Modelling collaboration of primary and secondary care for children with complex care needs: long-term ventilation as an example

The number of children assisted with long-term ventilation has consistently increased over the last 25 years. For example, some areas of the UK had a 30-fold increase in prevalence between 1994 and 2010, from 0.2 to 6.7 per 100,000 children [19]. This is consistent with other data available across the EU, with data from Italy showing a prevalence of 4.2 per 100,000 paediatric patients [20], and a report from Austria showing a prevalence of 7.4 per 100,000 [21]. This is supported by trends reported internationally. The number of children receiving long-term mechanical ventilation at home in Canada increased from 2 in 1991 to 156 children as of December 2011, with a twofold increase in the number of invasive ventilation initiations in the second 10 years, n = 45 (2001–2011) as compared to the first 10 years, n = 21 (1991–2000) [22]. Data from studies that have investigated long-term outcomes and quality of life for such children are scarce, but indicate reduced long-term health-related quality of life [23]. There is evidence of burn-out in parents within a short period of time of becoming the child’s primary care giver, with a recognised increase in physical or mental ill health issues, and a significant impact on family functioning [3]. Even though there are difficulties in providing appropriate mechanical ventilators for children and its management requires well trained care givers, many studies have shown that properly selected infants and children could be safely ventilated at home [24, 25]. The key factor of success is, however, the presence of a good team with dedicated members [24].

Following the MOCHA approach to fulfil the gap of data availability at European level, a set of ad hoc questionnaires sent to local experts in child health services (Country Agents (CAs)) were developed [14] to explore organisational and clinical aspects considering different tracer conditions at various children’s life stages. Some of them were specifically focused on children with CCNs to capture, in particular, how each EU/EEA country manages the interface between primary, secondary and social care. This is evident in the selection of questions that require information on the different parts of the child’s journey also including routinely health care service provision, such as preventive care. Each questionnaire is based on a vignette that captured the real-life context [26] and included sociocultural issues that can influence the family and care provision. As shown in the LTV vignette (Table 1), this approach allows to capture information on the basis of a non-experimental descriptive study and to perform both quantitative and qualitative analysis in the exploration of structures and processes of care provision for children [14]. This also allows to gain insights on the different levels of national management and treatment of children with CCNs providing a comparison across countries as well as across complex health conditions.

The level of collaboration among professionals was defined by analysing the answers provided by local CAs that collected comparative information across the 30 MOCHA countries to identify the different types of professionals involved in the performance of the child’s care activities. Answers from the CAs were also useful to understand the level of collaboration among settings, highlighting whether care professionals were part of a multidisciplinary team (MDT) belonging to a hospital and/or a primary care setting or individuals, each one carrying out specific professional-related activities. The questions adapted from the Standards for Systems of Care for Children and Youth with Special Health Care Needs (CYSCHN) [16] supported the identification of sub-processes that trace the care journey of children with CCNs, in particular:

Moreover, the access to specialised services was investigated to capture the interface between primary and hospital care focusing the attention on the following activities:

Within this process, the communication procedures adopted in each country to exchange information among professionals and settings were analysed as an important indicator of the level of care coordination in complex childcare management, where home care plays an important role.

To capture the level of collaboration in place in each country in the performance of the care process, the professionals involved in each activity were gathered and classified on the basis of the following categories:

The result of this analysis was subsequently described using the UML model as part of the framework proposed in [18] that facilitates the comparison between the different business processes performed in the MOCHA countries. In particular, two UML types of diagrams were selected as suitable means to describe the interaction between the different stakeholders as well as the activities performed in each identified scenario [27].

The UML use case diagram described the process as an ideal set of macro-activities to be performed providing a country-independent picture. For each activity, it highlighted the type of professionals involved. Moreover, the use of the composition relationships (depicted by the symbol ) outlined the configuration of the team associated with the settings they belong to (primary, secondary and social service). This interaction was also shown using green notes that were associated to the type of actor that performed the related activity in each country. On the basis of this diagram, countries that had similar procedures, used similar services, and were based on similar caregivers for the provision of care were grouped, and a set of activity diagrams were modelled.

The UML activity diagrams clearly represented how each country/group of countries carried out the care process capturing the activities performed and the messages exchanged by the different actors as well as triggering conditions taking also into account the location and timeline of each activity. Within the process flow interrupting connector represented by a zig-zag arrow ( ) was used to manage an exception that can occur during the execution of a relevant activity. Furthermore, this diagram adopted swim lanes for partitioning activities based on the actor that performs them. This partition was also useful to capture the interaction among actors and highlight the messages exchanged during the business process. In the paper, this partition was represented by vertical swim lanes. In our methodology, we adapted the UML convention adding a horizontal partition of the model to capture the different phases of the business process. It helped capturing the main sub-processes that constituted the whole process in line with the components of the CYSCHN [16].

A comparison of the activity diagrams was performed to detect differences and similarities among countries in the performance of the different parts of the care process also pointing out strength and weak points of each health system. This comparison can be also quantified using business process metrics [28, 29] that use the elements of the UML diagrams to assess the efficiency of the business process under investigation considering both the control flow (e.g. number of activities performed, the number of decision points, the hourglass within the process and the relevant waiting time, the information shared among professionals) and the organisational perspective (e.g. composition of teams, presence of systems adopted to exchange data and documents, role of the family within the process). Considering the purpose of this paper and giving the high-level description captured from the questionnaire, the analysis privileged the straightforwardness versus the fragmentation of the care process. Moreover, particular attention was given to the referral procedures related to the exchange of the personalised plan as an indicator of clinical integration.

The UML use case diagram reported in Fig. 1 shows the macro-activities performed by health and social care professionals before and after discharge, highlighting the high variability of team composition across countries. The use case Discharge may comprise the Organisation of the transition to home as shown by the extend relationship between the two use cases. As reported in the diagram, ten CAs indicate the presence of a discharge planning coordinator (DPC). Among them, Croatia and Estonia specified that this role is covered by a nurse. In Italy and in the Netherlands, this task is carried out by a MDT, while in Luxemburg, a social worker may, if needed, cover this role. Considering the development of the plan, this task is performed in the hospital setting either by a single professional (e.g. Czech R. and Iceland) or by a MDT (e.g. Cyprus and the Netherlands). In Luxembourg and in Poland, this activity is performed by the primary care physician. In Italy, this is done in collaboration with territorial services. Moreover, the plan can be developed in collaboration with child’s parents and/or members of the team. This is highlighted by the include relationship between the use cases Collaborate in the production of the plan and Develop of the plan. Once the child is discharged at home, health and social care professionals are mainly involved in the implementation of the plan as well as in the provision of general and screening services as identified in the plan. These use cases include the Discharge use case as a generalisation of Provide health care service. In the diagram, additional use cases are added to represent the care coordination in case of medical crisis (acute care), and community-based service and support (palliative care and psychological support to parents) providing information on the procedures facilitating the access to care and/or the different actors starting the process. These specific activities are further investigated in the following paragraphs highlighting the professionals involved in their management as well as the composition of the relevant team, if present.

Within the high variability of the health care settings and professionals providing care, the use case diagram also shows different patterns of collaboration depending on the type of activity performed. Moreover, it is possible to single out countries in which all types of macro-activities after discharge are performed under the supervision of hospital professionals. This is the case of Austria, Iceland and the Netherlands that are organised in a MDT, while in Belgium, Bulgaria, Lithuania and Romania, a single health care hospital professional provides these services. Primary care services monitor the child’s care after discharge, providing both prevention and curative services in Croatia. In other countries (Estonia and Ireland), both the provision of preventive screening and curative services are carried out via collaboration between primary and secondary care. In Norway and in Hungary, this collaboration is in place in the provision of curative services, while in Spain, this is managed through the provision of preventive screening. In Finland and Greece, general services are provided within primary care, while hospital professionals are accountable for the specialised ones. In the UK, there is a close collaboration between hospital and community professionals working in a MDT that also includes a social worker.

Starting from the UML use case diagram, countries have been classified on the basis of two features: (1) the presence or absence of MDTs for the care of children on LTV and (2) the type and composition of health care organisations in charge of the child’s care. These features have been adopted to determine the extremes of a continuum of care where the lowest degree of integration occurs when there is a high fragmentation of service provision carried out by individual professionals. On the other extreme, a full integration is achieved when hospital and primary care professionals manage the whole child pathway in a MDT. Between these extremes, we have identified two additional patterns of collaboration: (1) MDT composed by hospital professionals and (2) a composition of two different MDTs respectively based on hospital and primary care settings. Moreover, the asterisks indicate the presence of a social worker. A similar analysis [30] has been used to measure the integration of care in a single country.

Starting from this classification, the MOCHA countries have been clustered as shown in Fig. 2 and in the map reported in Fig. 3. Half of the analysed countries (n = 12) have a multidisciplinary team with different degrees of team composition, while in the rest of the countries (n = 11), care services are provided by individual professionals.

From the results of the clustering analysis, two activity diagrams are shown in Figs. 4 and 5. These two diagrams represent the countries falling in the two extremes of the model reported in Fig. 2.

In each diagram, the vertical partition is used to indicate the actors that perform group of activities and messages sent/received. In this case, actors are individual professionals (e.g. discharge planning coordinator, primary care professionals, hospital specialist) or a group of professionals working in a team (e.g. primary care team, multidisciplinary team, hospital team). Moreover, horizontal partitions are used to classify the activities within the following components of the CYSCHN [16]: (1) care coordination (i.e. implementation of the written plan); (2) access to care (i.e. acute services in case of medical crisis and provision of curative services); (3) screening, assessment and referral (i.e. provision of preventive care screening and developmental checks); (4) community-based services (i.e. psychological support to parents and provision of palliative care).