A Root Cause Analysis Using Text Mining for Patient No-Shows: A Case Study of Primary Care Centers Serving Rural Areas

Patient no-shows lead to a significant loss in the healthcare industry. Many researchers have focused on this problem in various healthcare settings using several methods. However, few studies used text mining to conduct a root cause analysis (RCA) and a Pareto analysis related to patient no-shows at outpatient primary care settings serving rural areas. This study uses a text-mining framework to investigate and analyze the factors and causes behind patient no-shows at Finger Lakes Community Health (FLCH), an outpatient primary care medical center serving rural areas in New York. The dataset used in this study was built from 5,400 attempted telephone interviews with patients who did not show up for their appointments. The text-mining framework was used to explore, analyze, and categorize information in the notes taken by interviewers. The novel framework uses lookup words that are weighted based on their order of appearance in a note to assign that note to a certain root cause. It was found that miscommunication, personal issues, and forgetfulness are key causes of patient no-shows. Also, it was found that almost 80% of the patient no-shows were caused by only five root causes. Based on this analysis, several interventions were proposed to reduce the incidence of patient no-shows.