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

Last registered on December 23, 2017

Pre-Trial

Trial Information

General Information

Title
Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools: A Randomised, Controlled Trial
RCT ID
AEARCTR-0002628
Initial registration date
December 21, 2017

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 23, 2017, 6:25 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
Nordic Institute for Studies in Inovation, Research and Education (NIFU)

Other Primary Investigator(s)

PI Affiliation
Norwegian Institute of Public Health
PI Affiliation
Centre for Learning Environment at the University of Stavanger
PI Affiliation
Centre for Learning Environment at the University of Stavanger
PI Affiliation
Nordic Institute for Studies in Innovation, Reseach and Education (NIFU)

Additional Trial Information

Status
In development
Start date
2017-09-01
End date
2020-07-01
Secondary IDs
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.
External Link(s)

Registration Citation

Citation
Bru, Lars Edvin et al. 2017. "Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools: A Randomised, Controlled Trial." AEA RCT Registry. December 23. https://doi.org/10.1257/rct.2628-1.0
Former Citation
Bru, Lars Edvin et al. 2017. "Increased School Nurse Resource in Systemic and Structured Collaboration with Norwegian Primary Schools: A Randomised, Controlled Trial." AEA RCT Registry. December 23. https://www.socialscienceregistry.org/trials/2628/history/24448
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
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.
Intervention Start Date
2018-01-01
Intervention End Date
2019-12-31

Primary Outcomes

Primary Outcomes (end points)
There are four primary in the study:
1. Student emotional well-being
2. School belonging
3. Bullying
4. Students attendance
Primary Outcomes (explanation)
The four primary outcomes measure a psychological dimension of well-being, a social dimension, and student absence. The social dimension comprises of two subdimensions.

The psychological dimension comprises of students’ perceived emotional well-being at school and focuses on affective states and emotional responses in class during the last week. The measures are inspired by the core affect scale developed by Russell (2003), however, we use a short version consisting of five items comprising of both positive and negative affect. A similar short version has been employed in previous studies such as in the Children’s World survey (Rees & Main, 2015) and has successfully been administered to Norwegian 5th to 7th graders in the Ungdata Junior project. In the present study, responses are given on a five-point Likert scale ranging from ‘never’ (1) to ‘always’ (5). In our analyses, the scale will be used as a composite measure indicative of students’ emotional well-being. Note that the measure has been subjected to quantitative and qualitative piloting with students in 5th and 6th grade. Moreover, the initial analyses will include confirmatory factor analyses to ensure scale reliability and validity. Items that do not reach statistical significance and other measures of goodness of fit statistics such CFI, IFI, TLI, RMSEA (Hoyle, 2012; Kline, 2011; Tabachnick & Fidell, 2007) in baseline data will be excluded. For the CFI, IFI, and TLI indices, values above .90 are typically considered as acceptable, whereas values greater than .95 indicate a good fit (Hoyle, 2012; Hu & Bentler, 1999; Kline, 2011). For well-specified models, an RMSEA of .06 or less reflects a good fit (Byrne, 2010; Tabachnick & Fidell, 2007).

The social dimension comprises of school belonging (or relatedness) and bullying. OECD defines a sense of belonging as a feeling of acceptance and being liked by the rest of the group, feeling connected to others and feeling like a member of a community (Baumeister & Leary, 1995; OECD, 2017). School belonging is measured using six trend items previously used in PISA 2012 and PISA 2015. Responses are given on a five-point Likert scale ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. In PISA 2015, the reliability of the scale (the Norwegian questions) was .86. Note that the answering format in PISA is a four-point Likert scale with the answering categories ranging from ‘strongly agree’, ‘agree’, ‘disagree’, and ‘strongly disagree’. We chose to include a ‘neutral’ option to increase reliability – a five-point Likert scale is the standard response format in the Norwegian Pupil Survey. Moreover, note that the question regarding students’ perceptions of loneliness at school is an item already included in the compulsory part of the Norwegian Pupil Survey.

Bullying is measured by means of one item and the question is compulsory in the Norwegian Pupil Survey. The students are asked whether they have been bullied by other students at school during the past few months. The response categories are ‘not at all’, ‘rarely’, ‘2 or 3 times a month’, ‘about once a week’, and ‘several times a week’. In the literature, there is a lack of consensus regarding the frequency of bullying that should occur to be defined as bullied. For instance, Olweus (2013) suggest 2 or 3 times a month, while Roland (1999) propose once or several times a week. In research reports analysing the Norwegian Pupil Survey data, a student is defined as bullied if he or she experience bullying 2 or 3 times a month or more (Wendelborg, 2017). Moreover, the research reports provide information regarding the criteria for excluding not reliable/unserious responses. Respondents who states that they experience bullying from others, from teachers, and cyberbullying several times a week, and in addition states that they bully others on the same questions, are excluded from the analysis. In the Norwegian Pupil survey 2016 this amounted to 0,1 percent (623 students) of the respondents (Wendelborg, 2017).

The final primary outcome is student absence. Each semester (spring/autumn) the participating municipalities will provide de-identified absence data for all pupils in the target grades for each school to the researchers. The data are structured on the individual level with a unique identification number for the individual student and comprise of registrations for each day that semester. Note that the researchers will not know the reasons for students’ absence.

Items (English)

Recall how you`ve felt last week in class. How often have you felt the following?
Been happy
Been sad
Been stressed
Been bored
Had fun

Other students seem to like me
I make friends easily at school
I feel like I belong at school
Do you sometimes feel lonely at school
I feel different than others and out of place in my school
I feel like an outsider (or excluded out of things) at School

Have you been bullied by other students at school during the past few months?

Secondary Outcomes

Secondary Outcomes (end points)
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.
Secondary Outcomes (explanation)
As with the primary outcomes, the secondary outcomes comprise of a psychological dimension, a social dimension, and a cognitive dimension. The psychological dimension comprises of three subdimensions whereas the social dimensions comprise of five subdimensions. The cognitive dimension is measured by the students’ results on the Norwegian National Tests. Also, we include student absence in physical education and a measure of implementation as secondary outcomes.

The psychological dimension comprises of students perceived motivation, academic self-concept and social well-being at school. The questions tapping students’ motivation are compulsory in the Norwegian Pupil Survey and focus on interest and liking for schoolwork – a conceptualisation of motivation corresponding to the theoretical framework of self-determination theory. The theory defines intrinsic motivation as the inherent pleasure and satisfaction derived from engaging in an activity and a main postulate is that social factors promote intrinsic motivation via satisfaction of individuals’ basic needs for autonomy, competence, and relatedness (Deci & Ryan, 2000; Ryan & Deci, 2000). The measure comprises of three items and previous analyses have shown a Cronbach’s alpha of .77 – .79 (Federici et al., 2016; Wendelborg, Røe, & Caspersen, 2016) . The responses are given on a 5-point scale ranging from ‘not in any subjects/not at all’ (1) to ‘in all subjects/to a large degree’ (5). Note that this measure is a composite of questions with different response categories. Therefore, one may question to what degree it functions as a scale or an index. Previous analyses of the Norwegian Pupil Survey (Wendelborg et al., 2016; Wendelborg et al., 2014) indicate that the measure function as a scale indicative of students’ motivation.

Self-concept is measured by a four-item scale representing a short version of the subscale of a Norwegian version of the Self-Description Questionnaire SDQ II (Marsh, 1992; Skaalvik & Rankin, 1992). In general, self-concept is a multidimensional construct referring to self-perceptions in different areas or domains (Bong & Skaalvik, 2003; Marsh, Byrne, & Shavelson, 1988). Academic self-concept is often defined as students’ perceptions of doing well or poorly in school in general (general academic self-concept) or domains, for instance, mathematics self-concept. In the present study, we measure general academic self-concept. Responses are given on a five-point Likert scale ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. Previous studies have revealed a Cronbach’s alpha of .77 (Skaalvik & Skaalvik, 2013).

The questions tapping into students’ social well-being is an index focusing on perceptions of well-being at school in general, in class, and between lessons (break/playtime). The responses are given on a 5-point scale ranging from ‘I don’t thrive at all/not at all’ (1) to ‘I always thrive/to a large degree’ (5). This measure is a composite of questions where one of the items have a different type of response categories. We will conduct confirmatory factor analyses to investigate the behaviour of this item and exclude it if it does not reach statistical significance and other measures of goodness of fit statistics (described earlier). Preliminary analyses of data from the Norwegian Pupil Survey 2016 reveals a Cronbach’s alpha of .81.

The social dimension of the secondary outcomes comprises of students’ perceptions of bullying, their work- and social environment, perceived emotional and instrumental support from teachers, and a question regarding school meals.

The Norwegian Pupil Survey includes additional items tapping into different aspects of bullying at school. The first additional item included in the present study is related to the item concerning bullying defined as a primary outcome. It asks the students whether the school provided help. Note that this question is only given to students who report that he or she experience bullying 2 or 3 times a month or more. The response categories are ‘no, no adults knew anything’, ‘the school knew, but didn’t do anything’, ‘yes, the school provided measures, but it didn’t help’, ‘yes, the school provided measures, and it helped a bit’, and ‘yes, the school provided measures and bullying stopped’. Moreover, all students are asked whether they have experienced cyberbullying or been bullied from teachers. Finally, two items ask the students whether they bully others, both at school and cyberbullying. The response categories for the latter items are ‘not at all’, ‘rare’, ‘2 or 3 times a month’, ‘about once a week’, and ‘several times a week’.

The students’ perceptions of their work environment are measured by three compulsory items in the Norwegian Pupil Survey. The questions focus on the students’ opportunities to work undisturbed, to what degree the class in general perceive working hard with school work is important, and to what degree the teachers consider mistakes to be part of the learning process. Responses are given on a five-point Likert scale ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. Previous studies have revealed a Cronbach’s alpha of 0,66 (Wendelborg et al., 2016).

Research identifies several dimensions of teacher support, such as emotional, informational, appraisal, and instrumental support (House, Umberson, & Landis, 1988; Malecki & Demaray, 2003). The number of dimensions and the labels used for them varies. However, in general, the categories of emotional and instrumental support are typically reported (Semmer et al., 2008). Emotional support is characterised by empathy, friendliness, encouragement, esteem, and caring, whereas instrumental support is characterised by tangible support, for instance, when teachers help students solve a problem or accomplish a difficult task. In the present study, students’ perceptions of the teachers as emotionally supportive is measured by three items. The scale is a shortened and modified version of a previously tested scale of emotional support (Federici & Skaalvik, 2014b; Skaalvik & Skaalvik, 2013). Previous studies have revealed a Cronbach’s alpha of .80 and .94. The responses are given on a 5-point scale ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (5).
Three items measure the students’ perceptions of teacher instrumental support. The questions are compulsory in the Norwegian Pupil Survey and focus on tangible support and to what degree students ask for help when needed. The responses are given on a 5-point scale ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (5). Preliminary analyses of data from the Norwegian Pupil Survey 2016 reveals a Cronbach’s alpha of .71.

Finally, the social dimension includes one item concerning school meals. The question is developed for the present study and the students are asked to what degree the meal break is characterized by ‘calm and order’. The responses are given on a 5-point scale ranging from ‘strongly disagree’ (1) to ‘strongly agree’ (5)

The cognitive dimension comprises of students results one the Norwegian National Tests. The National Tests were introduced in Norway in 2004 as part of a quality assessment system in education. These tests are run every autumn on 5th, 8th, and 9th graders and focus on core academic skills namely numeracy, literacy, and English. The main purpose of the tests is to provide educational authorities at local and national level with information on general student competency after the 4th, 7th, and 8th year of compulsory schooling.

We will use two types of test that are already implemented in schools. The national test in reading, English and mathematics for 5th grade students will be used to measure pre-intervention levels in students’ achievement. These national tests are available for the whole Norwegian student population, with a few exemptions. The cohort born in 2006 took the 5th grade test in the fall of 2016. They will take a national test in the same subjects in 8th grade in the fall of 2019 when students who attended the treatment schools have been exposed to 1.5 years of an extra school nurse resource. We will compare development in test results in 8th grade between treatment and control schools for this cohort. Depending upon additional funding, we will also conduct a study of the 2007 cohort who took the 5th grade test in 2017 and will take the 8th grade test in 2020, to measure the impact of the full two years of intervention, and studying effects on the two cohorts combined.

In addition to obtaining data of student absence in general, we wish to investigate the prevalence of absence during physical education classes. Note that there are some uncertainties about whether it is possible to get these data from the municipal authorities.

We also include a measure related to the implementation and process evaluation in the Norwegian Pupil Survey. The students are asked two questions regarding to what degree they know the school nurse and to what degree the school nurse is an ‘adult that is easy to talk with’. The main aim of these questions is to investigate possible differences in perceptions of the school health service between the treatment and control schools.

Experimental Design

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.
Experimental Design Details
Specifically, we estimate two equations in this project, specified in the pre-analysis plan. The first approach is an intent-to-treat effect. In equation (2), the nurse-to-student ratio is the explanatory variable which is instrumented by the treatment status of the school. Standard errors are clustered at the school level.

The outcome variables are constructed in the following way. For the questions on ‘emotional well-being’, the two positive emotions (been happy and had fun) will be given values from 1 (never) to 5 (always). The three negative emotions (been sad, stressed, and bored) will be given values from 5 (never) to 1 (always). An average will be calculated for each child, and we will conduct a log transformation. A similar outcome variable will be constructed for ‘school belonging’ where responses to the three positively loaded questions will be given values from 1 (completely disagree) to 5 (completely agree) whereas the three negatively loaded questions will be given values in opposite order. An average will be calculated per child and the variable will be log transformed. Bullying will be studied using a linear probability model. All these three outcomes will be studied using two-stage least squares, whereas a Poisson regression using generalised method of moments will be estimated for days of absence per semester as outcome (Wooldridge, 2010).

With four outcomes, we will correct the critical levels for rejecting null hypotheses for multiple hypotheses testing. However, the Bonferroni correction method is too restrictive if the hypotheses are correlated, which is an assumption in our study. We will therefore use the false discovery rate method developed by Benjamini and Hochberg (1995) (see also Fink, McConnell, and Vollmer (2014)).
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.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
123 schools
Sample size: planned number of observations
11,105 students per annum
Sample size (or number of clusters) by treatment arms
56 treatment schools
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The sample used in this study consists of an estimated 11,000 students in grade 5-7 at each point in time, distributed across 14 municipalities and 123 schools. The sample size was determined based on an initial power analysis. With access to more data, we present an updated analysis of the detectable effect size in the pre-analysis plan, which estimates this to be below 0.1 standard deviation. For most of the measures that we will use in the study, previous measures are not available. Furthermore, data on learning environment from the Pupil’s survey is only available for 7th grade. One of the measures we will use as a primary outcome which is available is the number of children who report being bullied 2-3 times per month or more frequently, a limit recommended by Olweus (1991). The intra-class correlation between schools was 0.02 on this variable, using data from 2016. For the power analysis, we will use another measure which consists of two questions about well-being in the Norwegian Pupil Survey, conducted among 7th graders in 2016. This is not one of our primary outcomes, as the primary outcomes other than bullying are not available in previous surveys that we have access to at the time of randomisation. However, we believe the answers will be highly correlated with our primary outcome variables. The responses to the questions on well-being are given values 1-5 where 5 is the most positive response. The statistic is then standardised and thereafter averaged across the two items. This average is then standardised again to create a combined indicator at the school level. The intraclass correlation of this variable is 0.05. We have used this two-item measure of well-being as the basis for the power analysis, since we have data on this indicator from 2015 and 2016. Unfortunately, we only had 23 schools and 913 7th graders in participating schools in our sample. Therefore, we selected an additional 190 schools with 8035 students from other municipalities in our simulation sample. Half the schools were assigned treatment status, and for the students in the treatment group, an additional term was added to the standardised combined response statistic. The term was a normally distributed random variable with mean 0.1 standard deviations, and 95% was added as an individual effect and 5% as a shared, school effect. This was simulated 400 times, and effect sizes were then calculated. The effect sizes were calculated in two models, one which included the school mean score in 2015 as a control. The 2015 school mean turned out to be highly correlated with the 2016 individual scores, with a correlation rate of 0.21. Furthermore, the predictive value of the previous score was very high, with a one standard deviation increase in 2015 predicting a 0.64 standard deviations increase in 2016. Table 9 shows the results of the simulations, after correcting the significance levels for multiple hypothesis testing using the false discovery rate method (see chapter 4 for details). We see that without the baseline control, the probability is 70% for finding at least one significant result with effect size 0.1 standard deviations when testing four hypotheses. With the baseline control, the probability rounds up to 100 %. In the 400 simulations, 98,5% of the p-values were lower than 0.0125 in this model. Though these simulations are very encouraging, it remains to be seen whether these results will reproduce in the actual study. It is clear that a discovery rate above the usual threshold of 80% as found through the simulation depends heavily on the strong intertemporal correlation and the weak intra-cluster correlation in this estimate. The primary outcomes may have a lower intertemporal correlation.
IRB

Institutional Review Boards (IRBs)

IRB Name
NSD - Norwegian Centre for Research Data
IRB Approval Date
2017-09-08
IRB Approval Number
55018
Analysis Plan

Analysis Plan Documents

Protocol

MD5: 34ca34c0d74fefb89f81fb3e337a3bba

SHA1: 875361477680f37d38c96f6a145eed0f4dc53bef

Uploaded At: December 23, 2017

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials