Evaluating neurorehabilitation: lessons from routine data collection

1
Evaluating neuro-rehabilitation lessons from
routine data collection

• J A Freeman
• (1) Faculty of Health and Social Work, School of
  Health Professions, University of Plymouth (2)
  Institute of Neurology, Queen Square, London

2
Background Clinical Databases

• Enable standardised clinical information to be
  collected
• Allow care to be examined as it occurs in routine
  practice rather than in controlled circumstances
• Have potential value for informing clinical
  practice, management of services and research
• Are being increasingly used within the UK to
  assess outcomes of healthcare services

3
In the UK..

• Few units have a structured way of collecting,
• storing, analysing or systematically
• interpreting and disseminating the information
• collected

Wastes resources
4
Aim of this presentation

• Describe the
• Introduction, development routine use of a
  clinical outcome database
• Clinical characteristics and outcome of a
  structured multidisciplinary inpatient
  neuro-rehabilitation programme
• Lessons learnt

5
Methods1. The Rehabilitation Unit Programme

• The Rehabilitation Unit
• 18 bedded specialist neurological inpatient unit,
  National Hospital for Neurology and Neurosurgery,
  London
• Specific admission criteria
• The programme
• Individually tailored, intensive, goal oriented
  programme
• Multidisciplinary input
• Goals broad ranging
• Process supported and monitored by an integrated
  care pathway

6
Methods2. Process of data collection

• Clinical data
• Diagnostic information, gender, age, length of
  stay, admission discharge destination
• Outcome measures
• Impairments (EDDS)
• Functional activity limitations (Barthel Index
  FIM)
• Restriction in participations (London Handicap
  Scale)
• Visual analogue scale (patient rated main problem
  / benefit gained)
• Measurable short and long term goals
• Scored by consensus following observation of the
  patient by
• the multidisciplinary team

7
Methods3. The database

• Custom designed
• Microsoft Access package
• Free text data minimised
• Database stores
• Basic demographic data
• Diagnosis coded according to five groups (stroke,
  spinal cord syndrome, MS, neuromuscular, other
  brain pathology) supplemented by free text
• Outcomes data (length of stay, total and item
  scores for all measures)

8
Methods4. The study sample

• All patients admitted to the Unit between May
  1993
• and December 2002 whose length of stay gt 10 days

9
Analyses

• Quality control of data
• Review of all diagnostic codes
• Missing data and out-of-range values checked
• Consistency of data checked
• Analyses
• Description of sample- descriptive statistics
• Determining effectiveness of programme
• Effect sizes calculated
• Investigation of potential predictors of outcome
• General linear model analysis

10
Results

• Of the 1463 consecutive patients admitted over
  nine year period, 1413 patients had length of
  stay gt 10 days.
• Complete diagnostic and demographic information
  available for 100 of these patients
• Functional limitations data available for 96 of
  these patients

11
Results 1. Diagnosis (n 1413)
12
Results 2. Barthel Index (n 1413)
13
Results 3. General linear model analysis
Barthel Index Scores R2 44
14
Discussion 1. Feasibility and validity

• Largest reported data set of inpatient
  neurological rehabilitation patients in the UK
• Demonstrates that systematic collection, analysis
  and interpretation of standardised clinical
  outcomes data can be successfully incorporated
  into routine clinical practice
• Methods robust and reproducible
• Data valid
• Could act as a model, thereby facilitating
  sharing of data at a national level

15
Discussion 2. The clinical benefits of using
the database

• Provides a focus for careful recording and
  monitoring of caseload and educational needs
• Systematic accrual of information over longer
  periods about range of conditions
• Facilitates a more performance oriented and
  accountable system of rehabilitation
• Provides supporting evidence to complement
  results of clinical trials

16
Discussion 2. The research benefits of the
database

• Provides objective data for determining sample
  size calculations for clinical trials
• Quick and accurate method for assessing
  feasibility of patient recruitment for clinical
  trials
• Adds to evidence base by complementing data from
  clinical trials
• Provides information to help aid in selection of
  outcome measurement instruments

17
Lessons learnt

• Key-workers are important to constantly monitor
  data collection
• Validated measures should be used, keeping the
  process simple, quick relevant
• Regular training in use of measures and feedback
  about data is essential
• A person with dedicated time is needed to manage
  the database
• Clear, dynamic leadership must be integral to the
  process