xRead Article 1 - Electronic Consults in Otolaryngology

Only the " rst patient visit was included was included during the study period. Responses for return and postoperative visits were excluded. Only complete PGOMPS that could be assigned a Total Score and those " lled out in English were included. Using the patient medical record number included in the Press Ganey, the Electronic Medical Record was utilized to identify patient demographics, including self-reported race. Ethnicity (ie, Hispanic/Latino) data was found to be inconsistently reported in the medical record and was not included. We also determined the clinic setting for each visit, whether it occurred at the main University Hospital or a satellite clinic. Insurance providers were categorized for each patient visit into the subcategories: Medicaid, Medicare, Private Insurer, Self-Payer, or Worker’s Compensation Fund. Furthermore, patients are also queried to self-report wait time. Total wait time was calculated by addition of patient-reported waiting room and examination room wait times. Due to the rightward-skew and high ceiling e # ect of the PGOMPS demonstrated in the literature, “satisfaction” in this study was de " ned as a patient giving a perfect score. This perfect score variable is used to mimic the “Top Box” score reported by the Press Ganey which represents the percentage of responses in the highest possible score for a question, section, or survey (either Very Good or Always responses). The Top Box score is reported and used by the Center for Medicare Services to compare hospitals and has been commonly used in the patient satisfaction literature using Press Ganey data. Categorical variables were summarized as count and percentages (%) using Minitab Statistical Software. Continuous variables were summarized as mean ± standard deviation. Univariate binary logistic regression analyses were used to identify independent predictor variables associated with patient satisfaction, as de " ned by receiving a perfect score. These studies yielded the following predictor variables: total wait time, <18 or ≥ 18 years old, race, clinic setting, and insurance payer. Following this identi " cation, multivariate binary logistic regression models were then conducted to determine if further associations with patient satisfaction could be seen. Results Application of inclusion and exclusion criteria in the aforementioned study period yielded a total of 3998 unique new patient visits with completed surveys and associated Total Scores. Of those included in the study, the mean age was 41.2 ± 25.8 years and 56.3% female. Patient gender was shown not be a statistically signi " cant 31,32

54