Microsoft Word search phase 3 Title Page Amendment

Aim 3:  Assess total and cause-specific mortality among 2002-2008 incident cases for the 
period from the date of diabetes diagnosis through December 31, 2010.   
Research Question 3.1:  Are the frequency and causes of mortality in youth with 
diabetes different than non-diabetic, age comparable youth?  Among youth with 
diabetes, does the frequency of mortality differ by type of diabetes, race/ethnicity or 
other socio-cultural factors? 
 
 

---

Section 4A - Study Objectives/Background and Significance (Phase 3 - 11/2010) 
Section 4A - Page 3 
 Registry 
Study
 
 
4.2.
 
BACKGROUND AND SIGNIFICANCE 
Diabetes mellitus is one of the most common severe chronic diseases of childhood.  Much of 
our knowledge of the epidemiology of diabetes in young people has been generated by large 
collaborative efforts based on standardized registry data, such as the DIAMOND Project 
worldwide 
(1, 2)
 and the EURODIAB Study in Europe 
(3, 4)
.  These registries showed that, 
while at the start of the 20th century childhood diabetes was rare and rapidly fatal, by the end 
of the century a steady increase in incidence had been reported in many parts of the world 
(2)
.  
However, epidemiological data for temporal trends in childhood diabetes are still lacking or 
are minimal for most of the global population of children, including in the U.S.  In addition, 
the epidemiology of diabetes in youth is changing.  As youth are becoming increasingly 
overweight, we are seeing more obese children with a clinical phenotype of T2D or “adult 
onset” diabetes.  As demonstrated by SEARCH for Diabetes in Youth 
(5)
, T2D is becoming 
the major form of diabetes in young people in several non-white populations, such as 
American Indians, Asian and Pacific Islanders, and African-Americans. 
4.2.1.
 
Importance of Diabetes Registries in Youth 
4.2.1.1.
 
Overview, Strengths and Limitations 
The need for, and utility of, long-term incidence-based registries of youth with 
diabetes has been recognized since the late 1970’s 
(6, 7)
.  Registries facilitate tracking 
of long-term trends in incidence in defined geographic regions 
(8, 9)
, providing 
potential clues to etiology, as well as allowing the determination of differences in 
presentation, patterns of treatment, access to care, acute and chronic complications , 
quality of life, quality of care and survival differences 
(10)
.  Such registries have 
proven utility in monitoring cancer 
(11, 12)
, and other chronic conditions.  Registries 
can also include the assessment of genetic predisposition, environmental and 
behavioral risk factors, treatment patterns, and quality of care from prevention 
through chronic complications that helps form the basis for new hypotheses about 
etiology and prognosis 
(11)
.  Linkage of registry information to census-based 
neighborhood and socio-demographic information through geographic information 
system methods also allows potential clusters and spatial locations to be identified 
(13)
.  Registries located at academic institutions with access to a cadre of well trained 
epidemiologists, clinicians, and biostatisticians also allows for the efficient use of 
registry data to form the basis for additional nested studies and in-depth research on 
hypotheses developed from the main incidence data 
(14 - 19)
. 
There are limitations to such diabetes registries, which include difficulties in 
complete ascertainment of cases, due to multiple providers caring for youth with 
diabetes which results in complex care patterns with few common ascertainment 
locations.  There is no requirement for hospitalization at onset, increasing the 
potential for cases to be missed in isolated primary care locations, and there is no 

---

Section 4A - Study Objectives/Background and Significance (Phase 3 - 11/2010) 
Section 4A - Page 4 
 Registry 
Study
 
 
centralized source for validation of case status (such as pathology laboratory reports 
for cancer).  In the U.S., increasing concerns about privacy and confidentiality have 
led individual hospitals or providers to refuse to participate in registries.  It is critical 
in the estimation of incidence rates and temporal tends in rates that case 
ascertainment be virtually complete, and if incomplete, that the degree of 
ascertainment be consistent over time.  The relatively low frequency of diabetes 
occurrence requires longer time periods and larger populations for stable rate and 
trend estimates.  There is also reluctance on the part of older youth and young adults 
to participate in registration activities.  Many of the limitations in developing 
registries of diabetes in youth in the U.S. are inherent in the underlying health care 
system (lack of common health identification system, multiple and fragmented care 
sites, privacy concerns, data systems designed to address only billing) and are not due 
to the registries themselves.  As demonstrated by the SEARCH for Diabetes in Youth 
registry, the majority of such limitations can be successfully addressed, resulting in 
unbiased estimates of rates and disease-risk factor associations. 
4.2.2.