Database research in pediatric urology

Descriptive epidemiology of children hospitalized for inflammatory bowel disease in Japan: Inpatient database analysis

BACKGROUND

Inflammatory bowel disease (IBD) - Crohn's disease (CD) and ulcerative colitis (UC) - are chronic inflammatory disorders of the intestine. Patients with IBD are at risk of hospitalization for disease exacerbation or IBD-associated complications. In the pediatric population, however, there are limited data on IBD hospitalizations. We therefore investigated the descriptive epidemiology of hospitalizations relevant to pediatric IBD.

METHODS

The national inpatient claims database in Japan was searched for children (≤ 18 years old) with a diagnosis of IBD. The study period was 2007-2010. Data on demographic characteristics and descriptive statistics of the hospital course were extracted and analyzed.

RESULTS

A total of 3559 admissions of 2175 patients met the definition of pediatric IBD: there were 1999 admissions for CD and 1560 admissions for UC. Internists were responsible for patient care in 56.6% of admissions, followed by pediatricians (27.5%). Of 3559 admissions, unscheduled hospitalizations accounted for 79.7%, and 7.6% of hospitalizations were attributable to complications of IBD, including intestinal, extraintestinal and other manifestations. The median age at first admission was 16 years (IQR, 13-17 years), in both the CD and UC groups. Compared with UC patients, CD patients had a higher number of hospitalizations (P < 0.001), but hospital stay was shorter (median: 6 vs 16 days, P < 0.001). There were seven fatal cases of IBD, accounting for 0.32% in the present series, and sepsis was the cause in five.

CONCLUSIONS

This study provides a description of pediatric inpatients with IBD and their hospital course in Japan.

Secondary data analysis of large data sets in urology: successes and errors to avoid

PURPOSE

Secondary data analysis is the use of data collected for research by someone other than the investigator. In the last several years there has been a dramatic increase in the number of these studies being published in urological journals and presented at urological meetings, especially involving secondary data analysis of large administrative data sets. Along with this expansion, skepticism for secondary data analysis studies has increased for many urologists.

MATERIALS AND METHODS

In this narrative review we discuss the types of large data sets that are commonly used for secondary data analysis in urology, and discuss the advantages and disadvantages of secondary data analysis. A literature search was performed to identify urological secondary data analysis studies published since 2008 using commonly used large data sets, and examples of high quality studies published in high impact journals are given. We outline an approach for performing a successful hypothesis or goal driven secondary data analysis study and highlight common errors to avoid.

RESULTS

More than 350 secondary data analysis studies using large data sets have been published on urological topics since 2008 with likely many more studies presented at meetings but never published. Nonhypothesis or goal driven studies have likely constituted some of these studies and have probably contributed to the increased skepticism of this type of research. However, many high quality, hypothesis driven studies addressing research questions that would have been difficult to conduct with other methods have been performed in the last few years.

CONCLUSIONS

Secondary data analysis is a powerful tool that can address questions which could not be adequately studied by another method. Knowledge of the limitations of secondary data analysis and of the data sets used is critical for a successful study. There are also important errors to avoid when planning and performing a secondary data analysis study. Investigators and the urological community need to strive to use secondary data analysis of large data sets appropriately to produce high quality studies that hopefully lead to improved patient outcomes.