Which women should get a mammogram for breast cancer screening?

What a great question!  This morning, Ted Ganiats, MD gave a great lecture on cancer screening at the KU Winter Symposium.  He compared the US cancer screening recommendations with their European counterpart reccomendations.

A few quick highlights:

• Europeans screen for colon cancer exclusively with FOBT and it has the same cost efficacy slope as colonoscopy
• 1 death is prevented from 13,000 mammograms in women over 50
• Women from 40-50 probably shouldn’t get a screening mammogram
• Europeans don’t even screen for prostate cancer because the evidence is lacking

The best evidence based guidelines have these qualities:

1. Level of evidence
2. Benefit
3. Harm
4. Cost
5. Value Judgments
6. Patient preference

Predicting a heart attack

Short of preforming a heart cath on every person, how do we assess risk of a heart attack?  For many years the best predictor has been a table generated from epidemological data from Framingham, Mass.

Last year the JUPITER trial suggested that asymptomatic, low risk persons undergo blood testing for high sensitivity C-reactive protein.

Now an Annals of Internal Medicine article from the US preventive task force rates nontraditional methods of screening for heart disease.

The results are both surprising and intuitive because the best predictive test turns out to be the one we use the most already: LDL.  They studied 9 tests: coronary artery calcium score as measured by electron-beam computed tomography, lipoprotein(a) level, homocysteine level, leukocyte count, fasting blood glucose, periodontal disease, ankle–brachial index, and carotid intima–media thickness.  There wasn’t enough evidence to make any of the tests significant in changing patient oriented outcomes.
I believe that LDL will continue to be the best predictor and that the new goals will try to push it as low as possible.  Does that mean that I should take a statin too? At age 28, male, and with an LDL of 83, I’m considering it.

My head is killing me…

Headache is a common presenting symptom to the primary care office.  When accompanied by nausea or photophobia the headache is classified as a migraine.  For acute migrane abortion, I have used any antinausea medication — usually promethazine.  This is a habit I will soon break.  Last month Journal of Emergency Medicine reported a randomized control trial of prochlorperazine vs promethazine.  The results are convincing — prochlorperazine just blows promethazine out of the water (69% to 39%). It also had less sedating (H1) effect and then next day people felt less groggy.  Now the knock in prochlorperazine in the past was akathisia.  However if you remember your physiology, when your patient shows up with a bit of a parkinsonian-like tremor — just add a little (H1 blocker) like Benadryl.  It can even be oral.  But don’t take my word for it…look at the analysis yourself.

Circumcision, Circumcision, Circumcision

Score one for the descendants of Abraham, a new study of men in Africa concluded that circumcision reduced the HIV rate by 50-60% in heterosexual males.  Importantly it didn’t reduce the male to female transmission rate, but in addition to condoms or abstinence – circumcision has proven to be the only method to slow HIV infection rates.

Researchers hypothesize the HIV virus lives in the foreskin mucosa.  UNAIDS and the WHO have set up a joint website promoting the benefits www.malecircumcision.org.  Other effects include reduced transmission rates of syphillis, chancroid, HPV, and herpies simplex.

Resident work hours.

December 2 was the fifth anniverery of the 80-hour work week implemented by the ACGME in 2003.  Now the Institute of Medicine released the update on the findings.  They recommend further work restrictions, however they come with a cost of 1.7 billion dollars.  They say that residents are still too fatigued. 

I’m not sure about these recommendations.  Of course they would need to be implimented by the ACGME to have any teeth, but I don’t think residencies will survive  with further restrictions.  One change might come to the residency I am in — night float.  We currently have a call system, but I am in favor of the night float system because I think it will reduce the 30 hours calls.

Abortion numbers

The CDC keeps track of many statistics, including the number of abortions preformed each year.  The article highlights the most recent data (from 2005).  Personally I hope that we can find ways to reduce the number of abortions, but this is a complex social problem requiring caring people, resources and information to the mother to allow her to make the best choice for her and family.

Being evidence based requires information

My favoriate tool is the internet. Within seconds, via Google or Wikipedia I can find the answer to many of my medical questions.  If I cannot locate the answer directly, my second resort is UpToDate or Pepid.

However the Iphone blends all these tools into the palm of my hand.  Today I’m stoked because I learned how to sync my calender with Google calendar with NuevaSync.  The instructions to this fantastic tool are here.   I can type in an event on my Iphone and it magically syncs with the calendar on my computer.

I have also used the PubMed app on the Iphone, it is a little clunky but if I really need an article, I will probably just use my laptop.

NEJM: JUPITER trial…or how the drug companies pulled the wool over our eyes

Earlier this month, the NEJM published the results of the JUPITER trial which took patients with normal LDL, but elevated levels hs-CRP and randomized them to placebo or rosuvastatin.  The authors state the trial of 17,000 patients was stopped early, at year because of five, because rosuvastatin reduced the number of major cardiovascular events.

I have to thank Colin P Kopes-Kerr for pointing out the the lead author in the study also holds the patent for hs-CRP test and thus would have another dog in the fight each time a hs-CRP test is ordered.  Don’t you doubt for a minute that the pharmacutal company, who sponsored the study and had their minions monitoring the 1316 sites of this multi-center trial, forgot about the results either.

Rosuvastatin

Rosuvastatin is more than $100 a month.  The results stated that the NNT is 95.  So you would need to give 95 patients rosuvastatin for 2 years to prevent one MI, stroke, hospitalization for UA or death.  That is just $228,000.  Chump change.

Maybe we should just put statins in the water.  However, then we wouldn’t be practicing medicine, just providing lining to the pharmaceutical companies without actually informing the patient of more effective less costly interventions like diet, exercise and smoking cessation (which have alway been unpopular anyway).  At the least when the pharmacutical companies through several million dollars at a study you get some pretty graphs.  I advise reading the article yourself to come up with your own conclusions.

Pretty graphs

Partner treatment in HIV and STDs

Chlamydia

Secondary syphilis

New recommendations hot of the press at the CDC emphasize the importance of partner treatment.  Usually the most difficult task in this portion is identifying all the partners involved.  The recommendation is to test for all other STIs if one is positive.  So, in a patient who test positive for chlaymdia, she and her partner should be tested for HIV, syphillis, and gonorrhea as well. Intrestingly enough, the states hold legal authority for identifying and notifying partners of infected patients.  The situation can become sticky rather quickly, as it is the duty of the medical provider or the state to warn partners of possible exposure.

Understanding the evidence

The Annals of Internal Medicine tackle this tough task in their lead article.  I recommend listening to the pod cast summary of the article.  But here is the summary: make sure you are evaluating POEs (patient oriented outcomes) rather than DOEs (disease oriented outcomes).   It doesn’t matter if we can lower LDL but it may matter greatly if we can reduce death. What is the risk stratification? Well, you can find out now.  Listen here.