Factors influencing length of stay amongst people with Down's Syndrome

Abstract

Several studies have demonstrated that children and adults with Down's syndrome (DS) have an increased average inpatient length of stay (LOS) compared with people without DS. This study assesses if the same relationship is observed in Hong Kong and explores the patient and admission characteristics that are associated with increased LOS amongst those with DS. The Clinical Data Analysis and Reporting System (CDARS) of the Hong Kong Hospital Authority (HA) was used to identify people with DS (ICD9 code 758.0) from inpatient and outpatient records of all HA hospitals from 1995-2012. A search was then performed on CDARS for the details of any admissions in 2010 by these patients. Inpatient data from 1995-2012 and outpatient data from 2010 were used to establish if patients were known to have congenital heart disease, other circulatory malformations, thyroid disease, diabetes, epilepsy, lymphatic or haematopoeitic malignancies or gastrointestinal malformations. Average LOS was calculated, stratified by 5 year age group and gender. For comparison with the general population, aggregate counts of admissions and total LOS for all admissions in 2010, stratified by 5 year age group and gender, were also obtained from CDARS. A negative binomial with log link regression model was then used to assess the relationship between LOS and admission age group, gender, ethnicity, known DS-associated co-morbidities, emergency admission, admitting speciality, surgery, ICU admission and Down's associated primary diagnosis in patients with DS. The average LOS was 1.75 times higher in those with DS than those without (8.49 days compared with 4.84 days). However, this was not consistent across age groups and between genders, especially in children and young adults. Children under 5 years with DS had a LOS 3.37 times longer (11.26 days compared with 3.34) but children aged 5-10 and 10-15 with DS had a lower LOS than those without DS. Admission factors significantly associated with increased LOS in DS included non-emergency admissions, surgery and admission to PICU/ICU. The effect of a primary diagnosis associated with DS was not significant. Patient characteristics significantly associated with increased LOS included infant age group, 3 or more known co-morbidities, circulatory malformations and thyroid disease. Gender and ethnicity had no significant effect on LOS. Further investigation is recommended to assess if the reason for decreased length of stay in children with Down's syndrome is the result of unnecessary emergency admissions.