Live-cell imaging techniques offer an analytical platform for investigating cellular structures and functions, significantly advancing our understanding of disease mechanisms and drug development. The real-time observation of morphology and dynamics of cellular structures is crucial for studying cell activity and cell-materials interactions. Dynamic changes in living cells can occur even at the subnanometer level, such as the minute deformation of the cell membrane during neuronal action potentials. Additionally, to meet the trend of next-generation atomic manufacturing, nondestructive and accurate in-line characterization is crucial for ensuring high yields in large-scale manufacturing. However, the widely used metrology tools, such as scanning electron microscopy (SEM) and atomic force microscopy (AFM), suffer from extremely low measurement throughput and may introduce invasiveness to the samples. Interferometric Quantitative Phase Microscopy (iQPM), a label-free wide-field imaging technique, has been widely employed to obtain morphology distributions and dynamic changes of samples quantitatively. To satisfy the application demands, the development of high-sensitivity iQPM is of great significance and importance.Firstly, the noise sources involved in the phase measurement process are introduced, such as camera noise (including photon shot noise, dark noise, readout noise), 1/f noise, instability of light source, mechanical vibrations, and air disturbances. By analyzing these noise sources, the key limiting factors affecting the phase sensitivity of the iQPM system are discussed, which provides a solid theoretical basis for developing strategies to improve the phase sensitivity of iQPM. The sensitivity improvement strategies are introduced from the suppression of environmental noise, detection noise, and speckle noise, respectively. In terms of environmental noise, the common-path interferometry-based iQPM techniques can effectively alleviate the optical path difference changes between the sample beam and reference beam due to mechanical vibrations and air disturbances, thus improving the temporal phase sensitivity. Meanwhile, the noise caused by mechanical vibrations can be further reduced through the optimized spatiotemporal filtering method. To reduce the influence of camera noise on phase sensitivity, the methods of increasing the effective well capacity and expanding the dynamic range have been proposed, enabling the measurement of subtle phase changes. From the perspective of speckle noise, the proposed strategies aim to superpose images with uncorrelated speckle patterns, which can be achieved by reducing the illumination temporal coherence and modulating the spatial spectrum of the illumination light. The high-sensitivity iQPM techniques have been applied in cutting-edge fields such as neuroscience, atomic-scale material metrology, and wafer defect detection. This review summarizes critical strategies that have driven substantial advancements in the phase sensitivity of iQPM. By employing these methods, the temporal phase sensitivity of iQPM has been pushed to an impressive 2 pm level. The primary objective of this work is to provide an important reference for further improvement of phase sensitivity. Notably, current noise suppression-based sensitivity enhancement strategies are constrained by the inherent photon shot noise, sensitivity improvement methods based on signal amplification may offer a path to break through the current limitations and achieve higher phase sensitivity.