Deb Tairas presentation

Health Services Research at HMSA: 

Health Services Research at HMSA Deb Taira, Sc.D. Research Manager February 16, 2006

Slide2: 

Why Research? “The burden of harm conveyed by the collective impact of all of our health care quality problems (overuse, underuse, misuse) is staggering. It requires the urgent attention of all the stakeholders…” Institute of Medicine National Roundtable on Healthcare Quality, 2000

Slide3: 

What is Research? Research is systematic investigation designed to yield generalizable results Goal: Improve quality and support cost-effective care

Slide4: 

Why Research at HMSA?

Slide5: 

Health Plan Benefit Structure, Cost Fee structure Treatment plan Cost and Quality Patient factors Age, gender, Ethnicity, Behaviors, Preferences, Income, Environment MD factors Age, gender, Ethnicity, Specialty, Evidence, Guidelines, CME Employer Government

Slide6: 

Why Research at HMSA? Improves quality of care Reduces inefficiencies through detection of overuse, underuse, and misuse Increases rigor in planning/documentation Supports key decisions Encourages evidence-based approaches Provides supplemental revenues through research grants

Data: 

Data Member Administrative data: age, gender, island or region, morbidity level (ACG), health plan type, days of enrollment, diagnoses, procedures, medications, cost, principal physician Satisfaction surveys: ethnicity, education, satisfaction Census: census tract income level Physician: specialty, practice region, board certification, health center affiliation

Some Examples: 

Some Examples

Slide9: 

Q1a. How does the physical health status of HMSA satisfaction survey respondents compare to that of the average American? Worse Same Better

Slide10: 

Q1b. How does the mental health status of HMSA survey respondents compare to that of the average American? Worse Same Better

Slide11: 

ANSWERS: Q1a=1; Q1b=3 national norm Citation: Taira et al. Managed Care Interface, 2001

Slide12: 

Q2. Which racial/ethnic group was least compliant with osteoporosis medications? Caucasian Chinese Filipino Japanese Hawaiian

Slide13: 

* Relative Days of Persistence on Osteoporosis Medication Relative to Caucasians ANSWER Q2= 3 Filipino

Slide14: 

Q3a. What percent of elderly PPO members report skipping medications due to cost? <5% 5%-10% 10-15% >15%

Slide15: 

Q3b. What percent of elderly 65cplus members report skipping medications due to cost? <5% 5%-10% 10-15% >15%

Slide16: 

Percent of Elderly Members Who Skip Medications Due to Cost ANSWER: Q3a=2; Q3b=3 Citation: Taira et al. American Journal of Managed Care, 2002

Slide17: 

Q4. Which group of patients tended to be more compliant with their anti-hypertensive medications? Patients seeing a MD of same age Patients seeing a MD of same ethnicity Patients seeing a MD of same gender All of the above None of the above

Slide18: 

  Impact of Seeing a Physician of Similar Age No significant differences Abstract submitted to the American College of Cardiology

Slide19: 

  Impact of Seeing a Physician of Similar Gender No significant differences Abstract submitted to the American College of Cardiology

Slide20: 

  Impact of Seeing a Physician of Similar Ethnicity No significant differences ANSWER Q4=5 None of the above

Slide21: 

Q5. Which Ethnic/Racial Group Had the Lowest Rate of Obesity (based on self-report height and weight from the MSS Survey)? Caucasian Chinese Filipino Hawaiian Japanese Portuguese

Slide22: 

ANSWER: Q5= 2 Chinese Percent Obese

Slide23: 

Data Strengths Large number of enrollees (n=677,000) Three types of coverage (HMO, PPO, Medicare cost contract) Clinical information includes diagnoses, procedures, medications Utilization measures include hospitalizations, ED visits, costs Information on patient and physician characteristics Data extract system near completion

Slide24: 

Data Limitations Not good for estimating prevalence of disease (excludes uninsured, most QUEST, most of Medicare, military, etc.) Limited clinical detail Limited hospital data (not know medications) No information on smoking, exercise, etc No information on uncovered services, pharmaceuticals (e.g. aspirin use)

Study Proposal: 

Study Proposal Background (1 paragraph) Objective(s) Study population (and how to identify them) Time period Variables Methodology Policy implications HIPAA considerations Benefit to HMSA members