Serum Lipid Profile as a Predictor of Recurrence in Chronic Suppurative Otitis Media: A Retrospective Electronic Health Record-Based Cohort Study.

BACKGROUND

Chronic suppurative otitis media (CSOM) is a major cause of preventable hearing loss, particularly in resource-limited, semi-urban healthcare settings where advanced radiological imaging is often unavailable for recurrence risk stratification. Systemic dyslipidemia has been mechanistically linked to mucosal inflammation and epithelial dysfunction, but its association with CSOM recurrence has not been directly examined. This study investigated the clinical utility of fasting serum lipid profiles as predictors of recurrence in patients with CSOM.

METHODS

We conducted a retrospective, electronic health record-based cohort study of 160 patients diagnosed with CSOM at Mirzapur Modern Hospital, Tangail, Bangladesh, between June 2023 and May 2024, followed for 12 months post-treatment. Patients were stratified into recurrent (Group A, n = 68) and non-recurrent (Group B, n = 92) cohorts. Continuous variables were compared with independent Student's t-tests; categorical variables were compared with chi-square tests incorporating Yates' continuity correction for 2×2 tables. Multivariable binary logistic regression identified independent predictors of recurrence, and receiver operating characteristic (ROC) curve analysis with Youden's index determined optimal diagnostic cutoffs.

RESULTS

The recurrent cohort had significantly higher mean total cholesterol (211.19 ± 24.55 vs. 183.81 ± 20.89 mg/dL), low-density lipoprotein cholesterol (LDL-C; 140.02 ± 22.51 vs. 110.71 ± 19.26 mg/dL), and triglycerides (180.26 ± 30.40 vs. 142.30 ± 28.61 mg/dL; all p < 0.001), and significantly lower high-density lipoprotein cholesterol (HDL-C; 35.12 ± 5.18 vs. 44.63 ± 5.97 mg/dL; p < 0.001) than the non-recurrent cohort. On multivariable analysis, LDL-C (adjusted odds ratio (OR) 1.088, 95% CI 1.051-1.127), HDL-C (OR 0.826, 95% CI 0.745-0.916), triglycerides (OR 1.029, 95% CI 1.009-1.049), and atticoantral (unsafe) disease type (OR 6.838, 95% CI 2.109-22.185) were independent predictors of recurrence (all p ≤ 0.005). LDL-C demonstrated the highest diagnostic accuracy (area under the curve (AUC) 0.850; optimal cutoff 138.5 mg/dL; sensitivity 64.7%, specificity 92.4%).

CONCLUSIONS

Baseline fasting lipid derangement is independently associated with CSOM recurrence, and the fasting lipid panel is a promising, low-cost candidate marker for recurrence risk stratification in resource-limited healthcare settings. As these findings derive from a retrospective, single-center cohort without external validation, the lipid panel should be regarded as hypothesis-generating rather than a validated clinical tool; prospective, multicenter studies with external validation and direct lipidomic analysis of middle ear effusion are needed before it can be positioned as a practical stratification instrument.