Jump to top menu Jump to main menu Jump to content
prepare-conversation
Research group/lab

Biostatistics Consultancy

The Department of Biostatistics provides statistical consultations within the CPO framework. CPO-stands for “Consultation centre for Patient Oriented research”

About our research group/lab

About our consultations

These consultations can be arranged on a case-by-case basis, but the department also offers in-depth collaborative consulting in larger projects. 

Our researchers provide statistical advice on, e.g.:

  • methodological aspects in the design of a study
  • numbers of subjects/patients required in a study
  • appropriate/optimal use of statistical methods
  • correct interpretation of statistical analyses
  • writing of the statistical analysis plan for clinical trials
  • preparation of randomization schedules
  • advice on the design of the database
  • elaborate statistical analyses of gathered data and state-of-the-art statistical modelling
  • writing of a statistical report and incorporating the results into articles

 

Funding and conditions 

Consultations that are funded by CPO
Sufficient funding is required for all consultation services offered by the department. For researchers from Erasmus MC, the *Consultation centre for Patient Oriented research (CPO) can provide funding for the consultations if the *conditions of the CPO are met (see useful links below). In that case, no additional fee will be charged by the Department of Biostatistics. 
The CPO only provides funding for consultations for patient-oriented research projects, and not for elaborate data analyses or consultations for research projects involving animals.

Consultations that are NOT funded by CPO
When you are a researcher from Erasmus MC, we charge an hourly rate of €85,00 and administration costs.

Consultations from researchers from outside Erasmus MC 
We have to charge hourly costs. Please contact us for more information and with the details you find under: 'Request a Consultation'.

 

Collaborations

In case of prolonged involvement, mutual arrangements must be made between the departments of the researcher and the consultant, and the consultations will then preferably be funded through grants and/or contracts. Our consultations manager will contact you for additional details of the collaboration.

Co-author
When the statistical consultant spends more than 5 hours on a particular project, it is expected that he/she will be included as a co-author in the scientific paper produced by this work, irrespective of what funding arrangements were made.

 

Time planning and availability

Statistical consultancy, in general, involves a careful analysis of the question(s) and the required statistical methodology. The time planning of the statistical analyses depends on the availability of the consultant. 
Note that our consultants are also researchers who primarily work on their research projects. For these reasons, considerable time may be required to finalize the analysis. Sometimes no consultant is available within a few weeks. Please tell us if you have a specific deadline, and do not wait to contact us until the last week before your deadline.

For undergraduate medical school students performing research at Erasmus MC, the initial contact with our department should preferably be made by the scientific mentor.

 

Data management and scientific integrity

In most consultation projects, patient data will be shared with the consultant. We will, of course, treat these data confidentially. In the context of scientific integrity, we require that all data files sent to the department of Biostatistics contain only anonymous data (no patient identifiers). For more information on how to submit data to the department, please refer to the *Rules & Guidelines.”

 

Useful links

- Consultation center for Patient Oriented research (CPO) 
- Conditions of CPO
- Common statistical issues in medical papers
Scientific integrity

Rules and Guidelines

Rules & Guidelines for Submitting Data to the Department of Biostatistics

Rules

  • The researcher that provides his/her data for analysis has the responsibility to ensure that the data are correct.
  • It is not allowed to store information in an analysis file that can be used to trace back individual patients or study participants, (e.g., names, home addresses, exact dates of birth, patient numbers from Erasmus MC and/or any other hospital, etc.). We can never accept data that contains this type of information.
  • Instead, each study participant should be assigned a unique identifier (i.e. a number) that is only used in this study, and this identifier should be included in all analysis files.
  • Members of the Department of Biostatistics will treat any received data as confidential and will store the data securely. The data will be deleted if so requested by the client.

Guidelines

It is essential that the data files can be imported into a statistical package without difficulties. To achieve this, the following guidelines should be followed:

  • Data should be delivered as SPSS or SAS databases, plain ASCII files (e.g., comma-separated, tab-delimited) or as Microsoft Excel files. In most cases, it is preferable to use the format of a statistical software package such as SPSS, as these packages allow you to specify the type of a variable and to assign value labels to categorical variables.
  • Variable names should be descriptive, should start with an alphabetic character (a-z), and a contain numbers. You should avoid using special symbols in variable names (e.g., “*”, “-”, etc.)

For SPSS files:

  • All variables should have a variable name/label
  • Ordinal/nominal variables should be provided with value labels
  • Missing values should be defined appropriately.

For Excel files:

  • The file should consist of a single table
  • The columns should correspond to different variables, and the rows should correspond to various patients/study participants and/or (for longitudinal studies) measurement occasions.
  • The first row of the table should contain the variable names
  • A column should consist either solely of numeric data or character data but the two should not be mixed. Do not use symbols such as “>” or “<” in numeric data. Notes on observations should be placed in a separate column.
  • When using categorical data, be careful that the category labels are spelt consistently (e.g., do not use all of these possibilities, ‘Group_1’, ‘Group1’, ‘Group-1’, ´group*1´, ‘Grp_1’, but just pick one and be consistent).
  • Do not use any special formatting to denote information needed in the analysis (e.g., do not use cells with colours, merged cells or hidden columns). Do not use blank rows or columns.

Data with repeated measurements: 

  • Prepare your data in the long format. This means that each patient can have more than one row/record, with each row corresponding to a measurement (e.g. baseline, 1st follow-up measurement, 2nd follow-up measurement, etc.).
  • Variables that do not change during the study (e.g., age at baseline and gender) must be given for each observation of a patient.
  • Include a numerical variable that denotes the timing occasion (e.g., 1, 2, 3, 4, 5) of the observation.

Common statistical issues in medical papers

Request a Consultation

To request a consultation, please read carefully:

our About our consultations and Rules and Guidelines section.

Then send an e-mail in English, to our Manager Eline van Gent, Biostatistics-Consultations at cpo.biostatistiek@erasmusmc.nl with the following information:
your name, micro section number and PaNaMa number
name(s) of principal investigator or supervisor
department name and department theme of the project (in Dutch)
title of the project/study/paper
a short description of the research project
a brief description of your question(s)
whether the study/project involves animals – YES / NO

 Please do not send any attachments with your request.

Useful links:
- Consultation center for Patient Oriented research (CPO)
- Conditions of CPO 
- Common statistical issues in medical papers
- Scientific integrity