Examining how customers perceive community pharmacies based on Google maps reviews: Multivariable and sentiment analysis.

Objective: This study aims to understand customer perceptions of community pharmacies utilizing publicly available data from Google Maps platform. Materials and methods: Python was used to scrape data with Google Maps APIs. As a result, 17,237 reviews were collected from 512 pharmacies distributed over Riyadh city, Saudi Arabia. Logistic regression was conducted to test the relationships between multiple variables and the given score. In addition, sentiment analysis using VADER (Valence Aware Dictionary for Sentiment Reasoning) model was conducted on written reviews, followed by cross-tabulation and chi-square tests. Results: The Logistic regression model implies that a unit increase in the Pharmacy score enhances the odds of attaining a higher score by approximately 3.734 times. The Mann-Whitney U test showed that a notable and statistically significant difference between "written reviews" and "unwritten reviews" (U = 39,928,072.5, p < 0.001). The Pearson chi-square test generated a value of 2991.315 with 8 degrees of freedom, leading to a p value of 0.000. Discussion: Our study found that the willingness of reviewers to write reviews depends on their perception. This study provides a descriptive analysis of conducted sentiment analysis using VADAR. The chi-square test indicates a significant relationship between rating scores and review sentiments. Conclusion: This study offers valuable findings on customer perception of community pharmacies using a new source of data.

Comparing Invasive Pulmonary Aspergillosis Mortality Between Liposomal Amphotericin B and Voriconazole in Patients With Hematological Malignancy or Hematopoietic Stem Cell Transplantation.

Objectives We evaluated liposomal amphotericin B versus voriconazole for the treatment of invasive pulmonary aspergillosis (IPA) in patients with hematological malignancy or hematopoietic stem cell transplantation (HSCT). Methods This retrospective cohort, single-center study included patients with compatible radiological diagnosis of IPA between 2016 and 2021. Results Forty-six patients with hematological malignancy or HSCT were diagnosed with IPA. Thirty-nine of them fulfilled the criteria for comparing liposomal amphotericin B (n=15) with voriconazole (n=24). Their median age was 48.5 years. Stem cell transplant recipients were 45.65%, and nearly half of the patients (47.83%) had acute myeloid leukemia. Twenty-six (56.52%) of the patients did not require oxygen therapy. The 12-week mortality was 13.33% (two out of 15) in patients who received liposomal amphotericin B compared to 25% (six out of 24) in patients who received voriconazole. There was no mortality judged to be related to IPA. Success or global clinical response was not different between the two drugs: 80% for liposomal amphotericin B versus 83.33% for voriconazole. However, the safety profile favored liposomal amphotericin B. Conclusion In this small cohort, there was an equipoise in the mortality and clinical and radiological outcomes obtained using liposomal amphotericin B or voriconazole for the treatment of IPA in hematological malignancy or HSCT.

Developing Machine-Learning Prediction Algorithm for Bacteremia in Admitted Patients.

PURPOSE: Bloodstream infection among hospitalized patients is associated with serious adverse outcomes. Blood culture is routinely ordered in patients with suspected infections, although 90% of blood cultures do not show any growth of organisms. The evidence regarding the prediction of bacteremia is scarce. PATIENTS AND METHODS: A retrospective review of blood cultures requested for a cohort of admitted patients between 2017 and 2019 was undertaken. Several machine-learning models were used to identify the best prediction model. Additionally, univariate and multivariable logistic regression was used to determine the predictive factors for bacteremia. RESULTS: A total of 36,405 blood cultures of 7157 patients were done. There were 2413 (6.62%) positive blood cultures. The best prediction was by using NN with the high specificity of 88% but low sensitivity. There was a statistical difference in the following factors: longer admission days before the blood culture, presence of a central line, and higher lactic acid-more than 2 mmol/L. CONCLUSION: Despite the low positive rate of blood culture, machine learning could predict positive blood culture with high specificity but minimum sensitivity. Yet, the SIRS score, qSOFA score, and other known factors were not good prognostic factors. Further improvement and training would possibly enhance machine-learning performance.