To determine whether there is any difference in infection rate
at 90 days between trauma operations performed in laminar flow and
plenum ventilation, and whether infection risk is altered following
the installation of laminar flow (LF). We assessed the impact of plenum ventilation (PV) and LF on the
rate of infection for patients undergoing orthopaedic trauma operations.
All NHS hospitals in England with a trauma theatre(s) were contacted
to identify the ventilation system which was used between April
2008 and March 2013 in the following categories: always LF, never
LF, installed LF during study period (subdivided: before, during
and after installation) and unknown. For each operation, age, gender,
comorbidity, socio-economic deprivation, number of previous trauma
operations and surgical site infection within 90 days (SSI90) were
extracted from England’s national hospital administrative Hospital
Episode Statistics database. Crude and adjusted odds ratios (OR)
were used to compare ventilation groups using hierarchical logistic
regression. Subanalysis was performed for hip hemiarthroplasties.Aims
Patients and Methods
The aim of this study was to define return to
theatre (RTT) rates for elective hip and knee replacement (HR and
KR), to describe the predictors and to show the variations in risk-adjusted
rates by surgical team and hospital using national English hospital
administrative data. We examined information on 260 206 HRs and 315 249 KRs undertaken
between April 2007 and March 2012. The 90-day RTT rates were 2.1%
for HR and 1.8% for KR. Male gender, obesity, diabetes and several
other comorbidities were associated with higher odds for both index
procedures. For HR, hip resurfacing had half the odds of cement fixation
(OR = 0.58, 95% confidence intervals (CI) 0.47 to 0.71). For KR,
unicondylar KR had half the odds of total replacement (OR = 0.49,
95% CI 0.42 to 0.56), and younger ages had higher odds (OR = 2.23,
95% CI 1.65 to 3.01) for ages <
40 years compared with ages 60
to 69 years). There were more funnel plot outliers at three standard deviations
than would be expected if variation occurred on a random basis. Hierarchical modelling showed that three-quarters of the variation
between surgeons for HR and over half the variation between surgeons
for KR are not explained by the hospital they operated at or by
available patient factors. We conclude that 90-day RTT rate may
be a useful quality indicator for orthopaedics. Cite this article:
