<p>Missing data is a common problem in data analysis and machine learning that adversely impacts model performance and accuracy. In the healthcare industry, where precise analysis and accurate forecasts guide clinical judgments, resolving this issue is very important. The sensitivity of the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) to parameter values, results in ineffective cluster detection, which is one of the shortcomings of the current imputation techniques highlighted in this work. Thus, we suggest an optimized DBSCAN that combines an optimized Random Forest with a Multi-Verse Optimizer (MVO) for efficient missing data imputation. Our approach preserves important information across numerical and categorical data formats in addition to the underlying distribution of healthcare datasets. Experiments on five benchmark healthcare datasets—Dermatology, Diabetes, Fetal Health Disease, Heart Disease, and Parkinson’s Diseases show that the suggested methodology outperforms state- of-the-art techniques in terms of accuracy. These datasets were chosen as they offer a strong evaluation of our methodology and are diverse and representative of actual healthcare situations. The findings confirm the effectiveness of our approach in addressing missing data in healthcare by showing notable improvements in Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R² by 16.85%, 26.35%, 16.03% and 4.06%, respectively, for the given datasets.</p>