Impact of Monitoring Quality Indicators in Maternity Departments on Mother and Infant Health

Overview

Hypothesis: Continuously monitor quality indicators with a specific method (CUSUM: Cumulative Sum) will increase the awareness of health care staff in maternity and permit rapid detection of a small dip in performance in order to enable prompt investigations and corrective measures when necessary , which decrease maternal and neonatal mortality and morbidity. Objective: To assess the impact of Cumulative Sum (CUSUM) charts used as a maternity dashboard to decreases maternal and neonatal mortality and morbidity. Design: Step-wedge cluster-randomized trial with prospective analyses of collected data. Setting: ten Maternity departments in France. Population: Data from 60 000 women and infants could be collected over 2 year's period. Method: Cumulative sum (CUSUM) charts were used to monitor the rate of quality indicators previously selected by Delphi method. Main Outcome Measures: Composite outcome that considers multiple clinical events : mortality, adverse outcomes for women and newborn.

1. RATIONAL Quality improvement has become a central tenet of health care. In obstetrics and gynaecology departments, each admission may affect the health of not one, but two individuals, making high quality of care particularly important. In addition, most women admitted for obstetric reasons are healthy individuals in whom the goal is full preservation of health. Measurement plays an important part in improving quality of care and promoting beneficial changes. Therefore, over the last decade much effort has gone into developing and using measures of quality. The first step towards measuring quality of care in obstetrical departments is to select quality indicators. Conventional quality indicators are severe events such as maternal death and neonatal death. However, advances in obstetrical management have made such events extremely rare, and other quality indicators must now be sought. Once quality indicators are identified, the best means of monitoring them must be determined. The use of statistical methods to monitor and control a process, known as Statistical Process Control (SPC), seems particularly promising for monitoring quality indicators. SPC uses rigorous time-series analysis methods, whose results are reported as a graph of changes in outcome rates over time. Moreover, SPC can help to determine whether these changes are real (i.e., related to a causative factor) or merely a manifestation of natural variability. Among SPC tools, CUmulative SUM (CUSUM) charts are widely used for industrial quality control and have been found effective for measuring and monitoring healthcare outcomes. In obstetrics, Cumulative Sum (CUSUM) charts have been used to monitor Apgar scores or the performance of sonographers in assessing nuchal translucency in the first trimester of pregnancy. Despite Establishing Cumulative Sum (CUSUM) charts for relevant quality indicators holds considerable promise for continuously assessing and improving the quality of care, nowadays, there are no studies that assess the impact of the CUSUM chart on mother and children health. The most published studies used CUmulative SUM (CUSUM) to see the behaviour of an indicator, but do not exploit the Cumulative Sum (CUSUM) as a SPC (statistical process control)tool 2. OBJECTIVES Primary objective: to assess the impact of Cumulative Sum (CUSUM) charts used as a maternity dashboard to decreases maternal and neonatal mortality and morbidity Secondary objectives * select a consensual set of quality indicator to assess the overall quality of obstetrical care and could be routinely monitored in maternity units. * estimate acceptable and unacceptable rates of quality indicators * assess Satisfaction of caregivers about the proposed tool * assess patient satisfaction about their care in maternity 3. OUTCOMES Primary outcome Indicator of mother health : is a composite outcome taking into account multiple clinical events: 1. \- mortality, 2. \- complications of pre-eclampsia * Hemolysis ,Elevated Liver enzymes, Low Platelet syndrome (HELLP): * Disseminated intravascular coagulation (DIC) eclampsia retro-placental hematoma, 3. \- postpartum hemorrhage: 4. \- severe bleeding 5. \- thromboembolism: Phlebitis Pulmonary embolism 6. \- postpartum infections 7. \- severe perineal injury 8. \- need of intensive care transfer for women Indicator of newborn health: is a composite outcome of mortality and morbidity. Morbidities take into account several clinical events: 1. \- postnatal deaths (8days 1 month) 2. \- prematurity: medical or spontaneous 3. \- low birth weight 4. \- maternal-fetal infection: occurring in the first hours of life 5. \- Transfer to neonatal intensive care unit (NICU) 6. \- chromosomal abnormalities with absence of screening 7. \- malformations without screening and diagnostic Secondary outcomes Care giver satisfaction Women satisfaction 4.STUDY POPULATION * Participating Center : Ten Maternity departments in France will participate to the study * Eligibility criteria Inclusion Criteria * All deliveries during the study period: Regardless of the mode of delivery: vaginal delivery, cesarean section, spontaneous or induced delivery Regardless of the pregnancy term * Women followed or not from the first trimester of pregnancy in the concerned maternity * New born during the study period Exclusion criteria * Pregnancies that ended with fetal death in UTERO or pregnancy termination for medical reasons 5.STUDY DESIGN It is a Step-wedge cluster-randomized, prospective multicenter study, evaluating the impact of continuous monitoring of quality indicators using CUmulative SUM (CUSUM) chart, on the health of mother and newborn in 10 participating centers. The advantage of this design is that the intervention is rolled out to all individuals or clusters in phases and this design may require fewer clusters than a parallel CRCT design. CUmulative SUM (CUSUM) will be set up in all centers, and for each quality indicator previously selected, with a delayed onset of disseminating the results of the CUmulative SUM (CUSUM) in group 2. A Delphi survey with an international expert panel in obstetrical care was used to develop a set of main quality indicators to be followed routinely in maternity. The basic performance of each maternity will be evaluated in the first period of data collection in order to have a perception of the performance of each maternity. As shown in Figure 1, the study will be conducted in three phases: Phase 1: is a 3 months testing and observation period. During this phase, rates of each quality indicator selected will be collected to adjust the acceptable and unacceptable rate set by practitioners of each maternity. Phase 2: In this phase, maternity will be randomly assigned to one of the following groups: the group to which the results of the CUSUM will be provided for a period of 12 months (Early CUSUM result group), or group to which the results of CUmulative SUM (CUSUM) will not be shared. Phase 3: In this phase, the maternity wards in the "early result group" will continue to receive the results of the CUSUM for a period of 12 months, while maternity in second group receive CUSUM result (CUSUM result Delayed group) during the same period 12 months. At the end of study period, maternity "early result group" will receive the results of the CUmulative SUM (CUSUM) for 24 months period, while maternity in "Delayed CUSUM result group" will receive CUSUM results for a period of 12 months. 6.STUDY STEPS 1. A modified Delphi Survey of an international multidisciplinary panel to select quality indicators A modified Delphi survey will be conducted involving international expert panel. The Delphi technique is a structured process that is widely used for quality indicators development. One of the main reasons for the popularity enjoyed by the Delphi technique is that a large number of individuals across diverse locations and areas of expertise can be included anonymously and do not interact directly with each other, so situation where the group is dominated by the views of certain individuals can be avoided. The modified Delphi technique is defined, among other, as the combined use of a self-administered questionnaire and of a physical meeting of the experts to enhance complex decision making process and clarify language and recommendations. The panel members will be asked to rate each of the quality indicators against 2 criteria: validity and feasibility. Validity was defined as the extent to which the characteristics of the indicator are appropriate for the concept being assessed and quality indicators was considered feasible if information needed to assess adherence was thought to be available in the medical record or from patient or simple to collect from any source whatsoever with a workload consistent with the activity of maternity. These will be both rated along a 9 point scale, from a score of 1 (definitely not valid or not feasible) to 9 (definitely valid or feasible). 2. Estimating acceptable and unacceptable rates of quality indicators: Maternity obstetricians will be asked to define acceptable and unacceptable rates for each quality indicator in accordance with their actual performances, activities and local needs. For each quality indicator, the acceptable rate was defined as the rate at and beyond which no corrective action was needed and the unacceptable rate as the rate at and beyond which an audit was needed to identify the cause of the change in performance and thereby allow corrective action. 3. Intervention is described as follows: Step 1: Implementation of the CUSUM control chart using available medical data Step 2: Submission of monthly results to the coordinator of each participating center (if CUmulative SUM (CUSUM) results period). Step 3: When an alarm is generated, the steering committee will meet with maternity staff to find the causes of these alerts and propose corrective measures. 7\. STATISTICAL ANALYSIS Sample size calculation Sample size calculation was based on the following assumption: * incidence of mortality / morbidity: 40 per 100 * Maternity activity over 1 year period: 3000 births on average per center. * A decrease in events of 30% In a cluster randomized trial, the outcome of individuals within clusters is not independent. The sample size should be increased to take into account these effects. The required increase is influenced by both the ICC and the size of the cluster. The investigators must increase the standard sample size by an inflation factor equal to \[(1 + (n-1) ρ)\] where n is the average size of centers in our study , estimated to 3000 patients per year, ρ the intraclass correlation coefficient (ICC). With a hypothetical ICC of 0.01, the number of subjects required is estimated to 7500 patients. The number of centers being set at 10 with an average inclusion of 3,000 women per year, the investigators will easily reach the number of women required. Statistical Analysis 1. Descriptive analysis Data are expressed as mean (SD), or median and 1st , 3rd quartiles, as appropriate for quantitative variables. Qualitative variables are expressed as total numbers and percentages Comparison of characteristics in each group was performed using χ2 test or Fisher exact test for qualitative variables. For quantitative variables, comparison involved the t.test or wilcoxon test. P\<0.05 was considered statistically significant 2. Statistical analysis Two analysis will be performed: First analysis: will compare the incidence of events taking into account in the Indicator of mother or newborn health from the 'Early result group' and the 'delayed start group' up 'at end of period 2 (see figure 2). Second analysis: to asses if there is a difference in event rates between the 'Early result group' and the 'delayed start group' for the duration of the study (Period period2 + 3). This analysis shows whether the benefit observed in the 'Early result group' observed at the end of the two continues until the end of the study when the two groups received the same intervention. The investigators use hierarchical models to account for possible correlations between observations in the maternity wards. The investigators can introduce variables dependent to maternity (level, status, number of medical staff, activity ...) and women (and newborn). All tests will be performed using the SAS software version 9.2 and ML-Win 3. CUmulative SUM (CUSUM) Chart CUSUM charts are constructed by calculating and plotting the sum St = max (0, St-1 + Xt - k+), where S0=0, to detect an upward change; or St = min (0, St-1 + Xt - k-), where S0=0, to detect a downward change. The CUSUM tests the null hypothesis, according to which the process is 'in control' (IC), that is, performance is acceptable; as opposed to 'out of control' (OC).