Long-term study of accommodative esotropia.

PURPOSE: This study aimed, using a large sample size, to determine the long-term results of standard treatment of accommodative esotropia and identify predictors of outcome while minimizing bias in data collection and analysis. METHODS: Data from all the files of a large, long-established pediatric ophthalmology practice were collected and analyzed using a masked protocol. The study included every esotropic patient who had been prescribed glasses. Criteria for patient inclusion were designed to conform to earlier studies by the authors. RESULTS: The database totaled 1307 patients, of who 354 met inclusion criteria. A greater difference between near and distance esodeviation (AC/A relationship) correlated with a higher rate of deterioration of accommodative esotropia control (P < 0.0001). Deterioration also positively correlated with earlier age of onset (P < 0.0001), inferior oblique overaction (P = 0.0005), and amblyopia (P < 0.005). CONCLUSIONS: This study demonstrates that a high AC/A relationship increases the likelihood of deterioration of accommodative esotropia, supporting the earlier studies, as well as the accuracy of this database. It also represents a new model for the utilization of clinical trials' bias-reduction principals in the analysis of retrospective data.

Long-term study of accommodative esotropia.

PURPOSE: Previous studies of accommodative esotropia have been hampered by bias-prone methods of data collection and analysis and by small sample size. The studies have conflicting conclusions, causing uncertain results. This study aims to determine long-term results of standard treatment of accommodative esotropia and identify predictors of outcome, while minimizing bias in data collection and analysis, using the largest possible sample size. METHODS: A research assistant collected data from all files of a large, long-established pediatric ophthalmology practice (M.M.P.). The assistant was given standardized collection forms that allowed inclusion of all patient data points over all visits. The assistant was masked as to study goals. She was instructed to include any patient with esotropia who had been prescribed glasses during treatment. Descriptive terms were converted to code numbers. A second, similarly masked research assistant entered data into a computerized database. Criteria for patient inclusion were designed to conform to earlier studies by I.H.L. and M.M.P. and were implemented by computer. RESULTS: The database totaled 1,307 patients (747,717 data points). Of these, 354 qualified for this analysis. A greater difference between near and distance esodeviation (AC/A relationship) correlated with a higher rate of deterioration of accommodative esotropia control (P<.0001). Deterioration also positively correlated with earlier age at onset, inferior oblique overaction, and amblyopia. CONCLUSIONS: This study agrees with our previous findings that a high AC/A relationship increases the likelihood of deterioration of accommodative esotropia, thus confirming the integrity of the database. This unique, unbiased dataset will be used for future analyses of esotropia.

Using dependency/association rules to find indications for computed tomography in a head trauma dataset.

Analysis of a clinical head trauma dataset was aided by the use of a new, binary-based data mining technique, termed Boolean analyzer (BA), which finds dependency/association rules. With initial guidance from a domain user or domain expert, the BA algorithm is given one or more metrics to partition the entire dataset. The weighted rules are in the form of Boolean expressions. To augment the analysis of the rules produced, we applied a probabilistic interestingness measure (PIM) to order the generated rules based on event dependency, where events are combinations of primed and unprimed variables. Interpretation of the dependency rules generated on the clinical head trauma data resulted in a set of criteria that identified minor head trauma patients needing computed tomography (CT) scans. The BA criteria contained fewer variables than were found using recursive partitioning of Chi-square values (five variables versus seven variables, respectively). The BA five-variable criteria set was more sensitive but less specific than the seven-variable Chi-square criteria set. We believe that the BA method has broad applicability in the medical domain, and hope that this paper will stimulate other creative applications of the technique.