaction with childbirth and consequently might influence satisfaction with the caregiver [21]. Management and experience of labour pain is different for women planning hospital birth, compared to women planning a home birth. The role of medical interventions, including augmentation, vaginal instrumental delivery and caesarean section was investigated [21]. Finally, in an explanatory analysis the extent to which transfer of care (yes/no) could explain possible differences in the association between planned place of birth and satisfaction with the caregiver was assessed.

Baseline and pregnancy related characteristics and labour outcomes were compared between low risk women who planned to give birth at home versus women who planned to give birth in hospital using student’s t-test for continuous and chi-square test for categorical characteristics. Mean scores of ‘general satisfaction with the caregiver’ and ‘quality of treatment’ were calculated for planned home and hospital birth, to calculate the minimal important difference (MID = 0.2 * SD_hospital) [23, 24]. It is a score that reflect differences in a client derived outcome of an intervention, that is meaningful to the client. The association between planned place of birth (home/ hospital) and satisfaction with the caregiver during labour (defined as 1: ‘general rating of satisfaction with the caregiver’ and, 2: the ‘quality of treatment by the caregiver’) was analysed using multilevel logistic regression analysis with two levels; the midwifery practice level and individual level, in order to account for clustering of women within midwifery practices. In similar models the association between planned place of birth and the 7 separate evaluative questions for ‘quality of treatment by the caregiver’ were analysed, as well as the association between planned place of birth and transfer (home-transfer, hosp-transfer) or not (home-home, hosp-home, hosp-hosp) with ‘general satisfaction with the caregiver’ and ‘quality of treatment’. To examine the effect of transfer in a planned home birth on satisfaction with the caregiver, as well as the overall effect of satisfaction in a planned home birth compared to a planned hospital birth, planned hospital was used as the reference.

Odds ratios and 95% confidence intervals were presented. Dummy variables were constructed in case of >2 categories. Adjustments were made for confounders using multivariable models. Additionally, potential explanatory factors were added separately in the models. Finally, a sensitivity analysis was performed including women with and without discrepancies in the definition for start of labour in primary care.

Table 3 shows that for primiparous women during labour both the 1: ‘general rate of satisfaction’ with the caregiver, and 2: ‘quality of treatment’, were significantly more often high among women planning a home birth compared to women planning a hospital birth, taking into account confounders and clustering of women within midwifery practices. For multiparous women, 2: ‘quality of treatment’ was statistically significantly more often high (equal to 4) among women who planned a home birth.

The separate items of the measure for 2: ‘quality of treatment by the caregiver’ in Table 4 shows that primi- and multiparous women who planned a home birth significantly more often had the feeling that they were ‘in safe hands with the caregiver’ and that the caregiver ‘took them seriously’, compared to women who planned a hospital birth. In addition, nulliparous women who planned a home birth significantly more often stated that ‘the caregiver explained things in an understandable way’. Multiparous women who planned a home birth significantly more often experienced that ‘the caregiver listened to them carefully’.

Compared to women who planned a hospital birth and who were transferred during labour as the reference group (hosp-transfer), ‘general rating of satisfaction with care’ and ‘quality of treatment by the caregiver’ during labour was significantly more often high for primiparous women who planned a home birth and who were transferred (home-transfer; Table 5). No differences were observed in any measure of satisfaction with care for multiparous women who planned a home or hospital birth and who were transferred (Table 5; transfer).

Table 5, no transfer, shows that multiparous women who planned a hospital birth but who actually gave birth at home (hosp-home) significantly more often gave a higher rate for ‘general satisfaction with the caregiver’ than multiparous women who planned a hospital birth and who actually gave birth in hospital in midwife-led care (hosp-hosp). Multiparous women who planned and had a birth at home (home-home) significantly more often rated ‘quality of treatment of care’ high compared to women who planned a hospital birth and who actually gave birth in hospital in midwife-led care (hosp-hosp).

The group who planned a home birth but actually gave birth in hospital in midwife-led care was too small for analysis.

Sensitivity analysis of 2286 low-risk women (the eligible study population including 85 women with discrepancies in the definition of start of labour in midwife-led care and known planned place of birth) for the association of planned place of birth and 1: ‘general rating of satisfaction with the caregiver’ and 2: ‘quality of treatment of care’, showed similar results.

Primiparous women in midwife-led care at the start of labour who planned a home birth were significantly more satisfied with the care they received from the caregiver during labour than women who planned a hospital birth. They rated ‘quality of treatment provided by the caregiver’ significantly higher when they had care transferred compared to primiparous women who planned a hospital birth and who had care transferred. Multiparous women who planned a home birth were significantly more satisfied with ‘quality of treatment’ and when they had care transferred they were equally satisfied with the care provided by the caregiver compared to multiparous women who planned a hospital birth and who had care transferred during labour.

For this study, we used data from a prospective cohort study. Although a randomised controlled trial would be preferred for planned place of birth this was shown not to be feasible [25]. Therefore, due to the observational nature of this study, we adjusted the association for confounders. In addition, clustering of women within midwifery practices was taken into account. Data were not complete for all eligible women, including data on planned place of birth. Some women do not choose their place of birth until they are in labour and in some cases the midwife might have forgotten to register this information. ‘Quality of treatment’ and ‘general rating of care’ by caregiver was not available for all women. Furthermore, women in the DELIVER study were higher educated and less often of non-Dutch origin than the general Dutch population [15]. In addition, the response rate for participation was only 62%. It seems, however, unlikely that among non-responders, the association between planned place of birth and quality of care would be in the opposite direction. In this study we used the CQ-index. This instrument has been validated among Dutch pregnant women [8]. A reliability analysis of the 7-items measuring ‘quality of treatment by the caregiver’ revealed a high internal consistency as well. However, measuring satisfaction or quality of treatment by the caregiver is a methodological challenge, and therefore we have interpreted the findings in the scope of clinical relevance in addition to statistical significance as described below.

To our knowledge, our study is the first to report on the association between planned place of birth and satisfaction with caregiver among women who were transferred during labour. It was shown previously that women who (planned to or actually) give birth at home have higher scores on ‘general satisfaction’ or ‘quality of care’ received from the caregiver [8, 26]. Furthermore it has been reported that women who planned a home birth were less satisfied with the caregiver and with their birth experience when they were transferred from home to hospital, compared to those who had a home birth and were not transferred [27, 28]. These studies, however, did not compare women who had transfer of care during labour and who planned birth at home versus in hospital. Another study found that transfer of care affected evaluation of the midwife similarly among women with a planned home birth compared to those with a planned hospital birth [22]. In our study primiparous women who planned a home birth and who were transferred during labour were more satisfied with the caregiver compared to primiparous women who planned a hospital birth and who were transferred. In multiparous women the results were in the same direction, although not statistically significant.

It is remarkable that our findings are not consistent with the assumption that transport and not meeting expectation would negatively influence satisfaction with caregivers among women who planned home birth [14]. Notably, women who plan hospital birth need to be transported during labour at some point as well. Furthermore, women who opt for a home birth have fundamental trust in their independent ability to give birth and have a trust in the birth process in general [20, 29, 30]. Possibly, this difference in motives might cause women to rate satisfaction with the caregiver differently, e.g. ‘feeling in safe hands with the caregiver’. Medical pain relief resulted in fewer women rating satisfaction with the caregiver highly among primiparous women. This partly explained a higher rating of satisfaction among women who plan home birth, since they less often receive medical pain relief. Possibly women who receive pain medication have had a longer and more exhausting labour which might affect satisfaction with caregiver.

Nevertheless, planned place of birth itself may also have some effect on satisfaction with the birth experience. Satisfaction with the caregiver differed between two different planned birth environments (home or hospital in midwife-led care), even after adjustment for confounders. It may be that midwives behave differently when they provide care in hospital than in a woman’s home. Another study showed an association between environment (e.g. ward design) and behaviour of care providers [31]. In our study, primi- and multiparous women more often stated that the caregiver took them seriously, or listened to them carefully (only multiparous women), or explained things in an understandable way (only primiparous women). This confirms reports that showed that environmental changes can influence the quality of interaction with the patient [31, 32]. Hospital rooms that were associated with improved patient outcomes were those that offered privacy, promoted social support and were calming [33]. It seems likely that the home environment satisfies these aspects more often than a hospital environment.

The cut-off point for satisfaction differentiated between very satisfied versus less than very satisfied. The majority of women were satisfied with the caregiver most of the time. Overall, the mean difference in both ratings of ‘general satisfaction’ and ‘quality of treatment’ of women who planned a home birth compared to women who planned a hospital birth was smaller than the minimally important difference for clinical relevance [23, 24].

Apart from health outcomes, experiences of women have become increasingly important aspects of good quality care. Previous studies in the Netherlands found no association between planned home birth and severe adverse maternal and neonatal health outcomes [19, 34, 35]. Recently, we showed that feelings of control were similar among women with planned home birth and who were transferred to the hospital, compared to women with planned hospital birth who were transferred [36]. Therefore, it seems justified that women in the Netherlands are offered a choice for place of birth, either at home or hospital.