L11 E2 Quasi Experimental design and MUE

Scientific Evidence pyramid and where Quasi @

Systemic review/meta analysis

RCT

Non RCT (quasi)

Observational

Qualitative

Pretest-post test description

Dependent variable measurement taken b4 and after intervention

Non eq control group design

Participant groups not randomized, so have differences at baseline

Have control and study group where you compare histories, retrospective

Time series design

Large series of observation made on same variable consecutively over time

Interrupted time series and multiple time series designs description

Is like pretest/post test with MULTIPLE measurement b4 and after (so more in between)

Regression Discontinuity Design

used to determine causal effects of interventions by assigning threshold

Retrospective expost facto design

Studying cause and effect relationships through retrospective search

How are Quasi designs related (3)

• ALL subtype aim to estimate a causal relationships btw intervention and its outcomes

• ALL rely on stat adjustments to reduce bias and enhance validity

• Intervention is NOT RANDOMIZED at each design

4 Times to use Quasi

Classic study design not feasible or ethical

investigator can't implement control group or randomize groups

Small sample size

not enough funding

Selection bias definition

Patients have difference between groups

Confounders and control techniques

Systematic error in way data is collected, creates false associations

Historical bias definition

Events that occur during study period that could influence outcome of the results

• ex. covid

Reporting bias

Publishing only positive results and ignoring negative ones

Temporal ambiguity definition

Undermine validity of prediction by overemphasizing features close to outcome of interest

Hawthorne effect

participants' behavior changes as a result of their knowledge of being in a study

Detection bias

Differences in how outcomes are measured in groups

Maturation bias definition

Changes in participants with passage of time

How to minimize selection bias, and reporting/detection bias

MUE definition

Performance/investigative improvement method focused on evaluating medication-use processes and improving patient or system outcomes

Can explore different processes and is quite flexible

Why do MUE? (3)

MUE vs DUE/DUR

MUE emphasis on improving patient outcomes and QOL

DUR: ongoing systematic assessment to ensure appropriate medication utilization at individual patient level

MUE team members

• core team member

• extenders

• subject matter experts

• other contributers

Objectives of MUE (8)

• Evaluate medication for safety/efficacy outcome, Reduce inappropriate prescribing

• practice standardization, Promoting guideline adherence

• Support interdisciplinary medication management,

• Support research for med alternatives, Evaluate new indications, Inform formulary decisions

Retrospective vs prospective MUE

Retrospective: look back, identify areas for improvement

Pros: Good for new drugs or processes (designed before a med is utilized), allow for real time monitoring (conducted while medication use)

5 parts of MUE life cycle

Design MUE

collect/analyze data

implement change plans

Assess effectiveness

Follow up/change/further research

Defining MUE 6 possible parts

Objective

Type of MUE

study design

setting and population

sample size

develop process for data collection

Criteria and collect data 6 things to consider when

Eligibility criteria

Endpoints

variables

measuring parameters

Is data collection possible

MUE external validity?

Analysis for MUE

Garbage in=garbage out: want high quality data

Analyze against standards

Any trends found?

Analysis inform improvement

Disseminating results for MUE

What stakeholders gotta share to?

COLLABORATE

4 parts of follow up for MUE

Document changes and write new processes

Ensure results is shared with ALL affected people

Set times for re evaluation

Ensure culture doesn't revert back

Mue Challenges (4)

Data availability/extractility

TIme constraints/urgency of MUE results

Garbage in=garbage out

Need to make sure variable descriptions are clear and measurable

4 MUE things to keep in mind

Interdisciplinary coordination

set clear objectives

schedule meeting check ins with other disciplines

REMEMBER STAKEHOLDERS

MUE 8 limitations

• Lack of organization, poor communication, no follow up

• data integrity

• Evaluation method disrupts patient care

• Lack of scope, Lack of education, Lack of hard wired corrective actions