A joint Australian, State and Territory Government Initiative
Experiences and lessons from benchmarking Older Persons
Mental Health Services
Dr Rod McKay .
Braeside Hospital
National Mental Health Benchmarking Project
27 November 2008
Goal of this presentation
Describe results of participation in national benchmarking of mental health services for older people
Demonstrate how KPI data was used to explore clinical practice
Method
Review of documents utilised for the benchmarking forums was conducted
Reflection upon the author’s experience of participation and discussion with other participants.
Who was involved? Seven older persons mental health services
from five Australian states Selected as representative of ‘good practice’ Expected to have a history of using data within
service Facilitator and 2 project staff 3 to 5 individuals from each service
Combinations of Senior clinicians (multiple disciplines) Service managers Information support staff Service project staff
What did we focus on?
Understanding indicators and how to use them Understanding each others services Length of stay and factors that may impact
upon thisDiagnosisOutcome measure profilesAllied Health PracticeECT practiceBed management
What did we find?
All services provided ambulatory and community services to consumers with mental illnesses including psychotic illnesses, mood disorders and BPSD
Consumer profiles on the HoNOS 65+ were consistent with nationally available data; and consistent with diagnostic mixes of services
We had sufficient similarity to compare practice and performance; despite significant differences in the proportions of consumers with different diagnoses Available resources, and Proportion of patients who were born overseas
All services had both strengths and weaknesses in profile and performance
What did we find- selected KPIsKPIKPI Mean of Mean of
services services (06-(06-07)07)
Range Range
2006-072006-07
28 day readmission 28 day readmission raterate
5%5% 3% to 7%3% to 7%
Mean Length of stayMean Length of stay 45.6 days45.6 days 36 to 65 days36 to 65 days
Treatment days / 3 Treatment days / 3 month community month community carecare
7.1 days7.1 days 4.6 to 8.8 days4.6 to 8.8 days
New Client IndexNew Client Index 61%61% 38% to 91%38% to 91%
Outcome readinessOutcome readiness
- inpatient- inpatient
-community-community
129%129%
57%57%89% to 193%89% to 193%
14% to 125%14% to 125%Post discharge care Post discharge care <7 days<7 days
59%59% 22% to 75%22% to 75%
Further KPIs
KPIKPI Mean of Mean of services services (06-07)(06-07)
Range Range
2006-072006-07
Av Acute Episode Av Acute Episode CostCost
$29,361$29,361 $21,748- $21,748- $36,556$36,556
Cost/ 3 month Cost/ 3 month community carecommunity care
$1,860$1,860 $893- $2,599$893- $2,599
% target pop. % target pop. community carecommunity care
0.8%0.8% 0.2-1.6%0.2-1.6%
% target pop. % target pop. inpatient careinpatient care
0.16%0.16% 0.08-0.26%0.08-0.26%
Local access Local access inpatient careinpatient care
72%72% 36-100%36-100%
More KPIs…..
KPIKPI Mean of Mean of services services (06-07)(06-07)
Range Range
2006-072006-07
New Client IndexNew Client Index 61%61% 38-91%38-91%
Area per capita Area per capita resources- resources- ambulatoryambulatory
$30$30 $9-$39$9-$39
Area per capita Area per capita resources- resources- inpatientinpatient
$57$57 $32-$78$32-$78
Preadmission Preadmission community carecommunity care
50%50% 22-81%22-81%
Post discharge Post discharge community carecommunity care
58%58% 22-75%22-75%
What did we do regarding differing Length of Stay between services?
All selected a sub set of patients staying less than, or more than 60 days to identify differences inAgeGenderLanguage and residential care statusHoNOS 65+ profileDiagnosisECT use
What did we find - Length of stay
Each service could identify groups who stayed longer, but the factors were not the same between all services
Trend for LOS over 3 years was a slow gradual increase
Extracts fromExtracts from benchmarkingbenchmarking workbookworkbook
One service demographics>60 days>60 days <60 days<60 days
malemale 44%44% 42%42%
Mean ageMean age 74yrs74yrs 77yrs77yrs
age <=70y.o.age <=70y.o. 32%32% 16%16%
CALDCALD 24%24% 49%49%
requiring interpreterrequiring interpreter 25%25% 26%26%
Admission Admission residence residence …………………… ……………………RACFRACF
18%18% 30%30%
…………………………………………homehome
82%82% 58%58%
……………………………………..othe..otherr
0%0% 12%12%
Discharge Discharge residence residence …………………… ……………………RACFRACF
47% (47% (26% hostel26% hostel)) 47% (8% hostel)47% (8% hostel)
…………………………………………homehome
53%53% 50%50%
……………………………………..othe..otherr
0%0% 3%3%
Change in Change in residenceresidence
41%41% 30% change30% change
One service synthesis
Factors that appear to be most associated with increased LOS are High scores on HoNOS65+ item 8 (‘other’) and item 9
(relationships’) Living at home at admission requiring hostel care on discharge Being aged <=70yrs Having a psychotic illness
Factors with weaker associations are High score on HoNOS65+ item 1 (behaviour) and item
11 (living conditions) Requiring a change in residence
What did we do - Length of Stay
Discussed findings of analysis Explored together different practice regarding
Allied Health staff resources and roles Discharge practices Flow to other services, where available ECT practice
Each service commenced their own improvement project based on their own needs Eg changes to bed flow, ECT practice, social worker practice,
identifying consumers ‘at risk’ of long length of stay for more intense early discharge planning
BUT these were mostly only starting implementation by the end of benchmarking
What was useful?
Using HoNOS 65+ (Routine Outcome Measure) to understand clinical profiles of consumers within teams
Extract fromExtract from benchmarkingbenchmarking workbookworkbook
What was useful?
Improving understanding of the similarities and differences in service provision between organisations
Identifying areas of key differences in service provision and performance
Establishing informal networks and contacts Sharing intellectual resources (eg job
descriptions, clinical tools uses)
What was useful?
collaborating in generating projects to explore differences
initiating local projects to improve practice. (with some limitations)
the national KPI set and data from routine outcome measurement collections have been valuable tools to assist these processes.
Improved understanding of the use, and limitations of the national Mental Health KPIs; and indicators and related data in general
What was problematic?
Poorly compatible IT systems and financial systems across states
Conducting ‘joint projects’ to change practice within services
Range of data literacy of participants Ability to recruit support/ project staff Providing feedback to staff from participating
organisations not attending forums Lack of rules regarding withdrawal from the forums Time frames of forums vs time for practice change Services finding time to do ‘homework’ Services integrating ‘benchmarking’ into ‘quality
improvement’
Were there time trends?
BUT…..
All this considers mental health services for older people as a ‘separate being’ to other mental health services
What happens when we look into the results of the adult mental health service benchmarking?
What are there lessons to be learnt?
For OPMH services? For Adult MH services? For the use and creation of KPIs?
Adult vs Older Persons mental health service performanceKPIKPI Older Older
Persons Persons Mean Mean (06-07)(06-07)
Adult Mean (06-Adult Mean (06-07)07)
28 day 28 day readmission ratereadmission rate
6%6% 12%12%
Mean Length of Mean Length of staystay
45.6 days45.6 days 13.913.9
Treatment days / Treatment days / 3 month 3 month community carecommunity care
7.1 days7.1 days 9.79.7
New Client IndexNew Client Index 61%61% 60%60%
Outcome Outcome readinessreadiness
- inpatient- inpatient
-community-community
129%129%
57%57%89% 89%
36%36%
Adult vs Older Persons mental health service performance
KPIKPI Older Older Persons Persons Mean Mean (06-(06-07)07)
Adult Mean Adult Mean (06-07)(06-07)
Av Acute Av Acute inpatient Episode inpatient Episode CostCost
$29,361$29,361($584/day)($584/day)
$9,472$9,472($680/day)($680/day)
Cost/ 3 month Cost/ 3 month community carecommunity care
$1,860$1,860 $1,975$1,975
% target pop. % target pop. community carecommunity care
0.8%0.8% 1.5%1.5%
% target pop. % target pop. inpatient careinpatient care
0.16%0.16% 0.3%0.3%
Local access Local access inpatient careinpatient care
72%72% 84%84%
Adult vs Older Persons mental health service performance
KPIKPI Older Older Persons Persons Mean Mean (06-(06-07)07)
Adult Mean Adult Mean (06-07)(06-07)
Area per capita Area per capita resources- resources- ambulatoryambulatory
$30$30 $57$57
Area per capita Area per capita resources- resources- inpatientinpatient
$57$57 $43$43
Preadmission Preadmission community carecommunity care
50%50% 60%60%
Post discharge Post discharge community carecommunity care
58%58% 89%89%
Conclusion Benchmarking can assist mental health services
for older people to improve their understanding of differences in practice
and performance; generate useful local actions based upon these
KPI sets and routine outcome measurement assist these processes
They can also assist discussions about differences between mental health services for different age groups; and their relative performance
Conclusion
Rewards for services requireStaff to be supported to develop appropriate skills Integrating benchmarking with service quality
processesA willingness to question established ideas and
practices Time