ARTICLE IN BRIEF
In a retrospective analysis of the density of neurosurgeons/neurologists per population as compared with the number of deaths from stroke over a three-year period, investigators found a significant reduction in mortality associated with an increase in density of neurology providers.
Is there an ideal ratio of expert care providers per population that would improve overall health outcomes? Does having more neurologists per capita decrease deaths related to stroke?
A Nov. 30, 2012, paper in the Journal of Neurosurgery attempts to address these questions through a retrospective analysis of the density of neurosurgeons/neurologists per population as compared with the number of deaths from stroke over a three-year period. The research team found a significant reduction in mortality associated with an increase in density of neurology providers.
Lead study author, Atman Desai, MD, chief neurosurgery resident at Dartmouth-Hitchcock Medical Center in Lebanon, NH, told Neurology Today that “nationally, the question of optimal resource and manpower allocation in health care has become increasingly important, but relatively little is known about the effect of specialists in improving national public health outcomes.”
The topic has been largely neglected in recent political discussions on health care reform, Dr. Desai said, and given that stroke is a major cause of death in the US, we wanted to investigate whether increased numbers of neuroscience specialists had a tangible effect on stroke mortality.
The main takeaway, he said, when we adjusted for a range of socioeconomic factors such as the rural and urban status of the county, its educational levels, wealth and poverty levels, and the density of general practitioners in a given county, is that “we still found that an increase of one neurologist or neurosurgeon was associated with .38 fewer deaths from stroke per million population.”
To the best of our knowledge, this is the first national-level study looking at the effect of specialist manpower on stroke outcomes, Dr. Desai said.
The study's primary outcome variable was the average in cerebrovascular disease deaths in the 3141 counties analyzed between 2004 and 2006, using the US Department of Health and Human Services Area Resource File (ARF). After adjusting for urban/rural status, density of general practitioners (GPs), and socioeconomic status, each increase of 1 neuroscience provider was associated with 0.38 fewer deaths from stroke per year (p < 0.001) per million population.
There was also a significant increase in stroke-related deaths in rural settings (p < 0.001) and with increased density of GPs (p < 0.001). “We found that neuroscience specialists were generally concentrated in and around cities that house tertiary care hospitals, so there was a large disparity in the density of these providers across the country,” Dr. Desai told Neurology Today.
The authors concluded that “the availability of specialists is an important factor in survival after stroke, and underlines the importance of promoting specialist education and practice throughout the country.”
However there were some limitations, Dr. Desai said, due to the lack of accessible information on the ARF. He noted that they couldn't adjust for such individual factors as race, smoking, hypertension, and obesity that have been implicated in stroke outcomes.
In addition, he said, the study didn't look at the impact of different aspects of stroke care on stroke mortality — for example, the role of surgical intervention. “Finally, the study didn't analyze the association of providers with stroke morbidity, but focused only on stroke mortality.”
Gustavo Saposnik, MD, associate professor of medicine (neurology) and director of the Stroke Outcomes Research Center at St. Michael's Hospital at the University of Toronto, told Neurology Today that this is an interesting study “reinforcing the well-known concept of providers/hospital volume-outcomes relationship. Higher density of neurologists and neurosurgeons is likely associated with prompt access to expert stroke care (especially in urgent situations), and consequently with better outcomes.”
The study's strengths include its county-level data, and adjustment for level of education and socioeconomic status, he said. However, the limitations are its “ecological design (that is, it's impossible to establish a cause-effect relationship), and lack of adjustment for patients and hospital level characteristics.” For example, Dr. Saposnik said, the two most important factors influencing stroke mortality — patients' age and stroke severity — were not included in the analysis. The results could have been influenced by variability in these factors, among others (such as ethnicity), he added. Furthermore, the definitions and codes for stroke were not provided in the analysis, he said.
S. Claiborne Johnston, MD, PhD, associate vice chancellor of research and director of the Stroke Service at the University of California, San Francisco, was not convinced by the study's assertion. “One would like to believe that this association is causal — that having more neuroscience clinicians available reduces stroke mortality. However, there are so many other potential explanations, I just don't think we can come to that conclusion from this study,” he told Neurology Today. These types of studies can suffer from something called the “ecological fallacy,” he added, where “association at a group level — for example, at the country level — does not necessarily translate to something at the individual level.”
“The main strength was the use of an established administrative database used by the US Department of Health and Human Services,” said Cheryl Bushnell, MD, associate professor of neurology and director of the Wake Forest Baptist Stroke Center at Wake Forest University Baptist Medical Center in Winston Salem, NC, but she acknowledged that this database was also a limitation.
“The administrative database can also be a weakness, because many assumptions often are made in analyses such as these. One major assumption is that the density of neuroscience specialists equals the density of high volume academic centers. This may be the case, but high vs. low volume stroke center was not included as a variable in the analysis. And, obviously, cause and effect cannot be established, as the authors stated.”
There may only be a small subset of neurologists in academic centers “who specialize mainly in stroke, and an even smaller subset of the neurosurgeons.” Adjusting in some way for counties with academic medical centers would also have been helpful, Dr. Bushnell told Neurology Today. “Stroke mortalities represent only about 10–15 percent of the population, and there is no way to know how far out from the stroke these deaths occurred (if this was known, the paper does not list it),” she added.
Nationwide, stroke mortality is decreasing, Dr. Bushnell said, “it is now the fourth leading cause of death and not the third. Therefore, I believe the focus should be more on improving outcomes, rather than just saving lives of patients who subsequently have a poor quality of life.”
IMPLICATIONS, NEXT STEPS
Dr. Desai told Neurology Today future studies should investigate “which aspects of stroke care have the greatest benefit, and the effect of providers and these interventions on morbidity from stroke and long-term functional outcomes.”
Future studies should also aim to develop an understanding of the challenges involved in increasing specialist providers in underserved areas, Dr. Desai added. This study demonstrates the potentially beneficial effect of boosting specialist providers in underserved areas, which is an inherently challenging task, he pointed out. “Issues such as these are clearly important in the ongoing discussion of optimal resource allocation in health care.”
In an accompanying editorial, Fred G. Barker II, MD, an associate professor of neurosurgery at Harvard Medical School, noted that the most powerful result in Dr. Desai's study was the effect of rural setting on stroke mortality. “For many US residents it is inevitable that the nearest hospital will lack sophisticated stroke care: most US towns smaller than 30,000 residents have no neurosurgeon, and most smaller than 20,000 have no neurologist. For residents of such small towns the average travel distance to a neurologist is 9 miles; for truly rural residents the average distance to a neurologist is 45 miles,” he wrote.
Given the time-sensitive nature of stroke, Dr. Barker suggested increased use of telestroke services and “the operation of market forces promoting the creation of more comprehensive stroke centers.”
Will the findings have an impact on health care policy? Dr. Saposnik said: “This novel study provides some evidence regarding the association between the combined density of neurologist/neurosurgeons and death for cerebrovascular disease.” However, he cautioned that further studies are necessary before any changes to health policy can be made, specifically supporting “decentralization of resources to facilitate access to neurology/neurosurgery in small hospitals or areas far from tertiary hospitals.”
LISTEN UP, TUNE IN: A Nov. 30, 2012, paper in the Journal of Neurosurgery offers a retrospective analysis of the density of neurosurgeons/neurologists per population as compared with the number of deaths from stroke over a three-year period. The research team found a significant reduction in mortality associated with an increase in density of neurology providers. Atman Desai, MD, chief neurosurgery resident at Dartmouth-Hitchcock Medical Center in Lebanon, NH, discusses the association of neurology providers and reduced stroke mortality: http://bit.ly/rCBryX.
FOR FURTHER READING:
• Desai A, Bekelis K, Erkmen K, et al. Association of a higher density of specialist neuroscience providers with fewer deaths from stroke in the United States population. J Neurosurg
2012; E-pub 2012 Nov 30.
©2013 American Academy of Neurology
• Barker FG. Editorial: Stroke, geography, and provider density. J Neurosurg
2012; E-pub 2012 Nov 30.
Neurology Today Quick Links