Going from CER to Patient-Centered Care: Implications of Heterogeneity (Text Version)
On September 28, 2010, David Atkins made this presentation at the 2010 Annual Conference. Select to access the PowerPoint® presentation (422 KB).
Going from CER to Patient—Centered Care: Implications of Heterogeneity
- Trial: Is treatment A better than treatment B?
- Clinician: Is treatment A better than B for this specific patient?
- Health care system: Is treatment A better than B, and for whom, in which settings?
Images: The Department of Veterans Affairs and Department of Defense logos. These two logos appear on all subsequent slides.
Heterogeneity and Policy
- Policies seek to promote use of "best" treatment option.
- "Best" treatment for population may not be same as that for individuals.
- Most important when variation is:
- Leads to big enough differences to change decision making.
- Treatment choices can't be adjusted.
- Is CABG the best option for all patients with diabetes?
- How might a health system encourage greater use of CABG in appropriate patients?
- Would it be appropriate to discourage CABG in groups where PCI produces equivalent outcomes?
Is CABG the "Best" Choice for Patients with Diabetes?
- Need to consider harms and complications.
- Patient preferences for different outcomes:
- e.g., short-term risks of CABG.
- Variation due to quality of surgeon:
- Applicability of trial evidence.
Policies Used To Influence Use of "Best" Treatments
- Audit and Feedback
- Coverage decisions:
- Conditional coverage
- Tiered coverage
- Quality Measurement:
- Incentives, Public reporting
Distinguishing Important from Unimportant Heterogeneity
- Does it change direction of NET benefit enough to alter decisions?
- Is it common?
- Is it predictable?
- Can it be detected and treatment modified in response to variation in benefits or harms?
Example: SSRIs for Depression
- Comparable effectiveness of most agents in depression responsiveness but individual variation.
- Affect decisions: YES—Variability in response and side effects.
- Common: YES
- Predictable: NO
- Can variable response be monitored? YES
Dealing With Variation in SSRI as Response in Policy
- Not possible to identify who will do better on a different agent.
- Cover only 1-2 SSRIs in formulary?
- Recommend starting all patients with a specific SSRI as initial therapy?
- Heterogeneity is a real and important phenomenon in research and policy.
- Examination of pre-specified factors in individual trials, SRs and meta-analysis can detect HTE:
- Be cautious about post-hoc sub-groups.
- Policies need to accommodate HTE.
- But doesn't mean that complete, unfettered clinician choice is best.