%0 Journal Article
%T A Comparative Statistical Study of Treatment Outcomes in Orthodontic Procedures Using Longitudinal Data Analysis
%A Percy Huata Panca
%A Vladimiro Ibañez Quispe
%A Juan Reinaldo Paredes Quispe
%A Lucero Danitza Mamani Chipana
%A Nelly Beatriz Quispe Maquera
%A Henry Quispe Cruz
%A Godofredo Quispe Mamani
%J Annals of Dental Specialty
%@ 2347-2022
%D 2026
%V 14
%N 1
%R 10.51847/0CR0iq1PtX
%P 102-111
%X Orthodontic treatment outcomes unfold over months or years, so repeated measurements are more informative than isolated pre-treatment and post-treatment comparisons. In routine clinical records, however, repeated observations are rarely complete or equally spaced. Missed appointments, delayed visits, incomplete cephalometric or cast records, changing compliance, and examiner variability are expected features of orthodontic data. These imperfections must be treated as part of the evidence rather than removed from the analysis. Many comparative orthodontic studies evaluate fixed appliances and clear aligners using complete cases or simplified before-and-after summaries. Such approaches can bias results when patients with poorer progress are more likely to miss visits or discontinue treatment. They also ignore that treatment groups often differ at baseline in age, severity, malocclusion type, and motivation. A longitudinal analysis must therefore address missing outcomes, attrition, irregular visit timing, measurement error, and non-random procedure assignment. This study compared longitudinal treatment outcomes between fixed appliances and clear aligners using a university orthodontic clinic dataset. The primary objective was to estimate change in Peer Assessment Rating score over treatment while explicitly accounting for irregular observation times, missing outcome measurements, patient dropout, and baseline imbalance. Secondary objectives were to compare complete-case and imputed estimates, evaluate propensity-score-adjusted treatment comparisons, and assess whether conclusions were robust to alternative longitudinal models. The analytic dataset included 215 patients treated in a university orthodontic clinic between 2017 and 2023, with 125 receiving fixed appliances and 90 receiving clear aligners. Planned records were baseline, 3 months, 6 months, 12 months, and end of treatment, but actual visit timing varied and some patients contributed only three or four usable outcome records. The primary outcome was PAR score, modeled using linear mixed effects regression with patient-specific random intercepts and time represented as actual months since baseline. Missing outcomes were addressed using multiple imputation, attrition was examined using pattern-mixture sensitivity analysis, and baseline imbalance was reduced using propensity score matching. Dropout before the final treatment record was 22% overall, with higher attrition in the clear aligner group than in the fixed-appliance group. After propensity score matching, 70 patients remained in each group, and both groups showed clinically meaningful PAR reduction of approximately 70%. The procedure-by-time interaction was not statistically significant, but the confidence interval was wide enough to include clinically relevant differences. Multiple imputation changed the estimated procedure contrast by 1.2 PAR points compared with complete-case analysis, and convergence problems required simplification of the random-slope structure in a small proportion of fitted imputation models. The study did not provide clear evidence that either fixed appliances or clear aligners produced superior longitudinal PAR improvement in this clinic cohort. This finding should not be interpreted as proof of equivalence, because attrition, measurement error, irregular follow-up, and reduced matched-sample size limited precision. The analysis demonstrates that longitudinal mixed models are useful for imperfect orthodontic records, but sensitivity analyses remain essential. Future comparative orthodontic studies require stronger retention strategies, standardized repeated measurements, and calibrated outcome assessment.
%U https://annalsofdentalspecialty.net.in/article/a-comparative-statistical-study-of-treatment-outcomes-in-orthodontic-procedures-using-longitudinal-d-pvnf54aukwtrkix