Clinical Research Oriented Workshop (CROW) Meeting: June 7, 2012

Present: Amanda Kennedy, Ben Littenberg, Connie van Eeghen

1.                  Start Up:  Integrated pharmacy project engaging pharmacists in outpatient care is going well – good feedback from many perspectives.

2.                  Presentation: Connie -  using NHANES III to analyze mortality outcomes based on reported use of prescribed medications took us on a long discussion about the long term objective of the researcher (Connie, in this case) and the tools that can move that objective forward.  Retained learnings (not to be confused with “retained earnings,” which is a different objective altogether):

a.       Long term objective: identify the determinants of successful change and adaptation in health care systems in order to facilitate uptake of new practices of care and management.

b.      This population-based data base does not support analysis of medical practice or health system operations, but it can be used to link population behavior that is influenced with both of these to mortality outcomes.

                                                  i.      Mortality outcomes include the date of death and, to some degree, cause of death

c.       The study will give the researcher practice with these data and analysis for a cross-sectional study that has mortality outcomes added, using a probabilistic match between NHANES III and NDI death certificate records for subjects who died up to 18 years after the survey.  Ben has linked NHANES data from the Household Adult, Examination, and Leptin data sets to the mortality data.  This study will be limited to the use of these data sets.

d.      Examples of the survey data that reflect medical practice/health systems are:

                                                  i.      Access to health provider

                                                ii.      Use of vitamins, minerals, and medicine: do they matter?

                                              iii.      Use of aspirin (which was asked in the context of pain relief): what is their role in the prevention of heart attack (identified by cause of death)?

                                              iv.      Validation of the Framingham risk score (which included cardiovascular mortality).

                                                v.      Public health topics:

1.      Oral health (for which there is extensive data in the survey)

2.      Obesity

3.      Tobacco use

4.      Hypertension…

e.       We reconvened in Ben’s office to test the data base for an indication that there is something of interest to study.  We wound up looking at aspirin use with the following results:

                                                  i.      33% of adult respondents (6,518) report having taken aspirin in the past month (this includes Anacin, Bufferin, Ecotrin, Ascriptin, and Midol) (starting pg. 27 in STAT log)  Note: the context of this question is under the heading of “non-prescription pain relief medications”

                                                ii.      The hazard ratio for ASA is 1.33 with a P value<0.001 and a CI of 1.26-1.41.  This means that those who took aspirin were 33% more likely to have died in the 18 years after participating in the survey. 

                                              iii.      When adjusted for age, HR = 0.96, P=0.186, CI .911-1.02  We know that people are more likely to take aspirin when they get older, and are also more likely to die when they get older.  Adjusting for age, taking aspirin is associated with a 4% lower likelihood of dying.

                                              iv.      When adjusted for age & sex, HR = .94, P=0.040, CI .893-.997.  In the 1980’s, men were more likely to be advised to take aspirin than women were, as a preventive against heart attack.  Adjusting for age and sex, taking aspirin is associated with a 6 % percent lower likelihood of dying.  

                                                v.      When adjusted for above plus heart disease, HR = .93, P=0.006, CI .876-.979.  Respondents who have been told they have heart disease are more likely to be advised to take aspirin.  Adjusting for all three characteristics, taking aspirin is associated with a 7% lower risk of dying.  Out of the 6,518 people who took aspirin, one could hypothesize that 456 did not die that otherwise would have.  And, the confidence interval range is below 1.0, indicating that if we collected data from many other, equivalent surveys, 95% of them would also indicate a relationship between aspirin and avoiding death when adjusted for age, sex, and heart disease. 

                                              vi.      For respondents with no heart disease and adj for age & sex, HR = .91, P=0.015, CI .837-.981.  Oddly, those patients who did not think they had heart disease (14,167 respondents) and took aspirin anyway, adjusting for age & sex, were even less likely to die (by 9%).  (Connie hypothesized that a lot of them really did have heart disease, and just didn’t know it, but the CDC currently reports the prevalence of heart disease (age adjusted) is around 6% of adults, so this is unlikely.  On the other hand, Hunink reported in a JAMA 1997 article that mortality from CHD from 1980 – 1990 dropped 34% - so maybe the prevalence back then was more like 9%.)

                                            vii.      For respondents with heart disease and adj for age & sex, HR = .95, P=0.216, CI .881-1.020.  So if you did think you had heart disease (5,703) and took aspirin, you were less likely to die by 5%.  So aspirin helped the apparently healthy more than the ill, with respect to heart disease. 

f.       Next steps:

                                                  i.      Connie will do a literature search (PubMed, OVID) on NHANES III (and will first check to see if this is a MeSH term), aspirin, and mortality.  She will review the results and summarize them for a future meeting.

                                                ii.      Connie, Amanda, and Ben will develop a research question to form the basis of a study design and bring to a future CROW meeting. 

3.                  Workshop Goals for 2012:

a.       Journal club: identify UVM guests and articles; invite to CROW ahead of time

b.      Research updates: share work-in-process

4.                  Next Workshop Meeting(s): Thursday, 11:00 p.m. – 12:00 p.m., at Given Courtyard Level 4. 

a.       June 14: Abby - research update (no Connie)

b.      June 21: Kairn - research update (no Connie)

c.       June 28:  Rodger and Connie - R03 feedback from reviewers

d.      July 5: ???

e.       July 12: (no Connie, Abby)

f.       July 19: (no Abby)

g.      July 26:

h.      Aug 2:

i.        Aug 9: (no Abby)

j.        Aug 16: (no Abby)

k.      Aug 23:

l.        Aug 30: (new schedule?)

m.    Future agenda to consider:

                                                  i.      Ben: budgeting exercise for grant applications

                                                ii.      Ben: Writer’s workshop on the effect of the built environment on BMI (Littenberg & Austin Troy)

                                              iii.      Journal Club: “Methods and metrics challenges of delivery-system research,” Alexander and Hearld, March 2012 (for later in the year?)

                                              iv.      Rodger: Mixed methods article; article on Behavior’s Influence on Medical Conditions (unpublished); drug company funding.  Also: discuss design for PCBH clinical and cost research.

                                                v.      Amanda: presentation and interpretation of data in articles

                                              vi.      Sharon Henry: article by Cleland, Thoracic Spine Manipulation, Physical Therapy 2007

Recorder: Connie van Eeghen