How the Data Were Organized
The PTSD-Repository contains over 300 variables extracted from RCTs of PTSD treatments published through May 2020. It will be updated annually. For ease of use, the variables are broken up into separate tables (called datasets), each with a focus designed to “chunk” related variables together.
Each dataset is organized at either the study level or the arm level. Study-level datasets correspond to the overall study, whereas arm-level datasets correspond to each specific treatment condition in the study. For example, a study might be classified at the study-level as a psychotherapy study, but at the arm level as having a Prolonged Exposure (PE) arm, a Cognitive Processing Therapy (CPT) arm, and a treatment as usual arm. Users are cautioned to pay attention to whether tables are organized at the study or arm level.
The organization scheme for tables and coding rules—decisions made about variables and response items—are outlined below. Each of the datasets also includes a series of repeated variables--the variables included in every dataset (also described below).
Understanding Study ID
Each dataset in the PTSD-Repository starts with a unique Study ID that helps the user understand key aspects of the study they are viewing. The Study ID always begins with the first author's name and year that the study was publised (i.e., AuthorXXXX). If an author has 2 or more studies published in the same year, a lower-case letter is used to distinguish between the studies (i.e., AuthorXXXXa and AuthorXXXXb).
In datasets that include arm-level data, Study ID may also include additional specifiers such as: measure name, intervention arm (or intervention arms being compared), assessment timepoint, and the analysis type. Specifically:
- Measure name: In the outcome datasets, an abbreviation of the measure name is included to differentiate outcomes associated with different measures (i.e., AuthorXXXX_CAPS and AuthorXXXX_PCL). A list of measure abbreviations can be found in the Abbreviations dataset.
- Intervention arm: Each treatment arm is identified
with a capital letter separated by an underscore (i.e., AuthorXXXX_A,
AuthorXXXX_B). For datasets that include between-groups data analysis in which
the conditions are compared, this could also look like AuthorXXXX_CAPS_AvB
(i.e., condition A versus B).
- Assessment timepoint: Each assessment time point
is indicated with a number separated by an underscore. Timepoints are either
coded as “baseline” or represent the number of weeks post-treatment where “0”
represents immediate posttreatment (i.e., AuthorXXXX_A_Baseline, AuthorXXXX_A_0, and
- Analysis type: Each study is coded with an underscore
of “ITT” to identify if the analyses are intention to treat (i.e., include
all participants who were randomized) or “Comp” to indicate completer
analyses in which only people who completed the treatment are included (i.e., AuthorXXXX_A_1_ITT and AuthorXXXX_A_1_Comp).
Important Note: Datasets organized at the arm level have multiple rows per study based on the study characteristics above. For example, a study that compares PE to CPT would have at least 2 rows corresponding to the 2 treatment arms. If there were 2 different measures (e.g., an interview and a self-report) administered to participants within 2 treatment arms, there would be 2 rows for each measure for each arm—equating to 4 total rows. If participants were measured at 2 different timepoints, that would add 2 more rows, etc. When creating visualizations, users are cautioned to be aware of occurrences in which individual studies may be contributing more than one data point.
Understanding Treatment Codes
The RCTs in the PTSD-Repository are organized in several ways to allow users to identify and sort those studies that interest them most.
Data are organized at the study level by treatment focus (i.e., PTSD or PTSD plus substance use disorder [SUD]) and study class that indicates the type of treatment studied. Then, treatments are coded at the arm level by study class subcode and treatment name.
Treatment Focus (study level)
Each RCT in the PTSD-Repository is coded as being a study that is focused on PTSD or both PTSD and SUD (PTSD+SUD). PTSD+SUD studies are treatment studies that are designed to target both PTSD symptoms and SUD symptoms. Studies that are coded as PTSD+SUD also have a subclass code for each arm to indicate whether the treatments are integrated or not integrated. Users can choose to access all studies in dataset or limit results to PTSD studies or PTSD+SUD studies by using a filter. For more information on how PTSD+SUD studies were coded, see our story, Treatments for Co-Occurring PTSD and Substance Use Disorder.
Study Class (study level)
Each RCT is coded with regard to 7 study classes:
- Complementary and Integrative Health (CIH)
- Nonpharmacologic Biological
- Nonpharmacologic Cognitive
- Collaborative Care
- Other Treatments
If a study included more than one type of active treatment (either within a single treatment arm or in different treatment arms), the study was coded as, mixed: psychotherapy and pharmacotherapy, mixed: psychotherapy and CIH, or other mixed.
Study Class Subcodes (arm level)
To help users view data by different types of treatments, the specific treatment arms are further classified into subtypes. See our Reference Guide: Treatments Found in the PTSD-Repository for more information about the organization of treatments.
- Psychotherapy subclass – Coded as trauma-focused (TF) or non-trauma-focused (non-TF).
- Pharmacotherapy subclass -- Coded as
Antidepressant: Selective serotonin reuptake inhibitor (SSRI);
Antidepressant: Serotonin-norepinephrine reuptake inhibitor (SNRI);
Antidepressant: Tricyclic antidepressant (TCA); Antidepressant: Monoamine
oxidase inhibitors (MAOI); Antidepressant: Other; Antiadrenergic;
Antianxiety: benzodiazepine; Antianxiety: nonbenzodiazepine; Antipsychotic;
Phytocannabinoid; Synthetic cannabinoid;
Mood stabilizer; Psychostimulants; Steroids; Mixed; Mixed SSRI and antipsychotic;
and Integrative Health (CIH) subclass—Coded as Acupuncture; Clinical
Meditation; Massage therapy; Tai
Yoga; Other physical activity
and recreational therapies; Animal-assisted; Natural products; Creative
Relaxation; Spirituality; Other
If a treatment arm included more than one type of treatment, both treatment types were coded in separate variables. These dichotomous variables are labeled according to the same study class categories noted above. For example, a treatment arm that includes the psychotherapy CPT and the CIH intervention, relaxation, was coded as both psychotherapy (Psychotherapy = yes) and CIH (CIH = yes).
- Because of this nuance, PTSD-Repository users are cautioned that if they create a visualization based on a specific treatment type using the individual treatment type variables like Psychotherapy or Pharmacotherapy, in a few cases it could include some other treatments as well.
- If users want to identify treatment arms that only include one specific type of treatment, they can limit the studies by first selecting those from the Study Class variable that focuses on the type of treatment of interest (e.g., selecting Psychotherapy as treatment class would ensure that a CPT plus relaxation study would not be included because it would have been coded as “Psychotherapy and CIH”) or by reviewing the treatment arms that are identified by searching psychotherapy and then omitting those that have another type of treatment indicated as well.
Treatment Name (arm level)
Treatments are assigned 2 types of names: the name used by the investigators in the research (Treatment Name) and the National Center for PTSD name (NCPTSD Treatment Name), which applies more consistent names across treatments to allow for easier identification and comparison of treatments. Most users of the PTSD-Repository will likely find the NCPTSD Treatment Name to be the one they want to use.
What Datasets Are Included in the PTSD-Repository?
An easy way to locate a dataset is to select the “Data Catalog” button in the top navigation. (If you are not signed in, look for the “Browse” button on the top left.) Then, in the left navigation under View Types, select “Datasets.”
Below is a short description of all datasets included in the PTSD-Repository. Click the dataset name to go to the primer page which provides a description of the dataset, a full list and description of variable columns, and metadata (keywords or phrases used to describe, tag or organize the data). Included in the description is information about whether the data are organized at the study or arm level.
Study Characteristics (study level). The Study Characteristics dataset includes basic information about study design. In addition to study class (defined above), the dataset includes clinical setting, whether the study looked at PTSD symptom clusters, whether there were any subgroup analyses, type of diagnostic measure used (e.g., diagnostic interview, self-report), and provider credentials. Use this dataset to better understand the variety of RCT designs employed, compare studies by study class or to identify gaps in the literature for potential new RCTs. Visualizations made from this dataset will be based on the 389 RCTs included in the PTSD-Repository.
Sample Characteristics (study level). The Sample Characteristics dataset includes information on the participants who are included in the studies. This information is provided at the study level—meaning, data are not broken down into the specific treatment arms such as the intervention or control groups. Use this dataset if you want to learn more about the number of participants in the study; inclusion and exclusion criteria related to substance use and suicidality; baseline clinical characteristics such as PTSD severity, trauma type, military status, and comorbidities; and, basic demographic information such as age, gender or race for the sample as a whole (not individual participants). Visualizations made from this table will be based on the 389RCTs included in the PTSD-Repository and answer questions such as Who Has Been Studied.
Study Interventions (arm level). The Study Interventions dataset includes information about each of 861 specific treatment arms that were studied in all 389 RCTs. Each study arm was coded to indicate the type of intervention or comparison condition. Download a list of study class and treatment categories. This dataset includes the study-level Study Class as well as individual variables for each category of treatment, coded as Yes or No for each arm. Study arm treatment category variables are as follows:
- Pharmacotherapy (as well as a subclass such as antidepressant, antianxiety, etc.)
- Psychotherapy (as well as a subclass to identify trauma-focused or non-trauma-focused therapy)
- Complementary and Integrative Health (CIH; as well as a subclass such as relaxation or meditation)
- Nonpharmacologic Biological
- Nonpharmacologic Cognitive
- Collaborative Care
- Other Treatments
The Study Intervention dataset also includes information on the format of the treatment (individual, group, couples, mixed); treatment delivery method (in person, by phone, by video, technology alone, technology assisted, written or mixed); dose or amount of treatment and, treatment completion and adherence. Use this dataset to learn about treatment studies of a particular type, including our story, Which PTSD Treatments Have Been Studied?
PTSD Outcomes Within Each Treatment Arm (arm level). The PTSD Outcomes Within Each Treatment Arm dataset includes information on how PTSD was measured and at which timepoint. Descriptive information is provided on which measure was used, the statistical approach to measuring change in PTSD, how PTSD was defined and the definitions of loss of diagnosis and clinically meaningful change, if these variables were included in the study.
Results in this dataset are provided for each treatment arm, and do not include comparisons across the arms (which can be found in the PTSD Outcome Comparisons Between Treatment Arms dataset described below). PTSD baseline mean score and standard deviation are provided as well as the within-treatment change from baseline. Within-group effect size is also included. The percent of participants who achieve loss of PTSD diagnosis and the percent who achieve clinically meaningful response are included for studies that reported these outcomes.
Use this dataset to learn about changes in PTSD within a specific treatment. For example, what is the change in Clinician Administered PTSD Scale score across all study arms of a specific treatment type? From the visualization, you can filter by PTSD measure of interest and type of treatment.
PTSD Outcome Comparisons Between Treatment Arms (arm level). The PTSD Outcome Comparisons Between Treatment Arms dataset includes the results of all between-group comparisons on PTSD outcomes. The omnibus test and effect size are provided, followed by pairwise comparisons for any study with more than 2 treatment arms. Results are provided for PTSD severity, loss of diagnosis and clinically meaningful change for studies that reported these outcomes. Use this dataset to look at changes in PTSD resulting from 2 or more treatments.
Datasets related to suicide outcomes (arm level). There are 4 suicide outcome datasets. Data on suicide is reported separately for continuous and dichotomous outcome measures. Datasets are further categorized as to whether the comparisons are made within or between arms. The name of the measure is included as well as the effect sizes, where applicable. Use this dataset to learn how about the effects of PTSD treatments on suicide outcomes.
Note. Other data on suicide can be found in the Sample Characteristics dataset which reports on inclusion/exclusion criteria and in the Harms dataset which reports on adverse events.
Other Outcomes (arm level). The Other Outcomes dataset includes information on whether the trial includes measures of depression, anxiety, substance use, sleep, anger, quality of life and functioning. Results in this dataset are provided for each treatment arm. The name of the measure is included as well as the between-group effect sizes. Use this dataset to learn how about the effects of PTSD treatments on other outcomes.
Risk of Bias (study level). Each study in the PTSD-Repository was coded for risk of bias in specific domains as well as an overall risk of bias for the study. Detailed information about the specific coding strategy is available in the Comparative Effectiveness Review [CER] No. 207, Psychological and Pharmacological Treatments for Adults with Posttraumatic Stress Disorder (PDF). Most domains were rated as Yes (i.e., minimal risk of bias), No (i.e., high risk of bias) or Unclear (i.e., bias could not be determined). The overall study rating is based on the domains and is coded as low, medium or high risk of bias.
The 4 domains that were included as components of the overall rating include:
- Selection bias: randomization adequate, allocation concealment adequate, groups similar at baseline and whether they used intention to treat analyses
- Performance bias: care providers masked and patients masked
- Detection bias: outcome assessors masked
- Attrition bias: overall attrition less than or equal to 20% vs over 20%; differential attrition less than or equal to 15% vs over 15%
Additional items assessed (but not considered as part of the overall rating) include:
- Reporting bias: all prespecified outcomes reported
- Reporting bias: method for handling dropouts
- Outcome measures equal valid and reliable
- Study reports adequate treatment fidelity based on measurements by independent raters
Harms (arm level). The Harms dataset includes information on serious adverse events and participant withdrawals from the studies due to adverse events. Adverse events are reported as the percent of participants within a treatment group that experienced an adverse event or who withdrew from the study. Detail on the specific adverse event or reason for withdrawal is provided when available. This dataset also reports on the percentage of patients who attempted or completed suicide (when available).
Other non-data datasets. There are 3 other datasets that are purely informational and do not contain specific data elements:
- References and Study Identifiers. Identifying information about each study, including citation, PubMed ID, PTSDpubs ID, ClinicalTrials.gov ID, location and funding source.
- Secondary Studies. A list of secondary studies included with primary studies reported in the PTSD-Repository. These secondary, referenced studies may or may not have contributed data to the abstraction of the listed primary study.
- Abbreviations. Definitions of abbreviations found throughout the PTSD-Repository.
Key Variables Repeated in Each Dataset
Repeated variables include a subset of variables that generally describe the treatment and sample; these are included in all of the datasets described above. Having some variables repeated in every dataset means that it is easier to create visualizations that include these variables.
In a few cases there are study-level variables included in datasets organized by arm. In these cases, the variable name notes (Study level) to remind the user that the data applies to the full RCT, not any particular arm of the study. For example, the Percent Female (Study level) variable is coded at the study level, so the values correspond to the overall study, not the specific treatment arm.
- Now that you understand how the data are organized, read our Getting Started with the PTSD-Repository story to learn tips on using the platform to create your own visualizations, download data and more.
- Using the Data Catalog button, choose "Files and Document" in the View Type menu to download our Reference Guide: Treatments Found in the PTSD-Repository that explains the classification of treatments included in the studies