Research News and Notes

1. Emergency duty and death from heart disease 

Many neurosurgeons, as part of their practice, take emergency calls. Sometimes, they cover more than 1 hospital and have to respond quickly to calls at all times of day and night. Often, a call means that a severely ill or injured patient requires immediate attention and intervention by the neurosurgeon, with a great responsibility placed on the neurosurgeon to perform complex judgment and proper intervention to save the patient's life and function. Reasonably, the stress of taking an emergency call is considerable, yet I am not familiar with any study on the effect of taking calls on the health of physicians, nor could I find any when searching Medline.

In the March 22 issue of the New England Journal of Medicine, Kales et al [2] reported a large study on duty-specific risks of death from coronary heart disease among firefighters. Similarly to surgeons on call, firefighters need to respond to emergency calls and deal with life-and-death matters. Arguably, they tend to face greater risk to their own life. The authors reviewed Federal Emergency Management Agency records of all the deaths of firefighters while on duty for the period 1994 to 2004, excluding the terrorist attacks on September 11, 2001. They obtained from a national database and from multiple large fire department estimates the proportion of time spent by firefighters on each of their activities. Most deaths from coronary heart disease were associated with suppressing fire (32.1%), although this activity took a very small portion of the firefighters' time. Interestingly, many died while responding to an alarm or returning from an alarm rather than when actually fighting the fire. Compared with mortality during performance of fire station duty, the odds of dying of heart disease while suppressing a fire were 12.1 to 136 times higher, 2.8 to 14.1 times as high during alarm response, and 2.2 to 10.5 times as high during alarm return. Interestingly, the death rate while responding to nonfire emergencies, including medical emergencies, was low. The authors did not comment on this disparity, but I assume that fire suppression was a more severe stress test, so that those firefighters who had heart disease tended to die while performing the most stressful activity. Moreover, the responsibility and involvement of a firefighter in a medical emergency tends to be limited.

It is impossible to extrapolate from this study to any other profession, but anecdotal stories that I hear indicate that an emergency call is a significant source of stress for many neurosurgeons and other physicians. A study on the health effect of this stress on physicians is long overdue.

2. Exclusion criteria in randomized trials 

In the era of “evidence-based medicine,” randomized controlled trials (RCTs) are thought to provide the best clinical evidence. Treatments for patients are routinely based on the results of RCTs. A careful look at many of them often reveals that a large number of patients in multiple institutions were screened, but only some of them were included in the study. A paradox may result, wherein the results of an RCT are applied by physicians to patients who would have been excluded from the original study.

In the March 21 issue of the Journal of the American Medical Association, Van Spall et al [4] reviewed the eligibility criteria of RCTs that were published in high-impact general medical journals. They found what many had observed for a long time—the RCTs do not always clearly report exclusion criteria. Trials involving multiple centers and those involving drug interventions are most likely to have extensive exclusions. The authors attempted to separate justifiable exclusions, such as unacceptable risk of known adverse reaction to the intervention in the excluded patient, and poorly justified exclusions—age, sex, chronic health condition, and others. The authors were mainly concerned that inappropriate exclusion criteria would impair the generalizability of the data and any conclusions from an RCT being applicable to the general population.

I am very happy to see a detailed analysis that addresses a major concern and cause of unease I always had about RCTs. If enough patients are excluded, the study may be “cleaner,” but it would no longer represent the universe of the patients who are candidates for the tested treatments. Clear reporting of exclusion criteria would be a major step toward greater transparency in RCTs.

3. Pharmaceutical company payments to physicians 

Five states and the District of Columbia passed laws that mandate disclosure of payments made by pharmaceutical companies to physicians. A recent article in the Journal of the American Medical Association attempted to access the data to learn about the extent of the phenomenon [3]. Only in Vermont and Minnesota were the data publicly available. In Vermont, 61% of payments were not released to the public because pharmaceutical companies designated them as trade secrets, and 75% of publicly disclosed payments were missing information necessary to identify the recipient. In Minnesota, only 25% of the companies reported in each of the 3 years of the study (2002-2004). In Vermont, there were 2416 payments of $100 or more to physicians; total, $1.01 million; median payment, $177 (range, $100-$20000). In Minnesota, among 6946 payments totaling $30.96 million publicly disclosed, there were 6238 payments of $100 or more to physicians; total, $22.39 million; median payment, $1000 (range, $100-$922239). In Minnesota, the authors could perform further physician-specific analysis. They identified 2388 distinct physicians who received payment of $100 or more. Some received multiple payments (up to 88). The median amount received was $1000 (range, $100-$1178203). Although reporting was very partial, the authors concluded that a large number of physicians received money from pharmaceutical companies, and a few received large sums of money. The actual effect of the gifts and payments on physicians' choices has not been studied, but even small gifts create the perception of obligation and diminish the stature of physicians in the public's eyes. Our professional societies have to take action before legislatures do so. Our strength is in science. Our professional societies can commission a study on actual physician behavior and the effect of gifts on the choice of drugs or equipment.

4. Hydrodynamics of normal-pressure hydrocephalus 

The aging of the population means an increased rate of dementia and increased referrals to neurosurgeons with the diagnosis of normal-pressure hydrocephalus (NPH). Recently, Bateman and Loiselle [1] from Australia used magnetic resonance measurement of intracranial hydrodynamics to distinguish between responders and nonresponders to shunting. They measured a variety of parameters—aqueductal flow, ventricular enlargement, sagittal sinus outflow, cerebral compliance ratio, and others. The study included 32 patients diagnosed with NPH, 12 age-matched controls, and 12 patients with Alzheimer disease (AD). The tests demonstrated a difference in most parameters between patients with NPH and controls, with patients with AD falling between these groups. This is already a discouraging result, indicating that patients with AD who were not diagnosed with NPH may have impaired cerebral hydrodynamics. The worst blow was the completely identical cerebral hydrodynamics between patients with NPH who responded and those who did not respond to shunt insertion. These results, albeit negative, are important. The inability to predict response to shunt insertion among patients with NPH raises a question about the validity of current methods of diagnosing this condition.