This month the Associated Press published a valuable report about the weak evidentiary basis for flossing as a way to reduce plaque and tooth decay, and to prevent gum disease. In response, the American Dental Association and the American Academy of Periodontology acknowledged the lack of evidence but said their recommendation stands: believe us, flossing helps.

But the AP article is less about floss and more about how medicine backs up its recommendations. What does it mean to have good evidence that something we do actually works?

The AP is right: the evidence for the efficacy of floss is not great. Floss is a 19th-century invention. Flossing recommendations are based on assumptions: that it removes plaque from between teeth and gums and that, since plaque leads to tooth decay and gum disease, flossing should improve or prevent these conditions. But this is largely anecdotal and does not rise to the standard we have for medical treatments in the 21st century.

Academic rigour has changed in the past several decades. Anecdotal evidence and tradition have given way to “evidence-based medicine,” which arises from well-designed, controlled studies. Once, angina sufferers had surgery to tie the mammary artery; evidence found this dangerous and unnecessary. Now, diet, exercise and relatively inexpensive medicine are used. Peptic ulcers required stress reduction and avoidance of spicy food until thorough evidence showed that the most common cause was bacteria. New studies are continuously published, improving what we know about a host of medical conditions and making treatment more effective and efficient.

The Cochrane Library, the gold standard in assessing medical research, has analysed the evidence for flossing. Cochrane says there is “some evidence” that flossing reduces the mild, reversible form of gum disease known as gingivitis, “weak, very unreliable evidence” that flossing plus brushing reduces plaque and no evidence for tooth decay prevention.

Terms such as “weak” and “very unreliable” have specific meanings directed at the quality of the studies that were conducted. The basis of evidence-based medicine is properly designed research protocols. Cochrane finds that even the best flossing studies have flaws, ranging from poor controls, gaps in the dental assessments of subjects or study periods that are too short. These flaws weaken studies and the results are, in turn, weak. We are left with little strong evidence of the efficacy of flossing.

I couldn’t tell you why, exactly, flossing research suffers so much, but I have my suspicions. Good studies are expensive; there is no fortune and glory to be found in flossing. It is inexpensive and a relatively easy habit to adopt, carrying no risk, so research does not carry a potential “big win” such as eliminating an unnecessary surgery or transforming pharmaceutical regimes.

Perhaps studies of flossing end up relegated to the lower levels of academics, where standards are not as rigorous. Regardless, the public and the profession are left with anecdotes and weak evidence. And maybe that just seems good enough.

It is worth emphasising that, as Cochrane reports, we have some evidence that floss helps to reduce gingivitis. In turn, gingivitis can lead to periodontal disease, although it does not necessarily do so. But periodontal disease is always preceded by gingivitis, which is the loss of the tissues that support the teeth, particularly bone. The Centres for Disease Control and Prevention reported in 2012 that about half of Americans over 30 and more than 70 per cent over 65 have some form of periodontitis — the more advanced, damaging type of gum disease — making it one of the most common diseases in the United States. Periodontitis is irreversible, making prevention that much more critical.

For an individual patient, treatment can be expensive, and nationally the cost is extraordinary. Like the efficacy of flossing, the exact dollar amount is unknown, but an attempt to calculate it was published in the journal Periodontology 2000 by L. Jackson Brown, Beverly Johns and Thomas Wall. In 1999, they calculate, US dentists performed $9.9 billion worth of preventive care for periodontal disease and $4.4 billion worth of treatment. In 2016 dollars, that is $14.3 billion and $6.4 billion. And this ignores the cost of tooth loss from periodontal disease.

Healthcare professions are obliged to develop evidence for what we do, ensuring that all we recommend to our patients is supported by high-quality research. We should be studying flossing, just as we should be studying all aspects of improving dental health.

In spite of the weak evidence, flossing remains a low-risk, high-reward habit. So until we have stronger, better-designed studies, we are left with the dental version of Pascal’s wager: if flossing is truly beneficial, then you win by flossing; if it is not beneficial, you don’t lose anything by flossing.

That is not really good enough, but at the moment it is all we’ve got.

• Timothy Levine is an orthodontist in private practice and a faculty member in the Department of Dentistry at Jacobi Medical Centre in New York