This study aimed to compare patients’ Salzmann Index scores for those who applied for Medicaid orthodontic coverage in Pennsylvania with their corresponding American Board of Orthodontics discrepancy index (DI) scores to assess if there is a correlation between Salzmann and DI scores. In addition, a threshold DI score was calculated that would correspond to Medicaid coverage approval. The study intended to answer the following questions: is there a correlation of 0.7 or greater between a patient’s Salzmann Index and their DI? If so, is there a particular DI score that can be used as the minimum score for approving Medicaid orthodontic coverage in the state of Pennsylvania?
Salzmann Index scores, DI scores, and approval and disapproval results for Medicaid orthodontic coverage were obtained from 104 subjects aged between 10 and 17 years. A linear regression model was generated to assess if there was a correlation between the Salzmann scores and DI scores. If a correlation coefficient of 0.7 or greater were found, a threshold Salzmann Index score would be determined for subjects who were approved for Medicaid orthodontic coverage. The threshold Salzmann score would be used in the linear regression formula to find the corresponding DI score, which would be designated as the threshold DI score for approval for Medicaid orthodontic coverage in the state of Pennsylvania.
A Pearson correlation of 0.453 was calculated between the 104 Salzmann scores and DI scores, demonstrating a moderate correlation. With the correlation coefficient being lower than 0.7, binary logistic regressions were calculated to assess the predictability between a given Salzmann score and approval and disapproval for Medicaid orthodontic coverage. The Salzmann score had an overall 68.3% success in predicting Medicaid orthodontic coverage approval/disapproval. Of the 58 subjects that were approved for Medicaid orthodontic coverage, 46.6% had Salzmann scores equal to or greater than 25. Of the 46 subjects that were disapproved for Medicaid orthodontic coverage, 78.3% had Salzmann scores equal to or less than 24.
With the lack of high prediction rates seen from the results of the regression models, the current system of Medicaid does not appear to show consistency for assessing the need for orthodontic treatment coverage. Multiple insurance companies that participate under Medicaid require a Salzmann score of 25 or greater for approval; however, the results show the Salzmann score is arbitrary in terms of approval and disapproval. There appear to be underlying factors apart from the Salzmann score that the Pennsylvania Medicaid system uses to justify whether a patient was approved or denied for coverage.
No meaningful correlation was found between Salzmann and discrepancy indexes.
Pennsylvania Medicaid was not consistent when assessing the need for orthodontics.
The Salzmann score was not predictive of Medicaid approval or disapproval in Pennsylvania.
A Salzmann score of 19 or greater may be a better threshold than 25 for Medicaid coverage.
A majority of insurance under the Medicaid plan reimburse patients in need of orthodontic treatment on the basis of orthodontic severity, also known as handicapping malocclusion and financial need. Each state has specific qualifiers that determine whether a patient’s orthodontic treatment will be covered at no cost. Medicaid was enacted as part of the Social Security Act signed by President Lyndon Johnson in 1965 under Title XIX. Title XIX was developed to provide health care coverage to the underserved who were in need of medical services, including dental and orthodontic care. Under the Early and Periodic Screening, Diagnosis, and Treatment Program established in 1967, a portion of Medicaid was mandated to provide access to orthodontic treatment for Medicaid-eligible children. As a result, Medicaid began covering orthodontic treatment costs for patients with handicapping malocclusions. Because of the federal government funding, 50% of Medicaid and each state funding the other 50%, the definition of handicapping malocclusion is defined at the discretion of each state.
The American Association of Orthodontists (AAO) defined medically necessary orthodontic care as “the treatment of a malocclusion (including craniofacial abnormalities/anomalies) that compromise the patient’s physical, emotional, or dental health.” The Salzmann Index was selected by the AAO to be used as the objective index to evaluate if a patient under Medicaid coverage was deemed to have a medically handicapping malocclusion.
The Salzmann Index is an assessment record intended to disclose whether a handicapping malocclusion is present and evaluate its severity according to the criteria and weighted point values assigned to them. The Salzmann Index assesses for missing teeth, crowding, rotated teeth, spacing, overjet, overbite, open bites, crossbites, and occlusion. A key advantage to the Salzmann Index is that no records are necessary to calculate a scoring. Intraoral photographs may help, but no casts or radiographs are typically needed to complete the assessment. However, the Salzmann Index has limitations because of the lack of inclusion of any assessment of the patient’s skeletal pattern and cephalometric values; additional treatment complexities; and anomalies such as impactions, transpositions, asymmetries, and it only scores permanent teeth. No psychological or health-related quality of life factors are assessed by either index.
In 1985, the AAO rescinded using the Salzmann Index and stated that they opposed using any index or classification system to determine the need for orthodontic treatment. However, because of a limited budget each state appropriates for Medicaid services, most states currently use an index such as the Salzmann Index to evaluate whether a patient has a handicapping malocclusion and qualifies for orthodontic coverage. The index of complexity, outcome and need; dental aesthetic index; and index of orthodontic treatment need are additional indexes that states use, which involve an esthetic component of malocclusion along with study cast analysis. Other indexes used include the handicapping labiolingual deviation and Peer Assessment Rating (PAR) index, which only includes cast analysis (occlusal) to determine the need for treatment. A validity study published in 1997 by Younis et al compared the reliability and validity of 3 occlusal indexes of orthodontic treatment need in predicting the opinion of treatment need from a panel of orthodontists. One-hundred-sixty casts were scored using the index of orthodontic treatment need, the handicapping labiolingual deviation index, and the handicapping malocclusion assessment record. The study concluded the occlusal indexes of treatment need were reliable and valid. However, the article concluded with a caveat that
an occlusal index by itself may not be the only instrument for prioritizing orthodontic treatment need in a public health system. Because malocclusion has generally not been associated with temporomandibular disorders, caries, and periodontal disease, but with negative socialization and self-image consequences, an occlusal index might be one component of a larger and more exhaustive index. An improved index might include a psychosocial component, a self-image component, and an overall facial attractiveness component to identify those individuals who would possibly benefit the most from receiving orthodontic treatment.
The American Board of Orthodontics (ABO) developed a discrepancy index (DI) to provide a better evaluation of a patient’s malocclusion severity on the basis of not only an occlusal component but also on a skeletal, radiographic, and complexity assessment. , Liu et al did a validation of the ABO DI and the U.S. PAR. Sixty-nine orthodontists rated malocclusion severity on a 5-point scale after reviewing a full set of pretreatment records. Their judgment was then compared with DI and U.S. PAR scores determined by 3 calibrated examiners. The results concluded there was excellent interexaminer reliability with the DI and U.S. PAR Index, validating both indexes for malocclusion severity. The DI provides a more holistic approach to diagnosing the severity of a patient’s malocclusion when compared with the Salzmann Index. The DI assesses a patient’s overjet; overbite; anterior open bite; lateral open bite; crowding; occlusal relationship; crossbites; cephalometric numbers; other anomalies such as impaction, missing teeth, supernumerary teeth, ankylosed teeth, asymmetries; and additional complexities. The DI does have some shortcomings as well, including the additional time needed to acquire a full set of records, such as a panoramic radiograph, cephalometric radiograph, cephalometric analysis, and dental casts.
In summary, there are multiple indexes currently available to assess the severity of a patient’s malocclusion. The Salzmann Index has been the standard for a majority of states when evaluating whether a patient will be accepted for Medicaid coverage in terms of orthodontic treatment. Currently, in the state of Pennsylvania as well as countless other states, the vast majority of insurance companies (ie, Pennsylvania Children’s Health Insurance Program and AmeriHealth Caritas) that participate with each respective state Medicaid program appear to require a minimum Salzmann score of 25 to qualify for orthodontic treatment coverage. As of 2015, 41 states use the Salzmann Index or something similar to determine Medicaid orthodontic coverage, which may lead to similar inconsistent approval and disapproval decisions. The Salzmann Index’s shortcoming as only an occlusal index only represents a patient’s malocclusion while overlooking other important factors such as skeletal discrepancies and condition complexity. There is a need for an index that is a comprehensive evaluation of a patient’s need for orthodontic treatment. Ghafari discussed how nearly all indexes except the DI are based on an occlusal assessment without evaluation of a cephalometric radiograph and therefore dismissing the skeletal component of malocclusion. He further stated that a medically necessary orthodontic care index should not just include the dental and skeletal components but also an assessment of physical and emotional factors referred to as health-related quality of life aspects.
Material and methods
Subject’s Salzmann Index scores, DI scores, and approval and disapproval results for Medicaid orthodontic coverage were obtained from records of 104 children who visited the Seton Hill University Center for Orthodontics clinic in Greensburg, Pennsylvania. These records were obtained from a convenience sample with all subjects being aged between 10 and 17 years who applied for Medicaid orthodontic coverage in the state of Pennsylvania between October 23, 2017 and October 23, 2019. The reliability of the Salzmann and DI scores was determined by rescoring approximately one third (33) of the subjects. Reliability coefficients for both the Salzmann and the DI were found to be 0.99. A linear regression model was generated to assess whether there was a correlation between the Salzmann scores and DI scores. If a correlation coefficient of 0.7 or greater was found, a threshold Salzmann Index score was determined for subjects who were approved for Medicaid orthodontic coverage. A threshold of 0.7 or greater was used because by convention, and according to Rumsey, “a general rule of thumb is that correlations close to or beyond +0.7 or −0.7 are considered to be strong,” or in other words, highly correlated. The threshold Salzmann score would then be used in the linear regression formula to find the corresponding DI score, which would be designated as the threshold DI score for approval for Medicaid orthodontic coverage in the state of Pennsylvania. If a significant correlation between a Salzmann score and DI score was not found (<0.7), binary logistic regressions were calculated to assess the predictability between a given Salzmann score and approval and disapproval for Medicaid orthodontic coverage.
The Pearson correlation between the 104 subject’s Salzmann and DI scores was 0.453, as shown in Table I , which is considered a low or moderate correlation. Because of the correlation coefficient falling below 0.7, which was needed to be considered an effective predictor for a DI score from a given Salzmann score, binary logistic regressions were calculated to assess the predictability between a given Salzmann score and approval and disapproval for Medicaid orthodontic coverage.