Shaare Zedek Medical Center


Date of Award

02.04.2013

Program

Crohn's Disease, Israel

Amount

$1,161,600

Term of Grant

51 months

Project Title

Population-based Analysis of IBD Using Administrative Data from Israeli HMOs

Project Description

Background: The incidence of inflammatory bowel disease (IBD) is increasing in many developed and developing countries. Health administrative data has the potential to allow population-based estimates of incidence and outcomes as well as to explore the association with environmental factors and predictive variables. There are very few international counterparts of complete centralized paperless electronic charts covering over 98% of the population, as in Israel. This makes Israel the ideal place for studying IBD in a macro-epidemiological design. We propose using the health administrative data collected by Israel’s HMOs to assess the burden of IBD in Israel over the years and to explore environmental factors associated with the disease and its outcome in population based analyses. Methods: The major obstacle in performing administrative research is accurately detecting the target population. Algorithm validation will be thus our first aim. We will use hospital charts of the medical centers participating in the Israeli SHARE to identify 1000 patients per HMO followed within the centers between 2002-2005, confirmed by chart review. These patients will serve as the “gold-standard” of true-positive (TP) IBD cases used to develop algorithms based on physician codes. Next, we will find the number of visits within each of the four HMOs with IBD-related codes sensitive enough to detect at least 90% of the TP. We will then search within each HMO for those meeting the above criteria. Charts of all suspected positives (estimated 3500-8000 for every 500,000 subjects searched) will be reviewed to confirm the TP. Those without any IBD related codes will serve as the true negatives (TN). Once the TP and TN have been determined, several algorithms will be explored and validated to maximize positive and negative predictive values. We will then use the validated algorithm to describe incidence of IBD in Israel and analyze temporal trends with age- and sex-standardized Poisson regression. Next, we will assess disease outcomes (e.g. hospitalization rates, outpatient physician visit rates, surgical resection rates, emergency department usage) in patients with IBD, in relation to different environmental exposures (e.g. rural vs. urban residence, previous exposure to medications and antibiotics etc) and disease management (use of the different medication, time to use of medications etc). Predictors of poor outcome will be explored. The funding will allow hiring an epidemiologist qualified in managing such a huge administrative database and experienced in coordinating the several IBD participating sites and the four HMOs. In addition, funding will also allow the massive amount of chart extractions and programming at the different sites and the HMOs. Importance: This study will generate meaningful hypotheses to explain changing patterns of incidence of IBD in Israel. The application of a validated algorithm to Israel’s administrative data will create one of the largest ongoing surveillance population based cohorts of IBD in the world and allow exploring causes of IBD incidence and factors contributing to disease outcome. Background: The incidence of inflammatory bowel disease (IBD) is increasing in many developed and developing countries. Health administrative data has the potential to allow population-based estimates of incidence and outcomes as well as to explore the association with environmental factors and predictive variables. There are very few international counterparts of complete centralized paperless electronic charts covering over 98% of the population, as in Israel. This makes Israel the ideal place for studying IBD in a macro-epidemiological design. We propose using the health administrative data collected by Israel’s HMOs to assess the burden of IBD in Israel over the years and to explore environmental factors associated with the disease and its outcome in population based analyses. Methods: The major obstacle in performing administrative research is accurately detecting the target population. Algorithm validation will be thus our first aim. We will use hospital charts of the medical centers participating in the Israeli SHARE to identify 1000 patients per HMO followed within the centers between 2002-2005, confirmed by chart review. These patients will serve as the “gold-standard” of true-positive (TP) IBD cases used to develop algorithms based on physician codes. Next, we will find the number of visits within each of the four HMOs with IBD-related codes sensitive enough to detect at least 90% of the TP. We will then search within each HMO for those meeting the above criteria. Charts of all suspected positives (estimated 3500-8000 for every 500,000 subjects searched) will be reviewed to confirm the TP. Those without any IBD related codes will serve as the true negatives (TN). Once the TP and TN have been determined, several algorithms will be explored and validated to maximize positive and negative predictive values. We will then use the validated algorithm to describe incidence of IBD in Israel and analyze temporal trends with age- and sex-standardized Poisson regression. Next, we will assess disease outcomes (e.g. hospitalization rates, outpatient physician visit rates, surgical resection rates, emergency department usage) in patients with IBD, in relation to different environmental exposures (e.g. rural vs. urban residence, previous exposure to medications and antibiotics etc) and disease management (use of the different medication, time to use of medications etc). Predictors of poor outcome will be explored. The funding will allow hiring an epidemiologist qualified in managing such a huge administrative database and experienced in coordinating the several IBD participating sites and the four HMOs. In addition, funding will also allow the massive amount of chart extractions and programming at the different sites and the HMOs. Importance: This study will generate meaningful hypotheses to explain changing patterns of incidence of IBD in Israel. The application of a validated algorithm to Israel’s administrative data will create one of the largest ongoing surveillance population based cohorts of IBD in the world and allow exploring causes of IBD incidence and factors contributing to disease outcome.

Shaare Zedek Medical Center


Date of Award

02.04.2013

Program

Crohn's Disease, Israel

Amount

$1,161,600

Term of Grant

51 months

Project Title

Population-based Analysis of IBD Using Administrative Data from Israeli HMOs

Project Description

Background: The incidence of inflammatory bowel disease (IBD) is increasing in many developed and developing countries. Health administrative data has the potential to allow population-based estimates of incidence and outcomes as well as to explore the association with environmental factors and predictive variables. There are very few international counterparts of complete centralized paperless electronic charts covering over 98% of the population, as in Israel. This makes Israel the ideal place for studying IBD in a macro-epidemiological design. We propose using the health administrative data collected by Israel’s HMOs to assess the burden of IBD in Israel over the years and to explore environmental factors associated with the disease and its outcome in population based analyses. Methods: The major obstacle in performing administrative research is accurately detecting the target population. Algorithm validation will be thus our first aim. We will use hospital charts of the medical centers participating in the Israeli SHARE to identify 1000 patients per HMO followed within the centers between 2002-2005, confirmed by chart review. These patients will serve as the “gold-standard” of true-positive (TP) IBD cases used to develop algorithms based on physician codes. Next, we will find the number of visits within each of the four HMOs with IBD-related codes sensitive enough to detect at least 90% of the TP. We will then search within each HMO for those meeting the above criteria. Charts of all suspected positives (estimated 3500-8000 for every 500,000 subjects searched) will be reviewed to confirm the TP. Those without any IBD related codes will serve as the true negatives (TN). Once the TP and TN have been determined, several algorithms will be explored and validated to maximize positive and negative predictive values. We will then use the validated algorithm to describe incidence of IBD in Israel and analyze temporal trends with age- and sex-standardized Poisson regression. Next, we will assess disease outcomes (e.g. hospitalization rates, outpatient physician visit rates, surgical resection rates, emergency department usage) in patients with IBD, in relation to different environmental exposures (e.g. rural vs. urban residence, previous exposure to medications and antibiotics etc) and disease management (use of the different medication, time to use of medications etc). Predictors of poor outcome will be explored. The funding will allow hiring an epidemiologist qualified in managing such a huge administrative database and experienced in coordinating the several IBD participating sites and the four HMOs. In addition, funding will also allow the massive amount of chart extractions and programming at the different sites and the HMOs. Importance: This study will generate meaningful hypotheses to explain changing patterns of incidence of IBD in Israel. The application of a validated algorithm to Israel’s administrative data will create one of the largest ongoing surveillance population based cohorts of IBD in the world and allow exploring causes of IBD incidence and factors contributing to disease outcome. Background: The incidence of inflammatory bowel disease (IBD) is increasing in many developed and developing countries. Health administrative data has the potential to allow population-based estimates of incidence and outcomes as well as to explore the association with environmental factors and predictive variables. There are very few international counterparts of complete centralized paperless electronic charts covering over 98% of the population, as in Israel. This makes Israel the ideal place for studying IBD in a macro-epidemiological design. We propose using the health administrative data collected by Israel’s HMOs to assess the burden of IBD in Israel over the years and to explore environmental factors associated with the disease and its outcome in population based analyses. Methods: The major obstacle in performing administrative research is accurately detecting the target population. Algorithm validation will be thus our first aim. We will use hospital charts of the medical centers participating in the Israeli SHARE to identify 1000 patients per HMO followed within the centers between 2002-2005, confirmed by chart review. These patients will serve as the “gold-standard” of true-positive (TP) IBD cases used to develop algorithms based on physician codes. Next, we will find the number of visits within each of the four HMOs with IBD-related codes sensitive enough to detect at least 90% of the TP. We will then search within each HMO for those meeting the above criteria. Charts of all suspected positives (estimated 3500-8000 for every 500,000 subjects searched) will be reviewed to confirm the TP. Those without any IBD related codes will serve as the true negatives (TN). Once the TP and TN have been determined, several algorithms will be explored and validated to maximize positive and negative predictive values. We will then use the validated algorithm to describe incidence of IBD in Israel and analyze temporal trends with age- and sex-standardized Poisson regression. Next, we will assess disease outcomes (e.g. hospitalization rates, outpatient physician visit rates, surgical resection rates, emergency department usage) in patients with IBD, in relation to different environmental exposures (e.g. rural vs. urban residence, previous exposure to medications and antibiotics etc) and disease management (use of the different medication, time to use of medications etc). Predictors of poor outcome will be explored. The funding will allow hiring an epidemiologist qualified in managing such a huge administrative database and experienced in coordinating the several IBD participating sites and the four HMOs. In addition, funding will also allow the massive amount of chart extractions and programming at the different sites and the HMOs. Importance: This study will generate meaningful hypotheses to explain changing patterns of incidence of IBD in Israel. The application of a validated algorithm to Israel’s administrative data will create one of the largest ongoing surveillance population based cohorts of IBD in the world and allow exploring causes of IBD incidence and factors contributing to disease outcome.