Introduction Osteomyelitis is an inflammatory infectious pathology that represents a public health problem due to the high morbidity associated with its high disabling potential and with a reserved short-term prognosis. The risk factors that favor the growth of a particular microorganism, in the maxilla or jaw after the placement of dental implants, include multiple general variables (age, sex, smoking and systemic diseases) as well as local variables (anatomical area, number of implants, type of osteomyelitis). The objective of this research is to perform a linear model to estimate the risk of osteomyelitis from the correlation of the multiple variables obtained from the review of the world literature.

Material and Methods Descriptive, retrospective, cross-sectional study. A multiple correlation model was constructed to estimate osteomyelitis (OM) from the general variables and local variables obtained from the review of world literature in the last 10 years.

Results The sample consisted of 38 patients with a minimum age of 43 years and a maximum age of 84 years and a mean of 61 years, 28 (73.6%) female patients and 10 (26.3%) males, 23 (60.5%) patients with systemic diseases and 15 (39.4%) healthy patients. Gender is related to osteomyelitis since the female sex has 13 times more risk of presenting osteomyelitis than the male gender. Age is related to osteomyelitis since at age over 60 there is 1 time more risk of osteomyelitis than those under 60 years of age. The presence of a systemic disease by itself was not statistically significant when performing the univariate logistic regression analysis, however, when performing the multivariate analysis arterial hypertension, diabetes, patients who smoke or consume immunosuppressive or immunomodulatory drugs if they are predisposed to osteomyelitis in 47% according to Cox and Snell and 64.7% according to Nagelkerke. With respect to the anatomical area, the jaw has 9 times more risk of osteomyelitis than the maxilla.

Conclusion The variables selected to perform our analysis, if they can behave as risk factors for osteomyelitis. In the multivariate model, the predictors of systemic diseases are independent, even without collinearity, asthma, bronchitis and smoking are related to respiratory diseases, as well as the combination with arterial hypertension, since although the individual effect of each of the variables cannot be precisely identified, The multivariate effect translates into an increase in the estimated regression coefficient (13% more likely) to present osteomyelitis, this is known as non-perfect multicollinearity. There is no specific statistical method to estimate osteomyelitis in patients who are candidates for dental implants, however, the present multiple correlation model allows us to determine to what extent the general, local and concomitant factors reported in the literature increase the risk of presenting osteomyelitis.