Temple Registry for the Investigation of African American Gastrointestinal Disease Epidemiology (TRIAGE)

• Epidemiologic factors associated with prevalent reflux disease. [ Time Frame: Two years ] [ Designated as safety issue: No ]
  Patients will complete a series of standardized questionnaires to determine the prevalence of reflux disease and risk factors for its development. We will query diet, depression, drug, tobacco, and alcohol use as well as a variety of other factors. We will study 750 African Americans and have 250 caucasian controls all from the same community in North Philadelphia.
  
  

• Longitudinal study on incidence of reflux disease [ Time Frame: 10 years ] [ Designated as safety issue: No ]
  We will follow the established cohort over 10 years to look at the incidence of reflux disease as well as the incidence of functional bowel disorders in African Americans.
  
  

Treatment Plan

The goal of our survey will be to obtain responses from at least 750 self-described African Americans,and 250 Caucasians.

Method From our previous study we found that relying on responses from surveys placed in resident mailboxes (which need to be mailed back to us) is ineffective. For this study we have divided the zip code into roughly four equal quadrants using Broad Street and Erie Avenue as dividers. We will place a flyer in 1000 mail boxes in each quadrant by hand delivery. Hand delivery assures that mailings to vacated buildings will not occur. We hand delivered surveys in our previous study and encountered no problems. This will allow for a balanced representation from the entire zip code (see data collection section for further discussion of this point). The flyer informs the participants that the Temple Gastroenterology department is performing a health survey and gives them a phone number to call if they are interested. This phone is in a locked lab run by the PI. As compensation for their time they will receive $25. Only one subject per household is eligible and interviews can be completed at the hospital or in the participant's home. Only those individuals willing to be contacted in future years for follow-up will be enrolled. Subjects must show an identification verifying that their home address is in the zip code (usually a driver's license).

The overall survey consists of several standardized questionnaires which have been organized into a logical, easy to complete, comprehensive unit. Patients will complete the survey on a laptop which contains all the questions in an easy to read and follow Microsoft Access database file. Essentially no computer skills are needed to complete it except for the use of a mouse. A member of the research team will sit in the room to help the subject if there are questions. This survey is password secured such that when the subject completes the survey they cannot access the responses of other subjects. Answering the last question automatically closes the file.

The individual questionnaires making up the overall survey are included in this packet. Initial questions ask the subject their name, address, contact phone numbers and demographic questions such as income and education level. Subsequent questions are the complete set of questions from standardized questionnaires such as AUDIT for alcohol use assessment. The cornerstones of the survey are the 17 question survey on GERD which covers all components of the Montreal classification, and the 65 questions which make up the Rome III functional bowel disease questionnaire. We estimate it will take participants 30-60 minutes to complete depending on age and educational level.

Risks.

The only potential risk is related to confidentiality. Sensitive personal information is assessed using the survey. Several levels of subject protection will be employed. These include use of a password protected file for database collection on a computer which is password protected when turned off. The computer will be kept in a locked office in the hospital. The phone used to interact with subjects is in lab #7 of the Parkinson Pavilion, 10th floor.

There are no interventions to be performed for this study. There will be, for example, no blood draws. Future studies may recruit individuals from this database for invasive diagnostic testing and further data collection but those will be covered under future IRB submissions.

Benefits

Subjects who participate in the study will develop a greater insight into the possible causes of their medical problem. In a relaxed environment will they have the opportunity to talk with a board-certified gastroenterologist about their health. They will be provided with accurate information and will be encouraged to engage the health care system by communicating with their primary care physician. The African American community as a whole will benefit tremendously as their health status within the US is presented in peer-reviewed literature. This has the potential to shape public policy. Future studies under separate IRB protocols may provide benefit to them such as free testing for potentially serious GI diseases.

Data Collection and Statistics

We will recruit a minimum of 100 African Americans from each quadrant of the zip code 19140 with the total adding to 750. Caucasian neighborhoods tend to be concentrated in one quadrant and therefore they will not undergo stratified recruitment. Comparison of continuous data between groups will be performed using t-tests or one-way analysis of variance as appropriate. Comparison of proportions will be analyzed using χ2. We will compare prevalence data between the racial groups with respect to GERD and the functional bowel diseases. A logistic regression will be performed to determine risk factors for these diseases. Race will be a pre-specified predictor variable along with those factors found significant (P< 0.05) on the univariate analysis. Model reduction will be performed using step-wise methodology. We will look at gender as a possible effect modifier. The results will provide odds ratios along with their respective 95% confidence intervals. We will calculate the population attributable risk (PAR) for significant independent variables from the regression analysis.