Dopamine D1 Agonists: First Potential Treatment for Late-Stage Parkinson's Disease.

Current pharmacotherapy has limited efficacy and/or intolerable side effects in late-stage Parkinson's disease (LsPD) patients whose daily life depends primarily on caregivers and palliative care. Clinical metrics inadequately gauge efficacy in LsPD patients. We explored if a D1/5 dopamine agonist would have efficacy in LsPD using a double-blind placebo-controlled crossover phase Ia/b study comparing the D1/5 agonist PF-06412562 to levodopa/carbidopa in six LsPD patients. Caregiver assessment was the primary efficacy measure because caregivers were with patients throughout the study, and standard clinical metrics inadequately gauge efficacy in LsPD. Assessments included standard quantitative scales of motor function (MDS-UPDRS-III), alertness (Glasgow Coma and Stanford Sleepiness Scales), and cognition (Severe Impairment and Frontal Assessment Batteries) at baseline (Day 1) and thrice daily during drug testing (Days 2-3). Clinicians and caregivers completed the clinical impression of change questionnaires, and caregivers participated in a qualitative exit interview. Blinded triangulation of quantitative and qualitative data was used to integrate findings. Neither traditional scales nor clinician impression of change detected consistent differences between treatments in the five participants who completed the study. Conversely, the overall caregiver data strongly favored PF-06412562 over levodopa in four of five patients. The most meaningful improvements converged on motor, alertness, and functional engagement. These data suggest for the first time that there can be useful pharmacological intervention in LsPD patients using D1/5 agonists and also that caregiver perspectives with mixed method analyses may overcome limitations using methods common in early-stage patients. The results encourage future clinical studies and understanding of the most efficacious signaling properties of a D1 agonist for this population.

Factors Driving Resistance to Clinical Decision Support: Finding Inspiration in Radiology 3.0.

PURPOSE: The effectiveness of evidence-based guidelines (EBGs) and clinical decision support (CDS) is significantly hampered by widespread clinician resistance to it. Our study was designed to better understand the reasons for this resistance to CDS and explore the factors that drive it. METHODS: We used a mixed-methods approach to explore and identify the drivers of resistance for CDS among clinicians, including a web-based multispecialty survey exploring clinicians' impressions of the strengths and weaknesses of CDS, two clinician focus groups, and several one-on-one focused clinician interviews in which individual participants were asked to comment on their rationale for choosing imaging utilization that might not be supported by EBGs. Additionally, a unique electronic learning and assessment module known as Amplifire was used to probe clinician knowledge gaps regarding EBGs and CDS. RESULTS: In both the quantitative and qualitative portions of the study, the primary factor driving resistance to CDS was a desire to order studies not supported by EBGs, primarily for the purpose of reducing the clinician's diagnostic uncertainty. CONCLUSIONS: Our results suggest that to enhance the effectiveness of CDS, we must first address the issue of clinician discomfort with diagnostic uncertainty and the role of imaging via educational outreach and ongoing radiologist consultation.

Developing and testing an automated qualitative assistant (AQUA) to support qualitative analysis.

Qualitative research remains underused, in part due to the time and cost of annotating qualitative data (coding). Artificial intelligence (AI) has been suggested as a means to reduce those burdens, and has been used in exploratory studies to reduce the burden of coding. However, methods to date use AI analytical techniques that lack transparency, potentially limiting acceptance of results. We developed an automated qualitative assistant (AQUA) using a semiclassical approach, replacing Latent Semantic Indexing/Latent Dirichlet Allocation with a more transparent graph-theoretic topic extraction and clustering method. Applied to a large dataset of free-text survey responses, AQUA generated unsupervised topic categories and circle hierarchical representations of free-text responses, enabling rapid interpretation of data. When tasked with coding a subset of free-text data into user-defined qualitative categories, AQUA demonstrated intercoder reliability in several multicategory combinations with a Cohen's kappa comparable to human coders (0.62-0.72), enabling researchers to automate coding on those categories for the entire dataset. The aim of this manuscript is to describe pertinent components of best practices of AI/machine learning (ML)-assisted qualitative methods, illustrating how primary care researchers may use AQUA to rapidly and accurately code large text datasets. The contribution of this article is providing guidance that should increase AI/ML transparency and reproducibility.