Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools: A Randomised, Controlled Trial

How to cite

Federici, Roger Andre; Helleve, Arnfinn; Midthassel, Unni Vere; Bru, Lars Edvin; Flatø, Martin. 2017. "Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools: A Randomised, Controlled Trial." AEA RCT Registry. AEARCTR-0002628. https://doi.org/10.1257/rct.2628-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Roger Andre Federici* , Arnfinn Helleve , Unni Vere Midthassel , Lars Edvin Bru , Martin Flatø

* Primary investigator

Registry ID
AEARCTR-0002628
DOI
https://doi.org/10.1257/rct.2628-1.0
First registered
Last updated
Status
In development
Trial dates
2017-09-01 – 2020-07-01
Countries
Private
JEL codes
I18, I22, I28, J13, J18, J24
Registry record
http://www.socialscienceregistry.org/trials/2628

Abstract

The project ‘Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools’ is part of a larger research programme financed by the Norwegian Directorate of Education. The main aim of the programme is to investigate how different professionals, working in collaboration with the school staff, may contribute to students’ learning environment and learning outcomes. The present study is a randomised, controlled trial whose principal aim is to investigate how an increased school nurse resource at randomly selected schools in 14 Norwegian municipalities affect the students’ psychosocial environment, and secondarily, their broader learning environment and academic outcomes.

The project is a cluster-randomised controlled trial where four schools in each of 14 municipalities are randomly selected to receive a 12.5% position increased school nurse resource from January 2018 to December 2019. The school nurse should work in a structured and systemic collaboration with the school and the resource should target 5th – 7th grade (age 10-12). In the project, systemic collaboration means that school nurses are involved in the school’s overall work with keeping overview of the health and well-being of the students, including initiating and implementing universal and preventive measures targeting psychosocial aspects of the learning environment. Structured collaboration means that schools and school nurses organise their collaboration by regular meeting/contacts in contrast to collaboration based on irregular and arbitrary contact between schools and the school health services only. Systemic work is described in the revised and recently passed guidelines for the school health service (The Norwegian Directorate for Health, 2017) and the project provides guidelines for the meeting series between the school nurse and staff at schools.

Intervention

The dosage in the project comprises of a 50 percent position increase in each participating municipality’s school health service. More specifically, the additional resource is to be used to increase the presence of the school health service at the four schools in the treatment group. This comprises of a 12.5 percent position additional availability of the school nurse at each school – targeting the students in 5th, 6th and 7th grade.

The additional school nurse resource should provide services in line with the revised and recently passed guidelines for the school health service (The Norwegian Directorate for Health, 2017). These guidelines highlight that the school health service should work systemic and in structured collaboration with the staff at school. The school nurse should collaborate with the schools in getting overview of the student population’s health and well-being, and identify possible areas where the school nurse could contribute with measures regarding health-related issues, prevention, and measures for all students or groups of students in 5th to 7th grade with an emphasis on improving the students' psychosocial environment. Thus, systemic means working in a universal and preventive manner with psychosocial aspects of the learning environment.

Research reports indicate that there is a lack of a clear theoretical conceptualisation of the construct of ‘systemic’ in the Norwegian school context (Bliksvær, Hannås, Hustad, & Strømsvik, 2015; Hustad, Strøm, & Strømsvik, 2013). Hustad et al. (2013) propose, based on evaluations of the Norwegian Educational and Psychological Counselling Service, to distinguish between three meanings of systemic depending on the system level at which the professional operates. First, systemic may mean work conducted within the children and adolescents’ psychosocial environment, i.e. the social system which the students are a part of. Interventions at this level may concern individuals, groups, or classes. However, these interventions should to a large degree be beneficial for the whole student population and include common goals, coordinated efforts and collaboration between the school health service and school staff. The second meaning is systemic work understood as conducted within the school as an organisational system. With this meaning, the school health service should contribute on a more strategic basis, for instance by initiating school-wide preventive measures or by introducing programs for reducing bullying. Finally, a third understanding sees systemic work as interacting with other public services related to the school. At this level, the school health service may, for instance, have a coordinating role for collaboration with municipal and national public services and other agencies. In the present study, the additional school nurse resource may contribute and act on all these three levels.

In addition to collaborate with the school staff in a systemic way and in accordance with the guidelines, the intervention is also structured by a set of criteria. That is, the municipal authority should have a plan for how to organise the additional resource, the schools should have plans for how frequently the meetings will take place and what will be discussed, and how measures are followed up. Although the municipal authority, the schools, and the school health service are given local autonomy, some structure is prerequisite for being able to work systemically and for collaboration not to solely rely on individual arrangements.

Experimental design

The project ‘Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools’ is a cluster-randomised controlled trial where four schools in each of 14 municipalities are randomly selected to receive a 12.5% position increased school nurse resource from January 2018 to December 2019.

The trial will test the hypotheses that additional school nurse resources contribute to the four primary outcomes defined in this protocol. These four outcomes will be measured at five points in time, before the trial begins at t0 and each semester during the trial (t1-t4). The main analysis will use all time points t1-t4 controlling for the levels in t0, but a separate analysis will also be conducted at each time point t1-t4. Specifically, we are interested in whether the nurse-to-student ratio affects the outcomes (Guttu, Engelke, & Swanson, 2004). Using the nurse-to-student ratio is a convenient way of handling three factors that may confound a reduced form estimate. First, the school size in our study varies considerably from 20 students in the 5th – 7th grade target group, to 238 students. We would expect that the same absolute increase in school nurse coverage would have a larger impact in a small school than in a large school. Secondly, school nurse coverage is decided at the municipal level and thus varies considerably between municipalities. Furthermore, the nurse coverage is likely to increase during the period under study. Thirdly, there is a potential issue of partial compliance (Glennerster & Takavarasha, 2013) and we want to be able to conduct the analysis even if compliance is sub-optimal. Our main specification thus uses the instrumental variables method to estimate a local average treatment effect (Angrist & Pischke, 2008), given that we obtain a first-stage F-statistic of 10 or above. Results will be interpreted as effects of an increase in the nurse-to-student ratio from the mean which corresponds to the increased resource.

Design details

Randomization method
Causal inference based on this trial relies on the comparison of schools that are randomly selected to receive an extra school nurse resource with schools whose nurse-to-student ratio is unaffected by the trial. For this trial, we will use a stratified, cluster-randomised design. First, the selection of schools is stratified by municipality. Within each municipality, four schools are selected to receive the treatment and all other schools that fulfil the selection criteria are followed as a control group. The reason for stratifying at the municipal level is both practical and purposeful. As the responsibility for the school health service lies with the municipality, an equal allocation of resources to each municipality was necessary for recruitment. Since factors that are likely to influence learning environment (such as socio-economic background, prior nurse coverage, and school quality) varies between municipalities, stratification at this level will improve balance. We then stratify a second time within each municipality, based on the measure of well-being and bullying. This is done to improve the balance between the control and trial groups along relevant dimensions. We here follow Athey and Imbens (2017) who argue that stratification until there are two treated units within each stratum is the method that leads to the smallest standard errors, tangent to other methods such as re-randomisation. Each municipality is thus divided into two strata, to balance both the combined well-being indicator and the bullying indicator. To do this, we follow a similar strategy as that used by Greevy, Lu, Silber, and Rosenbaum (2004) and King et al. (2007) for optimal multivariate matching before randomisation. Within each municipality, we rank each school according to student well-being and bullying. With these two rankings, we calculate the Mahalanobis distance to the top ranking (1,1). Each municipality has between 7 and 13 eligible schools, and the lowest performing half based on the Mahalanobis score become one stratum and the highest performing half another stratum. In the cases where there is an uneven number of schools in the municipality, a random school will be randomly placed in the lower or the upper stratum. A number of balancing tests will be carried out to assess whether the control and treatment groups are comparable. Variables that will be compare include the nurse-to-student ratio prior to the intervention, school nurse years of practice prior to intervention, and the primary outcomes.
Randomization unit
Schools.
Clusters
123 schools
Observations
11,105 students per annum
Arms
56 treatment schools
Primary outcomes
There are four primary in the study: 1. Student emotional well-being 2. School belonging 3. Bullying 4. Students attendance
Secondary outcomes
There are elleven secondary outcomes in the study, within the topics motivation, academic self-concept, social well-being, bullying, work environment, instrumental support, emotional support, meals, physical education attendance, relation to school nurse, and results on national tests.

Cited by

No citation data available for this registration yet.

Keywords