Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
This project seeks to identify and characterize features derived from digital data (e.g. social media, online search, mobile media) which are associated with coronary heart disease (CHD) and related risk factors, and develop models that use digital data and conventional predictive models to predict CHD risk and health care utilization.
Cardiovascular disease is the leading cause of death in the US. While secondary prevention approaches have improved longevity of patients, risk factors and adverse health behaviors (e.g., physical inactivity, smoking) are highly prevalent, and in most contemporary series, less than 1% of adults meet all factors of ideal CV health. The logistics and practicalities of meeting the goal of ideal CV health have not been clearly elucidated. Practice guidelines recommend using the Framingham risk score (FRS) or other risk prediction tools to classify patients' risk of CV disease. These models however are imprecise and there is increasing focus on identifying markers that provide better measures of risk. As digital platforms are increasingly used to document lifestyle and health behaviors, data from digital sources may provide a window into manifestations of novel risk factors and potentially a better characterization of existing risk factors. While it seems like a cliche to mention the profound impact of digital data on everyday lives, there is indeed great substance in the opportunities these new media provide for understanding behavioral, social, and environmental determinants of health. This project seeks to identify and characterize features derived from digital data (e.g. social media, online search, mobile media) which are associated with coronary heart disease (CHD) and related risk factors, and develop models that use digital data and conventional predictive models to predict CHD risk and health care utilization.
Not provided
Not provided
Not provided
Not provided
Not provided
| Label | Type | Description | Intervention Names |
|---|---|---|---|
| Case | Patients ages 30-74 with and without CHD (IICD 10: I63, I20-I25 ) within the last 5 years. |
| |
| Control | Patients aged 30-74 who have non-cardiovascular-related history |
|
| Name | Type | Description | Arm Group Labels | Other Names |
|---|---|---|---|---|
| Survey | Other | Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
| Measure | Description | Time Frame |
|---|---|---|
| Latent Dirichlet Allocation (LDA) Topics - Topics / Themes Discussed Between Patients With and Without Heart Disease | The primary outcome is topics and features (derived using the LDA method for clustering language data). For each participant, we included all available Facebook wall posts from the start of their account history through data collection, regardless of whether they occurred before or after a CHD diagnosis. We examined associations between linguistic features (unigrams, LIWC categories, LDA topics) and cardiovascular case status (CHD presence vs absence) using Pearson correlation and logistic regression. Latent LDA, a systematic method to identify text-based themes, was applied to generate 200 clusters of co-occurring words ("topics"). For each feature type (unigram, LIWC category, LDA topic), we fit separate logistic regression models and calculated Pearson correlation coefficients to assess predictive value for case status. Each language-derived feature was encoded as a normalized frequency count per user to enable consistent comparison across participants. | Through study completion, an average of 3 years |
Not provided
| Measure | Description | Time Frame |
|---|---|---|
| CHD Event | Reliability in predicting CHD related event in patient as measured by Framingham Risk Score. The Framingham Risk Score (FRS) is a validated means of predicting cardiovascular disease (CVD) risk. Input variables include age, cigarette smoking, total cholesterol, HDL cholesterol, systolic blood pressure measurement and treatment for hypertension. Point values are calculated based on each of these risks. A 10-year risk score can be derived as a percentage. Risk scores range from 0-20%. Low Risk: Less than 10% risk that you will develop a heart attack or die from coronary disease in the next 10 years. Intermediate risk: A 10 to 20% risk that you will develop a heart attack or die from coronary disease in the next 10 years. High Risk: A greater than 20% risk that you will develop a heart attack or die from coronary disease in the next 10 years. |
Inclusion Criteria:
Exclusion Criteria:
Not provided
Not provided
We will identify patients ages 30-74 with and without CHD (ICD 9:414.0, ICD 10: I63, I20-I25)
Not provided
Not provided
| Facility | Status | City | State | ZIP | Country | Contacts |
|---|---|---|---|---|---|---|
| University of Pennsylvania Health System | Philadelphia | Pennsylvania | 19101 | United States |
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
Not provided
| ID | Title | Description |
|---|---|---|
| FG000 | Case | Patients ages 30-74 with and without CHD (IICD 10: I63, I20-I25 ) within the last 5 years. Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
| FG001 | Control | Patients aged 30-74 who have non-cardiovascular-related history Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
| Title | Milestones | Reasons Not Completed | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Overall Study |
|
Not provided
Not provided
| ID | Title | Description |
|---|---|---|
| BG000 | Case | Patients ages 30-74 with and without CHD (IICD 10: I63, I20-I25 ) within the last 5 years. Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
| Units | Counts |
|---|---|
| Participants |
|
| Title | Description | Population Description | Parameter Type | Dispersion Type | Unit of Measure | Calculate Percentage | Denominator Units Selected | Denominators | Classes |
|---|---|---|---|---|---|---|---|---|---|
| Age, Continuous | Mean |
| Type | Title | Description | Population Description | Reporting Status | Anticipated Posting Date | Parameter Type | Dispersion Type | Unit of Measure | Calculate Percentage | Time Frame | Units Analyzed | Denominator Units Selected | Arm/Group Information | Denominators | Classes | Analyses |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Primary | Latent Dirichlet Allocation (LDA) Topics - Topics / Themes Discussed Between Patients With and Without Heart Disease | The primary outcome is topics and features (derived using the LDA method for clustering language data). For each participant, we included all available Facebook wall posts from the start of their account history through data collection, regardless of whether they occurred before or after a CHD diagnosis. We examined associations between linguistic features (unigrams, LIWC categories, LDA topics) and cardiovascular case status (CHD presence vs absence) using Pearson correlation and logistic regression. Latent LDA, a systematic method to identify text-based themes, was applied to generate 200 clusters of co-occurring words ("topics"). For each feature type (unigram, LIWC category, LDA topic), we fit separate logistic regression models and calculated Pearson correlation coefficients to assess predictive value for case status. Each language-derived feature was encoded as a normalized frequency count per user to enable consistent comparison across participants. | To evaluate predictive performance for ASCVD risk, we used regression and classification analyses. Pearson correlations assessed how language features predicted continuous risk scores. For classification, we measured AUC for binary risk (≥10% vs <10%) and standard risk categories. Three logistic regression models were tested: language alone, demographics alone, and combined. Arms/groups were combined because the goal was to predict ASCVD risk continuously and categorically across participants. | Posted | Number | 100% Confidence Interval | proportion probability AUC (Area Under t |
This study was cross-sectional and no adverse event data was collected.
This study was cross-sectional and no adverse event data was collected. Due to this study design feature the number of participants at risk for Serious Adverse Events is zero, the number of participants at risk for All-Cause Mortality is zero, and the number of participants at risk for Other (Not Including Serious) Adverse Events is zero.
Not provided
| ID | Title | Description | Deaths (Affected) | Deaths (At Risk) | Serious Events (Affected) | Serious Events (At Risk) | Other Events (Affected) | Other Events (At Risk) |
|---|---|---|---|---|---|---|---|---|
| EG000 | Case | Patients ages 30-74 with and without CHD (IICD 10: I63, I20-I25 ) within the last 5 years. Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
Not provided
Not provided
Not provided
| Title | Organization | Phone | Extension | |
|---|---|---|---|---|
| Director of Research | University of Pennsylvania | 2674280125 | digitalhealth@pennmedicine.upenn.edu |
Not provided
| Type | Includes Protocol | Includes SAP | Includes ICF | Document Label | Document Date | Document Uploaded Date | Document File Name |
|---|---|---|---|---|---|---|---|
| Prot_ICF | Yes | No | Yes | Study Protocol and Informed Consent Form | Dec 7, 2021 | Sep 5, 2025 | Prot_ICF_001.pdf |
| SAP | No | Yes | No | Statistical Analysis Plan | Dec 7, 2021 | Oct 1, 2025 | SAP_002.pdf |
Not provided
| ID | Term |
|---|---|
| D002318 | Cardiovascular Diseases |
Not provided
Not provided
Not provided
| ID | Term |
|---|---|
| D011795 | Surveys and Questionnaires |
| ID | Term |
|---|---|
| D003625 | Data Collection |
| D004812 | Epidemiologic Methods |
| D008919 | Investigative Techniques |
| D017531 | Health Care Evaluation Mechanisms |
Not provided
Not provided
Not provided
Not provided
Not provided
|
| Through study completion, an average of 3 years |
| Health Care Utilization | Prediction of cost for health care utilization between heart disease and non- heart disease subjects measured by insurance claims data | Through study completion, an average of 3 years |
| BG001 | Control | Patients aged 30-74 who have non-cardiovascular-related history Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. |
| BG002 | Total | Total of all reporting groups |
| years |
|
| Sex: Female, Male | Participants were allowed to select 'Prefer not to say' for their sex. 38 case and 43 control participants selected that option | Count of Participants | Participants |
|
| Race (NIH/OMB) | Count of Participants | Participants |
|
| Through study completion, an average of 3 years |
|
|
|
| Other Pre-specified | CHD Event | Reliability in predicting CHD related event in patient as measured by Framingham Risk Score. The Framingham Risk Score (FRS) is a validated means of predicting cardiovascular disease (CVD) risk. Input variables include age, cigarette smoking, total cholesterol, HDL cholesterol, systolic blood pressure measurement and treatment for hypertension. Point values are calculated based on each of these risks. A 10-year risk score can be derived as a percentage. Risk scores range from 0-20%. Low Risk: Less than 10% risk that you will develop a heart attack or die from coronary disease in the next 10 years. Intermediate risk: A 10 to 20% risk that you will develop a heart attack or die from coronary disease in the next 10 years. High Risk: A greater than 20% risk that you will develop a heart attack or die from coronary disease in the next 10 years. | Not Posted | Through study completion, an average of 3 years | Participants |
| Other Pre-specified | Health Care Utilization | Prediction of cost for health care utilization between heart disease and non- heart disease subjects measured by insurance claims data | Not Posted | Through study completion, an average of 3 years | Participants |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| EG001 | Control | Patients aged 30-74 who have non-cardiovascular-related history Survey: Interested participants may complete the informed consent online. After informed consent, the participant will be asked to share the digital data types that they use (Facebook, Instagram, Twitter, Google search, step data) and then participants will complete a cross-sectional survey. | 0 | 0 | 0 | 0 | 0 | 0 |
Not provided
Not provided
| D011787 | Quality of Health Care |
| D017530 | Health Care Quality, Access, and Evaluation |
| D011634 | Public Health |
| D004778 | Environment and Public Health |
| Native Hawaiian or Other Pacific Islander |
|
| Black or African American |
|
| White |
|
| More than one race |
|
| Unknown or Not Reported |
|