Background: Problem oriented medical record (POMR) is current standard for medical recording. Problem list is the central part of POMR. Complete problem lists may increase the quality of care. But problems lists are often inaccurate, incomplete, and out of date. Many clinical decision support systems were developed to increase completeness and accuracy of problem lists. Although these systems are good, they can be also the impetus for tremendous clinical documentation errors, which may carry risk for patient care. Moreover, there are still many hospitals without coded problem lists, even without problem lists. For physicians who are used to maintain problem lists with paper or computer free-text entry. Implementation of coded problem lists may have impact on completeness and accuracy of problem lists. In this thesis, we introduce the implementation of coded problem lists at one medical center in central Taiwan (this institution) and learn lessons from it. Method: We introduced the implementation and function of POMR-SYSTEM. It is replaced the paper base POMR in this institution. Healthy provider used this system to input coded problem lists and progress note for inpatient care. Two years after implementation, we retrospectively reviewed paper based problem lists and coded problem lists in department of internal medicine. We aimed to see the change of completeness and accuracy of problem lists after this transition. Efficiency and usability of POMR-SYSTEM were also checked. Result: These preliminary data suggested that implementation of the system significantly reduced completeness and accuracy of problem lists in some healthy providers. We discussed the possible reason for this result. We found that system usability, efficacy, incentives, leadership and education were important factors for successful implementation. Conclusion: This study suggested that implementation of coded problem lists carries risk of reducing completeness and accuracy of problem lists. Institutions should carefully develop their systems before full implementation of coded problem lists.