Hong Kong Med J 2010;16:440–6 | Number 6, December 2010
Factors associated with length of hospital stay in children with respiratory disease
LY Tsung, KC Choi, E Anthony S Nelson, Paul KS Chan, Rita YT Sung
Department of Paediatrics, The Chinese University of Hong Kong, Shatin, Hong Kong
OBJECTIVES. To explore factors associated with length of stay, and secondarily to explore the potential of enhanced diagnostics to address respiratory disease burden in children.
DESIGN. Prospective study.
SETTING. A university teaching hospital in Hong Kong.
PATIENTS. Data from 475 children with respiratory tract symptoms or fever admitted to Prince of Wales Hospital, Hong Kong from November 2005 to April 2007.
MAIN OUTCOME MEASURES. Aetiological diagnoses based on enhanced diagnostics and their association with clinical information.
RESULTS. Data from 469 subjects showed that major presentations were fever (84%), cough (72%), and runny nose (64%). The median length of stay was longest (3 days) for adenovirus, metapneumovirus and mycoplasma infections, while children with negative aetiological results had a median length of stay of 2 days. Fever duration during admission (P<0.001), the highest recorded temperature during admission (P<0.001), use of antibiotics during admission (P<0.001), ear pain before admission (P=0.019), and high white cell counts (P=0.021) were associated with increased length of stay (univariate analysis). Identifying an aetiological agent did not affect length of stay. Comparison of children with a positive immunofluorescence test result (rapidly available) with those in whom the test was negative though a positive multiplex polymerase chain reaction ensued (result not available to clinicians) also showed no association with length of stay.
CONCLUSION. Although rapid enhanced diagnostics may not have a major influence on length of stay, these data form an integral part of enhanced sentinel surveillance systems.
Key words: Influenza A virus; Length of stay; Nasopharyngeal diseases; Respiratory syncytial virus infections
View this abstract indexed in MEDLINE: