to the editor: Medicine has had many firmly held beliefs, traditions, and treatments that were accepted as absolute fact and routinely prescribed only to be later disproved and abandoned. Physicians constantly change their treatments in response to new information. Nothing in the past compares with the current high praise of evidence-based medicine.

The evidence-based approach is sound; however, the way it is applied is flawed, and it can also be misused. Few of the problems physicians deal with every day have had large, prospective, double-blind, placebo-controlled studies to provide best practices. For medical problems that have been studied, the study population may be different from the patient being examined. Study patients typically have only one defined problem, with no confounding diseases or multiple medication use. With a few exceptions, the only form of therapy that can satisfy the evidence-based medicine study criteria rigidly is drug therapy with a new patentable drug, for which the pharmaceutical companies are the major source of research funds.

If these studies show either a negative effect or no effect, then that is definitive proof that therapy should not be routinely performed. Even if there is a positive result, the studies do not discuss what treatment is most effective. Better approaches, especially nonpharmacologic, may exist. With large enough study populations, clinically useless, trivial effects can achieve statistical significance. Continuing to use treatments with minimal clinical effectiveness, but statistically significant results, raises medical costs. Simply because some medical treatments do not have studies does not mean a treatment is ineffective.

Health maintenance organizations (HMOs), insurance companies, and governmental programs, such as Medicare, use the absences of these types of studies to deny coverage for a treatment. For example, antibiotics for sinusitis are accepted even though many studies show no difference compared with placebo, while alternative medicine treatments can get nowhere with these agencies.

In 20 years of treating chronic pain, I have found that the best available treatment is Prolotherapy (an injection treatment that stimulates healing of injured ligaments or tendons using Dextrose or other mild irritants). It has a wide spectrum of use, is safe in experienced hands, is effective in a high percentage of appropriately selected cases, and has no drug side effects. Small, double-blind studies exist to support Prolotherapy; however, because the injectable solution for this treatment does not involve patentable drugs, there will likely never be any large studies of its efficacy conducted. The only treatment for chronic lower back pain that satisfies all the study criteria is nonsteroidal anti-inflammatory drugs (NSAIDs) that have been proven to have a positive analgesic effect compared with placebo. Long-term use of NSAIDs may have side effects. However, the underlying problem is not benefited, and suppression of the COX-2 enzyme could theoretically impair tissue repair.

These types of studies are best used to prove that a treatment is ineffective. The goal of studies that meet all these criteria being the standard of care is ideal, but rarely achievable.