Materials and Methods:
This study was reviewed and approved by the Institutional Review Board
at Nationwide Children’s Hospital prior to any data collection. Using
the International Classification of Diseases (ICD-9)diagnosis codes for DKA (250.1x), billing database records were reviewed
for children aged less than five years old between January 2002 and
December 2011 with a DKA admission to Nationwide Children’s Hospital in
Columbus, OH. DKA diagnosis was confirmed after chart review by
laboratory results: venous pH <7.3, hyperglycemia (blood
glucose >200mg/Dl), and ketonuria, as described by Woldorf
et al [16]. This patient group was then cross-referenced for CVC
placement (ICD-9 code 38.93 and 38.97) as well as VTE
(ICD-9 code 453.2, 453.4x, 453.8x, and 453.9) diagnoses. All VTE
codes were confirmed after chart review established that each case was
diagnosed by Doppler ultrasonography and was CVC-associated.
Data were recorded for two patient groups: “DKA with CVC” and “DKA
with CVC and VTE.” Demographic and hospital course information was
noted. Details of the CVC, including hospital day of placement, duration
of use, type, size, location, continuous fluid infusion, and
anticoagulation prophylaxis, were elucidated and compared between the
two groups. Laboratory findings at DKA presentation were also reviewed
to evaluate for predictive indicators for VTE in the setting of DKA with
CVC. White blood cell count (WBC), hemoglobin, hematocrit, platelet
count, blood urea nitrogen (BUN), creatinine, urine pH, venous blood pH,
and hemoglobin A1C were among those evaluated.
The patients that developed a VTE also had baseline coagulation studies
(Prothrombin Time (PT), activated Partial Thromboplastin Time (aPTT))
and thrombophilia testing performed at the time of VTE diagnosis.
Protein C activity, protein S activity, antithrombin activity,
Prothrombin G20210A mutation, Factor V Leiden mutation, anticardiolipin
antibodies, homocysteine levels, and quantitative D-dimer studies were
investigated at the discretion of the hematologist.
Billing database records were also searched using the same ICD-9 codes
for patients aged two months to five years of age for all CVC placements
during our study period for any reason other than DKA. We elected to use
two months as our lower age limit since our youngest patient with DKA
and a CVC was two months old. This patient group was subsequently
cross-referenced for VTE diagnoses up to three months after line
placement and was compared to the DKA group for the incidence of
CVC-associated VTE.
Descriptive statistics for patients with DKA and CVC are provided for
all variables. Reported demographic and laboratory variables between
groups were compared using Mann-Whitney rank sum tests. Fisher’s exact
test was used to compare the prevalence of CVC-associated VTEs between
children with DKA and those without DKA. Statistical significance was
determined by p < 0.05. Statistical analyses were performed
using Statistical Analysis System software 9.3 (SAS Institute Inc.,
Cary, NC, USA).