As a spine surgeon, I read Ms. Springen???s article with great concern. Debate of this nature deserves more than the seven paragraphs Newsweek has provided. The manuscript she references concludes that patient outcomes have not improved during the period of 1997 to 2005 despite increased medical expenditures for spine care. The methodology and statistical analysis of the data is so problematic as to make it, at best, uninterpretable, and at worst, misleading. There are numerous methodological problems with the data presented in this manuscript.
1) The Medical Expenditure Panel Survey (MEPS) is not a valid tool to assess a patient???s condition, their treatment, and certainly not their outcome. It was not designed for this use, and has not been subjected to the routine validity testing that is required of assessment tools for scientific publication. The nature of the data collected, i.e. self-report, is generally accepted as only a weak source of accurate data.
2) The ICD-9 codes to which the MEPS data are correlated are an entirely different set of patients. In other words, the individuals reporting the economic data in the MEPS survey are DIFFERENT individuals than those reporting the health outcomes data. How can conclusions on ???outcome??? be made on artificial correlations created from two unrelated populations of patients? An appropriate analogy might be to suggest that the crop of oranges was larger over this period of time because apple growers spent more time polishing their apples before sending them to market! Clearly there is no logical connection between these events.
3) A full 53 percent of the patients in the ICD-9 group were ???unspecified disorders of the back???. This makes it impossible to draw meaningful conclusions about back and neck problems because the population of patients was so poorly defined.
4) The difference in expenditure increase between the study groups (spine vs non-spine problems) over the years of the study was not statistically significant (p = .07). Nonetheless the authors ignore this and make broad sweeping conclusions regarding the higher expense of caring for spine problems.
5) The authors admit that when co-morbidity (other unrelated illnesses) was included as co-variate in the analysis, it ???weakened the diverging trend between the respondents???. (In other words, the patients with spine problems tended to be sicker patients in general, and therefore, more expensive to care for. p < .001) But the authors chose to not include that data in the analysis they reported and from which they made conclusions.
In summary, the conclusions of this manuscript are neither rational nor supported by the data reported.
Sincerely,
Richard G. Fessler, M.D., PhD.
Professor of Neurosurgery
Feinberg School of Medicine
Northwestern University









Discuss