HEPP

1
HEPP

• Revision 2003

2
What is a systems-based approach to managing
quality and safety in healthcare?

• Recognises that all humans make errors
• Most errors result from the system--inadequate
  training, long hours, ampoules that look the
  same, lack of checks, etc
• Swiss cheese model - when errors line up you
  get an adverse outcome

3
What is a systems-based approach?

• Active errors the errors that directly lead to
  an adverse event
• Latent errors failures built into the system (eg
  understaffing, error-tolerant culture, team
  conflict, poor management, value on status rather
  than knowledge)

4
Systems-based approaches

• Learn from other high risk, low error industries
  (e.g. aviation)
• Have organisational cultures that discourage
  latent errors
• Have reporting systems that promote no-blame
  learning

5
National Service Frameworks

• NSFs are part of the new means of managing
  quality in the NHS, where standards are set
  nationally, implemented locally, and monitored.
• Based on
• Evidence of clinical and cost effectiveness
• Views of users

6
National service frameworks

• For each specific service or care group
• set explicit, evidence-based national standards
• define explicit, evidence-based service models
• put in place programmes to support implementation
• attempt to address service variations across the
  country

7
National Service Frameworks

• establish performance measures and use them to
  assess performance
• specify timescales for delivery
• put in place supporting initiatives e.g.
  workforce training

8
Example Coronary Heart Diseasehttp//www.doh.gov
.uk/nsf/coronary.htm

• The CHD NSF set 12 standards for the prevention,
  diagnosis and treatment of CHD.
• The standards are to be implemented over a
  10-year period.
• They are underpinned by a number of fundamental
  values and guiding principles relating to
  access, efficiency, equity and quality.
• The NSF fixed immediate priorities and milestones
  against which progress would be measured.

9
National Service Frameworks

• The Commission for Healthcare Audit and
  Inspection ensures NSF standards are met and
  fulfils the monitoring function.
• Services with an NSF are likely to receive
  prioritisation for funding because PCTs and
  trusts are required to implement them.

10
Star ratings

• NHS Trusts are awarded stars based on how well
  they are doing, measured by performance
  indicators.
• Those that get 3 stars get more autonomy.
• Those that get no stars face tighter controls and
  will have to account to the NHS Modernisation
  Agency and be inspected more frequently by CHI.

11
What does clinical governance aim to do?

• Effective clinical governance should ensure
• continuous improvement of patient services and
  care
• a patient-centred approach that includes treating
  patients courteously, involving them in decisions
  about their care and keeping them informed
• a commitment to quality health professionals
  must be up to date in their practices and
  properly supervised where necessary
• the prevention of clinical errors wherever
  possible and the commitment to learn from
  mistakes and share that learning with others

12
Clinical governance covers

• NHS organisations systems and processes for
  monitoring and improving services, including
• consultation and patient involvement
• clinical risk management
• clinical audit
• research and effectiveness
• staffing and staff management
• education, training and continuing personal and
  professional development
• the use of information about the patients
  experience, outcomes and processes

13
Key components

• Clear lines of accountability
• Comprehensive programme of quality improvement
  activity
• Clear policies to manage risk
• Procedures for identifying and remedying poor
  performance

14
How to interpret a meta-analysis

• Meta-analysis involves combining the results of
  multiple studies to obtain more precise estimates
  of treatment effects.
• The "blob" in the middle of each line on the
  FOREST PLOT is the POINT ESTIMATE
• The width of the line represents the confidence
  interval of this estimate

15
Odds ratios

• An odds ratio is calculated by dividing the odds
  in the treated or exposed group by the odds in
  the control group
• When there is no difference between the two
  groups the odds ratio 1

16
Odds

• Odds are the ratio of the number of people in a
  group with an exposure to the number without an
  exposure. In a group of 100 people, if 20 people
  had the exposure and 80 did not, the odds would
  be 20/80 or 0.25.
• For undesirable outcomes an OR that is less than
  1 indicates that the intervention was effective
  in reducing the risk of that outcome.

17
Funnel plots

• Simple scatter plots of the odds ratios of trials
  against their sample size
• Used to detect publication bias
• Typically results from small studies scatter
  widely at the bottom of the graph
• The spread will narrow as the precision of the
  results increases with bigger studies
• When there is no bias it will look like a funnel

18
Heterogeneity

• Heterogeneity dissimilarity (from a statistical
  point of view)
• Fixed effects model sees variability as due to
  random variation ie if all the studies were big
  enough they would achieve the same results
• Random effects considers each study to be from
  a different pop, differences due to experimental
  error and pop diffs

19
Meta-analysis

• If the confidence interval of the result (the
  horizontal line) crosses the line of no effect
  (the vertical line), that can mean either
• there is no significant difference between the
  treatments OR
• the sample size was too small to be confident
  where the true result lies

20
Dimensions of quality of life
21
Dimensions of quality of life
22
Types of Instruments

• Disease specific Asthma Quality of Life
  Questionnaire, Arthritis Impact Measurement Scale
• Site Specific Oxford Hip Score, Shoulder
  Disability Questionnaire
• Dimension specific Beck Depression Inventory,
  McGill Pain Questionnaire
• Utility (seek to attach values to different
  health states) Euroqol

23
Types of Instruments

• Individualised (patient selects dimensions
  himself) SEIQoL
• Generic SF-36

24
Specific Instruments

• Advantages
• Very relevant content
• Sensitive to change
• Acceptable to patients

25
Specific Instruments

• Disadvantages
• Cant use them with people who dont have the
  disease so comparison is limited
• May not detect unexpected effects

26
Generic Instruments

• Try to capture broad range of aspects of
  health status
• Try to be relevant to wide range of patient
  groups

27
Generic Instruments advantages

• Can be used for broad range of health problems
• Can be used if no disease-specific instrument
• Enable comparisons across treatments for groups
  of patients

28
Generic instruments advantages

• Can be used to assess health of populations
• Can be used to detect unexpected positive or
  negative effects of an intervention

29
Generic Instruments disadvantages

• Loss of detail
• Loss of relevance
• May be less sensitive to changes that occur as a
  result of an intervention
• May be less acceptable

30
Properties any instrument must have

• VALIDITY
• Does it truly measure QoL? (and not anxiety
  instead, for example)
• RELIABILITY
• Does it measure QoL consistently and dependably?

31
Things you need to think about in selecting an
instrument

• Published work showing that the reliability and
  validity of this instrument have been
  established?
• Have there been other published studies that have
  used this instrument successfully?
• Is there anything about the way the instrument
  was developed that might affect its
  appropriateness for use by you?

32
Things you need to think about in selecting an
instrument

• Is it suitable for the area of interest?
• Does it adequately reflect patients concerns in
  this area?
• Is the instrument acceptable to patients?
• Is it sensitive to change?
• Is it easy to administer and analyse?

33
The Short-Form 36-item Questionnaire (SF-36)
34
SF-36

• Developed from instruments used in two
  large-scale studies conducted in USA
• RAND Health Insurance Experiment
• Medical Outcomes Study
• Short-forms (SF-36,SF-20,SF-12,SF-6) have been
  derived from longer (108 item) questionnaires of
  patient-assessed outcome

35
Two versions of SF-36

• Standard version uses 4 week recall period (can
  be used every 4 weeks)
• Acute version uses 1 week recall period

36
SF-36

• Now very widely used in research
• Growing number of publications
• Adapted and tested for British populations
• Reliability
• Validity

37
Suggested uses for SF-36

• Measure of general health
• Population surveys
• Patient management
• Resource allocation
• Audit tool
• Clinical trials

38
Description

• SF-36 contains 36 items which can be grouped
  into 8 dimensions. These are
• Physical functioning
• Social functioning
• Role functioning (physical)
• Role functioning (emotional)
• Bodily pain
• Vitality
• General health
• Mental health

39
SF-36 Scoring

• Responses to questions are scored
• Scores for items within each dimension are added
  together
• This score is transformed to give each
  respondents score for each dimension
• Scale from 0 (poor health) to 100 (good health)

40
SF-36 scoring

• You are NOT allowed to add up the dimensions
  to give an overall score
• Lots of arguments about whether overall scores
  are good or bad thing
• Limits usefulness of SF-36 in determining utility
  of different health states
• Makes it more difficult to interpret results in
  trials

41
Performance

• Acceptable
• 5-10 minutes completion
• Internal consistency good
• Test retest high
• Responsive to change
• Population data available

42
Problems with SF-36

• In (sick) elderly
• Sensitivity to change in very ill people
• Lack of single index

43
Specific vs Generic instruments
Specific
Generic

• Relevant
• Sensitive
• acceptable/understandable
• learn a lot about the effects of disease
• BUT
• No use if you dont have the disease
• cant compare effects of different diseases
• May not detect unanticipated effects

• Summarise effects of disease
• Can use for comparison
• Available if no specific instrument developed
• Can be used for whole populations
• May detect unanticipated effects
• BUT
• Not as sensitive or relevant as specific
• no detail
• acceptability problems

44
Patient satisfaction The policy background

• The NHS Plan (2000) emphasis on organising care
  around the patient and on accountability to
  patients
• National Service Frameworks - Use of patients
  views to make decisions about organisation and
  delivery of health care
• Important element of clinical governance.

45
The policy background

• Commission for Healthcare Audit and Inspection
  takes patients views very seriously
• Financial rewards for trusts linked to the
  results of the annual National Patients Survey
• Used in determining organisations star ratings

46
The policy background

• Involving patients and the public in healthcare
  published in Dept of Health in Sept 2001
• Set of proposals building on the NHS plan and the
  Kennedy principles a formal response to the
  Bristol inquiry

47
Proposals from Involving Patients and the Public

• Establishment of the Commission for Patient and
  Public Involvement in Health - an overseeing body
  which will set national standards and monitor
  local services, helping to ensure communities
  have an effective say in their local NHS.
• Locally based independent Complaints Advocacy
  Service (ICAS) in England, operating to core
  standards. http//www.doh.gov.uk/complaints/advoca
  cyservicelists.htm

48
Proposals from Involving Patients

• To introduce Patients Forums in every Trust, to
  bring the patients perspective in Trust
  management decision-making.
• Made up of local people, main role will be to
  provide input from patients on how local NHS
  services are run and could be improved. Each
  patients forum will have a representative on the
  trust board.

49
Proposals from Involving Patients

• Trust-based Patient Advice and Liaison Services
  to deal with patients concerns.
• Patients can use PALS to resolve or air concerns
  about treatment, care or support.
• PALS have direct access to the trust's chief
  executive and the power to negotiate an immediate
  solution.
• PALS will feed patients complaints back to
  ensure that lessons are learned and steps taken
  to ensure problems are tackled.

50
Proposals from Involving Patients

• Statutory local bodies, to be called Voice, will
  report patients' concerns and facilitate public
  involvement in the NHS.

51
Information for Informed Choice

• Information is seen as one of the key
  requirements of partnership
• Information for Health Strategy
• NHS Direct provides free information to
  patients and professionals

52
Information for Health Strategy

• Establishment of a national gateway site to
  health information for the public on the
  Internet.
• Strengthening the role of the Centre for Health
  Information Quality in accrediting patient and
  public information material.

53
Performance Indicators examples of patient
satisfaction at work

• National Cancer Performance Indicators
• Requires cancer centres to have an organised
  system for conducting patient satisfaction
  surveys
• Measures of trusts complying with this

54
Who measures patient satisfaction in the NHS?

• Department of Health through the national users
  survey
• Trusts, PCTs and other NHS organisations an
  obligation under the Performance Assessment
  Framework and under clinical governance

55
What is patient satisfaction?

• If we are going to measure it, it would help
  if we knew what it was!
• View of patient as consumer has been criticised
  is medical care a product?
• Definitions borrowed from world of commerce are
  clearly inappropriate

56
What is patient satisfaction?

• Lack of clarity in definitions of patient
  satisfaction.
• Probably better to think about patients
  evaluations or patients views rather than
  patient satisfaction.
• New NHS/DoH documents refer to patients views.

57
Hall and Dornans meta-analysis dimensions of
satisfaction

• Overall satisfaction
• access
• cost
• overall quality
• humaneness
• competence
• amount of information

• bureaucracy
• physical facilities
• providers of attention to psychosocial problems
• continuity of care
• outcome of care

58
Hall and Dornans meta-analysis dimensions of
satisfaction

• Understanding the significance of different
  dimensions has developed at different rates.
• Problems in interpreting some dimensions.
• It is particularly hard to conceptualise
  communication

59
Measuring patient satisfaction indirectly

• By looking at complaints and at patients who
  change doctor.
• http//www.nhs.uk/patientsvoice/how_to_complain.as
  p
• Performance indicators look at complaints (in
  addition to direct data).
• Health Service Commissioners report.
  http//www.ombudsman.org.uk/

60
Qualitative approaches to patient satisfaction

• Uses methods such as interviews, focus groups,
  observation
• Many qualitative studies have been very
  successful at identifying how patients evaluate
  care and what their priorities are.

61
Why use quantitative survey methods?

• Considered relatively cheap and easy to conduct
• Distrust of qualitative research
• Less researcher bias than interviews
• Less staff training required
• Anonymity more easily guaranteed
• Facilitates monitoring of performance

62
DIY instruments

• Local DIY instruments are developed by interested
  health professionals/ managers/researchers
• Can have advantages. However
• Many local instruments do not comply with basic
  standards for questionnaire design
• Many do not have proven reliability and validity
• Find higher levels of satisfaction than published
  instruments
• Lack of comparability

63
NHS National Survey

• CHAI runs national patient surveys to
• Provide feedback from patients which can be used
  for local improvement
• Provide information for star ratings and
  inspections
• Provide national data
• http//www.chi.nhs.uk/eng/surveys/index.shtml

64
What have studies of patient satisfaction found ?

• Patients tend to express satisfaction with health
  care.
• Lack of variation thought to be caused by
  patients reluctance to criticise NHS or health
  professionals.
• Greater dissatisfaction is expressed with
  specific aspects of care.

65
What have studies of patient satisfaction found ?

• Common causes of dissatisfaction
• delays in appointments or admissions,
• waiting around,
• receiving inadequate information,
• impersonal or unfriendly care.

66
Responding to patient satisfaction findings

• Are statements such as the following acceptable?
• Put simply, care cannot be of a high quality
  unless the patient is satisfied (Vuori, 1987)
• Should some aspects of the service be improved at
  the expense of others?
• How do we balance need to seek and act on
  patients views with other priorities?

67
Problems with responding to patient satisfaction
findings

• Some dissatisfaction may be unreasonable or
  unavoidable.
• Do we discount incompetent or distasteful
  views?
• Dissatisfaction may be less important than other
  outcomes in the long term.

68
Why change doesnt occur

• There are some things about which very little
  can be done because of
• the limitations of modern medicine
• the constraints imposed by limited resources
• institutional limitation
• conflict of priorities

69
Patient partnership

• Patient partnership underpins many recent policy
  initiatives
• Patient partnership is about more than
  satisfaction
• A model of patient care where patients are
  recognised as full partners
• Patients are offered full information and the
  opportunity to be involved in decision-making

70
Lots of unresolved problems about partnership

• People who dont want to be involved in
  decision-making
• Unknown consequences of involvement
• Do patients always know best?
• Under what circumstances should the power of
  patients be limited?

71
StBOP

• Abolition of
• 8 Regional offices
• 99 Health authorities
• Creation of
• 28 Strategic Health Authorities
• 304 Primary Care Trusts
• 4 departmental Directors of Health Social Care
• Alignment with 9 government regional offices

72
Local Structures

• Leicestershire, Northamptonshire Rutland (LNR)
  StHA
• 9 PCTs (6 in old LHA, 3 in old NHA)
• Leicestershire NHS Trusts
• Leicestershire Partnership Trust
• University Hospitals of Leicester

73
(No Transcript)
74
Primary Care Groups/Trusts

• All GPs are now members of a PCT
• Replaces fund-holding and take on many previous
  functions of Health Authorities
• Aim to involve primary care professionals (inc
  nurses etc) in the planning and commissioning of
  services

75
Key Responsibilities of PCTs

• Improving health and reducing inequalities
• Improving service provision
• Integrating health and social care

76
Improving health

• Assessing health needs of local communities
• Community development
• Health promotion and education
• Lead NHS organisations for partnership working
• Local strategic partnerships
• Sure Start, Quality Protects, Regeneration

77
Improving service provision

• As provider and commissioner
• Responsibility for all family health service
  practitioners
• Engagement of frontline staff and local
  communities
• Collaboration with other PCTs for commissioning
  some services
• E.g. tertiary care services

78
Integrating Health Social Care

• Work with Local Authorities to maximise
  opportunities for patients and users
• Potential for full integration through creation
  of Care Trusts

79
Eastern Leicester PCT

• Population of 180,000
• Younger than Leicestershire
• High proportion of ethnic minorities
• Diversity of languages, culture and religious
  observance
• High levels of deprivation and health needs
• 33 General practices
• High proportion of sole practitioners
• Lowest level of WTE GPs per 100,000 population in
  LNR
• Low levels of nurses and therapists
• Large number of Community pharmacists and General
  Dental practitioners

80
Resource allocation

• Necessary because of the growing pressures on
  health budgets
• A worldwide problem, not just a UK one
• Doctors are increasingly involved in making
  resource allocation decisions explicitly rather
  than implicitly

81
Possible criteria for allocating resources

• equality - everyone gets the same. Not useful as
  doesnt address different needs
• equity - equal treatment for equal need (need
  definition of need)
• need - ability to benefit from a health care
  intervention (difficult to define and measure)
• Still left with the problem of choosing
  between different treatments and different groups
  of treatments.

82
Cost effectiveness analysis

• Involves the measurement of benefits in a single
  natural or disease specific outcome such as
  number of strokes prevented
• You are interested in the least costly way of
  achieving a given outcome eg a reduction of 50
  strokes per annum

83
Cost minimisation analysis

• Here you are interested in which intervention
  costs the least, assuming the outcomes are the
  same (eg all drugs are equally effective at
  reducing blood pressure)

84
Cost benefit analysis

• Used when you can put a monetary value on the
  outcomes as well as the costs (valuing life and
  limb!)
• Benefits often defined as the costs avoided by
  the intervention (eg how many days in hospital
  you avoid by having a drug)

85
Cost utility analysis

• Involves comparing interventions according to
  their yields in terms of extended life and
  quality of life (often expressed as QALYs)

86
Quality Adjusted Life Years

• QALYs adjust life expectancy for quality of life.
• Try to combine quality and quantity of life into
  a single index.
• 1 year of perfect health 1 Quality Adjusted
  Life Year

87
QALY approach assumes that

• 1 year in perfect health
• is the same as
• 10 years with a quality of life 0.10 of perfect
  health

88
Use of qalys in priority setting

• Requires information on qalys for different
  procedures plus costs of these procedures
• These can then be ranked into a QALY league table

89
QALY league table

• Present value of
• extra cost per
• QALY gained ()
• Cholesterol advice and diet therapy 220
• Hip replacement 1,180
• Kidney transplantation (cadaver) 4,710
• Breast cancer screening 5,780
• Neurosurgery for malignant brain tumours 107,780

90
Using qaly league tables in priority setting

• An explicit, not implicit form of rationing.
• Does acknowledge population needs.
• Allows explicit comparisons between
  interventions.
• Allows prioritisation of interventions.
• May be useful at individual patient level.

91
Problems with QALY league tables

• Do not distribute resources according to need,
  but according to the benefits gained per unit of
  cost.
• Discriminate against elderly
• Discriminate against the already disadvantaged
• More appropriate for acute than chronic conditions

92
Problems with QALY league tables

• Dont distinguish between interventions that are
  life-enhancing vs life-saving
• Technical problems with their calculations
• QALYs may not embrace all dimensions of benefit
  Values expressed by experimental subjects may not
  be representative

93
Problems with QALY league tables

• May not be an acceptable form of rationing
• Evidence on costs is not good
• Assume that everyone perceives value of health in
  the same way

94
Sensitivity

• Sensitivity __a___
• a c

95
Specificity

• Specificity __d___
• b d

96
Positive predictive value (PPV)

• PPV - is the proportion of the people who are
  test positive who actually have the disease

97
Positive predictive value (PPV)

• PPV __a___
• a b

98
PPV the influence of the prevalence of the
disease

• The PPV is strongly influenced by the prevalence
  of the condition
• A low prevalence condition will have a lower PPV
  than a high prevalence condition, even if the
  sensitivity specificity of the tests are the
  same

99
Prevalence of the disease

• Prevalence __a c___
• a b c d

100
PPV the influence of prevalence for a high
prevalence disease

• Assume the following scenario
• Screening in a diabetic clinic for diabetic eye
  disease
• Diabetic eye disease (diabetic retinopathy)
  occurs commonly in diabetics and thus is a high
  prevalence condition in diabetic clinic
  populations
• Assume the prevalence of diabetic retinopathy is
  30.

101
PPV for a high prevalence disease

• Prevalence 30
• (Sensitivity 87 Specificity 96)

102
PPV for a high prevalence disease

• Prevalence 30 PPV 260 / 290 90
• (Sensitivity 87 Specificity 96)

103
PPV the influence of prevalence for a low
prevalence disease

• Assume the following scenario
• Now we are going to take the same test (same
  sensitivity and specificity) for diabetic eye
  disease and screen the general population where
  of course the condition is much less common (low
  prevalence)
• Assume the prevalence is 1

104
PPV for a low prevalence disease

• Prevalence 1
• (Sensitivity 87 Specificity 96)

105
PPV for a low prevalence disease

• Prevalence 1 PPV 9 / 49 18
• (Sensitivity 87 Specificity 96)

106
Impact on PPV of the same test for the same
disease when the prevalence is different

• High prevalence
• Sensitivity 87
• Specificity 96
• Prevalence 30
• PPV 90

• Low prevalence
• Sensitivity 87
• Specificity 96
• Prevalence 1
• PPV 18

107
Management roles for doctors

• Medical director (overall responsibility for
  medical quality)
• Clinical director (overall responsibility for
  directorate)
• Consultant (responsibility for team)
• General practitioner PCT/practice principal
• Any level

108
Current model of NHS management (acute trusts)

• Trust Board involved in strategy setting
• Chairman
• Chief Executive
• Executive Directors (inc Medical Director)
• Non-executive Directors
• http//www.uhl-tr.nhs.uk/about_uhl/thetrust.htmlC
  hairman

109
The Trust Board

• Involved in strategy setting
• Responsible for operational management
• Respond both to local needs and national
  requirements

110
Medical directors

• The Medical Director is separate from the
  Clinical Director
• Medical directors is responsible for quality of
  medical care
• Sits on the Board of the Trust and communicates
  between the board and the medical staff

111
Medical directors

• Expected to show leadership for medical staff
• Demonstrate appropriate values
• Have partnership with human resources functions

112
Role of Medical Director some tasks and
activities

• Approves job descriptions
• Disciplinary
• Induction training
• Interview panels equal opps
• Discretionary points
• Conducting strategic overview

113
Current model of NHS management (acute trusts)

• CLINICAL DIRECTORATES
• The trust is organised into clinical directorates
  (a bit like faculties in a university)
• Directorates are based on a speciality or group
  of specialities eg Radiology, Womens Health,
  Cardiology
• Headed up by a Clinical Director
• Usually a nurse manager and a business manager
  too

114
Role of Clinical Director - heads up a directorate

• Managerial
• e.g. Select staff Assign work and
  resources Appraise performance
• Co-ordinating and Liaising
• e.g. Propose actions Communicate to
  group Overcome problems

115
Role of Clinical Director

• Representative
• e.g. Present views advocate position
• Monitoring
• e.g. Standards of quality Expenditure Outputs
• Developing Service Relationships
• e.g. with PCTs and other directorates

116
The Clinical Directors role

• Provision of continuing medical education and
  other training.
• Directorate policies on junior doctors' hours of
  work, supervision, tasks and responsibilities.
• Implementation of medical and clinical audit.
• Development of management guidelines and
  protocols for clinical procedures.
• Induction for new doctors.

117
GMC guidance on managing teams

• Problems can arise when communication is poor or
  responsibilities are unclear.
• Each member of the team should know where
  responsibility lies for clinical and managerial
  issues and who is leading the team.
• Systems should be in place to facilitate
  collaboration and communication between team
  members.
• Systems should be in place to monitor, review
  and, if appropriate, improve the quality of the
  team's work.
• Teams should be appropriately supported and
  developed, and be clear about their objectives.

118
Problems of managing people

• Managing doctors has often been compared with
  herding cats.
• Doctors are highly intelligent, highly skilled
  people and high status people.
• Non-medical managers have great difficulty in
  asserting control over doctors.

119
Collegial Relations

• Culture can be hostile to clinical managers
• Evidence of hostile and difficult collegial
  relationships
• Positive evidence of performance management
• May be great difficulties in managing change