TeamBirth-SWE - a Care Process to Improve Safe and Person-centered Intrapartum Care

The primary objective of this project is to enhance understanding of intrapartum patient safety and person-centered care through the evaluation of an intervention designed to improve team communication and patient involvement during childbirth, thereby positively impacting maternal and infant health.

Even though childbirth generally is generally considered safe for women in high income countries, the time of birth is critical to the survival of women and their babies, as the risk of morbidity and mortality increases considerably if complications arise. Adverse events reported within maternity care varies from 4-12%, with as much as 57-74% of the incidents being deemed as preventable. According to the Swedish Health and Social Care Inspectorate (IVO), obstetrics and gynecology are one of the areas within somatic specialist care that receive the highest number of complaints. The repercussions of an adverse event or a near miss can have a major impact on both mother and infant, Consequences for women include post-traumatic stress syndrome, secondary fear of birth, and a prolonged interval to a subsequent pregnancy. For infants born with severe asphyxia, the consequences may be lifelong and the societal burden is significant.

Evaluations of teamwork programs or simulation-based teamwork interventions have demonstrated improved team performance and a reduction in morbidity. However, interventions aimed at improving teamwork within intrapartum care have primarily focused on emergency situations, not considering everyday practices. This oversight is significant as many adverse outcomes in intrapartum care develop gradually over time, providing several opportunities for the team to act. Interventions that include person-centered care, facilitate communication and support decision-making processes among interprofessional teams in their daily work have the potential to contribute to enhanced care and outcomes.

The TeamBirth care process was developed by Ariadne Labs, a joint center for health systems innovation at the Harvard T.H Chan School of Public Health (U.S), to improve communication and teamwork between care givers and women to address harmful variation in the quality of intrapartum care. TeamBirth is composed of brief team meetings "huddles" between care givers, women, and their partners, navigated by a shared planning board placed in the birthing room, visible to everyone. The board promotes transparent and reliable communication and includes four areas: (a) the names of the team members, starting with the woman; (b) the woman's preferences; (c) the care plan for the woman, infant, and labor progress; and (d) when the next team huddle is anticipated. TeamBirth has been evaluated for its feasibility and acceptability in the U.S.(25) Given that the intervention has solely been assessed for experiences and implementation, and exclusively within the U.S. context, further research on its effectiveness and whether it functions in other settings is required. We have adapted and tested TeamBirth in the Swedish context in a pilot study conducted at two labor wards in Stockholm during 2022-2023, and the insights have been utilized in the design of the proposed project.

The project addresses the overall aim and the specific research questions in 3 Work Packages (WP):

WP1. Effectiveness of the intervention on patient safety and person-centered care: Does thr TeamBirth-SWE lead to safer care, i.e., improved maternal and infant outcomes during labor, birth and the immediate postnatal period? Will the intervention be more beneficial for women with lower socio-economic status (SES), co-morbidities and high-risk pregnancies?

WP2. Patient reported experiences: Does TeamBirth-SWE improve person-centered care during childbirth? Does it contribute to individualized care, informed choices, shared decision-making, and perceived support from staff?

WP3. Process measures - Patient safety climate and interprofessional collaboration: Does TeamBirth-SWE lead to increased patient safety climate and interprofessional collaboration?

Effectiveness of TeamBirth in reducing adverse maternal and infant outcomes will be evaluated using an interrupted time series design (ITS). The ITS design is particularly suitable for interventions at a population level and over a clearly defined time period, also taking baseline and follow-up trends into account.

WP1. Patient safety outcomes and person-centered care Design, methods: ITS design with segmented regression analysis to study trends in the monthly rates of patient safety outcomes. Data from 2020-2028 will be used; two years before and after implementation at each participating site.

Setting: All labor wards participating in the project Variables/measurements: Patient safety is often measured by composite variables.(30) We decided to choose a primary outcome that reflects the intention of Swedish maternity care: a healthy mother, infant, and a positive experience. In this project we assume that patient involvement and shared decision-making which are cornerstones of person-centered care, contribute to a positive experience. Therefore, two primary outcomes were chosen.

Data sources: Register data from the Swedish Pregnancy Register and the National Swedish Pregnancy Survey 8 weeks after birth (NPS-8). The Swedish Pregnancy Register is a nationwide high-quality register collecting comprehensive data on pregnancy, labor and birth, including maternal health, prenatal care, obstetric interventions, and perinatal outcomes. The Register covers 99% of all births in Sweden and the majority of the variables are retrieved via automatic transfer from the medical record, updated every 24 hours. The NPS-8 is distributed to all pregnant women in gestational week 25, eight weeks and one year after birth. It is available in Swedish, English, Arabic, Farsi, Finnish, French, Spanish, and Somali.

Exposure: The TeamBirth intervention

Primary outcome:

1. Maternal and infant: The composite variable "Birth without major interventions and complications" which, will be used as primary outcome provided from the Swedish Pregnancy Register. It includes the following variables: no operative birth, no postpartum hemorrhage (bleeding>1000 ml), no severe perineal trauma (grade III-IV tears) and a 5-minute Apgar score ≥7.
2. Maternal: NPQ-8 question: "Were you involved in planning and decision-making during birth as much as you wanted to?" (Likert type, scale 1-5)

Secondary outcomes:

1. Maternal: cesarean section, instrumental birth, postpartum hemorrhage (bleeding >1000ml), severe perineal trauma (grade III-IV tears), 8 additional questions reflecting person-centered care from NPQ-8.
2. Infant: 5-minute Apgar score of ≤7, and pH umbilical arterial <7.05.
3. Nine additional questions reflecting person-centered care from the NPS-8.

Covariates: The patient mix before and after the intervention is assumed to be the same. This assumption will be investigated using descriptive time series analysis of potential risk factors for the outcomes, such as the proportion of first-time mothers, smokers, and pre-pregnancy BMI. Furthermore, since the intervention can be started at different times throughout the year, we will adjust for calendar month of birth. High workload, shortage of staff, major organizational changes, or other interventions targeting patient safety are hypothesized to be potential confounders. To account for high-workload, we will use the number of births per 24 hours. Another possible confounder is major long-term disturbances in the work situation, which will be addressed in sensitivity analyses, where a priori defined time periods of follow-up will be excluded, i.e., based on known periods of organizational turbulence, shortage of midwives, or other major changes.

Power calculation: In 2020, approximately 80% of the births at the study sites were without major interventions or complications, with 51% of women being completely satisfied with their participation in decision-making. Based on observed standard deviations from 2020, we estimate this study has 80% power (at 5% significance with a 2-sided test) to detect a 1.3 percent point increase in the proportion of births without major adverse events and a 2.5% percent point increase in the proportion of women completely satisfied with their participation during labor and birth. The latter outcome is measured on a Likert scale, and an analysis of ordinal data is expected to further increase statistical power. Since the intervention is complex, our hypothesis is a successive improvement of 1.5 percentage points each year for "births without major interventions and complications", and by 5 percentage points per year for the variable "satisfied with participation in decision-making".

Analysis: The proportion of women with the defined outcomes, per calendar month and study site, will be analyzed using multiple regression, weighted proportionally to the number of pregnancies per month. The basic model will include study site, calendar month, and a time variable divided into pre-intervention, 0-6 months, and 7-18 months after implementation. If trends in outcomes before the intervention are observed, analyses based on change of trends before and after intervention will also be performed. As the effect of the intervention may vary by population characteristics, we will perform stratified analyses based on parity, onset of labor, and study site.

WP2. Patient reported experience measures Design, methods: A web-based survey, including background questions, validated scales and study specific items, will be sent to women one month after birth. The survey will be available in Swedish and English.

Inclusion criteria: All women >15 years of age with spontaneous onset of labor or induction of labor after gw 22+0, and who are proficient in Swedish or English.

Exclusion criteria: Planned caesarean section, intrauterine fetal demise. Instruments: Validated scales will be used; Mothers Autonomy in Decision-making (MADM), Mothers on Respect Index (MOR), Mistreatment Index (MIST) and Childbirth Experience Questionnaire 2.0 (CEQ2) and Support and Control in Birth Scale (SCIB). Women will be asked about their experience of S-TeamBirth during their labor and birth.

Exposure: Women experiencing high use of TeamBirth-SWE (>5 huddles). Outcome: Primary outcome: Women's experiences of autonomy and shared decision-making (MADM). Secondary outcomes: Mothers on Respect Index (MOR), the Mistreatment Index (MIST), the Childbirth Experience questionnaire (CEQ2) and the Support and Control in Birth Scale (SCIB).

Exploratory factors: Mode of birth, adverse maternal or neonatal outcomes, fear of birth. Any imbalances between the groups related to background and pregnancy characteristics will be accounted for.

WP3. Patient safety climate and interprofessional collaboration Design, methods: A longitudinal cross-sectional design, using survey data measured at four time points.

Participants: All midwives, physicians and nurse assistants working with intrapartum care at the participating labor wards.

Instruments and measurements: Validated scales will be used; The Hospital Survey on Patent Safety (HSPOS 2.0) to measure patient safety climate and the Assessment of Collaborative Environments (ACE-15) to conceptualize the dimensions of interprofessional teamwork. Timepoints for measurement: One month before the start of the implementation (T0), 6 months (T1), 12 months (T2), and 24 months (T3) after the implementation of S-TeamBirth.

To assess the effect of TeamBirth-SWE on patient safety climate and interprofessional teamwork, data from pre- and post-intervention will be compared. Furthermore, the effect of the intervention will be analyzed over time. Data normality will be tested using the Shapiro-Wilk test, and ANOVA with repeated measures or Friedman's test will be conducted to compare the mean scores (dimensions and total scores) and assess differences between the time points (T0-T3). Posthoc tests will be performed to investigate differences in attitudes between professions. Differences in mean change scores will further be calculated for three samples to assess when (or if) changes in attitudes occur: Sample 1 (T1 compared to T0), Sample 2 (T2 compared to T1), and Sample 3 (T3 compared to T1) using t-tests or Mann-Whitney U-tests. If samples are proven to be dependent, a paired t-test or Wilcoxon signed-rank test will be employed.