Preliminary Findings comparative organizational analysis of 11 local health departments

1
Preliminary Findings comparative organizational
analysis of 11 local health departments

• Jacqueline Merrill
• Biomedical Informatics
• Columbia University
• AcademyHealth PHSR-IG June 7, 2008

2
Overview

• Sample
• Framework and data
• PH Work
• Preliminary correlations between network and
  performance

3
Sample

• National sample 11 local health departments
• 3 mixed
• 2 centralized
• 5 independent
• All but one served a designated metro area
• Persons/sq. mile 6 (AZ) 526 (NJ)
• Site selection
• Size 36 186 employees
• Latest date of NPHPS assessment and score
• Willingness
• Response rate 86 93

4
1240 workers surveyed
Nurse Environmental Admin Support
of sample 16 10 23
NACCHO 2005 24.4 10.2 25.8
Enumeration 2000 25 10 12
5
Employee characteristics
6
AGE
ALL
ENVIRONMENTAL
NURSE
7
The organization for network analysis
8
Data we collected the organization

• Employee x Employee
• Information Exchange/Communication network
• Employee x Task 44 tasks done for routine work
• 22 administrative tasks and 22 service tasks
• Employee x Knowledge 53 knowledge better than
  avg
• 15 administrative, 22 program and policy, 6
  analytic, 7 PH science
• Employee x Resource 54 resources adequate or
  not
• 18 general, 9 data and information, 15
  documents, 12 outside partners

9
Task Knowledge Resource Minimum Data Set

• Terms extracted from
• accreditation evaluation instruments
• workforce surveys
• competency sets
• operational definition
• PH business processes

10
Top 10 tasks
Use email 90
Use internet 85
Phone 84
Manage files 79
Meet w/ clients 65
Process requests 62
Educate public 59
Post information 48
Report data 43
Serve on committees 43
Top 10 knowledge items
Office skills 83
Job descriptions 83
Problem solving 77
Chain of command 77
Mission 76
Workplace safety 75
Policy and procedures 72
HIPAA regs 67
Data collection 65
Diversity 64
Top 10 adequate resources
Email access 92
Internet access 90
Computer 89
Desk space 89
List serve 87
Co-workers 86
Safe conditions 86
Management team 79
Staff directory 78
IT staff 76
11
Top 10 inadequate resources
Translator 27
PDA 25
Directory community groups 23
Consumer complaints 23
County emergency plan 23
Library of resources 22
Translated information 22
Staff development plan 21
CE 21
Medicare program staff 20
Bottom 10 knowledge items
Genetics 15
Biostatistics 15
Ecological model 16
Screening 23
Contract requirements 25
Strategies for partnership 26
Grant requirements 27
Risk commo 27
Env Health 27
Epidemiology 28
12
Top TASKS by occupation
ADMIN
Top 10 tasks
Use email 88
Phone 86
Manage files 82
Use internet 78
Process requests 63
Register enroll 50
Manage inventory 46
Post information 45
Meet w/ clients 45
Process billing 41
ENV
Top 10 tasks
Phone 96
Use email 95
Use internet 94
Enforcement 94
Manage files 92
Investigation 89
Schedule services 88
Conduct site visits 80
Process requests 75
Educate public 74
PHN
Top 10 tasks
Meet w/ clients 91
Use email 88
Phone 86
Use internet 86
Educate public 84
Deliver hlth services 83
Review records 83
Manage files 71
Perform screenings 66
Process requests 62
13
Task to employee drill down
14
Knowledge to employee drill down
15
Network measurements

• Centralization
• Hierarchy, decisions, dissemination
• Density
• Communication, cohesion
• Complexity
• Integration of organizational components
• Between group (divisions) density
• silos

FOUND A TWO MODE NETWORK STRUCTURE
16
Left Divisions strongest Density
.24 Centralization .29 Right Divisions all
ties Density .11 Centralization .52
Admin
Account
Dental
MCH
Rantoul
Inf Dx
Env
WH
17
Left Divisions strongest Density
.08 Centralization .06Right Divisions all
ties Density .19 Centralization .52
18
Strongest CORRELATIONS w/ Essential Service 4
EPHS4
0.79
0.82
0.52
0.71
-0.43
0.09
0.46
0.56
Ind 4.1 Ind 4.2
0.73 0.75
0.69 0.83
0.43 0.53
0.60 0.72
-0.47 -0.36
0.17 0.03
0.20 0.58
0.60 0.48
Density All
Density Strong
Complexity All
Complexity Strong
Centralization All
Centralization Strong
Reduced Block All
Reduced Block Strong
ES 4 Mobilize community partnerships and action
to identify and solve health problems Indicator
4.1 Identify key constituents encourage
participation in community health activities
maintain directory of community health
organizations Indicator 4.2 Establish community
partnerships local health improvement committee
assess effectiveness of partnerships
19
Are there commonalities between health
departments?
20
Correlation coefficients comparing network
measures in 10 local health departments Pink
correlation 0.7 Yellow
correlation 0.5
LHD1 LHD2 LHD3 LHD4 LHD5 LHD6 LHD7 LHD8 LHD9 LHD10
LHD 1 1.00
LHD 2 0.67 1.00
LHD 3 0.64 0.91 1.00
LHD 4 0.70 0.99 0.95 1.00
LHD 5 0.72 0.31 0.54 0.38 1.00
LHD 6 0.58 0.70 0.92 0.77 0.72 1.00
LHD 7 0.78 0.96 0.92 0.97 0.52 0.78 1.00
LHD 8 0.10 0.60 0.79 0.63 0.34 0.78 0.55 1.00
LHD 9 0.69 0.95 0.94 0.95 0.53 0.80 0.96 0.71 1.00
LHD 10 0.60 0.94 0.81 0.92 0.28 0.62 0.94 0.50 0.89 1.00
21
COCONINO Density .16 136 nodes
Champaign Density .24 115 nodes
Broome Density .19 122 nodes
Missoula Density .21 106 nodes
22
Orange Density .24 165 nodes
Putnam Density .27 97 nodes
Garland Density .53 42 nodes
Burlington Density .16 138 nodes
Champaign Density .24 115 nodes
StLucie Density .28 187 nodes
StJohn Density .31 115 nodes
Broome Density .19 122 nodes
Missoula Density .21 106 nodes
23
Management effectsIt has changed the way we
think about the organization

• COMMUNICATION--monthly all staff meeting, all
  staff emails, recognizing people, director
  reports at weekly meetings
• Re-organize space
• Hiring
• Data sharing
• Social initiatives

24
Ongoing

• Complete database
• Establish baselines
• Multiple regression to explain between group
  variance
• Non linear regression to explain variability in
  performance
• Repeat study w/ one LHD to determine network
  effect of managerial change
• Exploratory data analysis

25
Thank you
26
What we asked

• Information Exchange Give/get work info
• not routine
• routine not frequent
• routine frequent
• Knowledge better than average
• Resources for routine work
• relevant available
• relevant not available
• not relevant
• Tasks
• assigned
• could fill in

27
Organizational network analysis

• tool for understanding complexity
• empirical, relational, non-linear
• used for assessment management decision support

28
Network findings

• Correlations
• Performance general score with Density,
  complexity, centralization, block model density
• Performance factor scores with Density,
  complexity, centralization, block model density

29
PH Work Findings

• Top ten tkr backup and resources needed
• Mean median mode on of PHN Env Admin staff
• Mean median mode age, years in PH
• Range of prof, technicians, management admin
  staff
• Pairwise correlations TKR between PHN and Env,
  PHN admin, Env admin

30
Two mode network