TABLE OF CONTENTS
- Scope of use for this Data Extraction Form
- Overview of the Data Extraction Form
- Tab: Study characteristics
- Tab: Interventions
- Tab: Population
- Tab: List of outcomes
- Tab: Results
Template name
Intervention review_Drugs_v1.1
Scope of use for this Data Extraction Form
This data extraction form is designed for use in reviews that aim to:
obtain information about medical technologies (e.g. study characteristics, targeted population or intervention details)
estimate the effectiveness and safety of a medical technology that can be useful e.g. for policymakers, healthcare providers and researchers.
Overview of the Data Extraction Form
There are five main categories in the data extraction form, which are represented by tabs:
Study characteristics - covers general issues such as country in which the study was conducted, sponsors, methodology or timeline
Interventions - this tab contains details of the intervention - dosage, schedule, as well as subsequent medications
Populations - tab in which details of population, subpopulation, baseline characteristics and patient flow can be extracted
List of Outcomes - tab for outcomes mapping and their general description, e.g. units, definitions or timepoints
Outcomes - tab used to extract outcome values.
Extraction summary
You can export all the data, or specific types of data relevant to your needs. For example, you can export only safety data. You can also export data in two formats: simplified or statistical (useful for statistical analysis tools). The extracted data will be categorised by data type in tabs.
Tab: Study characteristics
Section: Study details
Description: This section defines the general characteristics of the extracted study, such as study ID, name, phase, etc.
Instructions: To extract data in the Study details section, either enter the data directly into the relevant extraction field (e.g. Study ID) or highlight the relevant data in the PDF and click the “+” button next to the appropriate extraction field, which creates a connection between the PDF and field.
Detailed description how to extract data in the text field are available here: https://help.laser.ai/support/solutions/articles/204000012759-how-to-extract-data-in-a-text-field-
Some fields are designed to contain vocabulary. In this case, you will see an arrow on the right-hand side of the extraction field. You can select a term from the vocabulary or add a new one.
A detailed description of how to extract data from a vocabulary field is available here: https://help.laser.ai/support/solutions/articles/204000012760-how-to-extract-data-in-a-vocabulary-field
You may notice that some fields are coloured yellow, which indicates that an AI model has been enabled for that field. Your task is to check the suggestion and either accept or reject it. After selecting the relevant field, you will be taken to the part of the text where the suggestion has been made so that you can check its accuracy. Next, accept or reject the suggestion in the data extraction form.
A detailed description of how to extract data using AI model suggestions is available here: https://help.laser.ai/support/solutions/articles/204000012762
The last part of the 'Study details' section is the 'Connections' panel, which is located below the extraction fields. In this case, all connections are done automatically while data extraction and are necessary for proper output visibility in the Extraction Summary. As a Researcher, you do not need to make any changes here.
- Section: Study type/design
Description: In this section, you can extract information related to the type and design of study.
Instructions: Fields in this section are text and vocabulary fields, all the details related to the extraction from this kind of fields are presented in the previous section.
Section: Methodology
Description: This section contains all the information related to the study methodology.
Instructions: This section contains two main elements for extracting data. The first element is a general part that describes the study's overall methodology, such as case number estimation or relevant changes to the methodology. The second element is a subsection where you can extract the analysed population in the study (e.g. ITT or per protocol). This section can be repeated if the study includes more than one population. To do this, click the 'Add' button. More details about subsections are available here: https://help.laser.ai/support/solutions/articles/204000069869-from-data-chaos-to-clarity-smarter-extraction-starts-here-in-data-extraction-form-creator
Section: Timeline
Description: Here, you can extract data that is common to the entire study, such as the study period, as well as multiple types of data, such as data cut-off or different types of study duration.
Instructions: To extract multiple data points for the same field, use the subsections as described in the previous methodology section.
Section: Locations
Description: The 'Locations' section is dedicated to extracting geographical data.
Instructions: The number of centres and the type of study (single- or multicentre) are extracted once per study, whereas the study countries and setting can be extracted multiple times in subsections. In the case of countries, subsections have been generated automatically because the AI model recognised many study countries and put them into separate subsections. You need to accept or reject the suggestions and match the country to the vocabulary if the model was unable to do so. You will then notice that, even after accepting a suggestion, the field remains yellow and the action still needs to be completed.
Tab: Interventions
Section: Study arm
Description: The 'Study arm' section is where you can find all the details related to interventions.
Instructions: In the first part of the section, you need to extract the study name and specify whether it is an intervention or a comparator. The 'Study arm' name is different from the 'Name of the intervention' concept. In the example in the screenshot, two interventions (gemcitabine and capecitabine) are elements of a single study arm. This arm should therefore reflect this information.
First, extract the study arm name by accepting or rejecting the model suggestions (or add a new one if the model is unable to extract it). Next, add all the interventions that form part of the study arm in separate subsections, extracting all their details.
Note: This step is crucial for generating the specific fields required for arm-level and comparison data in the 'Outcomes' section.
Section: Other interventions – details
Description: 'Other interventions' is dedicated to extracting information about interventions that were not the main ones.
Instructions: This section takes the form of a vocabulary list that defines the type of intervention in detail (during the study or pretreatment, and whether it is permitted or prohibited). You can enter additional information in the text field below.
Section: Subsequent therapy after withdrawal of the study medication
Description: In the 'Subsequent therapy' section, you should extract all interventions that were used after the main study phase.
Instructions: This section consists of the main field in which to enter the name of the therapy and a subsection dedicated to entering data per study arm (study arms are automatically generated if you have previously extracted interventions). To add a new section for the next therapy, click the plus button below the subsection.
Tab: Population
Section: Study population
Description: This section is dedicated to extracting all relevant information regarding the population, including its characteristics and the inclusion and exclusion criteria.
Instructions: All fields are text fields. You can accept or reject the model suggestions or add a new one using the plus button.
Section: Subgroups
Description: If the study population has any relevant subgroups, they should be extracted in this section.
Instructions: This section consists of two text fields: ‘Subgroup: Name' (where you enter the name of the subgroup) and 'Subgroup: Description' to enter its details. The fields are supported by an AI model. You can either accept or reject the suggested text or add your own using the plus button.
Section: Patient flow (per study)
Description: Here, you can extract the number of relevant patients per study at each stage overall.
Instructions: The section consists of two elements:
The first is a group of fields where you can specify whether this particular patient number relates to subgroups or the general study population, and also define the value type (i.e. n or % ).
The second part is a subsection that can be duplicated, in which you define the type of patient number by selecting an option from the two-level vocabulary. Note that these terms are consistent with the guidelines and terminology used in CONSORT diagrams.
Once you have defined the type of patient number and value, you can add an additional type by selecting the '+' button to add a new subsection.
Tip: As this is a two-level vocabulary, selecting a term from the first level reduces the number of terms that need to be selected from the second level.
Sections: Patient flow (per arm)
Description: Here you can extract the number of patients in each study arm at each of the study stages.
Instructions: This section consists of two elements:
the first is group of fields where you specify whether if this particular patient number is related to subgroups or general study population, select the value type (i.e. n or % ) and choose the type of patient number from the two-level vocabulary. The terms are consistent with the guidelines and terminology used in CONSORT diagrams.
The second part is an automatically generated subsection for each study arm, where you only enter the number of patients in a single arm for a particular number type.
Once you have defined the type of patient number and value, you can add an additional type by selecting the '+' button to add a new subsection.
Tip: as this is a two-level vocabulary, selecting the term from the first level will reduce the number of terms to be chosen in the second one.
Section: Baseline characteristics
Description: Here, you can extract all the concepts that describe the population at baseline level, such as sex, age, region, or specific disease-related variables.
Instructions: This section consists of three elements:
- A group of fields where you define the general concepts of the variable, including its name and category, whether it is a variable for the general group or a subgroup, the units, and the types of values and variables.
- A subsection for each study arm (automatically generated), where you only enter the number of patients in a single arm for a particular baseline variable.
- The optional comparison subsection allows you to extract the p value for comparisons between arms. To do this, define the comparison type in the last subsection and extract the relevant p value for this variable.
Note that you can extract baseline variables in two ways: directly from the text, or from table extracts. Further information on extracting data from tables can be found in our knowledge base: https://help.laser.ai/support/solutions/articles/204000012666.
Tab: List of outcomes
Description: The 'List of outcomes' tab is used to gather and map all relevant outcomes from the study, along with general information such as name, unit and definitions.
Instructions:
- Some fields are vocabularies; for example, 'Outcome: Primary/Secondary', where you need to select the appropriate term from the list.
- The final section is a subsection dedicated to timepoints at which the outcome is reported. Note that this can be added multiple times, as the outcome can be reported at several timepoints. To do this, use the 'Add' button.
Tab: Results
This tab is dedicated to extracting all the results data. For single outcome there are three main elements to complete:
Section ‘Outcome’
Subsection - Per study arm
Subsection - Comparison
You can multiply this group of fields as many times as the number of outcomes to be reported
Section: Outcome
Description: This section is dedicated to extracting details of outcomes, such as timepoint, value, variance type, etc.
The difference between this section and the previous one ('List of Outcomes') is that this section extracts more detailed information about a single outcome under certain conditions, whereas the 'List of Outcomes' section describes only general information. For example, in the 'List of outcomes' section, you extract the definition or unit, which are the same across outcomes. In the 'Outcomes' section, however, you need to extract data that varies for each outcome. For instance, a single outcome may be reported at various timepoints and have different subgroups and value types.
Instructions: You can extract outcomes in two ways: by extracting them directly from the text, or by using extractions from tables. More details about extracting data from tables are available in our knowledge base: https://help.laser.ai/support/solutions/articles/204000012666
Also, some fields are vocabularies, which means that you must choose an appropriate term from the list or add a new one. A detailed description is available here: https://help.laser.ai/support/solutions/articles/204000012760
Subsection: Per-arm values
Description: A group of fields for extracting per-arm data such as values and variances.
Instructions: To extract per-arm data, you first need to extract the study arms. Once this has been done, the subsection for per-arm values will appear automatically. Next, extract the data by highlighting the text and adding it to the form, or using table extraction.
Subsection: Comparison
Description: This subsection is used to extract data for comparison between the different groups.
Instructions: First, click the 'Add' button and select the arms to be compared. The appropriate fields will then appear. As before, you can then extract data from text or tables.
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article