Follow-up: Reported cost-savings with CAM-oriented primary care

In a previous entry it was reported that over 7 years, patients visiting chiropractors and other CAM-oriented primary care doctors had 60% fewer hospitalizations, 62% less outpatient surgery, and 85% lower pharmaceutical costs compared with total network HMO (health maintenance organization).

These findings are important, but there were some questions, which lead researcher, Dr. Richard Sarnat, graciously agreed to address.

If you haven’t do so already, first read the initial post. Then come back to read Dr. Sarnat’s answers.

Were the patient populations cared for by CAM and conventional healthcare providers of equal health, age, and social, economic, and educational background?

• “As all HMO’s are open enrollment without bias for pre-existing conditions, for the purposes of research each IPA is hindered by the self-selection of its enrollees.”
• “That is why we attempted to analyze this obvious study limitation through more creative means, such as looking at the distribution of disease states and standardized behavioral markers such as the tobacco utilization rates among the different control groups.”

Was the distribution of disease the same between the two groups?

• “CAM enrollees had far greater percentages of mental health and orthopedic problems and lower percentages of diabetes and heart disease when compared to the control groups. Again, due to lack of randomization, the self-selection bias could not be controlled.”
• “However, the more interesting finding to me is the wide spectrum of disease states that chiropractic PCPs effectively managed, which had never been described in the literature previously.”

What procedures were in place to track patients? use of CAM and conventional care?

• “Under NCQA [National Committee for Quality Assurance] guidelines, monthly meetings to review the utilization data were mandatory and provided medical management with the data points necessary in order to be proactive in the attempt to encourage behavior, which would promote wellness and avoid acute care crisis management whenever possible.”
• “Every member had every encounter tracked for both CAM and conventional utilization.”
• “Each encounter was tracked for both clinical outcome and cost. This included all physician encounters, paramedical provider encounters, pharmaceutical usage, and all diagnostics and treatments. The only variable not tracked would have been the members’ out of pocket expenses for noncovered benefits, such as supplements or nutraceuticals.”

Does lack of referrals to conventional care practitioners necessarily mean that referrals were not needed?

• “We made the assumption that the absence of a referral indicated the lack of need for the referral.
• “We feel this assumption was justified [because] these patients were followed very closely, on a proactive basis for prevention and wellness interventions [during] approximating bimonthly visits. This is in sharp contrast to conventional medical patients who routinely are seen an average of once annually for crisis intervention.”
• “When the physician/patient relationship is this familiar, it is unlikely that there is significant pathology which is unknown to the PCP. The fact that over a 7-year period we have never been cited for inappropriate readmissions (during the annual nursing review) or had a high percentage of surprise admissions (from undiagnosed or untreated disease) is testimony to this assumption.”

How do we know that lower drug cost means drugs could not have benefited these patient as much or more than the treatment they received?

• “As the study design was focused at looking at overall population statistics and not single variables this cannot be answered definitively.”
• “I believe we can infer the answer to some degree, however, from the fact that the utilization data from the network enrollment as a whole is based on medical management from a pharmaceutical/surgical model where members (similar to ours) would have been treated with pharmaceuticals as a first line treatment option instead of the CAM modalities utilized in our IPA.”
• “The failure of conventional medical strategies based on the pharmaceutical first model to control cost and improve clinical utilization over their present performance speaks for itself.”

The bottom line?
Others should follow Dr. Sarnat’s lead to undertake this type of comparative analysis.

8/17/07 18:44 JR