Implementation of a Regional Standardised Model for Perinatal Electronic Medical Records

Electronic recording of newborn health information contributes to improving the quality of care. Nonetheless, there is limited evidence on the implementation of perinatal electronic medical records models. We describe the development and implementation of an electronic recording model that includes data on the health care provided to both the mother and the newborn, standardised for six hospitals of a regional health care system. The implementation process was developed in 2 stages. During stage 1, the tool was introduced in hospitals to stablish first contact with the healthcare staff. The second stage consisted in designing a new strategy to stabilise the model. Technical issues were fixed, and a new version was drawn up based on multidisciplinary agreement. Indicators to monitor implementation were measured in both stages and compared using the chi-squared test. During stage 1, nearly every newborn got its electronic medical record with an appropriate connection to the mother’s data. However, certain forms that were meant to be filled in by staff were frequently neglected (completion rates: 36.7%-55.3%). In stage 2, there was a statistically significant increase in the completion rates of all these forms. As a result, a standardised discharge report was provided to every newborn at the end of stage 2. The PCR model implemented in the Region of Murcia is an innovative example of how the digitalisation and standardisation of data related to the care of healthy newborns at maternity wards is feasible across an entire network of hospitals.


Background
Although it is common to establish a basic minimum set of data for patients requiring admission at Neonatology Units, when newborns remains at the maternity ward with their parents, access to information about their care is often largely lost [1]. These data are often not recorded on electronic formats and sometimes they are simply included as part of to the mother's medical record [2]. On the other hand, the data collected from the mother's medical record is often limited when it comes to certain aspects of the birth and the puerperium period that have a direct impact on breastfeeding and the subsequent development of the newborn [3]. Thus, aspects such as early "skin-to-skin" contact, the initiation of breastfeeding and its difficulties, and feeding of the baby at the hospital are not recorded or become difficult to access once the episode is ends [4].
Another aspect to keep in mind is that the complexity of newborns seen in maternity wards has risen significantly over the past few years [5]. Many of the neonatal conditions that were previously attended at neonatal units are now handled at maternity wards with the aim of preserving the bond between parents and baby [6]. In addition, several procedures and screening tests are conducted on newborns in this area [7]. It would be convenient for all this information to be collected in a report that facilitates continuity of care, which is an unusual occurrence [8].
Despite the interest shown by scientific associations related to perinatology [9], there is limited evidence on the implementation of perinatal clinical records (PCR) for all births in a region, both nationally and internationally. Therefore, a significant amount of published models are based on data collected from birth certificate records or, at best, are systems that collect a basic minimum set of data including an incomplete set of indicators [1,10,11]. Thus, the lack of accessible data repositories has a negative impact on continuity of care for both mothers and infants [12].
Since 2017, the Murcian Health Service (MHS) has been developing a project to ensure health care for the first 1,000 days of a child's life [13]. This period includes conception, pregnancy, and life period until the infant's second birthday. This stage is critical for neurodevelopment and future health as an adult [14,15]. A key goal of this project is to implement structural enhancements that allow for continuity of care for all children from birth. One of the tools implemented to achieve this goal is a common PCR model for all its hospitals handling deliveries.
Prior to this project, in the MHS, data from the mother and the gestational period were collected in the mother's record and were not linked to the newborn's record. Records were mostly paper-based, the information provided at discharge was insufficient, and the minimum basic data set for the region was very limited in this regard. This new PCR integrates the data collected from the mother's medical record, as well as any assessments, tests, and procedures conducted on the infants during their stay at the maternity ward.
The aim of this study is to evaluate the implementation process of the PCR model in the Murcian Health Service's hospital network through the analysis of indicators related to its technical suitability, and the compliance of medical professionals.

Study design
Between December 2017 and June 2019, an observational prospective study was carried out in all the newborns admitted to maternity wards of the Murcian Health Service. The Murcian Health Service is a public health system that processes approximately 13,000 new births each year. It includes six hospitals with maternity ward that handle 53%, 20%, 12%, 9%, 4% and 4% of the births in the region. Of these pregnancies, 23% of mothers are foreign, predominantly Moroccan (74% of them). The study protocol was approved by the Ethics Committee of MHS's Health Areas II and VIII.

Description of the tool
Hospitals in the Region of Murcia manage the patients' medical record using the generic marketed software SELENE® (CGM Global. Woodbridge, UK) [15]. Mothers' and babies' medical record are comprised of several forms containing indicators related to the mother's clinical care, neonatal care, and other aspects such as skin-to-skin contact and breastfeeding (Table 1). These forms and the indicators included therein were designed by the heads of this project. A first version of the PCR was monitored for two years at the Santa Lucía General Hospital (part of the MHS) to verify its reliability and applicability to the remaining hospitals of the MHS. Forms are filled in by the different professionals involved in a predetermined order, from the admission of a pregnant woman for childbirth until she is discharged with the baby. Figure 1 summarises the clinical route of neonatal care, including the order of the interventions, the staff responsible for each of them, and the date and time of recording.
After childbirth, the neonate receives its own medical record number, and its first care process is started (newborn process). The aforementioned forms are created during this process. Then, an automatic linking form is generated ("data transfer from mother to baby") and transfers all relevant clinical information in the mother's record to the newborn's record. This information is summarised in an accessible report with portable format ("perinatal data report"), which speeds up subsequent medical examinations by the paediatricians.
At the time of hospital discharge, the family receives the "neonatal care report". This report retrieves the main information from the reports created beforehand for the parents and to ensure continuity of care. The care report and the forms from which it collects the information are standardised and identical in all hospitals handling deliveries within the MHS.

Implementation stages
Stage 1: diffusion, training and monitoring (from September 2017 to December 2018). During this initial stage of 16 months, appropriate interventions were carried out in order to introduce the new tool in hospitals, ensure that the staff was familiar with it, and encourage changes in working protocols. For this purpose, during the first month, project managers conducted several training sessions at each hospital. These sessions were attended by members of management, heads of departments, maternity supervisors and other selected professionals.
Once these sessions concluded, we started to gradually implement the PCR system at each hospital; due to technical difficulties, this process lasted until November 2017. The procedure followed in each hospital consisted in: conducting the necessary training sessions to ensure all personnel shifts were covered, implementation of the programme and physical attendance of software technicians until it was in a stable state.
Over the 13 following months, hospital departments were working with the new forms and reports, and continuously submitting feedback regarding issues observed during this stage. This information was gathered by the staff in charge of implementation. In June 2018, a face-to-face meeting was held with representatives of the different professional groups Fig. 1 Clinical route of neonatal care. Forms surrounded by a white circle are filled by midwifes; white hexagon: nurses; white rhombuses: paediatricians; white squares: automatic forms. 1 The first breastfeeding form may be recorded at the delivery room or at the maternity ward depending on where breastfeeding starts involved at each hospital. In this meeting, the problems detected were discussed, along with other suggestions and needs. Based on this discussion, different areas for improvement were identified and included in the PCR.
Stage 2, stabilisation, started in January 2019 and ended in June 2019. A different strategy was designed in January 2019 to achieve model stabilisation. In February 2019, a team comprised of a paediatrician, a midwife, a nurse and a technical support engineer of the application visited all six hospitals in order. The meetings at each hospital were attended by members of hospital management, supervisors, heads of departments and medical professionals (midwives, nurses, nursing assistants, obstetricians and paediatricians) with privileged knowledge about the development of the program locally. The implementation indicators of the hospital were thoroughly analysed in these meetings, the operational aspect of the application was verified, and many different suggestions were discussed.
Once these meetings concluded, in March 2019, technical issues were fixed, suggestions were analysed, and the team in charge of the PCR implemented the necessary changes for the new version of the PCR. Several interviews were also held with representatives of mothers' associations to inform and improve the update process. A project report was drafted including all the improvements implemented. It was then submitted to each hospital before implementing the changes. Lastly, in April 2019, the PCR's system was updated to the last version. There was a series of telephone calls to verify the proper operation of the system, solve minor issues, and asses the adhesion rates of medical professionals.

Procedures and data collection
Three assessments were conducted during the entire study: first, in April 2018, for an intermediate evaluation of stage 1; second, at the end of stage 1 (December 2018) to obtain a final value of the indicators during this stage; and third, in June 2019, to get a new measurement of indicators for stage 2 and compare them with stage 1. For these assessments, MHS's Technology Department provided the investigators with data from each hospital, broken down by month, to calculate the indicators. The data package included the number of infants admitted and discharged from each maternity ward and the number of them with each of the forms/reports recorded in their PCR. Patient's data were managed according to the Spanish Law on Data Protection.

Variables and statistical analysis
The implementation process of the PCR was evaluated in all its stages using a set of indicators aimed at evaluating its technical suitability, the compliance of medical professionals with the systems, and improvements implemented ( Table 2).
The information used to calculate the indicators was entered in a computer database. Statistical analysis was conducted using SPSS® 21.0 (SPSS Inc., Chicago, Illinois, USA). The numerical values are expressed with their absolute and relative frequency, in percentages. For cases in which variable comparison was conducted, the test used was the chi-squared test. A bilateral random value of lower than 1% (p < 0.01) is considered statistically significant.

Results
During the two stages of the project, 21,283 childbirths were included for the analysis. Of them, 2,046 (9.6%) required hospitalisation at neonatology units for the first few days of life, 1,427 (6.7%) of the referred from the delivery room. Thus, 19,856 neonates were admitted in maternity wards, of which 19,237 were then discharged from the hospital.
One of the project's aims was to ensure that infants haver their medical record with an adequate link to the data collected in their mother's record. This objective was achieved and maintained from the very first months of stage 1. Creating a record number is a must for any procedures that must be conducted on the infant, and the form "data transfer from mother to baby" is generated automatically. As a consequence, the value of the indicator "healthcare identity" was 100%, during both stage 1 and 2. The values for connection between mother's and infant's medical records were 100% in stage 1 and 98.3% in stage 2.
Regarding voluntary completion forms by the staff, "neonate" and "reception" achieved the highest adherence. Their adherence rates were above the target value set in both stages and there were no significant differences between them ( Table 3). The "discharge" form was close to the limit of the objective in both stages, just below the limit in the stage 1 and just above it in stage 2 (Table 3).
During stage 2, stabilisation, our efforts were mainly focused on the "first breastfeeding" and "care" forms, due to the low adhesion rates of these forms by professionals. For the first one, the visits to the hospitals showed confusion regarding the person in charge of filling them in when breastfeeding had not been started in the delivery room. For the second one, paediatricians communicated the issues that caused low completion rates along with their suggestions to solve them. Once the changes were applied, the value of these indicators rose progressively beyond the target (Table 3 and Fig. 2). The completion rate of the "subsequent breastfeedings" form increased during stage 2, but remained below target at the end of the stage (Table 3 and Fig. 2).
During the implementation process of the PCR, improvements were implemented based on a total of 35 incidents detected: thirteen in June 2018 evaluation and 22 in March 2019 evaluation. Most changes (21 out of 35) were related to form design issues detected during use. Ten changes were added with the purpose of increasing patient safety. Some examples are direct uploading of laboratory data in order to avoid transcription mistakes, and the application of valid ranges and alerts to avoid incorrect recording. Finally, four alert indicators were designed (risk allergy to cow's milk protein, hypoglycaemia risk, neonatal sepsis risk, and genital mutilation risk in girls) to help with the decision-making process. These indicators, based on certain established criteria, alert professionals and allow for the activation of the relevant action protocols.

Discussion
This PCR model provides hospitals not only with a piece of software, but also a clinical route for the care of healthy newborns, based on scientific evidence, which contributes to improving the health results of both mothers and infants [16,17]. The complete digital transformation of the records was feasible at our institution and makes information more accessible, adds security mechanisms to prevent mistakes, provides algorithms that alert clinicians in situations of risk, and, ultimately, contributes to improving patient safety [18][19][20][21].

Connection between mother's and infant's record
Newborns with "data transfer from mother to baby" form (quantitative) Newborns with "data transfer from mother to baby" form ______________________________ Total number of newborns > 95%  From an institutional point of view, having a standardised PCR model at hand allows us to move towards a decrease in clinical variability. This model makes it easier for professionals to adapt to the clinical route and promotes a more equitable and coordinated healthcare across different health areas. Furthermore, having key indicators such as the rates of skin-to-skin contact rates or exclusive breastfeeding at discharge, with a population base, in real-time, and broken down by health area, allows for a close monitoring of clinical practices at each hospital and provides opportunities for improvement [22][23][24]. On the other hand, care report is a key element, because it summarises the procedures conducted on newborns, guarantees that the information is transmitted to their parents, and ensures continuity of care. These aspects are vital in regions like ours, in which language barriers are common. As a whole, the PCR model presented yields benefits not only for healthcare professionals and institutions, but also for users.

Optimisation of electronic PCR
The project presented in this document endorses the feasibility of implementing a single PCR model for every childbirth in a group of hospitals of the same region. In July 2021, the Spanish Ministry of Health, Consumer Affairs and Social Welfare, as part of the call for "Good Practices within the National Health System" of 2019, classified this project as a practice in the category of reproductive health. This recognition supports our belief that this PCR model could inspire other hospitals or regions interested in improving their PCR models. The experience of the implementation described herein gave rise to useful mistakes and lessons than may be useful in future projects. Among them, we would like to highlight the collaborative and inclusive approach with healthcare professionals. Some of the most innovative developments came up from the contributions of the professionals at these meetings and the indicators reached its peak when the project adopted a closer approach.

Limitations
This project continues to evolve since the end of stage 2. The challenges encountered with the "subsequent breastfeedings" form motivated more in-depth update of the project and a new version, which is still under evaluation. Another line of action underway is currently analysing the quality of recorded data. Previous studies warned against potential mistakes made by PCR users when recording data [25,26]. In stage 3 of this project, we are evaluating data consistency with reality through random telephone surveys targeted at mothers.

Conclusions
The PCR model implemented in the Region of Murcia is an innovative example of how the digitalisation and standardisation of data related to the care of healthy newborns at maternity wards is feasible across an entire network of hospitals. The indicators included in it, the clinical route for