appendix 1: interim evaluation of inner pilot - Healthier North

Action 7.2
ANNEX: Initial evaluation of the
Inner Integrated Care Pilot
As part of the due diligence process to understand the impact of integrated care
on activity flows in acute care an initial review of the Inner pilot was conducted.
The focus of this was a lessons learnt approach using the frail elderly activity
for analysis- it is to be expected that one of the most comprehensive and
complex integrated care models in the UK would offer much learning to later
adopters. It is not a comprehensive analysis of the entire pilot.
OVERVIEW
The integrated care pilot (ICP) in Inner North West London (INWL) was
established to improve outcomes for patients, create access to better, more
integrated care outside of hospital, reduce unnecessary hospital admissions and
enable effective working of professionals across provider boundaries.
At the beginning of the pilot, INWL set targets for reducing emergency
admissions (including related non-PbR spend & subsequent follow up OP
attendances) and A&E attendances, with an overall target of avoiding 7
emergency admissions and 15 A&E attendances per 2,000 patients (which
translated to 1,771 avoided emergency admissions across the pilot in year 1) to
produce modelled savings.
To date,
– 93 case conferences were conducted at which 635 patients were
discussed;
– care planning has rapidly ramped up in January-March, with more than
7,200 plans completed to date (at a current rate of ~54 per day for
elderly and 43 per day for people with diabetes).
It is clear from analysis of other ICP’s (such as the recent DH report) that it is
too early to definitively declare success, however emerging data from Inner
NWL suggests that the ICP is having a positive impact. Given that other pilots
have not been able to show any changes in NEL activity over a two year period
(DH 2012), this is very encouraging.
In participating practices, July 2011 to January 2012, there has been an overall
reduction of 195 medical specialty NEL admissions among elderly, compared
to the same period a year ago (which translates to 334 on an annualised basis
comparing to absolute baseline, or 641 on an annualised basis comparing to the
originally modelled 6.1% growth trend assumption). This data does not take
account of the impact of the growth of care planning or people with diabetes
(data not available).
1
The pilot is beginning to show a significant difference in performance between
those practices participating in the ICP and those that are not. 195 avoided
admissions translate into a -6.6% growth rate in NEL medical specialty
admissions for participating practices compared to +0.3% trend for nonparticipating practices.
WHAT HAS BEEN DONE IN INNER ICP TO DATE?
The ICP in Inner NWL began in July 2011. The first 6 months (Jul to Dec
2011) centred on mobilising the pilot. This was the phase during which the
pilot started forming MDGs, began to hold case conferences and launched the
IT tool. To a limited extent, the MDGs started to complete the first care plans
during this phase as well.
MDG joined
Exhibit 1: Timeline
Apr
May
Jun
Acton
Case
conferences
Patients
discussed
Care
plans
IT
uploads &
services
Final Transitional Integrated Management Board
MDG
joining
date
Acton
Pre-launch
Jul
Aug
Sep
Oct
Nov
CLH
H&F C
Chiswick H&F NC
K&C N H&F SF
K&C S
Jan
Feb
Milestones
Mar
H&F SPC
Victoria
Chiswick H&F SF
K&C N
Dec
MDG 1st CC
K&C S
H&F NC
Victoria
H&F SPC
H&F C
CLH
25 CCs
100 pts
discussed
50 CCs
300 pts
discussed
600
care
plans
Prowell- Imperial,
ness
CLCH,
WLMHT,
Adastra
75 CCs
100 CCs
600 pts
discussed
~3,000
care
plans
~5,000
care
plans
7,500
care
plans
62
75
Services: CWH Services:
All
Ealing
prac- prac- CLCH, loaded CNWL
compatible ICO
tices tices Imperial,
practices1
CWH, GP
Mobilisation
Operational delivery
1 Excluding SystmOne and EMIS Web practices
Since the mobilisation phase, there has been a significant ramp up in care
planning. The number of care plans has increased from 600 in December to
~7,500 by the middle of March, which shows that in practice MDGs can
quickly ramp up work planning if the right enablers are in place. For the Outer
pilot setting appropriate ramp up rates will be a core requirement of the
performance management, to ensure full ramp up by the end of Q2, so that
impact can be seen earlier.
Case conferences have continued to ramp up with 94 conducted to date
discussing 635 unique cases (or 6-7 cases per conference). For the Outer pilot,
having clear expectations about the number of cases to be discussed each
month will help to trouble shoot any issues MDG’s are having with the
presentation of patients in the MDG.
2
There are two reasons for the prolonged mobilisation phase. First, the MDGs
did not all join the pilot on day one; instead they joined over a period of 7
months, with the last two joining in December. Secondly, the IT tool was only
available to enter care/work plans from the beginning of November. Since
November the rate of care planning has also increased. It is expected that
around 78% of elderly care plans will be completed by the end of the pilot’s
first year.
Actual historical trajectory
Exhibit 3: Care planning activity
Trajectory required to meet initial target
YTD (Jun 11 – Mar 12), indexed as % of total target care plans
Last 39 days trajectory
Elderly care planning target and rate
100
80
60
40
20
0
Dec
11
Jan
Feb
Mar
Apr
Care plan completion rate
(Care plans per day)
May
Jun
Initial required
36
First 28 days1
29
Last 39 days2
54
Jul
12
Diabetes care planning target and rate
100
80
60
40
20
0
Dec
11
Care plan completion rate
(Care plans per day)
Initial required
41
days1
13
Last 39 days2
43
First 28
Jan
Feb
1 Covers first three reporting periods
Mar
Apr
May
Jun
Jul
12
2 Covers most recent two reporting periods
WHAT HAS BEEN THE RESULT SO FAR?
The data collected to-date (up to January 2012) represents limited input by
MDGs (e.g., only the first few work plans created, still improving the
productivity of the case conferences) and, in line with national findings, the
effects of the pilot are still too early to tell definitively. Still, there is an
emerging distinct difference in the trend in non-elective medical and surgical
admissions between participating and non-participating practices1, for the frail
elderly pathway, as detailed below.
NEL admissions: Participating practices compared to non participating
practices: between July-January of 2011/12 (corresponding to the mobilisation
stage of the ICP) to the same period in 2010/11 shows that whilst nonparticipating practices saw a 5.7% growth in NEL activity, participating
1 The SUS data used in this report covers the period April 2009-Jan 2012. December 2011 and January 2012 have
not been post reconciled in line with reporting standards. However the magnitude of the difference in trends is
expected to be far larger than any change in data following reconciliation.
3
practices experienced a 0.4% reduction in non-elective medical and surgical
admissions
NEL admissions: Comparison of practices within the pilot over time:
Moreover, there is a positive change in the trend for participating practices for
NEL when compared over time. While they were growing at more than 13%
between a 7-month period in 09/10 and 10/11, this has now been reversed to 0.4% over the same period between 10/11 and 11/12. The non-participating
practices, however, have shown the continued trend of sustained growth in
NEL admissions that carried on over the past two years. This gives an
indication that there is no selection bias (i.e., the practices that joined the pilot
were not already admitting lower than average).
Exhibit 1: Trend in non-elective medical and surgical admissions
All non-elective medical and surgical admissions for patients aged 75 and over
Seven months Jul-Jan each year
Participating
INWL practices
+13.8%
Non-participating
INWL practices
-0.4%
+5.6%
4,964 4,943
-21
10/11 11/12
+0.9%
+5.8%
1,911
1,808
09/10
25,339
+1,398
+103
1,712
4,362
09/10
+5.7%
All NWL
23,722
10/11
11/12
23,941
09/10 10/11
11/12
ONWL’s growth rate over the same period was +7.7%
SOURCE: SUS data Apr 2009- Jan 2012
NEL medical admissions: as expected in year one, the greatest impact was
from reducing medical specialties. Participating practices in the three main
boroughs have seen a 6.6% reduction in NEL medical admissions in the 10/11
period (Jul–Jan) versus the same period in 11/12, which translates to 195
avoided admissions compared to the absolute baseline (Exhibit 2, 3). This
compares to a 0.3% growth rate for non-participating practices.
Similar to the trend for all NEL admissions, medical NEL admissions show
that practices which opted into the pilot, were, prior to the pilot, growing at
more than double the rate of non participating practices (Exhibit 2).
4
Exhibit 2: Trend in non-elective medical admissions
All non-elective medical admissions for patients aged 75 and over
Seven months Jul-Jan each year
Participating
INWL practices
11.3%
Non-participating
INWL practices
-6.6%
+5.4% +0.3%
2,937
1,097
All NWL
-1.5%
+6.5%
1,156 1,160
15,111
+4
-195
2,742
+916
14,416
14,195
2,638
09/10 10/11 11/12
09/10 10/11
11/12
09/10
10/11
11/12
ONWL’s growth rate over the same period was +11.0%
SOURCE: SUS data Apr 2009- Jan 2012
Exhibit 3: Trend in non-elective medical admissions by borough and
Participating practices
provider
Non-participating practices
Non-elective medical specialty admissions for patients aged 75 and over, resident in INWL
Growth rate, percent, 11/12 vs 10/11 seven months (Jul-Jan) each year
By ICP borough
By acute provider
-4.8
-9.6
Westminster
Imperial
0.8
Kensington
& Chelsea
Hammersmith
& Fulham
-10.1
-2.7
Chelsea &
Westminster
-0.3
-6.1
7.4
-5.0
3.4
Not statistically
significant
-1.61
Other
3.7
-6.6
-6.6
Total
Total
0.3
0.3
1 Not statistically significant due to small sample size
SOURCE: SUS data Apr 2009- Jan 2012
These reductions translate into 195 avoided non-elective admissions for
participating practices (2,742 in Jul-Dec 11/12, compared to 2,937 for the same
period in 10/11). It is important to note that this is an absolute comparison,
rather than a comparison to the 6.1% growth rate baseline. It also excludes the
impact on avoided Accident and Emergency (A&E) admissions (No data) and
all impact from Diabetes patients (No data). Annualising the 195 avoided nonelective admissions assuming the same growth trends leads to 334 expected
avoided admissions for a full year. It was projected in 2011/12, that
admissions would grow by 6.1%. This would translate to 641 avoided NEL
5
admissions2. For comparison, the total target reduction in NEL admissions for
the first year of the pilot in the Inner business plan across elderly and diabetes
NEL admissions (including medical and non-medical specialties) was 1,771.
Data to evaluate the impact of the ICP on non-elective surgical specialty
admissions, elective admissions, out-patient admissions and A&E admissions
across all providers is currently not available. We are also unable, at present, to
include any of the impact on diabetes patients due to the lack of data.
In Chiswick and Acton the trend is not well understood, as the nonparticipating practices are performing better than participating practices (8.3
and 5.1% for non participating practices in Hounslow and Ealing respectively
vs. 18.9 and 33.7% for participating practices). Further investigating into the
reasons behind this is required by the Inner team.
CONCLUSION
Overall the ICP in INWL has shown early promising signs in NEL activity
reduction during its first year despite the many challenges of implementing
such an ambitious change to practice, but it’s too early to definitively tell
(especially given that the available data doesn’t yet cover the phase when care
planning ramped up significantly)
In future evaluations, when the data is available, the reduction in admissions
from diabetes patients should also be included in performance evaluation.
2 641 = [(2,947 * 1.061) – 2,742] * 12/7 = [(2,947 admissions in 2010/11 * 1.061 growth rate) – 2,742 admissions
in 2011/12] * 12 months / 7 months
6
ADDITIONAL NOTE ON INITIAL PERSPECTIVE ON DEPARTMENT OF
HEALTH REPORT ON INTEGRATED CARE
The recent DH publication on integrated care aimed to analyse the effects of 16
integrated care pilots running in the UK, with a conclusion that integrated care
may not lead to significant financial savings. In particular, no significant
impact on NEL hospital admissions has been found, quite opposite to the
evidence emerging from the Inner ICP data. It is important to address the key
differences in the approaches of pilots cited by the DH publication and the
Inner NWL Integrated Care pilots that we believe have lead to DH-cited pilots
to not have any significant impact on emergency admissions:
■ Level of integration: Many of the pilots considered in the DH report are
integrated to a limited extent, compared to the wide integration across the
spectrum of care providers in Inner NWL ICP. As the executive summary
of the DH report states, “Most pilots concentrated on horizontal
integration – e.g., integration between community-based services such as
general practices, community nursing services and social services rather
than vertical integration – e.g., between primary care and secondary
care.” In 7 out of 16 instances secondary care was not involved. In 11 out
of 16 instances there was no clinical integration (i.e., no common clinical
pathways or agreed clinical protocols across multiple providers). Inner
pilot differs significantly in that it comprises of representatives from
across all settings of care from primary, secondary, community, social and
mental health care settings. In addition, the Inner pilot establishes
integrated clinical pathways across all settings of care. This is likely a
critical component to achieve the NEL reductions that many of the pilots
cited by the DH report fell short on.
■ Supporting infrastructure: While the approaches taken to establish each
of the 16 pilots cited in the DH report are different, the authors produce a
set of common theme recommendations around success factors. Upon
studying the facilitating factors and the barriers to making Integrated Care
successful among the pilots in the study, the DH report recommends a
well-supported infrastructure with strong leadership and clear
communications strategy in place: “Delivering integrated care can
require multiple preparatory activities, and perhaps most importantly,
clear, effective communication across different organisations, service
users, staff groups and professions. Integrated care interventions benefit
from performance management information and sentinel indicators to
both track and communicate progress. They might require not only
investment in IT but also changes to the policies and practices around its
use.” We believe that Inner and Outer IC pilot is well-positioned around
the areas of communication clarity and performance management, with the
well-resourced Operations team, a sound performance management
system in place and an information system that allows transparent yet
secure sharing of the data.
7
■ Focus on managing the complex cases: It has been noted in the DH
report that those ICPs with case management appear to be the most
promising. Through the MDG’s patients are effectively receiving case
management in its most expert sense, in that the range of expertise if from
across both the physical and mental health and social care spectrum. Inner
NWL Integrated Care pilot is designed with management of complex
cases at its core.
8