gms | German Medical Science

Measurement of health care quality in Germany is still largely based on indicators of medical and technical quality as well as financial performance. They are the main drivers for designing health care [28]. In many other countries, patient experience has long since been established as an indicator of care quality and is increasingly used in performance management systems alongside other indicators [46]. However, disregarding the importance of patient experience means ignoring the available evidence that recognizes patient experience as one of the three pillars of quality in health care alongside clinical effectiveness and patient safety [20].

The importance of patient experience may be even more pronounced in maternity care, since the birth of a child is not ‘just another day in a woman’s life’ [41]. Pregnancy and childbirth are major life events which may have long-lasting effects on women’s physical and mental health as well as their relationship towards their child [27]. A positive birth and care experience contributes to feelings of self-esteem and confidence in dealing with the new situation [16] and a better quality of life [57]. Women with a negative birth experience are more likely to feel insecure, anxious or even suffer from depression [5], [56]. This may in turn enhance the risk for impaired maternal-infant interaction and may negatively affect the child’s development [18], [39]. A negative or even traumatic birth experience has also been linked to post-traumatic stress disorder [1] as well as depression during [33] and after a future pregnancy [32]. Furthermore, the quality of the birth and care experience may influence the woman’s willingness to have another baby [29], [53], whether she fears giving birth to her next child [30] and her choice of mode of birth [31], [37].

The aim of this study is to explore the factors influencing the overall intrapartum care experience in the labour ward as well as the overall experience of postnatal care in the maternity ward.

This study is a secondary analysis of survey data collected for internal quality management purposes of the respective hospitals.

The questionnaire used for data collection is the validated translation of an existing instrument [44]. The translated version contains 101 items asking about experiences and reports during intra- and postpartum care in the hospital which may point to problems within the care process. The questions refer to the following themes: admission to the labour ward, midwives in the labour ward, doctors, interventions and care in the labour ward, care in the maternity ward, staff in the maternity ward, discharge management, facilities & service, overall impression and the first days at home. In addition, the questionnaire contains items about demographic characteristics of mother and child, the uptake of ante- and postnatal services, obstetric history, mode of birth and self-reported health (SF-1). The response options vary according to the type and subject of each item. The number of response options ranges from two to eight (mean=4). Respondents were asked to choose the most suitable answer option per question. Most of the questions ask respondents to report on specific care events and experiences. The risk of ceiling effects that are usually seen when patients are asked to rate their care, are less pronounced with this type of question, making them more useful for quality improvement purposes [14]. They are also less susceptible to the effects of expectations and response tendencies such as social desirability and acquiescence bias [26], [49] and problems can be more reliably revealed [10]. Furthermore, the variability of the answers is higher than with questions asking about satisfaction with care [2], [36]. However, since evaluation of certain processes and situations was explicitly desired, the questionnaire also contains a number of rating questions.

The original data was collected via postal surveys conducted after discharge from hospital. Women discharged from the respective hospital within eight to twelve weeks after a live birth, aged 18 years and older and who stayed at least two nights in the hospital were eligible for the survey. To maximise the response rate, two reminders were sent out to those who had not responded within two and four weeks after the initial mailing [19]. Each questionnaire had a unique coding that allowed the questionnaire to be assigned to the respective hospital and the woman’s home address in order to enable a mailing of the reminder to only those who had not yet returned their questionnaire. The list allowing the allocation of questionnaire code and address remained at the hospital and was destroyed after the second reminder was sent. The completed questionnaires were sent in opaque, prepaid envelopes to the institute responsible for data collection and analysis. In this way it could be ensured that completed questionnaires and code list were physically separated. The questionnaires were destroyed after completion of data entry.

The cover letter explicitly explained that participation in the survey was voluntary and data would be kept confidential. At the time of data collection hospitals were still allowed to use patient-specific data for internal quality management purposes without explicit consent of the patients. By signing the treatment contract at admission patients agreed to the use of their data for these purposes. Given that the mailing was conducted via the hospital, the completed questionnaire was sent directly to the institute responsible for data collection and analysis, and that both cover letter and questionnaire clearly pointed out that participation in the survey was voluntary, obtaining consent from the participants was not necessary. In this situation returning the questionnaire can be considered as consent.

The association between each explanatory variable and the intra- and postnatal care experience respectively was tested separately using a Chi-square-test. The effect size was calculated using Cramer’s V. Relative risks were calculated as the ratio between the percentages in each group and results were considered to be statistically significant at the 1% level.

To investigate the effect of possible predictors on care experience, logistic regression models (simultaneous model) were calculated using only those explanatory variables as independent variables that were significant in the bivariate analyses, and overall intra- and postnatal care experience as dependent variables respectively. For analysis purposes the outcome variables were dichotomized into women with a positive experience (excellent + very good + good) and those with a negative experience (fair + poor). Nagelkerke’s R² was calculated for each model to assess the proportion of explained variance. To investigate the association between predictors and outcome variables odds ratios were calculated. Results were considered significant at the 1% level. Statistical analyses were performed using IBM SPSS Statistics 22.

The data from a total of 9,597 women who participated in 121 patient surveys in 77 different hospitals conducted between 2015 and 2017 were used for analysis. The number of surveys being higher than the number of hospitals is due to the fact that a number of these hospitals conducted more than one patient survey in this three-year period.

The average response rate was 48% (range: 23% to 86%) which is considered a good rate for postal surveys and a prerequisite for a robust data basis [11].

The sociodemographic and obstetric background of participants as well as hospital characteristics are presented in Table 1 [Tab. 1].

Three quarter (75%) of the women reported an ‘excellent’ or ‘very good’ intrapartum care experience, only a very small percentage (7%) rated their experience as ‘fair’ or ‘poor’ (Figure 1 [Fig. 1]).

The overall postpartum care experience is somewhat less positive. Nearly one in six women reports that all in all, care in the maternity ward was ‘fair’ or ‘poor’ (Figure 2 [Fig. 2]).

Possible predictors of intrapartum care experience were chosen based on existing evidence. The following variables were tested in bivariate analyses for inclusion in the multivariate model. Variables marked with an asterisk (*) were not found to be significantly associated with overall intrapartum care experience in the bivariate analyses and were therefore excluded from the multivariate analysis: maternal age, education*, parity, antenatal hospitalisation*, mode of birth, preterm birth*, hospital size, lead professional in antenatal care*, participating in antenatal classes*, admission to neonatal nursery*, self-reported health, number of midwives during birth, availability of midwife, trust in midwife, involvement in decision-making, worries/fears could be discussed with midwife, adequate pain relief.

With 48% explained variance and 94% of correctly classified cases the model fit is considered good [3]. The multivariate analysis identified only variables as significantly associated with the overall intrapartum care experience that described some form of interaction with the midwife. Women who lacked trust in their midwife had an increased risk of more than five-fold for a negative intrapartum care experience than women who were able to establish a trustful relationship with their midwife (Odds Ratio (OR) 5.67; 99% confidence interval (CI) 3.32-9.69). A negative intrapartum care experience is more likely when the midwife is not available when the woman needed her (OR 4.77, 99% CI 2.25-10.11), when the woman was not involved in decisions about her care and treatment (OR 3.77, 99% CI 2.14-6.64) and when she did not receive adequate pain relief (OR 2.99, 99% CI 1.81-4.94) (Table 2 [Tab. 2]).

Possible predictors of postpartum care experience were also chosen based on existing evidence. The following variables were tested in bivariate analyses for inclusion in the multivariate model. Variables marked with an asterisk (*) were not found to be significantly associated with overall postpartum care experience in the bivariate analyses and were therefore excluded from the multivariate analysis: maternal age, education*, parity, antenatal hospitalisation*, mode of birth, preterm birth*, hospital size, lead professional in antenatal care*, participating in antenatal classes*, admission to neonatal nursery*, self-reported health, time spent with the newborn, intended feeding method, practical help with breast/bottle feeding, consistent advice on breast/bottle feeding, adequate postnatal pain relief, understandable information of test results, understandable answers to important questions, trust in maternity ward staff, availability of maternity ward staff, information on breast/bottle feeding, blood tests, check-ups and vaccination, information on baby care, on warning signs to look out for in mother and/or child and on uptake of daily activities.

With 51% explained variance and 86% of correctly classified cases the model fit is considered very good [3]. As with intrapartum care experience the multivariate analysis identified only variables as significantly associated with the overall postpartum care experience that described some form of interaction with midwives/nurses in the maternity ward.

Central to a positive postpartum care experience are trust in the midwives/nurses and their availability according to the needs of the new mothers. Lacking trust in the maternity ward staff increases the likelihood of a negative care experience by more than four and a half times (OR 4.49, 99% CI 3.05-6.59). Women for whom nurses/midwives were not available when needed were three times more likely to report a negative postpartum care experience compared with women for whom this was not the case (OR 3.13, 99% CI 1.86-5.26). Further important predictors of postpartum care experience are the practical help with breast or bottle feeding (OR 2.88, 99% CI 2.04-4.05) and receiving understandable answers to important questions (OR 2.50, 99% CI 1.76-3.56). Pain relief is not only an issue during labour and birth but also after the child is born. Women with inadequate postpartum pain relief are two times more likely to report a negative care experience than women whose pain was satisfactorily treated (OR 2.2, 99% CI 1.62-2.98) (Table 3 [Tab. 3]).

The analysis is based on survey data from more than 9.500 new mothers. Post-discharge, postal surveys of patient experience (as opposed to patient satisfaction) are considered a reliable method to assess care quality from the patients’ perspective [17]. The average response rate of 48% further substantiates the soundness of the data base [11].

The high proportion of women reporting a positive intrapartum care experience needs to be interpreted with caution. It is a well-known phenomenon that asking patients to rate their overall care experience will produce highly positive results and that problems with care can only be identified by asking them more detailed questions about specific aspects of their care [10], [42], [47]. It is assumed that social desirability, acquiescence bias and reluctance of criticising staff who has taken care of them, etc., are among the reasons for these so called ceiling effects in patient surveys [15]. They are also assumed to play a role in evaluations of maternity care [50]. In addition, in perinatal care the halo effect is thought to influence women’s overall care evaluation: the relief at having come through the experience safely as well as the overwhelming joy over the birth of a healthy baby may compensate for any negative experience either with staff or care processes and create a favourable environment for subsequent evaluations of care [34], [52]. Especially if the birth involved episodes in which the well-being of mother and/or child was depicted or experienced to be at risk, women tend to credit the professionals with any positive outcome [40], [48]. Another aspect that is discussed in this context is denial as the first stage of grief when expectations towards the birth have not been met [23].

Overall evaluation of postnatal care in the maternity ward is somewhat more negative than the overall intrapartum care experience, which is in line with the literature [8], [9], [55]. However, it has been argued that intra- and postnatal care cannot really be compared since the conditions under which care is provided are very different. Intrapartum care deals with an acute and potentially life-threatening situation where evaluations may be overshadowed by the halo effect. The nature of postpartum care is more diffuse, the emotional situation of the new mother is different and memory of exhaustion and the new demands of motherhood may influence overall care evaluation [55]. Nevertheless, both intra- and postpartum care should be woman-centered and caregivers should aim at facilitating the best possible care experience.

The results of this study show that a trusting relationship between woman and midwife is instrumental in facilitating a positive intrapartum care experience. This is in line with the existing evidence [23], [24]. With regard to the current discussion on one-to-one support during labour, it is interesting to note that continuous care by one midwife seems to be less important than one midwife being always available when needed. Similar results were reported by Knape et al. in their study on the association between attendance of midwives and workload of midwives with the mode of birth. They did not find a correlation between a specific number of midwives providing intrapartum care for one woman and mode of birth. However, the number of support measures provided by one midwife during birth and women’s satisfaction with the availability of the midwife were associated with mode birth [25]. This result also suggests that it is not so much the one-to-one care per se that is decisive for a positive care experience but rather the availability of the right person over the right time at the right place.

A sense of control is also an important contributing factor to a positive care experience. This can be achieved for example by adequately informing women about the childbirth process and/or planned interventions and by involving her in decisions about her care [16], [51]. With regard to the latter it is less important that women are actually taking the decisions but rather the way professionals communicate with her [6]. The results of this study show that a lack of involvement is associated with a 3.5-fold increase in risk for a negative care experience.

Labour pain had long been considered as one of the main factors influencing the birth and care experience [54]. Meanwhile studies have shown that although labour pain is an undeniable fact of childbirth and may cloud the experience, intense pain does not preclude an overall positive experience [6], [38], [51], [52]. The findings of this study also suggest that adequate pain relief is associated with intrapartum care experience, however, the effect is not as distinct as that of a successful interaction with the midwife.

Postnatal care encompasses more than routine observations and examinations to monitor physical recovery from birth. It involves emotional support as well as respectful, patient, understanding and sensitive treatment. It also involves facilitating the attachment between the mother and her newborn child and supporting the mother in adapting to the new situation. Therefore, it seems plausible that the postpartum care experience is also predominantly associated with trust in the nurses/midwives and their availability when needed. The existing literature also indicates that lack of emotional support and lack of understanding and sensitive care are among the strongest predictors for dissatisfaction with postnatal care [4], [8], [34], [55].

Practical support with feeding the newborn is the third most important predictor for a positive experience with care in the maternity ward. This is also in line with the existing evidence which indicates that new mothers not only need consistent information and counselling but also practical help with infant feeding, particularly breastfeeding [8], [55], not least because new mothers nowadays often lack role models for successful breastfeeding.

The results of this study show that receiving understandable answers to important questions is a predictor for the quality of the care experience, but whether women received information on the various specific postnatal topics is not. This may suggest that individually tailored information and advice may be more important than a comprehensive overview over the many issues coming up during the first weeks after birth. Given the fact that the length of postnatal hospital stay has declined considerably in the last decades it seems more important to ensure adequate care and support through home visits in the first days and weeks after birth as well as an effective transition from hospital to home.

Finally, the present study indicates that postpartum pain relief is an issue relevant to women’s postnatal care experience. Whereas there is a large body of literature on intrapartum pain relief, the evidence on postpartum pain relief is scarce. In two small qualitative studies on women’s experiences in the early postnatal period participants also reported that they felt that ward staff did not pay enough attention to their postpartum pain [22], [35]. A prospective, longitudinal multicentre study on pain after childbirth shows that nearly all women reported some kind of pain in the first 36 hours after birth, acute pain was reported by 11%. Severity of postpartum pain was an independent predictor for persisting pain eight weeks after childbirth and postpartum depression [21].

The results of this study are based on a secondary data analysis, meaning that the data analysed for this study were not collected to address the research question and possible relevant third variables may therefore not have been available for analysis. This has to be considered when interpreting the results. The data collection method (post-discharge, postal survey with two reminders) resulted in good response rates. However, in postal surveys on maternity care younger women, those with lower educational level as well as those with limited knowledge of the target language are typically underrepresented in this design. This needs to be taken into account when generalizing the sample results to a broader population [7], [12]. Finally, the cross-sectional design means that no causal relationships can be detected. The identified correlations and predictors need to be interpreted in light of the clinical context and the available evidence.

The results of the present study show that the majority of new mothers report a positive overall intra- and postpartum care experience. However, the interpretation of these very positive overall ratings needs to take into account the evidence on ceiling effects that are very common when asking patients for global ratings of their care experience.

The main predictors of a positive intrapartum care experience in the labour ward as well as a positive postpartum care experience in the maternity ward are a trustful relationship with the midwives and nurses and their availability when the woman needs them. Trust and individually tailored availability are therefore key components of care quality from the women’s perspective. In the last two decades, significant resources were devoted to modernizing labour and maternity ward facilities. Midwives, nurses and other health care professionals, on the other hand, have to perform and fulfil increasingly more tasks and duties with similar or even decreasing staffing levels. Against this backdrop, the results of this study clearly indicate the need to rethink maternity care in hospitals.

The results of the present study are not surprising, they are in line with a large international body of evidence. Rather, it is worrying that recent studies still indicate urgent need for action in this area [43], [45]. More than 15 years ago, evidence has already accumulated to a tipping point that Beverly Chalmers asked how much more do we need to implement humane and sensitive care that are respectful of women’s preferences and use only those interventions that are really needed and have proven to be effective [13]. Further research should therefore focus on what needs to be done to introduce the already existing evidence effectively into practice so that women and their children can finally benefit from the already existing knowledge rather than inventing the wheel again and again.

The author states that there is no conflict of interest. At the time of writing the author was employed by the institute that owns the data (Picker Institut Deutschland gGmbH).