

S148
Poster presentations / European Geriatric Medicine 6S1 (2015) S32
–
S156
P-432
Perioperative Comprehensive Geriatric Assessment is associated
with reduced inpatient length of stay
S. Singh
1
, R. Hodgkinson
1
, D. Shipway
2
, K. Moorthy
3
1
Imperial NHS Trust, London, United Kingdom;
2
United Kingdom;
3
Upper GI Surgery, St Mary’s Hospital, Imperial College Healthcare
Trust, United Kingdom, London, United Kingdom
Objectives:
Older surgical patients are frail and often have
multimorbidity. They develop more perioperative medical
complications and have longer inpatient length of stay (LOS).
Current guidelines state that best practice for older people
undergoing surgery should involve comprehensive geriatric
assessment (CGA) and access to a geriatrician. CGA reduces
inpatient LOS and medical complications in older patients
undergoing orthopaedic surgery, but limited data exists evaluating
its impact in other surgical populations. We aimed to
investigate whether a CGA approach for older patients undergoing
gastrointestinal cancer surgery could also reduce inpatient LOS in
an urban tertiary referral surgical unit.
Methods:
We established a geriatrician-led CGA liaison service for
patients aged 70 and over undergoing surgery for gastrointestinal
cancer. Pre-operative CGA was conducted in a rapid-access
outpatient clinic directly after diagnosis for patients deemed to
be high risk either as a result of frailty or multimorbidity. Proactive
post-operative multidisciplinary medical and therapy support was
embedded on the surgical ward to provide early post-operative
medical review and rehabilitation.
Results:
We calculated the LOS for patients undergoing gastro-
intestinal cancer surgery aged 70 and over in a 6 month period
both before and after the introduction of the CGA liaison service.
There was a 31% reduction in the LOS following the intervention.
Conclusions:
Perioperative CGA and post-operative geriatrician
support is associated with reduced inpatient LOS in patients aged
70 and over undergoing gastrointestinal cancer surgery.
Pre-intervention
(32 patients)
Post-intervention
(42 patients)
Total LOS
423
383
Mean LOS (days) 13.2
9.1
P-433
Differences in baseline characteristics and outcomes of older
persons requiring hospital admission after introduction of an
elderly care in reach service. Perioperative care of older people
undergoing surgery
–
Salford General Surgery (POPS-SG)
D. Houghton
1
, S. Krepple
2
, A. Vilches-Moraga
2
, J. Fox
2
, T. Thorpe
1
,
K. Wardle
2
, M.K. Peeroo
2
, E. Feilding
2
, Z.R. Alio
2
1
Salford Royal NHS Foundation Trust, Salford, United Kingdom;
2
Salford Royal NHS Foundation Trust, Salford, Manchester, United
Kingdom
Objectives:
The purpose of our study was to describe the impact
of an elderly care in reach service on patient outcomes.
Methods:
We compared all surgical patients (general, colorectal or
upper gastrointestinal) over the age of 74 discharged from hospital
between February 1st and March 31st 2014 with those assessed by
our in reach Service between February 1st and March 31st 2015.
Results:
Initial group (IG) (n 53) patients’ mean age was 81.3 years,
50% females, 81% emergencies, two in-hospital deaths, 3.7% deaths
and 20.7% readmissions before day 30. Length of stay was 10
days and 28% required review by generalists. Our in-reach team
(IR) assessed 63 individuals with a mean age of 81.4 years, 39%
females, 79% emergencies, 4 in-hospital deaths, 0 deaths and17.4%
readmissions before day 30. Length of stay was 12 days and 23%
required non-surgical reviews.
Present complains and diagnoses were similar (abdominal pain and
vomiting, cholecystitis and cancer). IG Patients had 3 comorbidities
on average compared to 5.1 and took 6.3 medications compared to
8.2. 49% of IG patients underwent medication review as opposed to
100% IR (with an average reduction of 2 medications). There was a
significant increase in the recognition of complications i.e. anaemia
(5.6 vs. 57.1%), acute kidney injury (1.8 vs. 31.7%) and constipation
(1.8 vs. 61.9%).
Conclusions:
An elderly care in reach service was able to increase
comorbidity and complication recognition, medication optimisation
and reduce out of hours reviews by non surgical specialists. There
was no significant change in clinical outcomes.
P-434
Predicting 30 day mortality after hip fracture: validating the
use of National Hip Fracture Database (NHFD) data
A. Johansen
1
, C. Tsang
2
, D. Cromwell
2
, C. Boulton
1
, R. Wakeman
1
,
V. Burgon
1
1
National Hip Fracture Database, London, United Kingdom;
2
Clinical
Effectiveness Unit, Royal College of Surgeons of England, London,
United Kingdom
Objectives:
The NHFD and Royal College of Surgeons of England
(RCS) have described a model with six predictive factors from the
NHFD dataset for casemix adjustment of 30 day mortality. Several
other outcome prediction tools have previously been described. We
set out to compare the NHFD-RCS model with the most widely used
of these – the Nottingham hip fracture score.
Methods:
We used the expanded dataset of our 2013 Anaesthetic
Sprint Audit of Practice (ASAP) – data for 7,906 patients aged
60+ years, who had hip fracture surgery in May–July 2013. We
linked to Office of National Statistics death data to identify patients’
mortality status 30 days after admission. We used the first six
weeks’ data (4,045 patients) to recalibrate the models, and the
next six weeks’ data (3,861 patients) to validate them. Multiple
imputation was used to manage missing data.
Results:
Several variables (AMT score, fracture type, some
individual comorbidities) were not significant predictors in
univariate analyses. After adjusting for other patient characteristics
we found age, sex, ASA grade (NHFD-RCS model), and number of
comorbidities (Nottingham score) to be the strongest predictors.
Both models displayed similar discriminative power; the highest c-
statistic achieved by each being 0.74. Both models over-estimated
mortality risk for patients in highest risk groups.
Conclusions:
Both models achieved moderate predictive
performance. In further work using NHFD data, we will explore the
scope for additional NHFD fields (eg. AMT score and deprivation)
to improve the NHFD-RCS model’s performance – for use with
individual patients and in hospital benchmarking.
P-435
Hip fracture following an inpatient fall: using the National
Hip Fracture Database (NHFD) to identify the true scale of
this challenge
A. Johansen
1
, C. Boulton
1
, V. Burgon
1
, F. Martin
2
, R. Stanley
1
,
R. Wakeman
1
, A. Williams
1
1
National Hip Fracture Database, London, United Kingdom;
2
Guys and
St Thomas’ Hospitals NHS Trust, London, United Kingdom
Objectives:
Hip fracture outcome is especially poor for people who
sustain this injury while an inpatient. Pre-existing medical and
psychiatric problems often prove challenging. In 2009 the UK’s
National Patient Safety Agency (NPSA) identified 840 hip fractures
after inpatient falls. We set out to identify the true incidence of
such presentations.
Methods:
During 2013 the NHFD collected data from all 182 trauma
units in England, Wales and Northern Ireland. We identified 64,838
hip fractures in people aged
>
60; over 95% of all such fractures. We