Sunday, April 25, 2004

The Expert PCP? 

Been thinking about the recent concept of the Expert Patient.

Just as Expert Patients are a threat to PCP's I suppose Expert PCP's are a threat to Non-expert specialists.

Friday, April 16, 2004

Virtual Colonoscopy 

Just when you thought it was safer to go inside. Recent Yahoo! News - Virtual Colonoscopy Method Questioned headlines point out the problems between the ideal and the real world.

I was troubled by this comment because it reinforces an ongoing misunderstanding:
The American Cancer Society recommends that people over 50 get a conventional colonoscopy every 10 years, calling it the "gold standard" of colon cancer screening. It said that there is not enough evidence to recommend the virtual test for those with average colon cancer risk.
The American Cancer Society actually endorses multiple screening choices ACS :: Can Colorectal Cancer be Found Early? and does not favor one over another. Besides, how can it be a "gold standard" in the absence of any outcomes data?

This JAMA -- Abstracts: Cotton et al. 291 (14): 1713 real world work hopefully places this intervention back on the drawing board. It will be interesting to see if we learn a lesson from this with future technologies.

Thursday, April 15, 2004

Considering Medicine Wasn't My Original Career Choice... 

Came across this from The Helix: October 2003 Archives.

My highest score: thoracic surgery.
My lowest score: family practice.

I wonder what this means since I've been happy doing family practice the last 15 years?

Sunday, April 11, 2004

PROVE-IT (Again) 

I have three reasons to revisit this issue.
1. My post had two errors which I corrected. One is the ARR for the primary endpoint is 3.9%, not 1.9%. The other was regarding figure 3 when I used a 60 instead of a 90 day interval. It is now updated.
2. A nurse and cardiologist recently both disappointed me with their understanding of what the research showed.
3. Medscape sends out an e-mail about the most read stories. This is what it said [my comments in italics]:
Aggressive Lipid-Lowering Decreases All-Cause Mortality

March 8, 2004 (New Orleans) — Compared with moderate [current goal] lipid-lowering therapy, aggressive lipid lowering therapy — reducing low-density lipoprotein (LDL) cholesterol to well below 100 mg/dL — is associated with a 3.9% [ARR] reduction in a composite end point of death from any cause as well as nonfatal myocardial infarction, unstable angina requiring hospitalization, revascularization or stroke, according to results reported here at the 53rd annual scientific session of the American College of Cardiology. [This is misleading and implies the death endpoint is separate.]

Aggressive therapy was also associated with a 16% reduction in the relative event rate compared with moderate treatment.

Christopher Cannon, MD, principal investigator of the Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 (PROVE-IT TIMI 22), told Medscape that the study results "finally answer the question: it is all about lower numbers. The lower the LDL, the better the protection." Moreover, he said that the benefit is "not about one drug versus another drug. It is about aggressive treatment versus standard treatment."

Steven Nissen, MD, from the Cleveland Clinic Foundation in Ohio, said the study results were compelling, but he told Medscape, "I disagree with the PROVE-IT investigators. I think this is a drug effect." Dr. Nissen was principal investigator of the Reversing Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial, which used the "exact same drug regimen." That study, which was reported in November 2003, found that 80 mg atorvastatin stopped progression of plaque as measured by intravascular ultrasound compared with pravastatin treatment. [A DOE outcome.]

Dr. Nissen, who was not involved in the PROVE-IT study, said "the findings are even more surprising when you consider that this study was carefully designed to show no difference between these two drugs."

The PROVE-IT study, which was presented here and simultaneously released online by the New England Journal of Medicine, enrolled 4,162 patients who were hospitalized for acute coronary syndromes within 30 days of study entry. [The actual method section says "...hospitalized for an acute coronary syndrome...in the preceding 10 days." This is secondary prevention in a high risk group.] Patients were randomized to 40 mg pravastatin (standard therapy) or 80 mg atorvastatin (intensive therapy). After treatment, median the LDL cholesterol level in the pravastatin group was 95 mg/dL while the use of intensive lipid-lowering therapy using atorvastatin lowered LDL cholesterol to a median of 62 mg/dL (P < .001).

All-cause mortality was reduced by 28% in the aggressive-treatment arm [Nope, figure 4 "Death from any cause" hazard ratio crosses unity and even the authors say: "The reduction in the rate of death from any cause was of borderline significance (28 percent, P = 0.07)".], while death from [or] MI was reduced by 18% [ARR = 1.7% or NNT = 59 but figure 4 95% CI isn't published, appears to touch unity.]. The only end point for which no benefit was reported was stroke, which "was infrequent, but the rates did not differ significantly between the groups," the authors write in their abstract. [Again, from figure 4, my interpretation is absence of benefit for all cause mortality, death from CHD, death from other causes, MI, death from CHD or MI and stroke. If stroke at 1% is "infrequent" (their words) why isn't death from CHD (1.4% to 1.1% also dismissed?]

At the time of randomization, the median LDL cholesterol level in each group was 106 mg/dL. About 25% of patients (n = 990) were receiving statin therapy prior to randomization and among these patients there was no difference in LDL cholesterol levels with pravastatin therapy, but randomization to atorvastatin "decreased LDL by an additional 32% compared to baseline," said Dr. Cannon.

These results "emphasize the central role of further lipid lowering of (LDL cholesterol) in reducing mortality [?] and morbidity [Mostly from reduction in revascularization and hospitalizations for unstable angina.]," Dr. Cannon told session attendees.

In a prepared statement, Robert H. Eckel, MD, spokesperson for the American Heart Association, said, "This is the first head-to-head trial comparing clinical events with two statins of proven effectiveness in cardiovascular disease prevention. The group with the greater reduction in LDL...cholesterol had the greater benefit.

Dr. Eckel, who is a professor of physiology at the University of Colorado Health Sciences Center in Denver, noted that it is unclear if higher doses of pravastatin would have achieved similar reductions in LDL cholesterol and similar mortality and morbidity benefits. "But this study represents a segment in a continuing line of research that lowering LDL cholesterol below the currently recommended goal of less than 100 mg/dL will be an important way to further reduce these patients' risk of cardiac death [Again with the death benefit that isn't true.]," he added.

The statin benefit was "evident very early, even in the first 30 [According to figure 3, isn't statistically true until 180.] days of therapy, and was consistent [But not statistically significant.] among all subgroups," Eric J. Topol, MD, writes in an accompanying editorial published in the New England Journal of Medicine. Dr. Topol, chairman of cardiology at the Cleveland Clinic Foundation in Ohio, noted that the result was surprising because the trial was designed to "demonstrate noninferiority of pravastatin, as compared with atorvastatin" and because the trial was of fairly short duration — just 24 months follow-up [But in a very high risk group.] — with mixed hard and soft end points [Cardiology literature thrives on combined endpoints.] that researchers believed would make it "impossible to discern differences between the effects of the two statins."

Dr. Topol adds that the results of PROVE-IT [POEM] and REVERSAL [DOE] taken together "herald a shake-up in the field." He concludes that aggressive therapy will result in a "sea change in the prevention and management of atherosclerotic vascular disease."

Another PROVE-IT investigator, Sidney C. Smith Jr., MD, professor of medicine and director of the Center for Cardiovascular Science and Medicine at the University of North Carolina in Chapel Hill, told Medscape that the study clearly [Maybe to him but not to me.] demonstrates the [marginal] benefit of aggressive treatment. But he said it may be difficult to convince clinicians to climb on the aggressive treatment bandwagon. [I'm pretty busy already with interventions that work better than this.]

That may be especially true given recent reports of excess rhabdomyolysis and kidney failure — including the report of fatal rhabdomyolysis in a 39-year-old woman — linked to rosuvastatin, which is considered the most potent statin. [Interesting but irrelevant since was not one of the drugs used in this study.]

Asked if these reports could dampen enthusiasm for aggressive statin treatments, Dr. Cannon said it was unlikely. He noted that the U.S. Food and Drug Administration "carefully reviewed Crestor [rosuvastatin] for more than a year and then studied it for an additional year before approving it — and it was approved in the face of Baycol." Bayer withdrew cerivastatin (Baycol) from the U.S. market on Aug. 8, 2001, because it was associated with excess and sometimes fatal rhabdomyolysis.

Dr. Cannon added that 3% of the patients in the atorvastatin group had abnormal liver function tests compared with 1% in the pravastatin arm. [With P < 0.001 for the stated DOE (LFT elevation) but with non-significant P = 0.23 for the more important POEM (discontinuation for myalgias).]

Asked about the rhabdomyolysis risk, Dr. Nissen [should have quoted the authors: "There were no cases of rhabdomyolysis in either group."] said, "Ask a patient if he or she is more concerned about the 2% [3.3% of either abnormal lab or discontinuation because muscle pain for atorvastatin] risk for an abnormal liver function test versus [an unproven reduction in death after two years] a 28% reduction in mortality. What do you think the answer will be?" [I don't know what the answer would be. No one asked any patients what they would do. I do know the answer would depend on how the question was worded.]
What if the question posed to patients was: "Immediately after a hospitalization for either a heart attack or admission for unstable angina, would you support spending $26,000 and taking a pill every day for two years to have a one in twenty-six chance of reducing your odds of either death, heart attack, hospitalization for unstable angina, revascularization (percutaneous coronary intervention or bypass surgery), or stroke?"

Some would and some would not.

Saturday, April 10, 2004

Fun With History #1 

Today's earlier post about prostate cancer screening made me wonder what clinical practice was like before PSA testing was around.

Consider this JAMA - Early detection of prostate cancer by routine screening abstract from 1984. It says:
In an attempt to detect prostate cancer when the disease was still localized, a free screening clinic was established for men over the age of 45 years. Digital rectal examinations were performed for 811 men. Prostate biopsy was recommended to 43 in whom abnormalities were found by digital rectal examination; only 38 complied. Prostate cancer was detected in 11 men. The patients with cancer ranged in age from 60 to 79 years, with the highest incidence of cancer in the group aged 70 to 79 years. The overall incidence in men between 51 and 80 years old was 1.7%. Staging evaluation revealed that none of the patients with prostate cancer had metastases to the bone or elevated serum acid phosphatase levels. Five men (45%) were found by clinical or pathological methods to have stage B disease. Two others (18%) showed radiographic evidence of lymph node metastases (stage D1). The cost of detecting each cancerous prostate tumor was approximately +6,300. Routine screening can be a cost-effective method for diagnosing prostate cancer in patients with less extensive disease. The ability of early detection to prolong survival of patients with this disease will require further investigation.
The chance for a male age > 45 of flunking the rectal exam was 43/811 = 5.3%.

How about this from the Journal of Urology in 1989?
An early detection study for prostate cancer was initiated to determine the effect of routine digital rectal examinations on the stage of prostate cancer at diagnosis. A prostate biopsy was recommended if induration, asymmetry or nodules were detected on the digital examination. During a 6-year period 4,160 examinations were performed on 2,131 men more than 45 years old. A prostate biopsy was performed on 144 men and 36 malignant tumors were detected, of which 68 per cent were clinically localized. Pelvic lymph node metastases were found in 6 per cent of the surgically staged cancer patients and in 10 per cent of the patients who had a high grade tumor. Surgical staging revealed that 50 per cent of the patients with clinical stage B disease were upstaged to stage C or D1 disease. These results suggest that mass screening programs using digital examination may not add sufficient benefit over conventional medical care to warrant the expense. Definitive proof that screening can lower the mortality rate from prostate cancer can be obtained only by a prospective randomized clinical trial.
This time and again for men age > 45, the chance of a rectal exam resulting in a biopsy was, 144/4160 = 3.5%.

Amazingly, it wasn't until 1998 that anybody bothered to do an RCT about the digital rectal exam (DRE). It showed [my italics]:
The utility of digital rectal examination (DRE) as a screening test for early detection of prostate cancer has not been established. Therefore, we evaluated the usefulness of DRE as a stand-alone screening test and in conjunction with measured serum prostate-specific antigen (PSA) levels of 0-3.9 ng/mL and transrectal ultrasonography (TRUS). METHODS: Our study population consisted of 10,523 men aged 54-76 years who were randomly assigned to the screening arm of the Rotterdam, The Netherlands, section of the European Randomized Study of Screening for Prostate Cancer. The underlying prevalence of detectable prostate cancer was estimated by logistic regression analysis and used for calculating the sensitivity of DRE as a test. Pathologic characteristics of 105 radical prostatectomy specimens were used to determine the aggressiveness of the tumors diagnosed (and missed) by DRE. RESULTS: The overall detection rate for prostate cancer in this population when serum PSA measurement, DRE, and TRUS were used was 4.5%, and the detection rate with DRE alone was 2.5%. The positive predictive value of DRE ranged from 4% to 11% in men with PSA levels of 0-2.9 ng/mL and from 33% to 83% in men with PSA levels of 3.0-9.9 ng/mL or more. Most tumors detected by DRE in men with PSA levels of less than 4.0 ng/mL were small (mean volumes = 0.24-0.83 mL), and most were well differentiated (Gleason scores of 6 or less). Minimal, moderate, and advanced cancers were seen in 42%, 42%, and 16% of men, respectively, with a PSA level of 4.0 ng/mL or less. DRE alone allowed detection of 264 (55.8%) of 473 cancers; 82 (17.3%) of the 473 cancers would have remained undetected by PSA-based screening alone (i.e., no follow-up procedures for PSA values of 0-3.9 ng/mL). CONCLUSIONS: For PSA values of 0-3.9 ng/mL, the positive predictive value and sensitivity of DRE, tumor volume, and tumor grade were strongly dependent on PSA level. DRE has a poor performance in low PSA ranges.
What about safety? So far, we're only seeing the possible benefits. Trends and risk factors for prostate biopsy complications in the pre-PSA and PSA eras, 1980 to 1997 from Urology 2002 says [my italics]:
OBJECTIVES: To assess the secular trends in postbiopsy complications and to identify the risk factors for complications. METHODS: Olmsted County residents who underwent a prostate biopsy between 1980 and 1997 were identified. All community medical records for the study subjects were reviewed to identify prostate biopsy-related complications, including gross hematuria, infection, pain, hematospermia, and acute urinary retention. RESULTS: Of the 2258 prostate biopsies, 377 (17%) were associated with at least one complication. The total complication rate per biopsy remained relatively consistent at about 17% from 1980 to 1986, 1987 to 1992, and 1993 to 1997 (P for trend = 0.8). The age-adjusted complication rate (per 100,000 men) increased from 26 to 60 in 1980 to 1986 and 1993 to 1997, respectively (P <0.001). This paralleled the increase in prostate biopsy use from 138 to 374 per 100,000 men in the same periods. The prevalence (per biopsy) of gross hematuria increased, 7.5% to 12.8% (P = 0.04); postbiopsy infection declined, 4.6% to 1.4% (P = 0.001); and hospitalization for infection declined, 1.2% to 0.2% (P = 0.06) between 1980 to 1986 and 1993 to 1997. A urogenital infection 6 weeks before biopsy was associated with an increased risk of a postbiopsy complication (odds ratio = 1.7, 95% confidence interval = 1.0 to 2.8) and an increased risk of a postbiopsy infection (odds ratio = 5.5, 95% confidence interval = 2.2 to 13.8). CONCLUSIONS: Although the complications per biopsy have stayed constant, the prevalence of postbiopsy complications in the community has increased tremendously because of the increased use of prostate biopsies. Specific strategies may be needed to reduce the incidence of postbiopsy infection in men with a recent urogenital infection before biopsy.
I'm not comfortable with "trend" concept and I don't agree that the hospitalization rate declined. Post biopsy hematuria is inconvenient but a hospitalization is important.
An alternative conclusion is: From 1980 to 1997, and in the absence of screening benefit on mortality, the biopsy rate went up without a decline in serious complications.

A quick Google finds Prostate Cancer -- Prostate Needle Biopsy a current site with stated risk:
Risks and complications

It is normal to expect some minor bleeding after needle biopsy, because the needle has entered areas that contain small veins. Blood in the urine, semen and with bowel movements may occur intermittently for a few days and possibly for a few weeks. The two primary risks of needle biopsy are severe bleeding and infection of the prostate gland or urinary tract. These risks are very rare, occurring in less than 1 percent of patients.
This seems like a far cry from the published risks and does not even mention the small but real chance of a hospitalization. I find another site Health Library - Prostate Biopsy equally misleading [my italics].
A prostate gland biopsy has a slight risk of causing problems such as: Infection. This is more common in men who have undiagnosed prostatitis. Usually taking antibiotics before the biopsy prevents an infection from developing. Bleeding into the urethra or bladder. This can cause a blood blister (hematoma), an inability to urinate, or a need to urinate often. Bleeding from the rectum. If you have had a transrectal biopsy, you may experience a small amount of bleeding from your rectum for 2 to 3 days after the test. However, contact your doctor if the bleeding persists. An allergic reaction to the medications (such as the anesthetic) used during the biopsy. A prostate gland biopsy does not cause problems with erections and will not make a man infertile.
I think patients deserve a number = risk, not language like "slight". The research shows the risk for post biopsy infection goes up with infection within the past 6 weeks. Taking antibiotics immediately before the procedure does not necessarily protect against this. Again, no mention is made of a potential hospitalization rate around 1%. Finally, while the risk profile is incomplete, they do make the effort to tell the patient what won't happen.

Fool's Gold Standard 

Am I the only one thinking this is irrational? They say:
In men with high PSA levels and negative biopsy results, a test for a protein may reveal undetected cancer, or confirm that the patient does not have the disease.
A negative biopsy in an individual with increased serum PSA "raises an important clinical dilemma, namely what to do with the patient at this time," investigators explain. Typically, these patients undergo repeat biopsies depending on subsequent PSA readings.
Here is the Abstract: Volume 171(4) April 2004 p 1419-1423 EARLY IDENTIFICATION OF INDIVIDUALS WITH PROSTATE CANCER IN NEGATIVE BIOPSIES from the Journal of Urology.

Let's assume a patient chooses PSA screening after a quality conversation using neutral language takes place. The PSA is elevated but the biopsy is negative.

These researchers are trying to invent another layer of testing? If a biopsy isn't a gold standard then what is?

There seems to be no limit to the early detection strategy.
I hope prostate cancer screening works. At least there are two studies (European Study of Screening for Prostate Cancer and Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial) in the pipeline to resolve the issue. Until then, why work on this stuff?

We already have: free PSA, density of PSA, age adjusted PSA, and velocity of PSA to improve the sensitivity and specificity. What if all the resources behind these technologies had instead been directed at meeting basic screening effectiveness criteria before the test was ever made available?

Tuesday, April 06, 2004


Nice to see The Onion also picked up on the same theme as yesterday's post from announcements by the National Sleep Foundation.

Sunday, April 04, 2004

Experts Agree #1 

There is no end to the "raise awareness" phenomenon by national organizations. The science section of today's paper has an article, Poll Finds Even Babies Don't Get Enough Rest, from the New York Times [registration required] that says:
"Infants average almost 90 minutes less sleep a day than the 14-hour minimum doctors recommend...".
I'm wary at this point because I'm skeptical that the right amount of sleep is even known. Things get worse:

"We've suspected that they are not getting enough sleep, but this is the first time we're really showing it," said Dr. Jodi A. Mindell, a professor of psychology at St. Joseph's University in Philadelphia and a member of the sleep foundation's board.
Dr. Mindell and other sleep experts say that young children who do not sleep enough do not function as well as their better-rested classmates at school and that their relationships with family and friends suffer. Infants and toddlers also need enough sleep to remain alert and open to the world around them, the experts say.

Dr. Lewis Kass, a pediatrician and the director of the sleep disorders center at the Children's Hospital at Montefiore in the Bronx, said the findings were significant because they provided firm data for a phenomenon familiar to pediatricians.

In many families today, he said, even small children take part in so many activities that sleep patterns are inevitably disrupted. "This puts in terms that people can grasp how big a problem this is," said Dr. Kass, who was not involved in the survey.

Pediatricians do not pay enough attention to the issue, according to the poll, with 52 percent of parents reporting that their children's doctors do not ask about sleep habits.

Dr. Kass and Dr. Mindell expressed surprise at the poll's finding that 69 percent of young children experienced sleep-related problems a few times a week. The problems included trouble falling asleep, heavy snoring, waking up at night, nightmares and restless legs syndrome, which involves unpleasant sensations in the legs, like itching and tingling.
At least the article points out:
Experts say they sense that small children get less sleep and have more sleeping problems than they did in the past, though they acknowledge they have little evidence.

My initial reaction was another April Fool's joke? For example, is it a coincidence that National Sleep Awareness Week (NSAW) is March 29 - April 4? And how did they poll all those babies anyway? I'm sure it is not a joke that the main sponsors of NSAW are: King Pharmaceuticals (Sonata), Sanofi-Synthelabo (Ambien), and Sepracor (Xopenex and Estorra, which is in phase III for insomnia). The sponsor page even says (my italics):
The National Sleep Foundation works closely with our sponsors to develop a customized program that achieves mutually beneficial goals. Sponsorship benefits include:
Year-round use of the National Sleep Awareness Week sponsor logo.
Targeted promotions, sampling and display at community events during National Sleep Awareness Week.
Recognition in nationally distributed National Sleep Awareness Week collateral materials, NSAW and NSF Web pages, NSF's annual report and other venues, as appropriate..

Here is the study [183 page pdf file]. Questions were asked of parents and caregivers with children <= 10 years of age (n = 1473). However, the infant (0- 11 months) n = 210 so I'm aready struggling with validity issues. The poll measures lots of things but what is the gold standard for the proper amount of sleep?

The American Academy of Pediatrics web site did not help when I tried to verify the "14-hour minimum doctors recommend" mentioned above. The American Academy of Family Physicians site dose have an article with a graphic that implies where the "14 hour minimum" comes from. As near as I can tell, Richard Ferber, M.D. bases ideal sleep on work as the director of Boston's Center for Pediatric Sleep Disorders at Children's Hospital from the 1980's.

Entrez PubMed didn't help when I tried to verify the experts credentials on the basis of research.

The issue seems long on opinion and short on outcomes to me.

One last thing. If "69 percent of young children experienced sleep-related problems a few times a week" then isn't that another way of saying normal?

Saturday, April 03, 2004

Mountain Sickness 

Ever since attending a Wilderness Medicine conference I've been wondering about my choices for altitude sickness prophylaxis.

The speaker was enthusiastic about ginkgo biloba as an alternative to acetazolamide.

Today I found a prospective, double blind, randomized, placebo controlled trial with intent to treat analysis: bmj.com Gertsch et al. 328 (7443): 797

The results are:
1. Ginkgo (120 mg twice daily) was not significantly different from placebo for any outcome.
2. The incidence of acute mountain sickness was 12% for acetazolamide (250 mg twice daily) and 34% for placebo (odds ratio 3.76, 95% confidence interval 1.91 to 7.39, number needed to treat (NNT) = 5.
3. Acetazolamide headache prevention NNT = 3.

Thursday, April 01, 2004

Beyond Tamoxifen? 

One of my breast cancer patients called about the recent NEJM article about Exemestane "better" than Tamoxifen . The information made me wonder about the absolute risk reduction of Tamoxifen therapy. This Lancet abstract is a place to start. Only considering estrogen receptor positive women and 5 year's of therapy:

1. The 10 year survival absolute risk reduction (ARR) for node positive patients = 10.9% (from 61.4% to 50.5%). So, number needed to treat (NNT) = 9.
2. The 10 year survival ARR for node negative patients = 5.6% (from 78.9% to 73.3%). Their NNT = 18.
3. The incidence of endometrial cancer was higher but the number was small and not statistically significant. Hmmm, I was under the impression the endometrial cancer risk was real.

Back to Exemestane (Aromasin) versus Tamoxifen (Nolvadex). The study seems valid (double-blind, randomized, intent to treat) with some limitations (primary endpoint is "disease-free survival instead of mortality and follow up is only 2-3 years). I'm impressed the study was designed to measure safety (quality of life, uterine thickness, bone metabolism, and bone mineral density) but disappointed again over the DOE > POEM outcomes. Also, the safety data is going to be reported later and separately.

The benefit of three years (mean 30.6 months) of Aromasin after 2 years of Tamoxifen instead of 5 years of Tamoxifen is:
1. Absolute risk reduction (ARR) 4.7% so number needed to treat (NNT) is 21 to prevent recurrence of breast cancer, diagnosis of secondary breast cancer or death.
2. There was no overall survival benefit.

The limited safety data include:
1. ARR 0.8% for thromboembolic events (NNH for Tamoxifen = 125).
2. No change between the two drugs for fractures.

I called a local pharmacy. Tamoxifen is about $200 for #100 or $730/year. Aromasin is $800 for #100 or $2920/year.
1. $2190 x 3 years x NNT (21) = $137,970 to prevent a non-mortality endpoint.
2. $2190 x 3 years x NNH (125) = $821,250 to prevent a thromboembolic event.

This page is powered by Blogger. Isn't yours?