Cooperation and coordination around children with individual plans

Abstract In this article we discuss the relationship between cooperation and coordination around six children with special needs who attend Norwegian day care centres. The children in focus have ‘individual plans’, in which goals, roles and activities at the centre of the cooperation are documented, and are as such a coordinating device. Each child has a cooperative team consisting of parents, representatives from day care centre, the child's special educator, the public health nurse and, for instance, their physiotherapist, family doctor etc. to support them. The cooperating teams are headed by a coordinator, who in most cases is the public health nurse. Methods used in the study are individual interviews with at least four representatives from each cooperative team, as well as two focus group interviews with representatives from two different specialist units. The research design is action oriented, and involves, in addition to data collection, two seminars with parents and practitioners from the day ca...


Introduction
During the last decades we have witnessed a substantial growth in research on childhood-related issues. One important field of interest within this research field has been the life conditions for vulnerable groups of children, such as children living in poverty, children in need of child welfare and children with special needs. Not surprisingly there is a tendency for this research to be concentrated almost solely on the familial dimension of the children's lives, especially as far as the youngest children are concerned. However, in pace with the growing awareness about childhood as a distinct life-phase, there has also been a growing awareness about the institutionalized character of modern childhood. Hence, research on children's lives in arenas such as schools and day care centres has grown in prevalence. Still, in spite of this increased interest, research on children's lives in day care centre remains to a large extent a neglected field.
The institutionalized character of modern childhood demands cross-institutional needs are concerned. In order to ensure the well-being of these children there is a need for substantial cooperation between the children's families, their day care centres, as well as bodies within different sectors of the welfare system (i.e. educational and medical bodies). However, research on the collaborative practice around children with special needs who attend day care centre is scarce in Norway (Borg, Kristiansen, and Backe-Hansen 2008). In this article we discuss findings derived from the project regarding experiences with different collaborative practices.
The key question we address is how individual plans may promote or restrain cooperation.

Background
The centre of our study is children with special needs who attend day care centres and who have so-called 'individual plans' (IP) in which goals, roles and activities at the centre of the cooperation are documented. The aim of the project is to address how the relations within the cooperative teams around these children are constructed and carried out from the perspectives of the different agents involved in the cooperation.
More than 85% of children between 1 and 5 years of age attend Norwegian day care centres (Statistics Norway 2010). During the last decade there has been a substantial increase in the use of day care centres, making them arguably the most important arena for peer socialization and learning for children under school age. It is the municipal authorities that are responsible for the administration of day care centres, although almost half of the day care centres are owned by private agents. The pedagogical practice in day care centres is universally defined in the national framework plan for day care (Ministry of Education and Research 2006). The plan emphasizes the unique responsibility of day care centres for the inclusion of children with special needs. Accordingly, day care centres have a responsibility to collaborate with external agents in order to ensure that children with special needs have the same opportunities to participate as other children.
Anyone who needs prolonged, combined services is entitled to an individual plan. The right to an individual plan is manifested in the social legislation, the legislation on the rights of the patient, and from 2010 in the legislation on child welfare. The purpose of an individual plan is to provide a complete, coordinated and individually tailored set of services and to ensure that one person (i.e. the coordinator) has the main responsibility for follow-up and coordination at all times. Due to the legislative basis of the individual plan the coordinator usually works in the sectors for health or social services. However, according to the health authorities' Guide to Individual Plans (Helsedirektoratet 2010) the user has the right to influence the choice of coordinator, and the coordinator, therefore, may be a person outside the health or social services.
An individual plan is intended to be a tool for facilitating cooperation, and does not in itself entitle the user to any welfare services. In other words, the plan is supposed to provide a common understanding of goals, roles and responsibility among the cooperating agents and is, as such, a documentation of a cooperative practice meant for internal use. The individual plan may contain other more specified plans, such as an individual education plan (IEP), a habilitation plan, a care plan etc. These plans are commonly connected to a certain area and provide more detailed descriptions of goals and activities within that area. Apart from contributing to a shared understanding within the cooperating unit, the individual plan may be an instrument that facilitates transfers from one sectorial set of services to another. For the children in our sample this is particularly relevant for the transition from day care centre to school, and in cases when the user moves from one municipality to another.

Data and methodology
The project design is an explorative and action-oriented case study. We follow six cases, each consisting of (1) the parents of the focus child, (2) the child's primary contact person in the day care centre (often an assistant), (3) the child's special educator, (4) the coordinator of the team (who most commonly works in the health sector) and/or (5) other relevant agents (e.g. physiotherapist, leader of the day care centre etc.). As mentioned previously, our focus of interest is the cooperative teams, and not the target children directly. However, the teams are constructed in order to meet the needs of the target children, and their situation is therefore of relevance. The requirements of maintaining participant anonymity limit our capacity to provide detailed information about the children. However, common traits shared by all six children are as follows: The children are twoÁfive years of age. There are both girls and boys present in the sample. The children have either physical, mental and/or oral handicaps, and are in need of long-lasting and coordinated effort from diverse agents within the welfare sector.
The cases were selected from two municipalities. In the preliminary planning of the project the researchers consulted diverse municipalities and the choice of which ones to use was based on both practical considerations (such as access) and representation. Although the two municipalities cannot be taken as representative of all municipalities their characteristics in terms of size and organization are regarded as quite 'common' in Norwegian standards. It should also be noted that representation was not requisite in the project. The selection of cases from two municipalities was intended to allow comparisons of the influences of different organizational forms and approaches to steering support services at the local level. However, the selection of cases within the two municipalities was determined largely by availability of participants. When contacting municipal administrators and day care centres we found that there appeared to be fewer children with individual plans than expected and it was not clarified whether children entitled to individual plans actually had one. In both municipalities, 'door-openers' contacted leaders of day care centres attended by children with individual plans. If consenting, the day care centre leaders then provided information about the project to the potential families. The families were then themselves required to contact the researchers if they wanted to take part in the project. With the families' consent we then contacted other members of the cooperative teams to invite them to participate.
The first phase of the project was dedicated to data collection, and can therefore be regarded as the 'research' phase. Our principal data source was individual interviews with three to six persons involved in each case. With a few exceptions the parents (representing the user) were interviewed first, followed by the coordinator of the cooperative team, the child's primary contact person in the day care centre, and the child's special educator. All participants provided written consent. In some of the cases the researchers also observed a cooperative team meeting, where they informed team members about the project and took field notes. Participating at these meetings was initially not defined as part of the data collection, but turned out to provide useful insight.
As a part of the research phase we also conducted two focus group interviews with representatives from two different specialist units in order to address some of the questions arising from the in-depth interviews. The focus groups thus provided supplementary information especially relevant for understanding the collaboration between first and second line services.
The data from phase one were analyzed by the five project team members together. First, the team members explored the data case-wise, identifying and comparing central findings and trends. Secondly, the data were categorized according to topic by means of Nvivo.
The next phase of the project can be regarded as the 'action' phase. This phase involved two seminars. At the first seminar, which was defined as internal, only informants from each case study were invited. The next seminar was external and was directed at agents from sectors that are often involved in this kind of cooperation, as well as students studying early childhood education. The seminars were intended to provide inspiration to a more reflective practice with regard to collaboration around children with special needs, and were both a part of the data collection and dissemination of results. Our aim with the seminars was to share knowledge and encourage reflections that can influence practice in a positive manner. The participants at the first seminar were invited to take part in both heterogeneous and homogeneous group discussions (focus groups) regarding their role in the cooperative team and their ideas about what makes qualitatively good cooperation. Also, the researchers shared some reflections and questions departing from the initial findings, and invited the participants to take part in the analysis. The second seminar was dedicated to dissemination of results and discussion of key topics arising from the findings. In addition to lectures by the researchers the seminar comprised both a short dramaturgic play by professional actors, addressing the challenges of communication for different roles in cooperative teams, and a panel discussion consisting of representative agents and experts (a parent, a public nurse/coordinator, a pre-school teacher, a representative from rehabilitation services and a special education administrator).
The knowledge generated from the project is intended to serve the purpose of improving cooperation within the teams, and the project is as such in line with action research. Action research can be understood as a form of research that derives from practice, serves practice and involves practitioners. In other words, it is considered a democratic form of research intended to provide development and change for the better, and as such it differs from academic research traditions that tend to be based on positivist ideas about objectivity and distance (Reason and Bradbury 2001;Greenwood and Levin 2007). There are numerous perspectives within action research that focus on, among other things, promoting learning, promoting organizational change and promoting emancipation (Herr and Anderson 2005;Freire 1970). These different perspectives do not have to be considered mutually exclusive, however. In our case we integrate aspects from diverse perspectives, acknowledging that research-based actions and practice-based questions might promote both learning, organizational change and emancipated action.
The project design obviously contains some ethical considerations. First, when information is derived from people who know each other and who know about each other's involvement in the research, the requirements of anonymity cannot be met. It is the responsibility of the researcher to ensure that sensitive information is handled carefully, which is a difficult task when the topic of research implies sensitivity. One way to deal with this problem is to generalize the tendencies that emerge from the data and raise causes and questions rather than stating 'facts'. For example, rather than stating the importance of the coordinator for the quality of the cooperation one might point at the circumstances of that role, and how these circumstances might influence the cooperation. Another issue worthy of ethical consideration is the social pressure involved in this multi-perspective approach. Since the quality of the research depends on the participation of specific persons it implies a kind of forced participation, although it is in principle voluntary. Furthermore, as the findings lead us to factors that ought to be improved, there is also a pressure to alter practice. If the project was initiated by the practitioners that would probably not have been a problem, but since our project is initiated by the researchers it might involve the same pressure that applies for participation. On the other hand, all of the informants who agreed to participate in the research project were informed of its goals and process and their participation was clearly optional. It is certainly possible that they agreed to participate because they felt that the objective of the project was important or because other key informants agreed to participate. The project was reported and approved by the Norwegian Privacy Ombudsman for Research at The Norwegian Data Inspectorate (NSD), thus reducing the risk of important ethical offences being made.
Coordination and cooperation: two sides of the same coin?
Coordination and cooperation have, according to Sagatun and Zahl (2003), different characteristics, though they are mutually dependent. Coordination takes place at the organizational level and on the initiative of the leadership, and is as such a 'topdown' phenomenon. Coordination is characterized by formality and duration, and is intended to secure long-lasting cost efficiency and a suitable sharing of workload. Implicitly it involves a reduction of autonomy for the bodies involved. Regular and settled meetings, intentional agreements and laws that regulate collaboration are examples of coordinating activities. We have previously described the individual plan as a coordinating instrument and as a tool for cooperation. Regulated by law, and usually presupposing regular meetings, the individual plan contributes to a certain coordination of the collaborative practice. The roles and responsibilities of the cooperating team are to a certain extent formalized in the plan-document, and the collaboration is intended to have a long-lasting duration.
Cooperation, on the other hand, takes place at the individual level, according to Sagatun and Zahl (2003), and has a more informal character than coordination. Put differently, whereas collaboration between organizations or entities can be understood as coordination, collaboration between persons can be understood as cooperation. Unlike coordination, cooperation is connected to a certain defined interest and has a more limited scope. It presupposes some kind of mutual trust and sharing of responsibility among the persons involved. Cooperation normally departs from a sense of consensus where there is agreement about goals and a shared consideration of efficiency and sharing of work. Within a cooperating team there is thus a potential for learning and development, but also a potential for conflict and inefficiency. Cooperation implies, in other words, a person-dependency which involves certain fragility in the cooperating practice.
To sum up, whereas coordination is formalized, time-limited and rooted in 'top', cooperation is less formalized, occurs more often on an ad-hoc basis and is rooted in 'bottom'. This understanding indicates that cooperation demands a stronger sense of personal involvement than coordination. More to the point, one might say that the framework for collaborative practice is coordination, whereas the content of the collaboration is the actual cooperation taking place among the persons involved. Hence, cooperation is a matter of adapting to the given coordination. Following this argument it becomes relevant to study which mechanisms, apart from coordination, that structure and have influence on cooperation. Based on diverse literature on cross-organizational cooperation, Knudsen (2004) describes three criteria for successful cooperation that we find useful in this respect: compatibility of domain, compatibility of ideology and quality and compatibility of mandate.
The first criterion deals with sharing of work between agents representing different domains. Domain, in this respect, refers to the different agents' mandatory scope and extension. The collaborating agents in our project represent different domains, such as parental socialization, institutional socialization, general supervision of health (public health nurse), special education, physiotherapy, etc. According to Knudsen (2004), cooperation between agents from domains that partly overlap is the most beneficial, i.e. agents that are supplementary to each other and make use of each others' competencies. In contrast, agents representing domains with no overlap have little to offer each other, whereas for agents with complete overlap there is a higher risk of conflict.
The second criterion deals with the correspondence in understandings and ideology between the cooperating agents. How do the different agents understand the 'problem', what do they regard as useful forms of treatment, etc? The ideology will often be based on professional belonging. However, correspondence in ideology is not a precondition for good cooperation, as the diversity in understandings most often is regarded as beneficial and educational. It is rather a question of how the different agents respect and appreciate this diversity. Grimen (2008) stresses the importance of trust between the cooperating agents. He argues that trust facilitates cooperation, as persons with mutual trust can support and supplement each other. Lauvås and Lauvås (2004) point out that cross-institutional cooperation tends to be multidisciplinary rather than interdisciplinary. Whereas multidisciplinarity implies a cacophony of voices and actions, interdisciplinarity integrates this diversity into a synthesized set of actions, and is as such a more beneficial form of cooperation.
The third criterion deals with the understanding of the mandate of the cooperation. A good cooperation presupposes a mandate that is clear and unbiased. Studies have shown that the cooperation benefits whenever there is a clear and mandatory user perspective (e.g. Gautun and Grønningsaeter 2002;Magnussen 2007). It is important to state that the understanding of 'qualitatively good cooperation' is biased in accordance with what the agents involved regard as the subject of their cooperation. For cooperation around a child with an individual plan to be beneficial there is a need for all these three forms of compatibility.
In this article we point at some challenges that contribute to restrain cooperation around children with individual plans, and discuss how cooperation might also be promoted by means of individual plans. We then question the importance of coordination for cooperation, and vice versa.

Experiences with cooperative practices
As our focus of interest in this paper is the relationship between coordination and cooperation we will in this section discuss some experiences with different coordinating practices, drawing on the findings in the project. We point to three areas of coordination and discuss how the different coordination practices influence cooperation. First, we look at some differences as to how the legislation on individual plan is comprehended and practiced in diverse municipalities. Then we look more specifically at the role of the coordinator of the cooperative team. Finally, we look at how the individual plan is used in comparison to other plans.

Challenges at the organizational level
Although the right to an individual plan is regulated by national legislation and national guidelines for individual plan use are readily available, understanding and practice varies across fields and municipalities, who are the executing authorities. According to the national guidelines, the service and/or officer who come in contact with a person who may be entitled to an individual plan has an individual responsibility to inform this person and/or his or her relatives about the individual plan and its availability. In many cases the user will be in contact with agents within the specialist sector (which includes hospitals, medical specialists and rehabilitation services) before the municipal sector intervenes. Hence, the potential applicant will be informed about the individual plan by the specialist unit. This is the case for all of the children in our sample. Many of them were diagnosed immediately after birth. Others have gone through long and intensive examinations by specialists in order to be diagnosed and treated. However, at some point, the users have to apply for an individual plan in the municipality where they live. From a principle perspective it has been argued that since the individual plan is defined as a right, and not as a special service for persons in need of long-lasting and coordinated services, it should be unnecessary having to apply for an individual plan.
The institutional transfer from the specialist sector to the municipal sector (which in this context covers family doctors, public health nurses, physiotherapists, nursing, habilitation and rehabilitation) involves some challenges. First, there seems to be a gap of knowledge about mandatory practices between the specialist agents and staff in the public administration. The specialists are often unfamiliar with the application process and routines for selecting coordinators, and have little possibility to follow an individual plan closely due to their wide zone of action and other responsibilities.
We are for instance not supposed to take part in all team meetings. It is simply a question of priorities. (Focus group with specialists) Public administrators, on their hand, are often left with a sense of having been designated the task of following up on a process that they have not been involved in and, consequently, sometimes perceive the specialists as too detached. She [the doctor at a specialist unit] obviously has many children and is busy and all that, but . . . and she has been here maybe once or twice a year, but I could have wanted her to be more available to us. (Coordinator) Secondly, there exists no national template for the individual plan. As a consequence, the municipalities have different formats and procedures, as do the specialist-units. Although the content of the individual plan is the same regardless of whether it is localized within the municipality or specialist institution, the adaptation of content to a different template often becomes a demanding and frustrating task for the coordinator. With reference to Knudsen's (2004) criteria for successful cooperation, the experiences reported in our project indicate insufficient ideological compatibility between the municipality and specialist sectors. There is reason to believe that the cooperation would benefit from an increased sharing of understandings, routines and possibilities.
The informants in our sample live in two different municipalities. The largest one practices a so-called purchaserÁprovider split model (Hansen and Ramsdal 2005) in which an independent and separate service-agency submits the application for an individual plan and is responsible for designating the coordinator, though in dialogue with the user. For children it is common practice that the public health nurse is appointed as coordinator. The purchaserÁprovider split model often involves a long application process, and, although the risk of refusal is small, the process is often perceived as tiresome and unnecessary. In the smaller municipality there is a so-called system coordinator who is responsible for the follow-up of all individual plans Á not necessarily in the role of a coordinator herself, but more as an administrator and advisor for the coordinators. In this municipality it is also common for the public health nurse to be appointed as coordinator when the user of the individual plan is a child. Thus, four of the children in our sample have a public health nurse as coordinator for their individual plan, whereas two children have respectively a physiotherapist and social worker from the specialist field. In the following section we look at the informants' experiences with the coordinator.

Coordination of individual plans
The role of the coordinator of the individual plan is to follow up the individual plan: The coordinator's most important task is to ensure that the planning work progresses properly, that you receive the information you require and that you, in accordance with the principle of user participation, have a say in the planning work. (Norwegian Directorate of Health 2010) According to the guidelines for individual plan use, the user's wishes should be given priority when the coordinator is appointed. However, as we have pointed out, it is common for the coordinating unit in the public administration to appoint a coordinator from the health or social services. For children that person is normally the public health nurse. Research on the coordinator role (Thommesen, Normann, and Sandvin 2006;Hansen 2007) stresses the importance of the relationship between the coordinator and the user for the quality of the individual plan. According to this research, the users often consider the coordinator as the most important person in the cooperative team. His or her role is multifaceted Á from being the users 'guide' in the welfare system, to being an advocate for the user in relation to other sectors in the welfare organization, to being a close and intimate contact person, or even confidant. Thus, the coordinator has to balance between closeness and distance to the user.
The four public health nurses who function as coordinators in our project all question their role. They have all been appointed as coordinators by the municipal coordinating unit, and not by the child's parents themselves. With one exception, they had all met the child and his or her parents at the regular health controls during the child's first year, but none of them knew the users well prior to being appointed coordinator. Apart from the regular meetings in the cooperative team they have irregular and infrequent contact with the children and their families. All of these coordinators find their role difficult, mainly due to their lack of knowledge concerning the user, but also because their role is undefined. Rather than being a close contact person and a constructive initiator, they perform the role of secretary (e.g. leading the meetings and writing the plan). In some cases they are the persons in the team who know the user the least. The following citation is taken from an interview with a coordinator: [. . .] It is [the child] we are supposed to cooperate for, in a way, so I feel that what I am doing is actually updating the plan and participating at the meetings and in the individual plan group. I am there a bit like an observer: listening to how he's doing, without me actually taking part in the treatment or anything like that [. . .] Our frustration has been that we are functioning a bit like secretaries for things we are not actively taking part in ourselves. [. . .] I think it is nice to know what's happening with the children, but maybe we could have received a summary Á that would have been just as useful for us as taking part and being secretaries for something we are not actively a part of. So yes, I think it would have been more appropriate if one of the other persons, who are actively involved in the treatment also had the responsibility for the individual plan. But I am aware that nobody else really wants to, so then it is left to the public health nurse [laughter].
The parents in these cases share this opinion. They also question the role of the coordinators due to their lack of contact with and knowledge of the child. Often they seem unaware of the fact that the coordinator has been appointed to the role. The same seems to apply for the day care centre personnel.
In these cases the meetings in the cooperative team are report-like. The different agents in the team report about their 'speciality' (i.e. domain). For instance, the agent may describe how the child interacts with peers in the day care centre, how his or her language develops, or how he or she is developing physiologically. The role of the coordinator is to lead the meeting, take notes/updating the individual plan and sometimes ensure that special concerns will be followed up on by the person in charge of that domain. All our informants ask whether it would have been more suitable for the day care centre personnel to be coordinators as they interact more closely with the child and are in position to follow him or her up on a daily basis. However, as individual plan is founded within the legislation on health and social affairs, and not the legislation on education, the day care centre, or alternatively school-personnel, has no obligation to take the responsibility as coordinator.
In the two remaining cases the coordinators are a physiotherapist and a social worker. Like the public health nurses, the physiotherapist has been appointed coordinator, though through a process of dialogue with the parents. Unlike the public health nurses she meets the child regularly in the day care centre and has regular contact with the child's parents. All the informants in this case are satisfied with the coordinator. In the last case the coordinator works in the specialist field, and proposed herself, in dialogue with the child's parents, to continue as coordinator after having initiated the individual plan. This is a clear break with the normal practice. It is interesting to notice that this coordinator regards her role slightly differently from the other coordinators, and seems less influenced by the normative expectations that apply for the other coordinators. In this case the parents are satisfied with the way she conducts her coordinating practice. The other informants in the case (i.e. day care centre teacher and special educator) on the other hand, are unaware of her being the coordinator and have attended few meetings in the cooperative team. It seems that whereas the parents benefit from having a 'close' coordinator, the cooperative team as a whole is less integrated than in the other cases. This might be due to the different manner in which the coordinators conduct their roles.
Based on our findings we question the practice of coordinators being appointed from an external agent that has little regular contact with the child and his or her parents. As research suggests (Thommesen, Normann, and Sandvin 2006), we find the relationship between user and coordinator to be an important aspect of cooperation as a whole and that cooperation benefits from coordination that is closely linked to the user's wishes and needs. We acknowledge that this represents some challenges at the organizational level. The personnel in the specialist sector are often in close contact with the user-family for a period and are often the ones to initiate an individual plan. However, follow-up in the municipal sector demands resources of time and space that the specialist sector tends to lack. Furthermore, coordination within the municipal sector requires knowledge, which is to a large extent tacit, about how this sector works. Likewise, personnel in the educational sector lack resources to follow-up the individual plan, nor are they in a legal sense obliged to do so. Another objection to them being coordinators has to do with the long-term perspective of an individual plan. As the child after a few years leaves day care centre for school, the primary contacts are replaced. An important aspect of the coordination of an individual plan is to ensure continuity and stability through such transitions. This is partly addressed by special educators who follow the child in their first year at school. However, follow-up in all domains requires a more long-term and holistic approach.
The fact that coordinators often lack the possibility of engaging in a close and long-term relationship to the user leaves a void of unrealized potential for highquality follow-up of a child with an individual plan. Our next question then is whether the plan-document might provide the information and 'closeness' that can replace the function of the coordinator?
The use and importance of the individual plan as a document Defined as 'a tool for cooperation' (Norwegian Directorate of Health 2010) the individual plan is intended to be an instrument that contributes to coordinating activity around the user. As mentioned previously it is common (but, according to the guidelines defined by Helsedirektoratet (2010), not a prerequisite) that the individual plan follows a certain template. However, it is not the document, as such, that is the purpose of the individual plan, but rather the process of defining goals, sharing responsibility and ensuring activity. As goals, roles and activities tend to change in line with the situation of the user, the individual plan, if used as intended, has to be evaluated and changed accordingly. However, there appears to be great variation as to whether and how the individual plan is being used. Based on our interviews we found that the individual plan often becomes a 'desk-document' Á once written it ends up in the drawer of the coordinator and remains there until someone is reminded about it. At that point it tends to be out of date and needs to be fully revised. Used this way, the potential for the individual plan to be a live and informative document is unrealized. On the other hand, if the individual plan becomes an integral part of the meetings and is updated accordingly the document might serve several purposes. First, it might replace accounts from meetings by centralizing them into that one document. Not only does this provide a more practical solution to accessing information, it also provides a process that in itself might have a pedagogical value. Furthermore, formulating shared knowledge and practice in writing contributes to minimize misunderstandings and refusal of responsibility. Furthermore, an updated and 'live' document is an important communicative devise that may contribute to ease the transition of the user from one institutional service to another, from one municipality to another, from one coordinator to another and so on.
However, it is obvious that for the individual plan to function as intended and fulfil the commitment as an informational and pedagogical tool for cooperation, someone has to be responsible to follow up the document, and generally that person is the coordinator. Hence, the individual plan cannot be regarded separately from the coordinator. Still, as a 'live' document the individual plan (and the adherent cooperation) becomes less dependent on a particular person, as all activities and plans are communicated in writing. As such it ensures a better cooperation also in absence of, or in case of change, of coordinator.

Cooperating across different domains
In the introduction we made it clear that the individual plan might consist of various more specified and detailed plans, such as an individual education plan (IEP), a care plan or a habilitation plan. These plans are usually in the hands of specialists and deal with a specific area or domain. For instance the special educators in the cooperative teams are in charge of the child's individual education plan, in which goals and activities related to the development of language, learning and educational activities are defined. In our study we have a special focus on the role of the day care centres in the cooperative teams. Prior to the recruitment of informants we made a request to several day care centres in the selected municipalities through the municipal day care centre offices. When we specified that we wanted to look at the issue of cooperation around children with an individual plan, it was generally mistaken for individual education plan, and there appeared to be little knowledge about the individual plan as a separate practice. Accordingly, there appeared to be relatively few children with individual plans in the day care centres in the two municipalities when considering the number of children with special needs who might be entitled to have an individual plan.
Our impression is that the individual plan and the individual education plan often are considered to belong to different domains, although the individual plan has a broad and holistic approach that is intended to synthesize all domains involved in the cooperation. This tendency came to the surface in the actual cases where the special educators interviewed talked about the individual education plan as 'their' plan and expressed unfamiliarity with the individual plan.
It is one of his plans Á the individual education plan is mine. [. . .] Because it is that one I deal with, and which is important to me.
Such separate understandings and practices contribute to a multidisciplinary perspective at the cost of the intentional interdisciplinary perspective of the individual plan. The individual plan is intended to integrate different perspectives from different fields in the cooperative team. As Knudsen (2004) points out, cooperation tends to be most useful when there is a certain overlap in domains. Our impression is that in order for the representatives from the different domains in the cooperative teams to have an integrated practice that allows for the full potential of 'the interdisciplinary' to be realized, it takes a readiness to appreciate this interdisciplinary need. In other words, it seems as if Knudsen's criterion of ideological compatibility is at stake. Our analyses of the cooperative teams show that there tend to be two major domains represented (Figure 1), and that at times there seems to be a missing link between them. Figure 1 is an example of cooperation based on experiences from different agents. The arrows in the illustration indicate cooperative relations; the 'thicker' the arrow, the more extensive is the cooperation. As we can see, there are only two relationships 'crossing' the two domains; that of the physiotherapist and the day care centre director and teacher respectively. As the physiotherapists tend to work with the children in the day care centre they come in contact with the educational domain. In this figure the parents (i.e. the users) are in the centre of the cooperation and have a relation to both domains, whereas the coordinator, who is supposed to be the node of the cooperation, is more peripheral. Although the picture varies from case to case, this illustration points at an important trend: the lack of contact between the two domains.
One important potential of the individual plan and the coordinator might be to fill the missing link between the two domains, thus ensuring a more holistic approach. When this is successful the individual plan might be a coordinating instrument that makes the conditions favourable for cooperative and joint action across different domains.

Conclusion
We find that good coordination might promote cooperation, under the presumption that the cooperation is cherished and regarded as useful for the persons involved. Our findings, thus, underpin Knudsen's criteria for successful cooperation (2004). On the other hand, we find that 'forced' coordination, illustrated by the practice of delegating individual plan coordinators to families, is experienced as overly demanding and is likely to be a barrier to cooperation. It appears, then, that cooperation benefits from a bottom-up perspective, in which it is the aim of the cooperation that defines the coordination, and not vice versa.
Drawing on empirical findings, our study provides a contribution to research on institutional collaboration. However, the scope of the study does not allow for extensive analyses of how the coordinationÁcooperation gap might be traversed. In order to fully understand collaborative practice there is a need to explore the understandings and the knowledge upon which the collaboration is based, involving the different actors' knowledge about each other and their perceptions of successful collaboration and obstacles. This study partly meets this request, but does not allow for extensive comparison. Furthermore, there is a need for more quantitative approaches seeking to map collaborative practice and understandings from different perspectives, and particularly in relation to children. Our study points at some tendencies and raises questions that ought to be further investigated, and as such it provides valuable knowledge about cooperation and coordination around children with individual plans. Furthermore, the study puts day care centres on the map as collaborative agents, thus recognizing their role as important actors in the collaboration around children with special needs.