UCL logo

UCL Discovery

UCL home » Library Services » Electronic resources » UCL Discovery

Monitoring blood stream infection in neonatal intensive care units

Leighton, P.H.; (2011) Monitoring blood stream infection in neonatal intensive care units. Doctoral thesis, UCL (University College London). Green open access

[img]PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader


Comparisons of the incidence of blood stream infection (BSI) between neonatal intensive care units (NICUs) can promote sharing of potentially better practices for infection control. Comparisons should take into account differences in babies’ vulnerability and the invasive procedures which can introduce infection. I carried out a systematic review of methods reported in the literature, or used by regional monitoring systems, for comparing the incidence of BSI among NICUs. I found substantial variation, especially in the risk factors used to adjust incidence estimates. The use of routinely recorded administrative data would minimize and accelerate staff workload for BSI monitoring. I investigated which risk factors recorded in routine data should be adjusted for when comparing BSI incidence between NICUs. I linked microbiology laboratory records with administrative records collected over four years for three London NICUs. I analysed rates of BSI using various methods, including Poisson regression and logistic regression assuming a matched case control design. With both approaches, National Health Service level of care was the strongest predictor for BSI incidence. Using Poisson regression models, the rate ratio for BSI, adjusted for birth weight, inborn/outborn status and postnatal age, was 3.15 (95% confidence interval (CI) 2.01, 4.94) for intensive care and 6.58 (95% CI 4.18, 10.36) for high dependency care, relative to special care. The case control study gave slightly larger estimates of effect than the Poisson regression models. Total parenteral nutrition was significantly associated with BSI incidence but explained less of the variance among babies than level of care. Using the results from the risk adjustment model, I demonstrated how routine data can be integrated into a method for prospective, risk adjusted monitoring. This method involved standardised infection ratios and a sequential probability ratio test. The method can evaluate changes in BSI rates over time and between NICUs. It could also be used to quantify improvements following infection control interventions.

Type:Thesis (Doctoral)
Title:Monitoring blood stream infection in neonatal intensive care units
Open access status:An open access version is available from UCL Discovery
UCL classification:UCL > School of Life and Medical Sciences > Faculty of Population Health Sciences > Institute of Child Health

View download statistics for this item

Archive Staff Only: edit this record