Exploring drivers of patient satisfaction using a random forest algorithm.

BACKGROUND: Patient satisfaction is a multi-dimensional concept that provides insights into various quality aspects in healthcare. Although earlier studies identified a range of patient and provider-related determinants, their relative importance to patient satisfaction remains unclear. METHODS: We used a tree-based machine-learning algorithm, random forests, to estimate relationships between patient and provider-related determinants and satisfaction level in two of the main patient journey stages, registration and consultation, through survey data from 411 patients at a hospital in Abu Dhabi, UAE. Radar charts were also generated to determine which type of questions-demographics, time, behaviour, and procedure-influence patient satisfaction. RESULTS: Our results showed that the 'age' attribute, a patient-related determinant, is the leading driver of patient satisfaction in both stages. 'Total time taken for registration' and 'attentiveness and knowledge of the doctor/physician while listening to your queries' are the leading provider-related determinants in each model developed for registration and consultation stages, respectively. The radar charts revealed that 'demographics' are the most influential type in the registration stage, whereas 'behaviour' is the most influential in the consultation stage. CONCLUSIONS: Generating valuable results, the random forest model provides significant insights on the relative importance of different determinants to overall patient satisfaction. Healthcare practitioners, managers and researchers can benefit from applying the model for prediction and feature importance analysis in their particular healthcare settings and areas of their concern.

Publisher Correction: The Scales Project, a cross-national dataset on the interpretation of thermal perception scales.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

The Scales Project, a cross-national dataset on the interpretation of thermal perception scales.

Thermal discomfort is one of the main triggers for occupants' interactions with components of the built environment such as adjustments of thermostats and/or opening windows and strongly related to the energy use in buildings. Understanding causes for thermal (dis-)comfort is crucial for design and operation of any type of building. The assessment of human thermal perception through rating scales, for example in post-occupancy studies, has been applied for several decades; however, long-existing assumptions related to these rating scales had been questioned by several researchers. The aim of this study was to gain deeper knowledge on contextual influences on the interpretation of thermal perception scales and their verbal anchors by survey participants. A questionnaire was designed and consequently applied in 21 language versions. These surveys were conducted in 57 cities in 30 countries resulting in a dataset containing responses from 8225 participants. The database offers potential for further analysis in the areas of building design and operation, psycho-physical relationships between human perception and the built environment, and linguistic analyses.

Experiences with a prototype tracking and verification system implemented within an imaging center.

RATIONALE AND OBJECTIVES: Most health care facilities currently struggle with protecting medical data privacy, misidentification of patients, and long patient waiting times. This article demonstrates a novel system for a clinical environment using wireless tracking and facial biometric technologies to automatically monitor and identify staff and patients to address these problems. MATERIALS AND METHODS: The design of the location tracking and verification system (LTVS) was based on a workflow study which was performed to observe the physical location and movement of patient and staff at the Healthcare Consultation Center II (HCC II) running hospital information systems, radiology information systems, picture archive and communication systems, and a voice recognition system. Based on the results from this workflow study, the LTVS was designed using a wireless real-time location system and a facial biometric system integrated with the radiology information system. The LTVS was tested for its functionality in a laboratory environment, then evaluated at HCC II. RESULTS: Experimental results in the laboratory and clinical environments demonstrated that patient and staff real-time location information and identity verification can be obtained from LTVS. Warning messages can immediately be sent to alert staff when patient's waiting time is over a predefined limit, and unauthorized access to a security area can be audited. Additionally, patient misidentification can be prevented during the course of examinations. CONCLUSIONS: The system enabled health care providers to streamline the patient workflow, protect against erroneous examinations and create a security zone to prevent, and audit unauthorized access to patient health care data required by the Health Insurance Portability and Accountability Act mandate.

Predicting clinical image delivery time by monitoring PACS queue behavior.

The expectation of rapid image retrieval from PACS users contributes to increased information technology (IT) infrastructure investments to increase performance as well as continuing demands upon PACS administrators to respond to "slow" system performance. The ability to provide predicted delivery times to a PACS user may curb user expectations for "fastest" response especially during peak hours. This, in turn, could result in a PACS infrastructure tailored to more realistic performance demands. A PACS with a stand-alone architecture under peak load typically holds study requests in a queue until the DICOM C-Move command can take place. We investigate the contents of a stand-alone architecture PACS RetrieveSend queue and identified parameters and behaviors that enable a more accurate prediction of delivery time. A prediction algorithm for studies delayed in a stand-alone PACS queue can be extendible to other potential bottlenecks such as long-term storage archives. Implications of a queue monitor in other PACS architectures are also discussed.