Karen Palmer & Agortsas: Systematic Review of ABF of Hospitals: Potential Effects on Quality, Access, Cost, Efficiency, and Equity

50 %
50 %
Information about Karen Palmer & Agortsas: Systematic Review of ABF of Hospitals:...
Health & Medicine

Published on February 24, 2014

Author: informaoz

Source: slideshare.net


Karen Palmer, Adjunct Professor, Simon Fraser University & Thomas Agoritsas, MD, Department of Clinical Epidemiology & Biostatistics,Faculty of Health Sciences McMaster University
delivered this presentation at the 2014 Activity Based Funding conference at Toronto Convention Centre. Presentations at the event explored the risks, benefits and experiences of activity-based funding from around the world. For more information about the annual event, please visit the conference website: http://www.healthcareconferences.ca/activitybasedfunding

Systematic Review 
 Activity-Based Funding of Hospitals Impact on Quality, Access, Equity, Cost, Efficiency “ABF Systematic Review” Team Karen Palmer, Thomas Agoritsas, Taryn Scott, Sohail Mulla, Ashley Miller, Arnav Agarwal, Andrew Bresnahan, Afeez Hazzan, Rebecca Jeffery, Arnaud Merglen, Ahmed Negm, Reed Siemieniuk, Neera Bhatnagar, Irfan Dhalla, Bob Evans, John Lavis, John You, Stephen Duckett, Danielle Martin, Gordon Guyatt

Rationale for SR •  Opinions about ABF are divided •  Would risks outweigh the benefits? •  Is the intervention effective and safe? •  Until now, not a single systematic review!

What is ABF •  Hospitals paid fixed amount per patient episode of care •  Hospital services •  prospectively classified into clinically meaningful “bundles” of care •  ordinarily provided to patients admitted with particular diagnoses •  Activity bundles account for •  patient diagnosis •  complexity of care •  type •  volume •  Intensity •  Money follows the patient

What is NOT ABF •  Not Pay for Performance (P4P) •  Not BC’s unique 
 “patient-focused funding” (PFF) •  Not per Diem

Research Question •  How does ABF affect ? •  Mortality •  Quality of care (e.g. readmission) •  Discharge destination (post-acute care) •  Access to care (e.g. severity of illness, volume of care) •  Equity (e.g. patient characteristics) •  Hospital costs, and health care system costs •  Efficiency

Guess how many citations? •  Retrieved by a comprehensive search of the literature (medical librarian) •  Since 1980, all languages with English title, all countries •  Potentially address research question ? 16,565

How many studies answer “yes” 
 to this question? Does the study include original quantitative data that allows an evaluation of the relative impact of ABF versus alternative funding systems in hospitals and/or non-hospital medical/surgical facilities on cost or quality or access or equity or efficiency? ? 260 1.6% of citations Finding a needle in a hay stack!

Methodology •  To answer our research question, we did a systematic review and meta-analysis. •  Non-systematic reviews highly prone to bias •  “The more passionate scientists are about their work, the more susceptible they are to bias.” •  High stakes: ABF is not an intervention at the patient-level but at the population level for a whole province, or even a whole country.

Systematic Review – 9 steps Define Research Question Define Eligibility Criteria & Methods (steps, rules, manuals) Develop and Conduct Search Strategy Screen Titles and Abstracts: Potential Eligibility Review Full Text: Eligibility Abstract Data (Quantitative and Narrative) Assess Credibility of Findings (Risk of Bias) Analyze and Summarize Outcome Data (Quantitative and Narrative) Look for Explanation in Variability of Findings

Eligibility Criteria •  Original quantitative data •  Compared the impact of ABF vs. alternative funding systems of acute care settings •  Excluded studies that did not include a comparator group where ABF was not implemented •  “0% ABF” period or “no ABF” jurisdiction •  Excluded studies where comparison was only based on hypothetical modeling •  Included all analytical study designs

Study designs Single Jurisdiction Multiple Jurisdictions Before-After Study Parallel controlled study Parallel with Before-After Study 0% 100% (50%) Jurisdiction 1 0% Jurisdiction 2 xx% J 1 0% xx% J 2 0% 0% E.g., difference-in-difference analyses, time-series

Search strategy (1) •  Collaboration with health librarian •  Searched 1980 onwards, all countries,
 all languages •  9 electronic bibliographic databases •  •  •  •  •  •  •  •  •  OVID MEDLINE EMBASE OVID Healthstar CINAHL CENTRAL Health Technology Assessment NHS Economic Evaluation Database Cochrane Database of Systematic Reviews Business Source Complete •  Grey literature: personal files, reviews, websites, and consultation with experts

Search strategy (2) •  Complex search strategies for ABF and related terms: 
 92 lines of search terms= very comprehensive search! •  Iterative process to refine and optimize search

Screening for eligibility •  2 steps: •  all Title and Abstracts •  then Full-text •  2 independent reviewers
 calibration for inter-rater reliability •  Conflict resolution 
 3rd adjudicator if necessary

Flow Chart (1) Citation retrieved n=28,939 Title and Abstract screened n = 16,565 Duplicates removed n=12,374 Excluded 15,620 Full-Text Screened N = 945 Eligible Studies N = 260 Excluded 685 US Studies n=156 International Studies n=71 64 Countries

Outcomes LOS n=143 Quality of care n=75 Cost n=71 Mortality following acute care n=12 Discharge to PAC n=36 Efficiency n=20 Mortality in PAC
 n=4 Access or Equity n=172 Administrative Burden n=8 Readmission 

Outcomes LOS n=143 Quality of care n=75 Cost n=71 Mortality following acute care n=12 Discharge to PAC n=36 Efficiency n=20 Mortality in PAC
 n=4 Access or Equity n=172 Administrative Burden n=8 Readmission 

Current first review
 n=65 studies Mortality following acute care n=12 Volume of care (Access) n=26 Mortality in PAC
 n=4 Patient characteristics (Equity) n=31 Readmission 
 n=30 Discharge to PAC n=36 + . Severity of illness (Access) n=30 Indicator of risk selection but also differences in coding 
 (upcoding, DRG creep)

Discharge to PAC •  Broadly defined PAC: any “intermediate level care” needed after acute hospital stay; anything but “home” with self-care •  Different in each country
 e.g. US Medicare includes: •  skilled nursing facilities (SNF) •  in-patient rehabilitation facility (IRF) •  long-term care hospital (LTCH) •  home health agency (HHA) •  We dichotomized discharge destination to: •  PAC •  Home without any PAC

Study demographics 
 n=65 studies —  Data from 10 different countries —  Australia (3) —  Austria (1) —  England (1) —  Germany (2) —  Israel (1) —  Study Design —  Before vs. After (59) —  Parallel groups (3) —  Combined (3) —  Italy (4) —  Scotland (1) —  Sweden (1) —  Switzerland (2) —  US (50) —  Sampling —  Random (11) —  Convenience (36) —  All eligible institutions (19)

Standardized Data Abstraction •  Need rules for •  reproducibility •  protection against bias •  abstraction manual (85 pages) •  In pairs, reviewers independently abstracted: •  •  •  •  country and year of ABF implementation data source, sampling methods study population outcomes assessed and results… •  Standardized forms adapted to each study design

Credibility of Findings •  Although many decisions are supported by evidence of limited quality, some evidence is more credible than other 
 (less prone to bias) •  Assessed again in duplicate for 3 domains •  Documentation quality of data sources (3 pts) •  Number of eligible outcomes simultaneously examined (1 pt) •  Comprehensiveness/ appropriateness of the adjustment (2 pts) •  Credibility score •  from 0 to 6 •  <4 lower credibility •  >=4 higher credibility

Analysis of Outcome Data •  Meta-analysis combines finding across studies •  Not all studies provide enough data to do this •  Systematically assessed which studies could be pooled Poolable data? Yes No Meta-analysis Random effect models Structured vote counting Direction Magnitude of effect Statistical significance

Forest plots and conclusion not shown Embargoed until publication

Add a comment

Related presentations

Related pages


slide 1 systematic review on the effects of anti- inflammatory asthma medication on feno values class 19 fmup 2008/2009 adviser: ...
Read more

Jennifer Cowen - Minter Ellison Lawyers - Liability Risks ...

... Requirements for Public Hospitals ... Karen Palmer & Agortsas: Systematic Review of ABF of Hospitals: Potential Effects on Quality, Access, Cost, ...
Read more

Early Effects of Resident Work-Hour Restrictions on ...

... A Systematic Review and Plea for Improved Studies ... .push({}); Slide 1 Early Effects of Resident Work-Hour Restrictions on Patient Safety: ...
Read more

Ivan Szpakowski - Citi Research - slidesearch.net

Iron Ore Market Outlook: Deep Cyclical Correction or Continued Buoyancy? See Appendix A-1 for Analyst Certification, Important Disclosures and non-US ...
Read more

The Effects of Pre- And Postoperative Fibrinogen Levels on ...

The Effects of Pre- And Postoperative Fibrinogen Levels on Blood Loss After Cardiac Surgery a Systematic Review and Meta-Analysis_files. The Effects of Pre ...
Read more